JPH10327591A - Motor controller - Google Patents

Abstract

PROBLEM TO BE SOLVED: To secure the optimum gain of a motor over the wide rotating domain of the motor and to obtain stable output torque from the motor by generating gain data by means of a gain generating means based on at least either the rotating speed of the motor or a current command. SOLUTION: A driving section 10 supplies electric power to stator windings IU, IV, and IW based on stator current commands iu*, iv*, and iw* and stator currents iu, iv, and iw. A drive controller 11 in the driving section 10 controls the gate voltages across upper IGBTs 14U, 14V, and 14W and lower IGBTs 16U, 16V, and 16W based on the stator current commands iu*, iv*, and iw* and stator currents iu, iv, and iw. The positive and negative electrodes of a power source 12 are respectively connected to the collectors of the upper IGBTs 14U, 14V, and 14W and the emitters of the lower IGBTs 16U, 16V, and 16W. An electrolytic capacitor 13 is connected in parallel with the power source 13. When weak field control is performed highly efficiently on a motor used for an electric automobile, etc., a vibrationless output is realized.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a motor control device for obtaining a stable output torque in a wide rotation range when a motor used in an electric vehicle or the like is subjected to field weakening control with high efficiency.

[0002]

2. Description of the Related Art A brushless motor having no mechanical commutation mechanism controls a stator current flowing through a stator winding in synchronization with a rotation position of a rotor, rotates the rotor in a predetermined direction, and outputs a predetermined output. Generate torque. Since the field is formed by the permanent magnet, the efficiency is high, and it is widely used as a motor for electric vehicles.

[0003] Hereinafter, the features of the brushless motor and the field weakening control will be described, and then an example of the prior art will be described.

[Features of Brushless Motor] A permanent magnet is arranged on the rotor of the brushless motor to form a field. The same direction as the field is called the d-axis, and the direction shifted by 90 electrical degrees from the d-axis is called the q-axis. Here, the q-axis has the same direction as the back electromotive force excited in the stator winding by the field, and when a current is applied in the d-axis direction, the current phase advances. Hereinafter, FIG.
The magnitude of the current is I, the current component flowing in the d-axis direction is d-axis current Id, and the current component flowing in the q-axis direction is q
The axis current Iq and the current phase are represented by β.

When a current is applied to the stator winding, the q-axis component Vq and the d-axis component Vd of the terminal voltage are represented by the following equation (1). Further, the output torque Trq of the brushless motor is expressed as (Equation 2). Where ωe is the electrical angular velocity, R
Is the stator winding resistance, ψ is the number of armature interlinkage fluxes by the permanent magnet, Ld is the d-axis inductance, and Lq is the q-axis inductance.

[0006]

(Equation 1)

[0007]

(Equation 2)

In a surface magnet type brushless motor having a permanent magnet disposed on the surface, the d-axis inductance Ld and the q-axis inductance Lq are equal (Ld = Lq), so the second term of (Equation 2) is 0, and the magnitude of the current is When the current I is constant, the output torque Trq has a maximum value when the current phase β = 0 °. On the other hand, in the embedded magnet type motor in which the permanent magnet is embedded in the rotor, the d-axis inductance Ld is smaller than the q-axis inductance Lq (Ld <Lq). Takes the maximum value when
The second term takes the maximum value when β = 45 °. Therefore,
As shown in FIG. 21, a certain value between β = 0 ° and 45 ° (β
At 0 °), the output torque reaches the maximum value T0.

[Weakening field control] When the brushless motor runs in power, as shown in a vector diagram of the field weakening control (FIG. 22A), when the rotation speed ω of the brushless motor is increased, the induced voltage ωe · ψ increases. ωe ・ ψ 、 R ・ I
q and ω · Lq · Iq as vector addition V
However, when the voltage limit circle is reached, the brushless motor cannot increase its rotation speed beyond the rotation speed ω of the brushless motor when the voltage value V reaches the voltage limit circle.

When the power source is a battery or the like,
Due to the deterioration of the battery, the terminal voltage and the current value of the battery change. However, for simplicity, it is assumed here that the terminal voltage of the battery (the radius of the voltage limiting circle) is constant.

Next, consider increasing the rotation speed of the brushless motor. As shown in FIG. 22, by flowing Id, the voltage ω · Ld · I
d occurs. As a result, a voltage margin is generated in the brushless motor to increase the rotation speed (FIG. 22B).
When the number of rotations of the brushless motor is constant, the q-axis current Iq can flow by the generated voltage margin, and the brushless motor can further generate an output torque. When the current value is constant, the rotation speed ω can be increased by the generated voltage margin (FIG. 22C). As described above, the d-axis current I
The control for flowing d and generating a voltage margin is called field weakening control.

Further, when the magnitude of the current is kept constant and the current phase is changed, the d-axis current Id increases, so that a voltage margin occurs and there is an effect of field weakening control. At the same time, since the q-axis current Iq decreases, Lq · Iq decreases and the voltage value V decreases, so that a voltage margin is further generated.

Note that the d-axis current Id causing the voltage margin
May be the minimum necessary current for the terminal voltage supplied to the brushless motor to return within the voltage limit circle. When a d-axis current Id that is greater than the required minimum is applied, copper loss increases and the efficiency of the brushless motor deteriorates.

When the voltage has a margin, the stator current follows the stator current command well as shown in FIG. 7A, but when the voltage margin is lost, the stator current follows the stator current command as shown in FIG. 7B. do not do.

[Example of Conventional Technology] Conventionally, as a method for field-weakening control of a brushless motor, a method described in IEA-92-30, a meeting of the Institute of Electrical Engineers of Japan, has been known. d-axis current command Id * (Equation 3)
And the field weakening control is performed.

[0016]

(Equation 3)

Here, ωbase is the base rotation speed, ωma
x is the maximum rotation speed, and Id * max is the d-axis current at the maximum rotation speed ωmax.

Further, in the 1991 IEEJ Industrial Application Division National Convention Lecture Paper No. 74, PP. In steps 310 to 315, the d-axis current command Id * is calculated using the target rotation speed, d-axis winding reactance, q-axis winding reactance, stator winding resistance, and no-load induced voltage at unit speed. ,
Perform field weakening control. However, in an actual brushless motor, since the motor constant changes due to a change in resistance due to operating conditions, a change in inductance due to magnetic flux saturation, and a change with time, the d-axis calculated from an arithmetic expression as in the above-described conventional technique is used. Current command Id * is not optimal.

Further, the motor control device described in Japanese Patent Application Laid-Open No. 8-266999 detects a current error which is a difference between a stator current command and a stator current, and when the current error is large, a d-axis current command Id * Is increased, and when the current error is small, feedback control is performed to decrease the d-axis current command Id *. In this way, the minimum d-axis current command Id * is always realized, and the increase in copper loss due to the increase in the d-axis current Id is minimized.
Can be converged in real time to the operating point of highly efficient field-weakening control.

[0020]

When the above-described feedback control using the current error is used, in order to operate stably in all the operation regions, the gain indicating the rate of change of the d-axis current command Id * and the current error are shown. The reference value indicating the reference,
The design was made to have a marginal value. Therefore, when used in a wide rotation range, an optimum gain or reference value cannot be given in some operation regions, and the response or output torque may not be optimum.

An object of the present invention is to further improve the technique described in JP-A-8-266099, and particularly when used for electric vehicles, in any driving situation. An object of the present invention is to generate a smooth torque that does not cause an uncomfortable feeling in response to an operation command, and realize a safe and stable driving operation.

[0022]

In order to solve the above-mentioned problems, a motor control device according to the present invention comprises a driving means for supplying electric power to a stator winding of a motor based on current command data (hereinafter referred to as a current command). A stator current detecting means for detecting a stator current flowing through the stator winding; and a saturation generator for generating saturation data indicating how much the stator current departs from the current command based on the current command and the stator current. Degree creation means, reference value creation means for creating the reference value of the saturation, gain creation means for creating gain data indicating a rate at which the current command is changed, and the saturation, the reference value, and the gain data. In the motor control device having a current command creating means for creating the current command based on the, the gain creating means,
The gain data is created based on at least one of the number of rotations of the motor and the current command. As a result, the gain decreases as the current phase advances, so that an optimum gain can be maintained in a wide rotation range, and a stable output torque can be obtained. By decreasing the gain when the number of rotations of the motor increases, the gain becomes optimal in a wide rotation range, and a stable output torque can be obtained. Further, by decreasing the gain when the absolute value of the current command increases, the gain becomes optimal and a stable output torque can be obtained.

A motor control device according to another aspect of the present invention includes:
A driving unit that supplies power to a stator winding of a motor based on a current command, a stator current detection unit that detects a stator current flowing through the stator winding, and the stator current is set based on the current command and the stator current. Saturation creating means for creating saturation data indicating how far away from the current command, reference value creating means for creating a reference value of the saturation, and gain data indicating a rate of changing the current command. In a motor control device including a gain creating unit and a current command creating unit that creates the current command based on the saturation, the reference value, and the gain data, it is determined whether the motor is in a steady state. Steady state determination means, the current command, the saturation, the stator current, the number of rotations of the motor, and the output torque of the motor; Is composed of a vibration index calculating means for calculating a vibration index indicating the degree of vibration based on the at least one vibration, the gain generated means is to create the gain data based on the vibration index when the steady state. When the vibration index is large, the gain is reduced to reduce the vibration of the system. On the other hand, when the vibration index is small,
Increase the gain to speed up the response of the system. As described above, by increasing or decreasing the gain according to the vibration index, the optimum gain can be obtained even if the operating environment is different, if there is an individual difference of the motor control device, or if the type of the motor is different. , Making the design easier.

A motor control device according to another aspect of the present invention comprises:
A driving unit that supplies power to a stator winding of a motor based on a current command, a stator current detection unit that detects a stator current flowing through the stator winding, and the stator current is set based on the current command and the stator current. Saturation creating means for creating saturation data indicating how far away from the current command, reference value creating means for creating a reference value of the saturation, and gain data indicating a rate of changing the current command. In a motor control device, comprising: a gain creating unit; and a current command creating unit that creates the current command based on the saturation, the reference value, and the gain data. And generating reference value data based on at least one of the current command and a drive voltage that is a voltage applied to the drive unit. That. By continuously increasing the reference value when the current command increases, safer and more natural control can be obtained when used as a motor control device. Further, by increasing the reference value when the rotation speed increases, the reference value becomes optimal in a wide rotation range, and a stable output torque can be obtained. By decreasing the reference value when the power supply voltage further increases, the reference value becomes optimal in a wide rotation range, and a stable output torque can be obtained.

A motor control device according to another aspect of the present invention comprises:
A driving unit that supplies power to a stator winding of a motor based on a current command, a stator current detection unit that detects a stator current flowing through the stator winding, and the stator current is set based on the current command and the stator current. Saturation creating means for creating saturation data indicating how far away from the current command, reference value creating means for creating a reference value of the saturation, and gain data indicating a rate of changing the current command. A preparation period for determining the reference value in a motor control device including a gain creating unit and a current command creating unit that creates the current command data based on the saturation, the reference value, and the gain data. A preparation period signal generating means for generating a preparation period signal indicating the reference value, and a determination timing signal indicating a timing for determining the reference value. A determination timing signal generating unit that generates the determination timing signal, wherein the current command generation unit keeps the current command constant when the preparation period signal is generated, and the reference value generation unit generates a determination timing signal. Then, a reference value is created based on the degree of saturation. The reference value is set to an intermediate value between the saturation when there is a voltage margin and the saturation when there is no voltage margin. This allows
The reference value can be automatically set, and the design becomes easy.

A motor control device according to another aspect of the present invention includes:
A driving unit that supplies power to a stator winding of a motor based on a current command, a stator current detection unit that detects a stator current flowing through the stator winding, and the stator current is set based on the current command and the stator current. Saturation creating means for creating saturation data indicating how far away from the current command, reference value creating means for creating a reference value of the saturation, and gain data indicating a rate of changing the current command. Gain creating means, current command creating means for creating the current command data based on the saturation, the reference value, and the gain data, and state determining means for creating state index data indicating an operation state of the motor. In the motor control device, the gain creating means creates the gain data based on the state index data. When the operation state does not change, a stable operation is realized by reducing the gain.
In addition, when the operating state changes, a quick response is realized by increasing the gain.

