JPH0884405A - Control device for electric rolling stock - Google Patents

Abstract

PURPOSE: To provide a control device for an electric rolling stock which can provide high stickiness without having a detector for idle-running slide to be able to reduce a quantity of change of torque. CONSTITUTION: The rotational speed of an electric motor is detected by a rotational speed detector 3 to supply a signal S2 indicating the rotational speed to an acceleration deviation computing means 4. An acceleration deviation quantity ▵dfm/dt is computed in the acceleration deviation computing means 4, and an acceleration deviation signal S3 indicating the acceleration deviation quantity ▵dfm/dt is supplied to a stickiness deduction part 5. In the stickiness deduction part 5, fuzzy deduction is performed by adopting the acceleration deviation signal S3 as an input to deduce stickiness. A stickiness signal S4 outputted from the stickiness deduction part 5 is supplied to an electric motor drive means 1. The electric motor drive means 1 controls the electric motor according to the stickiness signal S4.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control device for an electric vehicle, and more particularly, to a drive control device for an electric vehicle which can improve adhesive performance in an electric vehicle driven by an induction motor (hereinafter referred to as an electric motor). It is about.

[0002]

2. Description of the Related Art FIG. 10 shows, for example, JP-A-1-286702.
FIG. 6 is a block diagram showing a conventional electric vehicle control device disclosed in Japanese Patent Publication No. In the figure, 10 is the body of an electric vehicle, 11 is a track (rail), and 12a to 12d are driving wheels driven by electric motors (not shown).

Reference numerals 13a to 13d denote moving wheels 12a to 1 respectively.
It is a rotation frequency detector that detects the rotation frequency of 2d.
The rotation frequencies f1 to f4 output from the rotation frequency detectors 13a to 13d are supplied to the arithmetic circuit 15. The acceleration α output from the accelerometer 14 is also supplied to the arithmetic circuit 15. In the arithmetic circuit 15, the reference speed V of the moving wheels 12a to 12d is calculated based on the acceleration α during power running or regeneration.
r and the reference frequency fr are calculated. In this case, the reference frequency fr is the reference speed Vr converted as the rotation frequency of the electric motor. Further, the arithmetic circuit 15 corrects the reference frequency fr based on the rotation frequencies f1 to f4 during coasting.

The rotation frequencies f1 to f4 output from the rotation frequency detectors 13a to 13d are supplied to the speed calculation circuit 16 and the speed V1 of the moving wheels 12a to 12d.
~ V4 is calculated. The speeds V1 to V4 output from the speed calculation circuit 16 are supplied to the speed difference detection circuit 17. The reference speed Vr output from the arithmetic circuit 15 is supplied to the speed difference detection circuit 17. Speed difference detection circuit 17
Then, the relative speed difference ΔV from the speeds V1 to V4 and the reference speed Vr.
1 to ΔV4 are detected. In this case, the relative speed difference ΔV1
~ ΔV4 is the reference speed V from the speeds V1 to V4 during power running.
r is subtracted and formed, and at the time of regeneration, speeds V1 to V4 are subtracted from the reference speed Vr and formed. The relative speed differences ΔV1 to ΔV4 output from the speed difference detection circuit 17 are supplied to the slipping detection circuit 20.

Reference numeral 18 denotes a speed difference sensitivity setting device for slipping detection, and the speed difference detection sensitivity 19 output from the speed difference sensitivity setting device 18 is supplied to a slipping detection circuit 20. In this slipping detection circuit 20, relative speed differences ΔV1 to ΔV4
And a slip detection signal are formed based on the speed difference detection sensitivity. For example, the relative speed difference ΔV1 to ΔV4 is compared with the detection sensitivity (for example, 2 km / h), and the relative speed difference ΔV1 to ΔV1
If ΔV4 is equal to or higher than the detection sensitivity, it is determined that the vehicle is slipping and the slipping / sliding detection signal Sd is output. The detection signal Sd output from the slipping detection circuit 20 is supplied to the subtractor 21. The subtractor 21 is supplied with the standard pattern fsr of the slip frequency corresponding to the adaptive load of the electric vehicle.

From the subtracter 21, the standard pattern fsr of slip frequency is output as it is when the slip and slip detection signals Sd are not output, while the slip frequency is output when the slip and slip detection signals Sd are output. Standard pattern fsr is narrowed down by a fixed amount and output. The slip frequency fs of the electric motor output from the subtractor 21 is supplied to the adder 22. The reference frequency fr is supplied from the arithmetic circuit 15 to the adder 22. Then, the adder 22 adds the slip frequency fs and the reference frequency fr and outputs the inverter frequency fi.