According to another aspect of the present invention, there is provided a motor control device for supplying electric power to a stator winding of a motor based on a current command, a stator current detecting unit for detecting a stator current flowing through the stator winding, Saturation creating means for creating saturation data indicating how much the stator current departs from the current command based on the current command and the stator current, and reference value creating means for creating a reference value of the saturation. A gain creating means for creating gain data indicating a rate at which the current command is changed; a current command creating means for creating the current command data based on the saturation, the reference value, and the gain data; In a motor control device comprising: a state determination unit that creates state index data indicating a state, the reference value creation unit includes: It is intended to create the reference value based on the data. When the operation state changes in the direction in which the output increases, the current command is quickly increased by reducing the reference value to realize a quick response. On the other hand, when the operation state changes in the direction in which the output decreases, the current command is rapidly reduced by increasing the reference value, and a quick response is realized.

A motor control device according to another aspect of the present invention comprises:
In a motor control device having a driving unit for supplying electric power to a stator winding of a motor based on a current command, and a current command creating unit for creating the current command data, the current command creating unit includes a q-axis current command data And d-axis current command data, and q-axis current correction amount data based on at least one of the motor rotation speed and the current command. The q-axis current command is corrected by the q-axis current correction amount. q
A shaft current command correction means is added. By performing the q-axis current command correction, a flat output torque can be obtained with respect to the rotation speed.

A motor control device according to another aspect of the present invention comprises:
A motor for supplying electric power to the stator winding of the motor based on the current command, and a current command creating means for creating the current command, wherein the motor controls the generated magnetic flux by changing the current phase of the current command. In the control device, the current command creating means creates q-axis current command data and d-axis current command data, and controls the current so as to keep a product of the q-axis current command data and the d-axis current command data constant. This is to change the current phase of the command. When the q-axis current command data is generated as described above, the q-axis current command data decreases even if the rotational speed increases and the saturation increases, and the d-axis current command data increases due to the field weakening control. The output torque decreases as the q-axis current command data decreases. As described above, since the output torque decreases as the rotation speed increases, safe and more natural control can be realized.

A motor control device according to another aspect of the present invention includes:
A driving unit that supplies power to a stator winding of the motor based on a current command; and a motor control device including a current command creating unit that creates the current command. Adding a phase compensation amount calculating means for calculating a phase compensation amount indicating a current phase delay, wherein the phase compensation amount calculating means increases the phase compensation amount when the rotational speed of the motor increases, and the current command creating means , Generating q-axis current command data and d-axis current command data, compensating the q-axis current based on a value obtained by multiplying the phase compensation amount and the d-axis current command data, The d-axis current is compensated based on a value obtained by multiplying the current command data. The current vector is rotated by a current phase delay amount. As described above, by performing the phase compensation, safer and more natural control without a sense of incongruity is realized.

[0031]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to FIGS.

<< First Embodiment >> [Overall Configuration] FIG. 1 is a block diagram showing the operation of a motor control device according to a first embodiment. The stator windings 1U, 1V, 1W of the brushless motor are connected to the drive unit 10. The current detection units 2U, 2V, and 2W respectively include stator currents iu and i flowing through the stator windings 1U, 1V, and 1W, respectively.
v and iw are detected and output to the drive controller 11 and the saturation creation unit 20 in the drive unit 10. The rotary encoder 3 detects the rotational position θ of the rotor of the brushless motor, and outputs the detected rotational position to the rotational speed calculator 4, the 2/3 phase converter 7, and the saturation generator 20. The rotation speed calculation unit 4 calculates the rotation speed ω based on the rotation position θ, and calculates the saturation degree creation unit 20, the state determination unit 25, the reference value creation unit 30, the gain creation unit 40, and the current command creation unit 50.
And output to The drive voltage measurement unit 5 measures the voltage of the power supply 12 in the drive unit 10 as the drive voltage Vba, and outputs it to the reference value creation unit 30. The accelerator unit 6 has an accelerator value A
cc is output to the state determination unit 25 and the current command creation unit 50.

The saturation creating unit 20 determines the rotational position θ and the rotational speed ω
A saturation degree Sat is created based on the stator current command iu * and the stator current iu, and is output to the current command creation unit 50. The state determination unit 25 creates a state index Fcond based on the rotation speed ω and the accelerator value Acc, and a reference value creation unit 30
And the gain creation unit 40. The reference value creation unit 30 creates a reference value Ref based on the rotation speed ω, the power supply voltage Vba, the state index Fcond, and the current commands Id * and Iq *,
Output to the current command creation unit 50. The gain creating unit 40 determines the rotation speed ω, the state index Fcond, and the current commands Id * and Iq *
And a gain Gai is generated based on
Output to The current command creation unit 50 determines the rotation speed ω, the accelerator value Acc, the saturation Sat, the reference value Ref, and the gain Gai.
Then, a d-axis current command Id * and a q-axis current command Iq * are created and output to the 2/3 phase converter 7, the reference value creator 30, and the gain creator 40. The 2/3 phase converter 7 creates stator current commands iu *, iv *, iw * based on the rotational position θ and the current commands Id *, Iq *, and outputs them to the drive controller 11 in the drive unit 10.

The drive unit 10 receives the stator current commands iu *, i
Power is supplied to the stator windings 1U, 1V, 1W based on v *, iw * and the stator currents iu, iv, iw. The drive controller 11 in the drive unit 10 controls the stator current command iu.
*, Iv *, iw * and stator currents iu, iv, iw based on upper IGBTs 14U, 14V, 14W and lower I
The gate voltages of the GBTs 16U, 16V, and 16W are controlled. The positive electrode of the power supply 12 is connected to the upper IGBTs 14U, 14V, 1
Connected to a 4 W collector, and the negative electrode of the power supply 12 is connected to the lower IG
Connected to the emitters of BT16U, 16V, 16W.
The electrolytic capacitor 13 is connected to the power supply 12 in parallel. Upper IGBT 14U, 14V, 14W emitter and lower I
The collectors of the GBTs 16U, 16V, and 16W are respectively connected, and further, the stator windings 1U, 1V, and 1W, respectively.
Connected to. Upper diode 15U, 15V, 15
W, lower diode 17U, 17V, 17W is upper IB
GT14U, 14V, 14W, lower IGBT16U, 1
The cathode and the collector, and the anode and the emitter are connected to 6 V and 16 W, respectively.

[Overall Operation] The rotation speed calculating section 4 rotates the rotation angle θ [degre] at every sampling interval ΔT [sec].
e] is sampled, and the rotation speed ω [r /
min]. Here, it is assumed that the rotation angle θ is represented by a mechanical angle.

[0036]

(Equation 4)

The accelerator unit 6 creates an accelerator value Acc based on the accelerator pedal angle, brake pedal angle, rotation speed, shift position, and the like. The shift (not shown) has forward, reverse, parking, and neutral positions. When the brake is not depressed, the accelerator value Acc is increased in proportion to the accelerator depression angle. The sign of the accelerator value Acc at this time is defined as positive, and the power running mode is defined. When the accelerator pedal depression angle increases, the accelerator value increases, and the motor control device increases the positive output torque. When the accelerator is not depressed, the absolute value of the accelerator value Acc is increased in proportion to the brake depression angle. The sign of the accelerator value Acc at this time is negative,
Defined as regeneration mode. When the brake step angle increases,
The accelerator value is reduced (increase the absolute value of the accelerator value), and the motor control device decreases the negative output torque (increases the absolute value of the negative output torque). Further, when the shift is neutral, the accelerator value Acc is set to 0, this is defined as the coasting mode, and the output torque is set to 0 as described later. In addition, when not included in the above, for example, when the shift is forward or backward and the accelerator and the brake are not depressed, the output torque may be set to 0 as coasting. In addition, a negative accelerator value of a small magnitude may be output to generate a negative regenerative torque, which may serve as an engine brake in the internal combustion engine, provided that the user using the motor control device determines it appropriately. Good.
Note that, at this time, a configuration may be adopted in which coasting is performed when the rotation speed is low, and regeneration is performed when the rotation speed increases.

The saturation creating unit 20 integrates the absolute value | iu * -iu | of the difference (current error) between the stator current command iu * and the stator current iu for one period as shown in (Equation 5). The result of multiplying the rotation speed ω of the brushless motor by the saturation degree Sat. Here, t is an arbitrary time, and T is a cycle of the stator current command iu *.

[0039]

(Equation 5)

The integration period is obtained from the rotational position θ. The rotation position θ indicates a mechanical angle, and one cycle of the current error is one cycle of the electrical angle. Therefore, a period obtained by dividing one cycle of the rotation position θ by the number of pole pairs is one cycle of the current error (iu * -iu). Become. Specifically, when the number of magnetic poles of the brushless motor is 2p,
The integration period is set to 0 [degree] to 360 / p [degr].
ee], 360 / p [degree] to 2.360 / p
[Degree],..., (P−1) · 360 / p [d
eg) to 360 [degree]. Here, the integration period is started from 0 [degree], but may be started from a different angle. In addition, the current error (iu *-
Although the cycle of iu) is one cycle of the electrical angle, the cycle of the absolute value of the current error (| iu * -iu |) is a half cycle of the electrical angle. Therefore, the integration period may be a half cycle of the electrical angle. When there is an offset in the current error (iu * -iu), the accuracy of the integral value deteriorates. Furthermore, the integration period may be a multiple of the period of the electrical angle, but the integration period becomes longer and the saturation S
It takes time to create at, and the response of the field control deteriorates. If the integration cycle is one cycle of the mechanical angle, it is possible to eliminate the influence of uneven magnetization of the rotor of the brushless motor, but the integration cycle becomes longer. Here, for example, when the rotation speed ω is small, integration is performed only for one electrical angle period,
When the number of rotations ω is large, the number of cycles occupied by the integration period may be increased as the number of rotations ω increases, such that the integration is performed only during one mechanical angle period. Also, the integration start time t
Is given arbitrarily, but the absolute value of the current error | iu * -iu |
Is given at or near the minimum location, when the configuration of the integration operation takes a long time to reset the integration result, the effect of the reset is reduced, and a highly accurate integration result is realized.

Since the integration period is inversely proportional to the rotation speed ω, the integration value of the absolute value of the current error differs depending on the rotation speed ω even if the current error is the same. Therefore, the multiplication result of the integral value of the absolute value of the current error (| iu * -iu |) and the rotation speed ω is used as the saturation degree Sat. The state determination unit 25 monitors the rotation speed ω and the accelerator value Acc, sets the state index Fcond = 1 when the operation state changes in a direction in which the output increases, and sets the state index Fcond to 1 when the operation state changes in a direction in which the output decreases. , State index Fc
otherwise, state index Fcond
= 0. Specifically, when the rate of increase of the rotational speed ω is equal to or greater than a set value or when the rate of increase of the accelerator value Acc is equal to or greater than a set value, the state index Fcond = 1. On the other hand, when the decrease rate of the rotation speed ω is equal to or more than a set value or when the decrease rate of the accelerator value Acc is equal to or more than a set value, the state index Fcond is set to −1. Here, when the direction of the increase or decrease of the output is not known such that the accelerator value Acc is decreasing while the rotational speed ω is increasing, the state index Fcond = 0.

The reference value creation unit 30 stores the reference value R
Create ef. The gain creating section 40 creates the gain Gai by a method described later.

Next, the operation of the current command generator 50 will be described. FIG. 2 is a flowchart showing the operation of the current command creating unit in the first embodiment, and the details will be described below.

In step (110), the creation of a current command is started. In a process (111), an accelerator value Acc, a rotation speed ω, a saturation degree Sat, a reference value Ref, and a gain Gai are input.

In step (112), a current index Iin is created. As shown in (Equation 6), the result of multiplication of a certain set constant Kacc and the accelerator value Acc is set as the current index Iin.
Here, the constant Kacc is set so that when the absolute value of the accelerator value Acc becomes the maximum, the absolute value of the current index Iin becomes the maximum value of the absolute value of the current command in the current index circumferential operation mode described later. . Further, the maximum value of the absolute value of the current command in the current index circumferential operation mode is IGBT
Is set based on the maximum allowable current and the maximum value of the torque output.

[0046]

(Equation 6)

In the processing (113), the minimum value Id * min of the d-axis current command Id * and the maximum value Id * of the d-axis current command Id *
Create max. As shown in (Equation 7), the product of the absolute value | Iin | of the current index and sinβ0 is the minimum value Id * mi.
n. Here, β0 is a current phase β that gives the maximum value of the output torque Trq of the brushless motor as shown in FIG.
It is 0 ° for the surface magnet type brushless motor, and is a certain value between 0 ° and 45 ° for the embedded magnet type brushless motor. On the other hand, a certain set constant Kidm
ax2 and the absolute value | Iin | of the current index plus a set constant Kidmax1 in the result of multiplication by the maximum value I
Let d * max.

[0048]

(Equation 7)

In step (114), a d-axis current command Id * is created. As shown in (Equation 8), the reference value R is obtained from the saturation degree Sat.
The result of subtracting ef is multiplied by the gain Gai, and the result of adding the previous value Id * old of the d-axis current command is set as the d-axis current command Id *. Here, the previous value Id * old stores the previous value of the d-axis current command.

[0050]

(Equation 8)

In step (115), the d-axis current command Id * is limited. When the d-axis current command Id * is smaller than the minimum value Id * min (Id * <Id * min), the d-axis current command I
d * is changed to the minimum value Id * min (Id * = Id *
min). Also, the d-axis current command Id * is the maximum value Id * m
When it is larger than ax (Id *> Id * max), the d-axis current command Id * is changed to the maximum value Id * max (Id *).
= Id * max).

In the process (116), the operation branches into a powering mode and a coasting / regeneration mode depending on the operation state. Since the sign of the current index Iin matches the sign of the accelerator value Acc, powering is performed when the current index Iin is positive, coasting is performed when the current index Iin is 0,
Regeneration is performed when the value is negative. Therefore, when the current index Iin is positive, the process (117) is executed next in the powering mode. When the current index Iin is 0 or negative, coasting
The mode is set to the regeneration mode, and then the process (120) is executed.

In the process (117), the operation branches into a current index circumferential operation mode and a d-axis operation mode depending on the magnitude of the d-axis current command Id *. When the d-axis current command Id * is smaller than the current index (Id * <Iin), the process (118) is executed next as the current index circumference operation mode. When the d-axis current command Id * is equal to or greater than the current command (Id * ≧ Ii
n) Next, the process (119) is executed as the d-axis operation mode.

In the process (118), a q-axis current command Iq * (power running, current indicator circumferential operation mode) is created. The square root of the result of subtracting the square value of the d-axis current command Id * from the square value of the current index value Iin is defined as a q-axis current command Iq *.

In step (119), a q-axis current command Iq * (power running, operation mode on d-axis) is created. q-axis current command Iq
Set * to 0.

In step (120), a q-axis current command Iq * (coasting / regeneration mode) is created. Current index Iin and cos
The multiplication result of β0 is set as a q-axis current command Iq *.

In step (121), q-axis current command correction (Iq * correction) is performed by a method described later. In step (122), the phase of the current command is compensated by the method described later.

In step (123), a d-axis current command Id * and a q-axis current command Iq * are output. Processing (124)
Saves the d-axis current command Id *. Since the d-axis current command Id * is used in the next process, the d-axis current command Id * is set to a new previous value Id * old of the d-axis current command. here,
0 is given to Id * old as an initial value.

In step (125), the generation of the current command is completed.

Next, a method of q-axis current command correction (Iq * correction) in the process (121) will be described. FIG. 3 is a flowchart showing the operation of the q-axis current command correction in the first embodiment, and the details will be described below.

In step (130), q-axis current command correction is started. In the process (131), the q-axis current command Iq * and the rotation speed ω are input.

In the process (132), the q-axis current correction amount Iqco
Create r. As shown in (Equation 9), when the rotation speed ω is larger than a certain constant Kiq2, the result obtained by multiplying the rotation speed ω by the constant Kiq2 and the certain constant Kiq1 is multiplied by the q-axis current correction amount Iqcor. I do. Also,
When the rotation speed ω is equal to or less than the constant Kiq2, the q-axis current correction amount I
qcor is set to 0.

[0063]

(Equation 9)

In the process (133), the operation branches into the powering mode and the coasting / regeneration mode depending on the operation state. q-axis current index Iq
Since the sign of * matches the sign of the current index Iin, when the q-axis current command Iq * is positive, power running is performed, and when it is negative, regeneration is performed. Here, when it is 0, both the d-axis operation mode of power running and the coasting time are applicable. The result of the corrected q-axis current command Iq * is the same regardless of which of the two branches, so here, for the sake of convenience, both are set to the coasting mode without any problem. Therefore, when the q-axis current command Iq * is positive, the process (134) is executed next in the powering mode. When the q-axis current command Iq * is 0 or negative, the process is executed in the coasting / regeneration mode, and the process (136) is executed next.

In step (134), the flow branches to a q-axis current correction mode and a q-axis current non-correction mode according to the magnitude of the q-axis current command Iq *. The q-axis current command Iq * is the q-axis current correction amount I
When it is smaller than qcor (Iq * <Iqcor), the process is executed next in the q-axis current correction mode (135). When the q-axis current command Iq * is equal to or larger than the q-axis current correction amount (Iq * ≧ Iqcor), the q-axis current command Iq * is not corrected in the q-axis current non-correction mode.
Execute 7).

In step (135), the q-axis current command Iq * (power running, q-axis current correction mode) is corrected. The q-axis current correction amount Iqcor is defined as a q-axis current command Iq *.

In step (136), the q-axis current command Iq * (coasting / regeneration mode) is corrected. q to q-axis current command Iq *
A value obtained by adding the axis current correction amount Iqcor is set as a new q-axis current command Iq *.

In step (137), a q-axis current command Iq * is output. In the process (138), the q-axis current command correction ends.

In this embodiment, since the q-axis current command correction is independently operated, the operation of branching from the operating state to the powering mode and the coasting / regeneration mode overlaps with the processing (116) and the processing (133). Although it appeared, the operation of the q-axis current command correction was divided and included in the operation of the current command creation, and the branch was set to 1
It may be one.

Next, the method of phase compensation in the process (122) will be described. FIG. 4 is a flowchart showing the operation of the phase compensation in the first embodiment, which will be described in detail below.

In step (140), phase compensation is started. In the process (141), the rotational speed ω, the d-axis current command Id *, and the q-axis current command Iq * are input.

In step (142), a phase compensation amount Pcom is created. As shown in (Equation 10), a certain set constant Kp
The result of multiplication of the rotation speed ω.com and the rotation speed ω is defined as a phase compensation amount Pcom. Here, the unit of Pcom is radian.

[0073]

(Equation 10)

In the process (143), the phase is compensated. (Equation 1
As in 1), the phase compensation amount Pcom and the d-axis current command Id
* And the q-axis current command Iq * to perform phase compensation.

[0075]

[Equation 11]

At step (144), a d-axis current command Id * and a q-axis current command Iq * are output. Processing (145)
Ends the phase compensation.

Then, the 2/3 phase converter 7 is given by (Equation 12)
, The d-axis current command Id * and the q-axis current command Iq * on the two-phase rotational coordinates are replaced with the stator current command i on the three-phase stationary coordinates.
Convert to u *, iv *, iw *. Here, θe is an electric rotation angle, and is appropriately converted from the rotation angle θ. The cycle of the electric rotation angle θe is 1 / the cycle of the rotation angle θ.
It is p times. (The number of magnetic poles of the brushless motor rotor is 2
p-pole).

[0078]

(Equation 12)

The driving unit 10 includes stator windings 1U, 1V, 1
A current represented by the stator current commands iu *, iv *, iw * is passed through W. Hereinafter, the details will be described.

The power supply voltage 12 supplies power to the drive unit 10,
Electrolytic capacitor 13 smoothes power supply voltage 12.

FIG. 5 is a circuit diagram of a drive controller according to the first embodiment.
U, 61V, 61W, comparators 62U, 62V, 62W,
And a triangular wave generating circuit 63. The operational amplifiers 61U, 61V, and 61W calculate the stator current i from the stator current commands iu *, iv *, and iw * as shown in (Equation 13).
The result of subtracting u, iv, and iw and the result of multiplication of the current minor gain Ke, which is a constant, are created as PWM current errors eu, ev, and ew, respectively. Here, (Equation 1)
In 3), only the proportional operation is performed, but a proportional / integral operation or a proportional / integral / differential operation may be performed. The triangular wave generation circuit generates a triangular wave (several kHz to several tens kHz). The comparator 62U compares the PWM current error eu with the triangular wave, and
When the current error eu for M is large, the upper IGBT 14U is energized and the lower IBGT 16U is de-energized. When the current error eu for PWM is small, the upper IBGT 14U is de-energized.
The lower IBGT 16U is energized. Where the upper IGB
When the energized state of T14U and lower IGBT 16U transitions, both upper IGBT 14U and lower IGBT 16U are de-energized, and a short time (dead time) is provided to prevent short circuit of drive power supply 12. The same operation is performed for the other two phases.

[0082]

(Equation 13)

FIG. 6 is a waveform diagram showing PWM signals at the time of saturation and at the time of non-saturation. FIG. 6 shows a triangular wave, a current error eu for PWM, the operation of the upper IBGT 14U and the operation of the lower IGBT 16U. It is shown. In addition,
FIG. 6 shows a waveform when a dead time is not provided for simplicity. FIG. 7 is a waveform diagram showing the stator current command and the stator current at the time of saturation and at the time of non-saturation, and shows the stator current command iu * and the stator current iu.

At the time of non-saturation, the amplitude of the PWM current error eu is smaller than the amplitude of the triangular wave, and there is a margin in the power supply voltage.
As shown in FIG. 7A, the stator current iu flows according to the stator current command iu *. On the other hand, at the time of saturation, the amplitude of the PWM current error eu is larger than the amplitude of the triangular wave, and the upper IGBT 14U or the lower IGBT 16U
Is in a state where there is no room for a voltage that is always energized,
As shown in FIG. 7B, the stator current iu does not flow according to the stator current command iu *.

That is, at the time of non-saturation, the voltage has a margin and the current error is small, and at the time of saturation, the voltage has no margin and the current error is large. Further, as described in [Field-weakening control] of the related art, increasing the d-axis current Id increases the voltage margin,
As the d-axis current Id decreases, the voltage margin decreases.

Therefore, by using a value indicating the magnitude of the current error as the saturation Sat, d is calculated by increasing or decreasing the saturation Sat.
The shaft current Id is increased or decreased to control the voltage margin. Where P
Due to the characteristic characteristic of the WM control, even when there is a voltage margin, the stator current iu is not equal to the stator current command iu * as shown in FIG. Therefore,
When the saturation Sat at this time is set as the reference value Ref, and the saturation Sat is larger than the reference value Ref (Sat> Re
f) It is determined that there is no voltage margin, and the d-axis current is increased. When the saturation Sat is smaller than the reference value Ref (Sat <R
ef), it is determined that the voltage margin is sufficient, and the d-axis current is reduced, and feedback control is performed to make the saturation Sat equal to the reference value Ref. As described above, by flowing the minimum necessary d-axis current Id, the copper loss is minimized.

Processing in current command creating section 50 (114)
Since FIG. 2 performs the above-described operation, the optimum Id is realized, and the field weakening control is performed with high efficiency.