Next, the operation will be described. Driving wheel 12a-
12d slips or slides, the relative speed difference ΔV1 to ΔV4
Is equal to or higher than the detection sensitivity, the slipping / sliding detection circuit 20 outputs a slipping / sliding detection signal Sd.
Is continuously output until the moving wheels 12a to 12d re-adhere, that is, until the relative speed differences ΔV1 to ΔV4 become smaller than the detection sensitivity. While outputting the slipping and sliding detection signal Sd,
From the subtracter 21, a standard pattern fsr of slip frequency narrowed down by a certain amount is output as the slip frequency fs of the electric motor.

This slip frequency fs accelerates re-adhesion because the current flowing through the electric motor is narrowed down by a certain amount during slipping. The slip frequency fs acts positively during power running and negatively during regeneration. Therefore, by controlling the inverter of the electric vehicle at the inverter frequency fi obtained by adding the slip frequency fs to the reference frequency fr, the reference frequency f is obtained even during slipping.
r becomes stable.

[0009]

Since the conventional electric vehicle control device is constructed as described above, when the idling skid occurs on the driving wheels 12a to 12d driven by the electric motor, the idling skid amount is predetermined. If it does not exceed the detection level of, it will not operate and control will be delayed. Therefore, the amount of torque fluctuation from slipping to re-adhesion becomes large,
There were problems such as deterioration of acceleration.

The present invention has been made in order to solve such a problem, and it is possible to obtain a high adhesion control without detecting slipping and control an electric vehicle capable of suppressing a torque fluctuation amount to a small value. The purpose is to obtain the device.

[0011]

According to a first aspect of the present invention, there is provided a control device for an electric vehicle, comprising: a rotation speed detector for detecting a rotation speed of the electric motor; and an electric motor of the electric motor based on an output of the rotation speed detector. An arithmetic unit for calculating information on acceleration deviation, an adhesiveness inference unit for performing fuzzy inference based on the output of the arithmetical unit to output an adhesiveness, and an electric motor based on the adhesiveness output by the adhesiveness inference unit. And a motor driving means for driving the.

According to a second aspect of the present invention, there is provided a control device for an electric vehicle, wherein a rotational speed detector for detecting rotational speeds of a plurality of electric motors and accelerations of the electric motors based on outputs of the rotational speed detectors. An arithmetic unit for calculating information on deviation, an adhesiveness inference unit for performing fuzzy inference based on the output of the arithmetic unit to output an adhesiveness, and a plurality of units based on the adhesiveness output by the adhesiveness inference unit And an electric motor drive means for driving the electric motor.

An electric vehicle controller according to a third aspect of the present invention is the electric vehicle controller according to the first or second aspect, wherein the electric motor driving means outputs a torque command signal for driving the electric motor. And a torque command calculator for driving the electric motor based on the torque command signal and the degree of adhesion.

An electric vehicle controller according to a fourth aspect of the present invention is the electric vehicle controller according to the first or second aspect, wherein the electric motor driving means outputs a torque command signal for driving the electric motor. And a torque command calculation unit that outputs the torque command signal based on the degree of adhesion.

[0015]

In the electric vehicle control apparatus according to the present invention, the acceleration deviation calculating means obtains information about the acceleration deviation based on the information about the rotation speed, and the adhesion degree inference unit performs fuzzy inference from the information about the acceleration deviation. The electric motor is driven by the electric motor driving means according to the obtained adhesiveness. And
When the tackiness is high, the motor is controlled so that it will rotate in a manner commensurate with the planned running.On the other hand, when the tackiness is low and the tackiness is low, the rotation speed of the motor is reduced so that the rotation speed is low. Control to minimize slipping. As a result, the electric motor can always be optimally driven according to the state of the driving wheels and the rails, and the slipping amount can be suppressed to a minimum even during slipping.