As described above, the degree of saturation Sat may be any value as long as it indicates the magnitude of the current error. Therefore, the saturation degree Sat is not limited to the multiplication result of the integral of the absolute value of the current error (| iu * -iu |) and the rotation speed as in the present embodiment (Equation 5). For example, the current error of another phase is used, the average value of the integrated values of two or more phases is used, the sum of the square values of the current errors of each phase is used, and the absolute value and the square value of the current error are passed through the LPF. A value may be used. Further, the stator currents iu, iv, iw are converted into 3/2 phases, and the d-axis current I
The saturation Sat may be created based on the d- and q-axis currents Iq, and a comparison result with the d-axis current command Id * and the q-axis current command Iq *. For example, the q-axis current command Iq * and the q-axis current I
q error (proportional q-axis current error: Iq * -Iq), q
The integral of the error between the axis current command Iq * and the q-axis current Iq (integration of q-axis current error 誤差 (Iq * -Iq) dt), or the sum of both at a certain set ratio (q Proportional integration of shaft current error: KPQ (Iq * -Iq)
+ KIQ∫ (Iq * −Iq) dt) may be used. Also, not only the q-axis current but also the d-axis current may be noted.
An error between the d-axis current command Id * and the d-axis current Id (proportionality of the d-axis current error: Id * -Id) or the d-axis current command Id * and d
The integral of the error of the axis current Id (integration of the d-axis current error: ∫ (Id * −Id) dt), or the sum of both at a set ratio (the proportional integral of the d-axis current error) : KPD (Id * −Id) + KID∫ (Id
* -Id) dt) may be used. Further, both the q-axis current and the d-axis current may be considered. For example, as the degree of saturation, the sum of the square of the integration of the q-axis current error and the square of the integration of the d-axis current at a certain ratio ({KIQ} (Iq * −
Iq) dt)｝ ² + ｛∫ (Id * −Id) dt)｝ ² )
Or the sum of the square of the proportional integral of the q-axis current error and the square of the proportional integral of the d-axis current at a certain ratio (｛KPQ (I
q * -Iq) + KIQ∫ (Iq * -Iq) dt)} 2 +
{KPD (Id * -Id) + KID} (Id * -Id)
dt)｝ ² ) may be used.

[Operation of Reference Value Generating Unit 30] The reference value generating unit 30 performs a predetermined constant Ref as shown in (Equation 14).
0 and the correction coefficient Krefa for the q-axis current command Iq *
A correction coefficient Krefb for the rotation speed ω, a correction coefficient Krefc for the power supply voltage Vba, and a state index Fco.
The result of multiplication by the correction coefficient Frefd for the nd is created as a reference value Ref.

[0090]

[Equation 14]

Correction coefficient Kr for q-axis current command Iq *
8a, the q-axis current command Iq * is Iq *
When it is 1 or more (Iq * ≧ Iq * 1), Krefa1 is set. At the time of regeneration (Iq * <0), Krefa2 is set. Further, when the q-axis current command is positive and smaller than Iq * 1 (0 ≦ Iq * <Iq * 1), the q-axis current command Iq * is a linear function, and continuously when Iq * = 0 and Iq * 1. To be. Here, Krefa1 is Krefa
Greater than 2. Thus, the correction coefficient Krefa is continuously and monotonically increased with respect to the q-axis current command Iq *.

As shown in FIG. 8B, the correction coefficient Krefb for the rotational speed ω is such that the intercept is Krefb for the rotational speed ω.
1. A linear function in which the slope is monotonically increasing of Krefb2.

Correction coefficient Kref for power supply voltage Vba
8C, as shown in FIG. 8C, with respect to the power supply voltage Vba, the intercept is monotonically decreasing with Krefc1 and the slope being −Krefc2.
The next function. Correction coefficient K for state index Fcond
refd is the state index Fcond as shown in (Equation 14).
= 1, the correction coefficient Krefd = Krefd1, and when the state index Fcond = 0, the correction coefficient Krefd = Kr.
efd2, and the correction coefficient Krefd = Krefd3 when the state index Fcond = −1. Where Kref
It is assumed that d1 <Krefd2 <Krefd3.

[Operation of Gain Creator 40] The gain creator 40 calculates the constant Gai0, the correction coefficient Kgaia for the current phase command β *, the correction coefficient Kgaib for the rotation speed ω, and the current command as shown in (Equation 15). Size | I * |
The multiplication result of the correction coefficient Kgaic relating to the state index Fcond and the correction coefficient Kgaid relating to the state index Fcond
Create as i. Here, the magnitude of the current command | I * |
And the current phase command β * are respectively the q-axis current command Iq
The magnitude and current phase of the current command represented by * and the d-axis current command Id * (FIG. 20).

[0095]

(Equation 15)

Correction coefficient Kga for current phase command β *
When the current phase command β * is β0 (β * = β0) as in FIG. 9A, ia is set to Kgaia1. When the current phase command β * is equal to or more than β * 1 (β * ≧ β * 1), Kg
aia2. Further, when the current phase command β * is equal to or larger than β0 and smaller than β * 1 (β0 ≦ β * <β * 1), the current phase command β * is a linear function, and β * = β0, β *
When 1, it becomes continuous. Here, Kgaia1
Is greater than Kgaia2. Thus, the correction coefficient Kg
aia is decreased continuously and monotonously with respect to the current phase command β *.

As shown in FIG. 9B, the correction coefficient Kgaib for the rotational speed ω is such that the intercept is Kgaib for the rotational speed ω.
1. A linear function of -Kgaib2 with a monotonically decreasing slope.

The correction coefficient Kgaic for the magnitude of the current command | I * | is, as shown in FIG.
Regarding I * |, the intercept is Kgaic1 and the slope is -Kga.
Let ic2 be a monotonically decreasing linear function. State index Fco
When the state index Fcond = 0, as shown in (Equation 15), the correction coefficient Kgaid for the nd is
= Kgaid1, and when the state index Fcond = −1 and the state index Fcond = 1, the correction coefficient Kgaid =
Kgaid2. Here, Kgaid1 <Kgai
d2.

[Operation and Effect of Current Command Vector] In the case of power running, processing (117) in current command creating section 50 (FIG. 2)
In the current index circumference operation mode (processing (11
8)) and a d-axis operation mode (process (119)). FIG. 10A is an explanatory diagram showing the locus of the current vector in the first embodiment, and the circumference indicates the size of the current index. When the rotational speed ω of the brushless motor increases, the voltage margin eventually disappears, and the saturation Sat increases. Since the d-axis current command Id * increases by the process (114), the current commands (Iq *, Id *) change in the direction of the arrow. First, the current commands (Iq *, Id *) move on the current index circle until the current phase command changes from β0 to 90 ° (current command circumferential operation mode), and thereafter, the d-axis current command moves on the d axis. Move while increasing Id * (operation mode on d-axis). Here, β0 is a current phase (FIG. 21) for realizing the maximum output torque Trq of the brushless motor,
In the current command circumferential operation mode, the current phase β advances with an increase in the rotational speed ω, and the output torque Trq decreases (the lower right portion in FIG. 10B). In the d-axis operation mode, since Iq * = 0, the output torque Trq is obtained from (Equation 2).
= 0 (Trq = 0 in FIG. 10B).

When the current index proportional to the accelerator value increases in the order of (1) → (2) → (3), the output torque Trq also increases in the order of (1) → (2) → (3).

When used as a control device for a motor for an electric vehicle, when the accelerator pedal is depressed at a constant angle and the rotational speed increases and the output torque increases, the electric vehicle is accelerated beyond the driver's request. I feel uncomfortable. In addition, since the driver drives the electric vehicle on the premise that the output torque increases when the accelerator is depressed, when the output torque decreases when the accelerator is depressed and the accelerator value is increased, the driver feels discomfort and feels discomfort. Riding comfort may be worse.

In this embodiment, when the rotational speed ω increases, the output torque Trq decreases. Since the direction in which the accelerator value increases and the direction in which the output torque Trq increases are equal in all the rotation ranges, a safer and more natural operation is possible. Can be realized.

[Effects of Gain Creating Section 40] Gain Gai
Is increased, the settling time of the d-axis current command Id * becomes earlier from (Equation 8), so that the gain Gai should be larger. However, when the gain Gai is increased, the saturation S
The d-axis current command Id * greatly fluctuates due to noise included in at, a calculation time delay, and the like, and accordingly, the output torque Trq greatly fluctuates. Then, when used as a control device for a motor for an electric vehicle, the riding comfort of the electric vehicle deteriorates. Therefore, it is necessary to reduce the gain Gai to some extent and suppress the vibration of the d-axis current command Id *.

FIG. 21 shows that the maximum output torque Tr is obtained.
In the vicinity of the current phase β0 that realizes q, the slope of the output torque Trq with respect to the fluctuation of the current phase β is slight, but the current phase β advances and the slope is large near 90 °. Output torque Tr
q fluctuates greatly. Further, as shown in (Equation 16), when the current phase β is around 90 °, (I · cos β) becomes small. Therefore, even if the d-axis current Id is slightly changed, the current phase β largely changes. Therefore, if the gain Gai is constant regardless of the current phase command β *, the current phase β becomes β
Gain G at which d-axis current command Id * does not oscillate when it is near 0
Even if ai is given, the current command Id * may oscillate when the current phase β is around 90 °. Here, (Equation 15)
When the current phase command β * increases, the correction coefficient Kgaia related to the current phase command β * decreases.

As described above, when the current phase β advances, the gain G
By reducing ai, an optimum gain Gai is obtained in a wide rotation range, and a stable output torque Trq is realized.

[0106]

(Equation 16)

When the rotational speed ω of the brushless motor increases, the integration period of the absolute value of the current error for generating the saturation Sat becomes shorter (Equation 5), and the accuracy of the saturation Sat deteriorates. Therefore, if the gain Gai is kept constant irrespective of the rotation speed ω, the d-axis current command Id * may oscillate in the high rotation range even if the d-axis current command Id * does not oscillate in the low rotation range. Here, as shown in (Equation 15), when the rotation speed ω increases, the correction coefficient Kgaib regarding the rotation speed ω decreases.

Thus, the rotational speed ω of the brushless motor
As the gain increases, the gain Gai is decreased, so that the optimum gain Gai is obtained in a wide rotation range, and a stable output torque Trq is realized.

Further, when the magnitude | I * | of the current command is increased, noise generated by the drive unit 10 may increase the noise corresponding to the saturation degree Sat. Therefore, when the gain Gai is constant regardless of the magnitude | I * | of the current command, when the magnitude | I * | of the current command is small, even if the d-axis current command Id * does not oscillate, the current command When the magnitude | I * | is large, the d-axis current command Id * may vibrate. Here, as shown in (Equation 15), when the magnitude | I * | of the current command increases, the correction coefficient Kgaic related to the magnitude | I * | of the current command decreases.

As described above, when the absolute value | I * | of the current command increases, the gain Gai is reduced, so that the optimum gain Gai is achieved, and a stable output torque Trq is realized. Further, when the operating state such as the accelerator value Acc or the rotation speed ω changes, the value at which the d-axis current command Id * converges changes. Therefore, if the gain Gai is kept constant irrespective of the operation state, the response may deteriorate when the operation state changes. Here, as shown in (Equation 15), when the operation state does not change (state index Fcond = 0), the correction coefficient Kgaid is reduced, and when the operation state changes (state index Fcond = 1 and Fcond =).
-1) Increase the correction coefficient Kgaid. Thus, when the operating state does not change, a stable operation is realized by reducing the gain Gai. Further, when the operating state changes, the gain Gai is increased to realize a quick response.

In this embodiment, the current phase command β *, the rotation speed ω, the magnitude of the current command | I * | and the state index Fcon
The correction has been made with respect to d, but it is effective to correct at least one of them. In particular, the motor control device of the present embodiment does not have a voltage margin when the rotation speed ω increases, and operates so as to increase the voltage margin by advancing the current phase command β *. There is a strong correlation, and using only one of them has a great effect.

Further, the method of correction is not limited to the function of the present embodiment, and various modifications are possible.