In the electric vehicle controller of claim 2,
Based on the information about each rotation speed in the acceleration deviation calculation means,
Information on the acceleration deviation is obtained, and the electric motor is continuously controlled by the adhesion degree obtained by fuzzy inference by the adhesion degree inference unit using the information on the acceleration deviation as an input. Then, when the adhesiveness is high, the electric motors are controlled so that each electric motor rotates so as to correspond to the planned running,
On the other hand, when slippage occurs and the adhesiveness is low, the electric motor is controlled so that the rotation speed of the electric motor, which is determined to be low, becomes low, and the slipping is suppressed to a minimum. As a result, the electric motor can always be optimally driven according to the state of the driving wheels and the rails, and the slipping amount can be suppressed to a minimum even during slipping.

In the electric vehicle controller of claim 3,
In the control device for an electric vehicle according to claim 1 or 2,
The electric motor driving means optimally drives the electric motor based on the torque command signal and the degree of adhesion.

In the control device for an electric vehicle according to claim 4,
In the control device for an electric vehicle according to claim 1 or 2,
A torque command calculation unit that outputs a torque command signal calculates and outputs a torque command signal based on the degree of adhesion, and drives the electric motor by this torque command signal.

[0019]

【Example】

Example 1. FIG. 1 is a block diagram showing the basic configuration of the present invention. Example 1 will be described with reference to FIG. FIG.
In the figure, reference numeral 1 is an electric motor drive means for supplying the input S1 to the electric motor 2. Reference numeral 2 denotes an induction motor (hereinafter simply referred to as an electric motor) driven by the electric motor drive signal S1. Reference numeral 3 is a rotation speed detector for detecting the rotation speed of the electric motor 2, and the rotation speed detector 3 is a signal S indicating the rotation speed of the electric motor 2.
2 is supplied to the acceleration deviation calculating means 4. The electric motor drive means 1 outputs a drive signal S1 to the electric motor 2 so as to accelerate or decelerate at a constant rate during acceleration or deceleration. In the embodiment, the electric motor is an induction motor, but the present invention may include an electric motor based on other operating principles such as a DC motor and a synchronous motor.

In the acceleration deviation calculating means 4, the rotation speed fm
Is differentiated to obtain the acceleration. It also has a function of sequentially storing the obtained acceleration. In this past history, the acceleration at the time when it is determined that the acceleration is sufficiently constant is set as the reference acceleration. A signal S3 indicating an acceleration deviation amount Δdfm / dt obtained by subtracting the reference acceleration from the acceleration, for example, using the reference acceleration set in this way and the above acceleration.
Is supplied to the adhesion degree inference unit 5.

In the adhesion degree inference unit 5, the acceleration deviation amount Δdfm
The adhesiveness ad_l is estimated by fuzzy inference based on the signal S3 indicating / dt.

A method of performing control using fuzzy inference is called fuzzy control. The fuzzy control in the adhesion degree inference unit 5 will be described below with reference to FIGS.
In the fuzzy control, an input value and an output value are vaguely linguisticized by a function called a membership function, which is used as an input / output amount, and a control rule called a fuzzy rule serves as a control gain to obtain an operation amount.

This fuzzy control rule associates the input / output amounts with words as shown in FIG. For example,

"If dfm / dt is ZR, then ad_1 is B." If the acceleration deviation amount Δdfm / dt is zero,
The stickiness is high. “If dfm / dt is PB, then ad_1 is S.” If the acceleration deviation amount Δdfm / dt is positively large, the degree of adhesion is small.

Then, the control rule can be read. Regarding the membership function that makes it possible to apply the input / output amount to this control rule, FIG. 3 shows the acceleration deviation amount Δdfm / dt, and FIG. 4 shows the adhesiveness ad_1. These FIG. 3 and FIG.
In, the vertical axis indicates each rate. The symbols in these figures are

NB = minus large, NS = minus small, ZR = zero, PS = small to plus, PB = large to plus, S = small, LS = small to small, M = medium, LB = small Big, B = big,

Is meant. As an actual means of fuzzy control, for example, a method called max-min composition-centroid method or algebraic product-addition-centroid method ("fuzzy control", Nikkan Kogyo Shimbun Co., P74-P84) can be used. Both of these are generally widely used means, and it is said that the latter is preferably used when the rotation speed detection for obtaining the acceleration of the input amount includes many errors, for example. Among these execution methods, fuzzy inference is performed first and a data table corresponding to input values-output values is provided, a fuzzy chip is used, or a CPU is used for inference operation. .

The adhesiveness inference unit 5 uses the result obtained by the fuzzy inference as an adhesiveness signal S4 to drive the motor driving means 1
Supply to. The electric motor driving means 1 uses the adhesiveness signal S4.
The motor drive signal S1 is controlled based on

Next, the operation will be described. When the electric vehicle is traveling in a state of high adhesion, there is almost no difference between the acceleration of the electric motor and the reference acceleration, and the acceleration deviation signal S3 has a value close to zero. By inputting the acceleration deviation amount Δdfm / dt based on the acceleration deviation signal S3, the adhesion degree inference unit 5 shows the result of "the adhesion degree is large" by the fuzzy inference by the control rule shown in FIG.
d_1 is output to the electric motor drive means 1 as an adhesion signal S4. The electric motor drive means 1 which has received the adhesiveness signal S4 controls the electric motor drive signal S1 to a magnitude commensurate with the intended drive of the electric motor based on the adhesiveness signal S4.

When the tackiness between the driving wheel driven by the electric motor 2 and the rail becomes low, for example, in rainy weather, the electric motor 2 rotates faster and the rotational speed value output by the rotational speed detector 3 is increased. Rises. Then, the acceleration deviation calculating means 4 increases the acceleration obtained by differentiating the rotation speed signal S2. Therefore, a deviation occurs between the reference acceleration and the acceleration, and the acceleration deviation amount Δdfm / dt indicated by the acceleration deviation signal S3 also increases. The adhesiveness inferring unit 5 infers the adhesiveness with respect to the electric motor 2 based on the magnitude of the acceleration deviation signal S3, and outputs the adhesiveness signal S4 indicating the adhesiveness lower than that at the time of adhesion.

When the electric motor 2 starts sticking, the rotation becomes slow, and the acceleration deviation calculating means 4 causes the acceleration deviation amount Δdfm.
The acceleration deviation signal S3 indicating / dt is output as a negative value. The adhesiveness inference unit 5 outputs an inference result that "adhesiveness is a little higher", and outputs an adhesiveness signal S4 indicating an adhesiveness higher than when idling occurs to the electric motor drive means 1.

The electric motor drive means 1 controls the electric motor drive signal S1 by the adhesiveness signal S4 inferred by the adhesiveness inference unit 5 as described above, thereby realizing the electric motor drive always in the state of the wheel and the rail. To do. Further, in the embodiment, since only one variable is used as the inference input, the control rule can be easily constructed.

Example 2. FIG. 5 is a block diagram showing a second embodiment of the present invention. The second embodiment includes a plurality of electric motors.
Since it is often controlled by one driving means, for example, an inverter at present, the case where the present invention is applied to a plurality of electric motors is also shown.

In FIG. 5, reference numerals 2a to 2d denote a plurality of electric motors driven by the electric motor driving means 1. 3a-3
Reference numeral d denotes a rotation speed detector that detects the rotation speed of each of the electric motors 2a to 2d. These rotation speed detectors 3a to 3d
The signals S11 to S14 indicating the rotation speed output by
The acceleration deviation calculation unit 4A supplies the acceleration deviation amount Δd.
fm / dt is calculated. This acceleration deviation calculation means 4A
The acceleration deviation signals S21 to S24, which indicate the acceleration deviation amount Δdfm / dt, are supplied to the adhesiveness inference unit 5A.

The electric motor drive means 1 outputs an electric motor drive signal S1 to the electric motor so as to accelerate or decelerate at a constant rate during acceleration or deceleration of the electric vehicle.

The acceleration deviation calculating means 4A differentiates the rotation speed signals S11 to S14 to obtain accelerations.
Further, the acceleration deviation calculating means 4A uses the rotation speed signal S11.
Among S14, it has a function of time-differentiating the minimum value during power running and the maximum value during regeneration and sequentially storing them. In this past history, the acceleration at the time when it is determined that the acceleration is sufficiently constant is set as the reference acceleration. Acceleration deviation amount Δdfm / dt between the reference acceleration and the acceleration set in this way
The acceleration deviation signals S21 to S24 indicating the above are supplied to the adhesion degree inference unit 5A.

In the adhesiveness inference unit 5A, the acceleration deviation amount Δd
Fuzzy inference is performed from the signals S21 to S24 indicating fm / dt to obtain the adhesiveness ad_1 for each axis. The value obtained by adding and averaging the respective adhesivenesses ad_1 of the obtained results is the adhesiveness signal S4.
Is output to the motor driving means 1. Fuzzy reasoning
This is the same as the means used in the first embodiment described above.