Further, instead of the current phase command β *, d
The ratio of the axis current command Id * to the q-axis current command Iq * (Id * /
Iq *) may be used. Current command magnitude | I
Instead of * |, the magnitude | Iq * | of the q-axis current command may be used. Further, the state index Fcond is (−1,
Although it has three values of 0 and 1), it may have a continuous value and the correction coefficient Kgaid may be continuously changed. In that case,
More precise control. The state index Fcond is not set to 0 only when the state changes, but after the state changes, the state index Fcon is set for a certain set time.
d = −1 or the state index Fcond = 1 may be held. Even if the accelerator value Acc changes, the d-axis current command I
d * does not immediately converge to a certain value and takes some time. By holding the value of the state index Fcond for this time, the convergence time of the d-axis current command Id * is shortened, and a quick response is realized.

[Effect of Reference Value Creation Unit 30] When the reference value Ref during regeneration and the reference value Ref during power running are constant and the reference value Ref during regeneration is smaller than the reference value Ref during power running, The d-axis current command Id * tends to be larger. Therefore, when the current index Iin for the slight power running and the current index Iin for the small regeneration are given, the d-axis current command Id * is larger in the regeneration and the output torque Trq
Becomes larger. Since the driver drives the electric vehicle on the assumption that the output torque increases when the accelerator is depressed, if the output torque decreases when the accelerator is depressed and the accelerator value is increased, the driver feels discomfort and rides the electric vehicle. May be uncomfortable. Here, as shown in (Equation 14), when the q-axis current command Iq * decreases, the q-axis current command Iq
The correction coefficient Krefa for * is continuously reduced.

As described above, when the q-axis current command Iq * increases, the reference value Ref is continuously increased, so that when used as a control device for a motor for an electric vehicle, safer and more natural driving is realized. .

When the rotation speed ω increases, the delay of the stator current iu with respect to the stator current command iu * increases,
Since the current error (iu * -iu) increases, the saturation S
at increases. Therefore, even if an optimum reference value is set in a low rotation speed region, the saturation level increases in a high rotation speed region, and the reference value becomes relatively small, so that an optimum d-axis current command Id * may not be created. Here, as shown in (Equation 14), when the rotation speed ω increases, the correction coefficient Kref related to the rotation speed ω
b is increased.

As described above, by increasing the reference value Ref when the rotation speed ω increases, the optimum reference value Ref is obtained in a wide rotation range, and a stable output torque Trq is realized.

Further, when the power supply voltage Vba increases, the stator current iu follows the stator current command iu * well, and the current error (iu * -iu) decreases, so that the saturation degree Sat decreases. Therefore, even if an optimal reference value is set at a certain power supply voltage, the power supply voltage Vb
When a increases, the saturation degree Sat decreases and the reference value relatively increases, so that an optimal d-axis current command Id * is not created. Here, as shown in (Equation 14), when the power supply voltage Vba increases, the correction coefficient Kref related to the power supply voltage Vba.
Decrease c.

As described above, when the power supply voltage Vba increases, the reference value Ref is reduced, so that the reference value Ref is optimal in a wide rotation range, and a stable output torque Trq is realized. Further, when the operating state such as the accelerator value Acc or the rotation speed ω changes, the value at which the d-axis current command Id * converges changes. Therefore, if the reference value Ref is kept constant irrespective of the operation state, the response may deteriorate when the operation state changes. Here, when the operation state changes in the direction in which the output increases as in (Equation 14) (when the accelerator value Acc or the rotation speed ω increases: the state index Fc)
the correction coefficient Krefd is reduced. On the other hand, when the operation state changes in the direction in which the output decreases (when the accelerator value Acc or the rotation speed ω decreases: the state index Fcond = −1), the correction coefficient Krefd is increased. Now, when the operating state changes in a direction to increase the output, it is necessary to increase the d-axis current Id in order to increase the effect of the field weakening control. On the other hand, when the operation state changes in a direction in which the output decreases, it is necessary to reduce the d-axis current Id in order to reduce the effect of the field weakening control. As described above, when the operation state changes in the direction in which the output increases, the d-axis current command Id * is quickly increased by decreasing the reference value Ref (Equation 8).
Achieve fast response. On the other hand, when the operation state changes in the direction in which the output decreases, the d-axis current command Id * is quickly reduced by increasing the reference value Ref (Equation 8), and a quick response is realized.

Although the q-axis current command Iq *, the rotation speed ω, the power supply voltage Vba, and the state index Fcond are corrected in the present embodiment, it is effective to correct at least one of them. Here, when the power supply voltage Vba increases, the saturation Sat decreases, but the voltage margin increases and the rotation speed ω increases, so that the operation region shifts to the high rotation side and the saturation Sat increases. Thus, the power supply voltage Vba
Increases, the saturation Sat decreases due to the increase in the power supply voltage Vba itself, and the saturation Sat increases due to the increase in the rotation speed ω.
Increases simultaneously, and therefore, it is different depending on each system whether the reference value should be increased or decreased. Therefore, caution is required when the correction is performed based on the power supply voltage Vba instead of the correction based on the rotation speed ω.

Further, the method of correction is not limited to the function of the present embodiment, and various modifications are possible.

Further, when the rotational speed ω increases, the voltage margin disappears, and the saturation Sat increases, so that the d-axis current command I
In order to increase d * and advance the current phase command β *, the current phase command β * may be used instead of the rotation speed ω. Further, the state index Fcond has three values of (-1, 0, 1), but may have a continuous value, and the correction coefficient Kgaid may be continuously changed. In that case, more precise control is performed. Although the state index Fcond is not set to 0 only when the state changes, the state index Fcond = −1 or the state index Fcond = 1 may be held for a certain set time after the state changes. Accelerator value A
Even if cc changes, the d-axis current command Id * does not immediately converge to a certain value, and it takes some time. Only for this time,
By holding the value of the state index Fcond, the convergence time of the d-axis current command Id * is shortened, and a quick response is realized.

[Effect of q-axis Current Command Correction] The brushless motor generates a negative output torque when rotated, even if the stator current commands iu *, iv *, iw * are set to 0. Less than,
This will be described in detail.

For simplicity, a single-phase inverter will be described. FIG. 11A is a circuit diagram of a single-phase inverter. Driven by the drive voltage Vba, the IGBT (A,
A ', B, B') and a diode form a full-brick circuit, and the back electromotive voltage Vbemf is connected to the load. Here, the load is composed of a motor inductance Lmot and a motor resistance Rmot.

When a current opposing the back electromotive voltage Vbemf is applied, the IGBT (A, A ') is energized and the IGBT (B, B,
B ′) is de-energized. At this time, the current is indicated by a dotted line (AA ′).
on), the load has a difference (Vb
a-Vbemf). On the other hand, when a current flows in the direction of the back electromotive voltage Vbemf, the IGBT (A,
A ′) is de-energized, and the IGBT (B, B ′) is energized. At this time, since the current tends to flow in the direction of the dashed line (BB'on), the sum (Vba + Vbem) is applied to the load.
The voltage of f) is applied. Here, the gradient of the current when the energization is switched is proportional to the voltage applied to the load. Therefore, as shown in FIG. 11B, when the current that is opposed to the back electromotive voltage Vbemf flows (AA′on). The gradient of the current is smaller than when the current flows in the direction of the back electromotive voltage Vbemf (BB'on). Now, at the time of power running, the back electromotive voltage Vb
Since a current is applied to the emf and a current flows in the direction of the back electromotive voltage during regeneration, the positive side in FIG. 11B is a power running side and generates a positive torque, and the negative side is a regeneration side and a negative torque. Occurs. As shown in FIG. 11B, since the average of the current waveform is negative, the average generates a negative torque.

As described above, during coasting (when the current index Iin is 0), the driving unit 10 controls the stator current i
When u, iv, and iw are set to 0, a negative torque is generated as shown by a broken line in FIG. Here, when the rotation speed ω increases, the induced voltage increases, so that the absolute value of the generated negative torque also increases.

Now, in an automobile driven by an internal combustion engine, the clutch can be disengaged and transmission of torque to the drive wheels can be eliminated. On the other hand, in order to improve the transmission efficiency of the motor and the drive system and reduce the cost of the electric vehicle, a transmission-less and clutch-less structure in which the drive wheels are directly connected via gears or the like is desired. For this reason, the transmission of the clutch disengaging torque cannot be eliminated. Therefore, when the brushless motor rotates, a braking force is generated by the negative torque.

Therefore, as in the process (121) (FIG. 3), the correction is performed by adding the positive correction amount Iqcor to the q-axis current command Iq *. Since the generated torque increases in the positive direction when the q-axis current command increases, the output torque Trq is corrected as shown by the solid line in FIG. In addition, since the absolute value of the negative torque increases as the rotation speed ω increases, the correction amount Iqcor increases as the rotation speed increases, thereby realizing a flat torque.

As described above, by performing the q-axis current command correction, a flat output torque is realized with respect to the rotation speed ω.

Here, a description will be given of the meaning of branching into the q-axis current correction mode processing (135) and the q-axis current non-correction mode (no processing) in the processing (134) in FIG. The q-axis current correction is peculiar to coasting and regeneration, and it is not necessary to reduce the output torque as the number of revolutions increases during power running, since this does not pose a problem as a characteristic of the electric vehicle motor. Now, the current correction amount Iqcor = 20A
At this time, consider the case of coasting (Iq * = 0 before correction) and the case of slight power running (Iq * = 1 before correction). The corrected q-axis current command Iq * during coasting is 20A. on the other hand,
If the q-axis current is not corrected during power running, the q-axis current command Iq
* Is 1A, the q-axis current command Iq * is larger and the output torque Trq is larger during coasting when the accelerator value Acc is smaller than during power running, and the magnitude relationship between the accelerator value Acc and the output torque Trq is reversed. May feel uncomfortable. Therefore, when the q-axis current command Iq * is smaller than the q-axis correction amount Iqcor during power running, the q-axis current command Iq *
Let q * be the q-axis correction amount Iqcor.

As described above, the q-axis current command Iq *
Is small, the q-axis current command is corrected, and the q-axis current command I
When q * increases, the safer and more natural operation is realized by eliminating the q-axis current command correction.

The correction amount Iqco is represented by a linear function of the rotational speed.
Although r is given, it may be given by another function.

Also, at the time of regeneration, only the d-axis current command Id * is increased by the field weakening control.
The term increases and the negative torque increases. Here, the q-axis current command Iq * may be corrected from (Equation 2) so that the output torque Trq is constant. Here, the q-axis current command I
As q * is smaller (absolute value is larger), the second in (Equation 2)
Since the increment of the negative torque of the term is large, the q-axis current command Iq
The correction amount of * increases.

Further, in the present embodiment, the q-axis current correction during power running is configured not to correct the q-axis current command when the q-axis current command Iq * increases. The q-axis correction amount Iqcor may be reduced. This embodiment shows an embodiment in which the q-axis correction amount Iqcor is reduced with the increase of the q-axis current command Iq *.

In this embodiment, the field-weakening control and the q-axis current correction are combined. However, the field-weakening control of another method using a map or the like is combined with the q-axis current correction,
It is effective to use the q-axis current correction alone.

[Effect of Compensation for Current Phase Delay] Drive Unit 1
0 (FIG. 1) indicates the stator current commands iu *, iv *, iw *.
, The stator currents iu, iv, iw are delayed from the stator current commands iu *, iv *, iw * due to the inductance of the stator windings 1U, 1V, 1W. . When the rotation speed ω increases, the stator current commands iu *, iv *, i
Since the cycle of w * is short, the delay is relatively large. This delay is called a current phase delay.

When the current phase delay increases, the driver of the electric vehicle may feel uncomfortable. For example, in the coasting mode in the current command creating unit 50, the current command has only the d-axis current command Id * and the q-axis current command Iq * is 0 in the high-speed rotation range by the field weakening control. FIG. 13 shows this state, and the magnitude of the current command is | I * | and the current vector having only the Id * component is shown on the Iq * -Id * system. If there is a current phase delay, the Iq * -Id * system is rotated on the Iq * -Id * system by the current phase delay amount Pdel. Does not have a q-axis component (Iq * component) like (0, | I * |), but in an effective Iq * '-Id *' system, (| I * | sinPde
l, | I * | cosPdel) as q-axis component (Iq
* ′ Component) | I * | sinPdel. Therefore, from (Equation 2), a positive torque is generated.

As a result, when the accelerator value Acc is 0 (in the coasting mode) in the high-speed rotation range, a positive torque is generated, so that the driver may feel uncomfortable and the riding comfort of the electric vehicle may deteriorate. Here, the current vector (Iq *,
Id *) is rotated by the current phase delay amount Pdel, and converted into a corrected Iq * "-Id *" system as shown in (Equation 17). Here, the second equation assumes that Pcom is small,
This is a second-order approximation, and is used as (Equation 11) in the phase compensation process (122) in the current command creating unit 10.

[0139]

[Equation 17]

As described above, by performing the phase compensation,
Realize safe and more natural driving without a sense of incongruity in driving.

In the present embodiment, the phase compensation amount Pcom is approximated by the second order. However, the first order approximation or the third-order approximation can also be used.

The phase compensation amount Pcom may be proportional to the rotational speed ω as shown in (Equation 10), or may be given by another function or table.

Further, in this embodiment, the saturation Sat
Although the field-weakening control is performed by using the method described above, even when the field-weakening control is performed by another method using a table or the like, it is effective to perform the phase compensation by the above-described method.

<< Second Embodiment >> In the second embodiment, the q-axis current command Iq * in the powering mode of the current command creation unit 250 is used.
The other configuration is the same as that of the first embodiment, except that the method of creating is changed, and redundant description is omitted.

FIG. 14 is a flow chart showing the operation of the current command generator in the second embodiment. The current command creation unit 250 of the present embodiment performs the operation of the process (201) in place of the processes (117) to (119) in the current command creation unit 50 in the first embodiment, and other parts are included. Are duplicated and will not be described.

[Operation of Current Command Generating Unit] Processing (201)
Creates a q-axis current command Iq * (powering mode). As in (Equation 18), a value obtained by dividing the result of multiplication of the current index Iin, cos β0 and the minimum value Id * min of the d-axis current command by the d-axis current command Id * is defined as a q-axis current command Iq *.

The other parts are the same as those of the first embodiment, and the description is omitted.

[0148]

(Equation 18)

[Operation and Effect of Current Command Vector] As described above, when the q-axis current command Iq * is created, the rotation speed ω increases, the saturation degree Sat increases, and the field-weakening control creates the (d-axis current command Id *). (By the operation of the process (114))
Even if the d-axis current command Id * increases, the q-axis current command Iq *
Is reduced, and the result of multiplication of the d-axis current command Id * and the q-axis current command iq * is kept at Iin · cosβ0 · Id * min (Id * · Iq * = Iin · cosβ0 · Id * m).
in = const). From equation (2), the q-axis current command I
Due to the decrease of q *, the first term of the output torque Trq decreases,
Since the second term is more constant than Id * · Iq * is constant, the output torque Trq decreases.

When used as a motor for an electric vehicle, if the output torque increases as the number of revolutions increases, the driver may feel uncomfortable and the riding comfort of the electric vehicle may deteriorate.

On the other hand, in this embodiment, when the rotational speed ω increases, the output torque Trq decreases, so that safer and more natural driving can be realized.

In this embodiment, the current index circumference operation mode (process (118), FIG. 2) of the first embodiment is eliminated, but the current index circumference operation mode and Id * .Iq * are fixed. There is also an effect when used in combination with operating the current vector while keeping the current value, and is included in the present invention.

In this embodiment, the field weakening control is performed using the saturation degree Sat. However, even when the field weakening control is performed by another method using a table or the like, it is effective to operate the current command by the above-described method.

Third Embodiment A motor control device according to a third embodiment obtains a gain Gai in real time from a vibration such as a saturation degree Sat.

FIG. 15 is a block diagram showing the operation of the motor control device according to the third embodiment. In this embodiment,
Compared to the first embodiment, the first embodiment differs from the first embodiment only in that the state determination unit 25 is deleted, the steady state determination unit 350 and the vibration index calculation unit 360 are added, and the operation of the gain creation unit 340 is different. The description of the same configuration as that described above is omitted.

[Overall Configuration and Operation] Steady State Determination Unit 35
0 generates a steady state signal Sta indicating whether or not the apparatus is in a steady state based on the rotational speed ω and the accelerator value Acc by a method described later, and outputs the signal to the gain creating section 340. The vibration index calculation unit 360 creates a vibration index Vib based on the saturation degree Sat by a method described later, and outputs the vibration index Vib to the gain creation unit 340. The gain creating unit 340 calculates the steady state signal Sta and the vibration index Vi
b, a gain Gai is created by the method described below,
Output to the current command creation unit 50.

Also, as compared with the first embodiment, the rotation speed calculation unit 4 outputs the rotation speed ω to the steady state determination unit 350.
It is not output to gain creation section 340. Accelerator unit 6 also outputs accelerator value Acc to steady state determination section 350. The saturation generator 20 also outputs the saturation Sat to the vibration index calculator 360. The current command creating unit 50 does not output the current commands (Id *, Iq *) to the gain creating unit 340.

Since other configurations are the same as those of the first embodiment,
Description is omitted. [Operation of Steady State Judgment Unit 350] Steady state judgment unit 350
Monitors the rotation speed ω and the accelerator value Acc, and during a certain set period, the difference between the maximum value and the minimum value of the rotation speed ω is equal to or less than a certain set value, and the maximum value of the accelerator value Acc is When the difference between the minimum values is equal to or less than a set value, it is determined that a steady state has occurred, and the steady state signal Sta is set to the H level. Otherwise, it is determined that it is not in a steady state,
The steady state signal Sta is set to L level.

In this embodiment, the rotational speed ω and the accelerator value Acc are monitored. However, if at least one of the saturation Sat, the current command, the stator current, and the like is used and a steady state can be determined, whatever is monitored. Good.

[Operation of Vibration Index Calculation Unit 360] As shown in (Equation 19), the vibration index calculation unit 360 calculates the maximum value max (Sat) and the minimum value min (Sat) of the saturation degree Sat in a certain set period. Is referred to as a vibration index Vib.

[0161]

[Equation 19]

[Operation of Gain Creating Section 340] FIG. 16 is a flowchart showing the operation of the gain creating section in the third embodiment, and the details will be described below.

In the process (341), the generation of gain is started.
In the process (342), the steady state signal Sta and the vibration index Vib are input.

In the process (343), the operation branches into a gain change mode and a gain non-change mode according to the steady state signal Sta. When the steady state signal Sta is at H level (Sta =
H) Then, the process (344) is executed as the gain change mode. Further, when the steady state signal Sta is at the L level (Sta = L), the process is executed next in the gain non-change mode (346).

In the process (344), a gain Gai is created (gain change mode). As shown in (Equation 20), a value obtained by subtracting a certain constant Rgai from the vibration index Vib is multiplied by a predetermined constant Ggai, and a value obtained by adding the previous value Gaold of the gain is defined as the gain Gai.
Here, the previous value Gaiold stores the previous value of the gain Gai.

[0166]

(Equation 20)

In the process (345), the gain Gai is limited. Set minimum value Gami with gain Gai
n (Gai <Gaimin), the gain G
ai is changed to the minimum value Gaimin (Gai = Gai = Gaimin)
min). Also, when the gain Gai is larger than a certain set maximum value Gaimax (Gai> Gaimax)
x), the gain Gai is changed to the maximum value Gaimax (Gai = Gaimax).

In the process (346), the gain Gai is created (gain non-change mode). The previous value Gaiold is set as the gain Gai.

At step (347), the gain Gai is stored. In order to use Gai in the next processing, the gain Gai is set as a new previous value Gaiold of the gain.

It should be noted that a predetermined constant is given to Gaold as an initial value. In processing (348), the gain Gai is output.

In the process (349), the creation of the gain Gai ends. [Effects of Third Embodiment] In the first embodiment, the gain Gai
Is the current phase command β *, the rotation speed ω, and the magnitude of the current command | I
* | And the state index Fcond. Also, due to noise on the saturation degree Sat, etc.,
The gain Gai could not be increased so much. Here, the noise varies depending on the operating environment and individual differences between the control devices of the individual motors. For this reason, the control response may be delayed in order to allow a margin Gai smaller than the limit. Further, when the type of the motor changes, the gain Gai must be changed, which may make the design difficult. In this embodiment, when the vibration index Vib indicating the vibration of the system is large as in (Equation 20), the gain Gai is reduced to reduce the vibration of the system. On the other hand, when the vibration index Vib is small, the gain Gai is increased to speed up the response of the system.

As described above, in this embodiment, by increasing or decreasing the gain Gai based on the vibration index Vib,
Even if the operating environment is different, there is an individual difference in the motor control device, or the type of motor is different, the optimum gain G
ai is realized. This facilitates the design.

In this embodiment, the vibration coefficient Vib is calculated from the saturation Sat. Here, when the saturation Sat oscillates, the d-axis current command I created based on the saturation Sat
d * also oscillates and the stator current commands iu *, iv *, iw
* Also vibrates. Then, the effective values of the stator currents iu, iv, iw also oscillate, the output torque Trq also oscillates, and the rotational speed ω
Also vibrate. Therefore, not only the saturation degree Sat but also the current command, the stator currents iu, iv, iw, and the output torque Tr
The vibration coefficient Vib may be calculated based on at least one of q and the rotation speed ω.

Further, in the present embodiment, the difference between the maximum value and the minimum value in a certain period is used as the vibration coefficient.

The steady state may be determined as long as the vibration coefficient Vib can be sufficiently obtained from the saturation degree Sat or the like, and the system may move macroscopically.

The initial value of the gain Gai is given by the method of the first embodiment, or the gain Gai given by the method of the present embodiment is given.
i may be corrected by the method of the first embodiment.

Fourth Embodiment A motor control device according to a fourth embodiment obtains a reference value Ref in real time from the relationship between the saturation degree Sat and the rotation speed ω.

FIG. 17 is a block diagram showing the operation of the motor control device according to the fourth embodiment. In this embodiment,
As compared with the first embodiment, the drive voltage measurement unit 5 and the state determination unit 25 are deleted, and a preparation period signal creation unit 470 and a decision timing signal creation unit 480 are added. This embodiment is different from the first embodiment only in the operation of the reference value creating unit 430 and the current command creating unit 450, and therefore, the description of the overlapping configuration will be omitted.

[Overall Configuration and Operation] The preparation period signal generation unit 470 uses the preparation period signal S based on the rotation speed ω, the accelerator value Acc, and the determination timing signal Mak in a manner described later.
tp is generated, and this is determined by the decision timing signal generator 480.
And the current command creation unit 450. The determination timing signal creation unit 480 creates a determination timing signal Mak based on the rotation speed ω, the saturation degree Sat, and the preparation period signal Stp by a method described later, and uses this as a preparation period signal creation unit 470.
And the reference value creation unit 430. Reference value creation unit 43
A value 0 creates a reference value Ref based on the saturation degree Sat and the determination timing signal Mak by a method described later, and outputs this to the current command creation unit 450.

Also, as compared with the first embodiment, the rotation speed calculator 4 outputs the rotation speed ω to the preparation period signal generator 470 and the decision timing signal generator 480, but the reference value generator 430 Do not output. Accelerator unit 6 also outputs accelerator value Acc to preparation period signal generator 470. The saturation generator 20 also outputs the saturation Sat to the determination timing generator 480 and the reference value generator 430. The current command creation unit 450 creates a current command by a method described later, but does not output the current command to the reference value creation unit 430.

Since other configurations are the same as those of the first embodiment,
Description is omitted. Next, the operation of the current command creation unit 450 will be described. The current command generation unit 450 outputs the preparation period signal Stp
Is at the H level during the preparation period, the current command (Id *,
Iq *) is kept constant. On the other hand, when the preparation period signal Stp becomes L
If the level is not the preparation period, the current commands (Id *, Iq *) are created in the same manner as in the first embodiment. FIG. 18 is a flowchart showing the operation of the current command creating unit in the fourth embodiment, and the details will be described below.

In step (451), the creation of a current command is started. In the process (452), the preparation period signal Stp is input.

In step (453), the process branches into a current command change mode and a current command non-change mode according to the preparation period signal Stp. When the preparation period signal Stp is at the L level (Stp =
L), as the current command change mode, next processing (111)
Execute Further, when the preparation period signal Stp is at the H level (Stp = H), the current command non-change mode is set, and then the process (454) is executed.

Steps (111) to (123) are the same as those in the first embodiment, and a duplicate description will be omitted.

In the process (454), the current commands (Id *, Iq
*) Is created (current command non-change mode). Previous value Id
* Old is the d-axis current command Id *, and the previous value Iq * ol
Let d be a q-axis current command Iq *.

In this embodiment, as described later, the preparation period signal generator 470 initially sets the preparation period signal Stp in the initial state.
To H. Therefore, when the current command creation unit 450 is first executed, the preparation period signal is not H, and the current command change mode is executed.
It is not necessary to set the initial values of * old and Iq * old.

At step (455), the d-axis current command Id * and the q-axis current command Iq * are stored. Since the d-axis current command Id * and the q-axis current command Iq * are used in the next process, the d-axis current command Id * is set to a new previous value Id * old of the d-axis current command, and the q-axis current command Iq * is set to A new previous value Iq * old of the q-axis current command is set.

In the process (456), the creation of the current command is completed. The other configuration is the same as that of the first embodiment, and the description is omitted.

[Operation of Preparation Period Signal Generating Unit 470] In the low rotation range where the field weakening control is not performed and the accelerator value A
When cc is constant, the preparation period is set, and the preparation period is continued until the reference value is determined (until the determination timing signal Mak becomes H). If the accelerator value Acc is changed during the preparation period, the preparation period is canceled. Specifically, the following is performed.

The initial state is not the preparation period, and the preparation period signal Stp is set to L (Stp = L).

When the rotational speed ω is smaller than a set value ωstp (ω <ωstp) and the difference between the maximum value and the minimum value of the accelerator value Acc is smaller than a set constant for a set period. , And the preparation period signal Stp is set to H (Stp = H).

In the state where the preparation period signal is H (Stp = H), when the accelerator value Acc is changed and when the decision timing signal Mak becomes H, the preparation period is canceled, and Signal is L (Stp = L)
And

[Operation of Determination Timing Signal Generating Unit 480] During the preparation period, the rotation speed ω and the saturation Sat are monitored, and when the inclination of the saturation Sat with respect to the rotation speed ω suddenly changes, a pulse-like determination timing signal is output. (Positive logic). Specifically, the following is performed.

When the preparation period signal Stp becomes H, the operation is started, and the inclination of the saturation degree Sat with respect to the rotation speed ω is calculated and stored (stored as Ssat0). Hereinafter, the slope (Ssat) of the saturation degree Sat with respect to the rotation speed ω is continuously calculated, and the calculated slope Ss with respect to the stored slope Ssat0.
When the ratio with at becomes larger than a set value Rsat (Ssat / Ssat0> Rsat), a decision timing signal is generated. Then, the operation ends.

If the preparation period signal Stp becomes L during the operation, the operation is interrupted and terminated. [Operation of Reference Value Generating Unit 430] Decision Timing Signal Ma
The saturation degree Sat when k is input is updated as the reference value Ref. Specifically, the following is performed.

In the initial state, a set initial value is set as a reference value Ref. Then, the decision timing signal Mak
And the saturation Sat0 at that time is set as a reference value Ref (Ref = Sat0).

[Effect of Fourth Embodiment] Saturation degree creating section 20
The saturation Sat varies among individual brushless motors or changes over time due to the variation of component parts and aging.
Therefore, when the brushless motor is operating, the optimum reference value R
If ef is set, highly efficient field-weakening control can be realized without being affected by variations and aging.

As shown in FIG. 19A, when the rotational speed ω is increased while the current command and the load are kept constant, the saturation Sat gradually increases due to a phase delay or the like while there is a voltage margin in a low rotational speed range. Then, when the voltage margin is lost in the high rotation region, the stator current iu does not follow the stator current command iu *, and the saturation degree Sat rapidly increases. Therefore, the place where the gradient of the saturation degree Sat with respect to the rotation speed ω changes suddenly is an intermediate place between when there is a voltage margin and when there is no voltage margin. Therefore, the saturation S is larger than the saturation Sat0 at this time.
When at is small, there is a voltage margin, and when the saturation Sat is larger than the saturation Sat0 at this time, there is no voltage margin. Therefore, the saturation Sat0 at this time is set to the reference value Ref.
(Ref = Sat0), the d-axis current command Id * is increased or decreased, and the operation is continued at an intermediate position between when there is a voltage margin and when there is no voltage margin.

As described above, the reference value Ref can be automatically set, and the design becomes easy.

As shown in FIG. 19B, when the rotational speed ω is increased while the current command and the load are kept constant, the output torque Trq decreases when the voltage margin is lost, so that the slope of the output torque Trq with respect to the rotational speed ω changes suddenly. The saturation Sat at the place where the operation is performed may be used as the reference value Ref. Further, since the angular acceleration with respect to the rotational speed ω also changes, the saturation degree Sat at the place where the angular acceleration changes may be used as the reference value Ref.

Further, in the present embodiment, the saturation Sat0 when the inclination of the saturation Sat with respect to the rotational speed ω changes is set as the reference value Ref, but any value near this value is included in the present invention. In particular, when a value Sat1 (FIG. 19) slightly larger than the saturation degree Sat0 when the inclination changes is set as the reference value Ref, the output torque Tr slightly decreases, but when the rotation speed ω changes from a large state to a small state ( B → A), the difference between the reference value Ref and the degree of saturation is large (Ref-Sat), d
A quick convergence of the shaft current command Id * is realized.

The initial value of the reference value Ref may be given by the method of the first embodiment, or the reference value Ref given by the method of this embodiment may be corrected by the method of the first embodiment. .

In this embodiment, when the reference value Ref is low and the accelerator value Acc is constant, the operation of creating the reference value Ref is continued. However, the operation is continued until the reference value Ref is created once after starting. Alternatively, the reference value Ref may be stored by operating only at the time of shipment.

In the first to fourth embodiments, the current command creating units 50, 250, and 450 use (Equation 8)
As described above, the d-axis current command Id * was increased or decreased depending on the degree of saturation Sat. Increasing the d-axis current command Id * is equivalent to advancing the current phase command β *. Therefore,
Appropriately design the current command creation unit, as shown in (Equation 21),
The current phase command β * may be increased or decreased. Where β * ol
d is the previous value of the current phase command β *.

[0205]

(Equation 21)

Further, in the above-described embodiment, only the integration operation was performed in the creation of the d-axis current command Id * in (Equation 8) and the creation of the gain Gai in (Equation 20). But,
Any device that performs at least one of a proportional operation, an integral operation, and a differential operation, as well as the integral operation, is included in the present invention. For example, a d-axis current command I including a proportional / integral / differential operation
The creation of d * is shown in (Equation 22). Where Id *
(I) is the i-th d-axis current command Id *, and e (i) is the difference (Sat-Re) between the i-th saturation Sat and the reference value Ref.
f), KI is integral gain, KP is proportional gain, and K
D is a differential gain, and (i-1) and (i-2) show the operation one before and two before the i-th, respectively. The above embodiment shows a configuration in which only KI is not 0 and KP and KD are 0. The present invention includes at least one of KI, KP, and KD, if at least one of them is corrected or determined from the vibration index Vib.

[0207]

(Equation 22)

In the first to fourth embodiments, the current command generators 50, 250, and 450 increase or decrease the d-axis current command Id * according to the saturation Sat as shown in (Equation 8). Increasing the d-axis current command Id * is equivalent to decreasing the q-axis current command Iq *. Therefore,
The current command creation unit may be appropriately designed to increase or decrease the q-axis current command Iq *. In the first, second, and fourth embodiments, the reference value creation units 30, 430 change the magnitude of the reference value Ref. Here, increasing the degree of saturation Sat has the same meaning as reducing the reference value Ref. Therefore, since changing the magnitude of the saturation Sat instead of changing the magnitude of the reference value Ref is essentially equal, changing the magnitude of the saturation Sat is also included in the present invention. Note that a specific method corresponding to the reference value creation unit 30 is as shown in (Equation 23).

[0209]

(Equation 23)

In the first to fourth embodiments, the current command generators 50, 250, and 450 increase or decrease the d-axis current command Id * according to the saturation Sat as shown in (Equation 8).
The size to be increased / decreased per calculation may be limited.
Further, a dead zone may be provided. Further, hysteresis may be provided. These have the effect of stabilizing the control.

In the drawings for explaining the above-described embodiments, the brushless motor is an embedded magnet type brushless motor, and the current phase β0 for realizing the maximum torque is not 0. The present invention is not limited to a brushless motor, and may be used in a surface magnet type brushless motor or a brushless motor having saliency.

The motor control device described above does not control only a brushless motor, but is also included in the present invention when used for controlling another motor such as a synchronous reluctance motor.

The stator current commands iu *, iv *, iw
*, And three phases are used for the stator currents iu, iv, and iw. However, only two phases may be used, and a value obtained by adding these two phases and changing the sign may be used as another one phase. According to the stator current command iu *, iv *, iw *, the stator current i
Although the feedback control is performed so that u, iv, and iw flow, the d-axis current Id and the q-axis current Iq are obtained by using two or more of the stator currents iu, iv, and iw. Feedback control may be performed so that the current flows according to the q-axis current command Iq *.

Although the IBGT is used in the driving unit 10,
Other driving devices such as bipolar transistors and MOS-FETs may be used. Although the power supply is a DC power supply, a DC power supply obtained by rectifying an AC power supply may be used.

[0215]

As apparent from the detailed description of the embodiments, the present invention has the following effects. That is, when the motor used in an electric vehicle or the like is subjected to field control with high efficiency, an output torque without vibration is realized, the output torque does not increase with respect to the number of revolutions, and the output torque coincides with the increase and decrease of the accelerator pedal depression angle. Thus, a motor control device that generates a stable output torque in a wide rotation range even when there is a variation in set values, a change over time, and an environmental change even when the output torque becomes 0 during coasting.

[Brief description of the drawings]

FIG. 1 is a block diagram of a motor control device according to a first embodiment.

FIG. 2 is a flowchart illustrating an operation of a current command creating unit according to the first embodiment;

FIG. 3 is a flowchart showing the operation of q-axis current command correction in the first embodiment.

FIG. 4 is a flowchart showing an operation of phase compensation in the first embodiment.

FIG. 5 is a circuit configuration diagram of a drive controller according to the first embodiment.

FIG. 6 is a waveform chart showing PWM signals at the time of saturation and at the time of non-saturation;

FIG. 7A is a waveform diagram showing a stator current command and a stator current at the time of non-saturation, and FIG. 7B is a waveform diagram showing a stator current command and a stator current at the time of saturation.

FIG. 8A is a diagram illustrating a relationship between a reference value correction coefficient and a q-axis current command according to the first embodiment; FIG. 8B is a diagram illustrating a relationship between a reference value correction coefficient and a rotation speed according to the first embodiment; ) Shows the relationship between the reference value correction coefficient and the power supply voltage in the first embodiment.

9A is a diagram illustrating a relationship between a gain correction coefficient and a current phase command according to the first embodiment. FIG. 9B is a diagram illustrating a relationship between a gain correction coefficient and a rotation speed according to the first embodiment. FIG. 7 is a diagram showing the relationship between the gain correction coefficient and the magnitude of the current command in the first embodiment.

10A is a locus diagram of a current vector in the first embodiment, and FIG. 10B is an explanatory diagram showing torque characteristics when the current vector is set as shown in FIG. 10A.

11A is a circuit diagram of a single-phase inverter, and FIG. 11B is a current waveform diagram of FIG.

FIG. 12 is an explanatory view showing torque characteristics before and after q-axis current command correction in the first embodiment.

FIG. 13 is an explanatory diagram showing a current vector when there is a current phase delay.

FIG. 14 is a flowchart illustrating the operation of a current command creating unit according to the second embodiment.

FIG. 15 is a block diagram of a motor control device according to a third embodiment.

FIG. 16 is a flowchart illustrating an operation of a gain creating unit according to the third embodiment.

FIG. 17 is a block diagram of a motor control device according to a fourth embodiment.

FIG. 18 is a flowchart illustrating an operation of a current command creation unit according to the fourth embodiment.

FIG. 19A is a diagram illustrating the relationship between the rotation speed, the reference value, and the saturation in the fourth embodiment. FIG. 19B is a diagram illustrating the relationship between the rotation speed, the reference value, and the output torque in the fourth embodiment. Illustrated illustration

FIG. 20 is a relationship diagram of a d-axis current, a q-axis current, and a current phase.

FIG. 21 is a diagram showing a relationship between a current phase and an output torque in an interior magnet type brushless motor.

FIGS. 22A to 22C are vector diagrams of field-weakening control;

[Explanation of symbols]

 1U, 1V, 1W Stator winding 2U, 2V, 2W Current detection unit 3 Rotary encoder 4 Rotation speed calculation unit 5 Drive voltage measurement unit 6 Accelerator unit 7 2/3 phase conversion unit 10 Drive unit 20 Saturation degree creation unit 25 State judgment Unit 30, 430 Reference value creation unit 40, 340 Gain creation unit 50, 250, 450 Current command creation unit 350 Steady state determination unit 360 Vibration index calculation unit 470 Preparation period signal creation unit 480 Decision timing signal creation unit

 ──────────────────────────────────────────────────の Continuing on the front page (72) Inventor Satoshi Tamaki 1006 Kadoma Kadoma, Osaka Prefecture Matsushita Electric Industrial Co., Ltd. (72) Inventor Mineaki Isoda 1006 Kadoma, Kadoma City, Osaka Prefecture Matsushita Electric Industrial Co., Ltd.

Claims (26)

[Claims] A driving means for supplying electric power to a stator winding of the motor based on a current command; a stator current detecting means for detecting a stator current flowing through the stator winding; A saturation creation unit that creates saturation data indicating how much the stator current deviates from the current command; a reference value creation unit that creates a reference value of the saturation; and a rate at which the current command is changed. A gain control means for generating gain data, and a current command preparation means for preparing the current command data based on the saturation, the reference value, and the gain data. Generating the gain data based on at least one of a rotation speed of the motor and the current command. Control device. 2. The motor control device according to claim 1, wherein said gain creating means decreases said gain when a current phase of said current command advances. 3. The motor control device according to claim 1, wherein said gain creating means decreases said gain when the rotation speed of said motor increases. 4. The motor control device according to claim 1, wherein said gain creating means decreases said gain when an absolute value of said current command increases. 5. A driving unit for supplying electric power to a stator winding of a motor based on a current command, a stator current detecting unit for detecting a stator current flowing through the stator winding, and based on the current command and the stator current. A saturation creation unit that creates saturation data indicating how much the stator current deviates from the current command; a reference value creation unit that creates a reference value of the saturation; and a rate at which the current command is changed. A motor control device comprising: gain creating means for creating gain data shown; and current command creating means for creating the current command data based on the saturation, the reference value, and the gain data, wherein the motor is in a steady state. Steady-state determining means for determining whether or not the current command, the current command, the saturation, the stator current, and the number of revolutions of the motor. Vibration index calculating means for calculating a vibration index indicating a degree of vibration based on at least one vibration with the output torque of the motor, wherein the gain data is based on the vibration index when the gain creating means is in the steady state. A motor control device, characterized in that a motor is created. 6. The vibration index calculating means sets the amplitude of the saturation vibration as the vibration index, and the gain creating means creates the gain so as to decrease the gain when the vibration index increases. The motor control device according to claim 5, wherein 7. The current command creating means creates q-axis current command data and d-axis current command data, and the vibration index calculating means sets an amplitude of vibration of the d-axis current command as the vibration index, 6. The motor control device according to claim 5, wherein the gain creating unit creates the gain so as to decrease the gain when the vibration index increases. 8. A driving unit for supplying electric power to a stator winding of a motor based on a current command, a stator current detecting unit for detecting a stator current flowing through the stator winding, and based on the current command and the stator current. A saturation creation unit that creates saturation data indicating how much the stator current deviates from the current command; a reference value creation unit that creates a reference value of the saturation; and a rate at which the current command is changed. A motor control device comprising: gain creating means for creating gain data shown; and current command creating means for creating the current command data based on the saturation, the reference value, and the gain data, wherein the reference value creating means Is a reference based on at least one of the number of rotations of the motor, the current command, and a drive voltage that is a voltage applied to the drive means. A motor control device for generating value data. 9. The current command creating means creates q-axis current command data and d-axis current command data, and the reference value creating means increases the reference value when the q-axis current command increases. The motor control device according to claim 8, wherein: 10. The motor control device according to claim 8, wherein the reference value creation unit decreases the reference value when the current phase of the current command advances. 11. A driving means for supplying electric power to a stator winding of a motor based on a current command, a stator current detecting means for detecting a stator current flowing through the stator winding, and based on the current command and the stator current. A saturation creation unit that creates saturation data indicating how much the stator current deviates from the current command; a reference value creation unit that creates a reference value of the saturation; and a rate at which the current command is changed. A motor control device comprising: gain creating means for creating gain data shown; and current command creating means for creating the current command data based on the saturation, the reference value, and the gain data, wherein the reference value is determined. A preparation period signal generating means for generating a preparation period signal indicating that the preparation period is to be performed; and a timing for determining the reference value. Decision timing signal creation means for creating a decision timing signal, wherein the current command creation means keeps the current command constant when the preparation period signal is generated, and the reference value creation means has a decision timing A motor control device, wherein when a signal is generated, reference value data is generated based on the degree of saturation. 12. The preparation period signal generating means generates the preparation period signal when the number of rotations of the motor is smaller than a predetermined number of rotations and a torque command indicating an output torque to be generated by the motor is constant. 12. The motor control device according to claim 11, wherein the generation of the preparation period signal is canceled when the determination timing signal is generated and when the torque command is changed. 13. The determination timing signal generating means,
12. The motor control device according to claim 11, wherein the determination timing signal is generated when a gradient of the saturation degree with respect to the rotation speed of the motor changes. 14. The determination timing signal generating means,
12. The motor control device according to claim 11, wherein the determination timing signal is generated when the acceleration of the motor changes. 15. A driving means for supplying electric power to a stator winding of a motor based on a current command, a stator current detecting means for detecting a stator current flowing through the stator winding, and A saturation creation unit that creates saturation data indicating how much the stator current deviates from the current command; a reference value creation unit that creates a reference value of the saturation; and a rate at which the current command is changed. Gain creating means for creating the gain data shown; current command creating means for creating the current command data based on the saturation, the reference value, and the gain data; and state index data showing the operating state of the motor. A motor control device having a state determining means, wherein the gain creating means creates the gain data based on the state index. A motor controller, characterized by. 16. A driving unit for supplying electric power to a stator winding of a motor based on a current command, a stator current detecting unit for detecting a stator current flowing through the stator winding, and based on the current command and the stator current. A saturation creation unit that creates saturation data indicating how much the stator current deviates from the current command; a reference value creation unit that creates a reference value of the saturation; and a rate at which the current command is changed. Gain creating means for creating the gain data shown; current command creating means for creating the current command data based on the saturation, the reference value, and the gain data; and state index data showing the operating state of the motor. In the motor control device having state determination means, the reference value creation means creates the reference value data based on the state index. A motor controller, characterized by. 17. The motor control device according to claim 15, wherein the state determination unit creates the state index data based on a rate at which the number of rotations of the motor changes. 18. The motor according to claim 15, wherein the state determination unit creates the state index data based on a rate at which a torque command indicating an output torque generated by the motor changes. Control device. 19. A motor control device comprising: a driving unit that supplies electric power to a stator winding of a motor based on a current command; and a current command generating unit that generates the current command. Creating d-axis current command data and d-axis current command data; creating q-axis current correction amount data based on at least one of the number of rotations of the motor and the current command; A motor control device, further comprising a q-axis current command correction means for correcting an amount. 20. A driving unit for supplying electric power to a stator winding of a motor based on a current command, a stator current detecting unit for detecting a stator current flowing through the stator winding, and based on the current command and the stator current. A saturation creation unit that creates saturation data indicating how much the stator current deviates from the current command; a reference value creation unit that creates a reference value of the saturation; and a rate at which the current command is changed. A motor control device comprising: gain creating means for creating gain data shown; and current command creating means for creating the current command data based on the saturation, the reference value, and the gain data. Creates q-axis current command data and d-axis current command data, and sets at least one of the motor rotation speed and the current command Hazuki creates a q-axis current correction amount data, a motor controller, characterized in that the q-axis current command q-axis current command correction means for correcting only the q-axis current correction amount is added. 21. The motor control device according to claim 19, wherein the q-axis current command correction means increases the q-axis current correction amount as the rotation speed of the motor increases. . 22. The q-axis current command correction means,
21. The motor control device according to claim 19, wherein the q-axis current correction amount is reduced when the shaft current command increases. 23. A driving device for supplying electric power to a stator winding of a motor based on a current command, and a current command creating device for creating the current command data, wherein a current phase of the current command is changed to generate a magnetic flux. In the motor control device for controlling the amount, the current command creating means creates q-axis current command data and d-axis current command data, and keeps a product of the q-axis current command data and the d-axis current command data constant. A motor control device, wherein the current phase of the current command is changed so as to keep the current phase. 24. A driving unit for supplying electric power to a stator winding of a motor based on a current command, a stator current detecting unit for detecting a stator current flowing through the stator winding, and based on the current command and the stator current. A saturation creation unit that creates saturation data indicating how much the stator current deviates from the current command; a reference value creation unit that creates a reference value of the saturation; and a rate at which the current command is changed. A motor control device comprising: gain creating means for creating gain data shown; and current command creating means for creating the current command data based on the saturation, the reference value, and the gain data. Creates q-axis current command data and d-axis current command data, and calculates the product of the q-axis current command data and the d-axis current command data. A motor control device, wherein the current phase is changed so as to be kept constant. 25. A motor control device comprising: driving means for supplying electric power to a stator winding of a motor based on a current command; and current command creating means for creating the current command. Phase compensation amount calculating means for calculating a phase compensation amount indicating a current phase delay of a stator current flowing through the motor, wherein the phase compensation amount calculating means increases the phase compensation amount when the rotation speed of the motor increases, Current command creating means creates q-axis current command data and d-axis current command data, and compensates the q-axis current based on a value obtained by multiplying the phase compensation amount by the d-axis current command; A motor control device for compensating for the d-axis current based on a value obtained by multiplying the d-axis current by the q-axis current command. 26. A driving unit for supplying electric power to a stator winding of a motor based on a current command, a stator current detecting unit for detecting a stator current flowing through the stator winding, and based on the current command and the stator current. A saturation creation unit that creates saturation data indicating how much the stator current departs from the current command; a reference value creation unit that creates a reference value of the saturation; and a rate at which the current command is changed. A motor control device comprising: a gain creating unit that creates gain data; and a current command creating unit that creates the current command data based on the saturation, the reference value, and the gain data. The current phase delay of the current is calculated so that the current phase delay of the current command increases as the rotation speed of the motor increases. The current command creating means creates q-axis current command data and d-axis current command data, and calculates the q based on a value obtained by multiplying the phase delay by the d-axis current command. A motor control device for compensating for the axis current and for compensating the d-axis current based on a value obtained by multiplying the phase delay and the q-axis current command.