The adhesiveness inference unit 5A supplies the result obtained by the fuzzy inference to the electric motor drive means 1 as an adhesiveness signal S4. The electric motor drive means 1 controls the electric motor drive signal S1 based on the adhesiveness signal S4.

Next, the operation will be described. When the electric vehicle is running in a highly sticky state, there is almost no difference between the acceleration of the electric motor and the reference acceleration, and the acceleration deviation signals S21 to S2
4 is a value close to zero. This acceleration deviation amount Δdfm / d
With the input of t, the stickiness inference unit 5A shows the result of "the stickiness is large" by the fuzzy reasoning according to the control rule shown in FIG. 4, and outputs the stickiness ad_1 to the electric motor drive means 1 as the stickiness signal S4. . The electric motor driving means 1 which has received the adhesiveness signal S4 controls the drive signal S1 to have a magnitude corresponding to the planned driving of the electric motor based on the adhesiveness signal S4.

Also, for example, in the case of rain, the electric motor 2a
When the adhesiveness between the driving wheel and the rail driven by the motor becomes low, the rotation of the electric motor 2a becomes faster and the rotation speed detector 3
The rotation speed value output by a increases. Then, the acceleration deviation calculating means 4A increases the acceleration obtained by differentiating the rotation speed signal S11. Therefore, a deviation occurs between the reference acceleration and the acceleration, and the acceleration deviation amount Δdfm / dt indicated by the acceleration deviation signal S21 also increases.

Based on the magnitude of the acceleration deviation signal S21, the adhesiveness inferring unit 5A infers the adhesiveness with respect to the electric motor 2a, and the adhesiveness signal S4 showing the adhesiveness lower than that at the time of adhesiveness.
Is output. The electric motor drive means 1 uses the adhesiveness signal S
4 controls to reduce the rotational speed of the electric motor 2a. When the electric motor 2a starts sticking, the rotation becomes slow, and the acceleration deviation calculating means 4A causes the acceleration deviation amount Δdf.
A negative signal S21 indicating m / dt is output. The adhesiveness inference unit 5A outputs an inference result that "adhesiveness is a little high", and outputs an adhesiveness signal S4 indicating an adhesiveness higher than when idling occurs to the electric motor driving means 1.

The electric motor drive means 1 controls the electric motor drive signal S1 based on the adhesiveness signal S4 inferred by the adhesiveness inference unit 5A as described above, thereby always realizing the electric motor drive adapted to the state of the wheel and the rail. To do.

Example 3. Next, another embodiment of the present invention will be described with reference to FIG. In FIG.
Portions corresponding to those in FIG. 1 are designated by the same reference numerals, and detailed description thereof will be omitted. In the present embodiment, the electric motor driving means 1 in FIG. 1 has a current detector 8 for detecting an electric motor current, a torque command calculation unit 6 for generating a torque command, a motor current signal S6 output by the current detector 8, and a torque command. The electric motor drive signal S1 is based on the torque command signal S5 output from the calculation unit 6, the rotation speed signal S2 output from the rotation speed detector 3, and the adhesiveness signal S4 indicating the adhesiveness output from the adhesiveness inference unit 5.
Is provided to the electric motor 2 and the electric motor drive controller 7. The electric motor drive controller 7 uses, for example, a VVVF inverter that can drive the electric motor at a variable speed.

The acceleration deviation calculating means 4 has a function of differentiating the rotation speed to obtain an acceleration and storing the acceleration. A reference acceleration is set from the stored acceleration, and the acceleration deviation amount Δdfm /
dt is calculated, and the acceleration deviation signal S3 is supplied to the adhesion degree inference unit 5 as a signal indicating the acceleration deviation amount Δdfm / dt. The adhesiveness inference unit 5 receives the supplied acceleration deviation amount Δd.
The adhesiveness ad_1 is estimated by fuzzy inference based on the acceleration deviation signal S3 indicating fm / dt. An adhesiveness signal S4 indicating the adhesiveness output from the adhesiveness inference unit 5
Is supplied to the electric motor drive controller 7.

Next, the operation will be described. While the electric vehicle is accelerating or decelerating, when the driving wheel and the rail are in an adhesive state, the rotational speed is accelerated almost constant and the electric vehicle runs. In the acceleration deviation calculation means 4,
The acceleration is obtained by differentiating the rotation speed. Further, the acceleration storage function is used to call the acceleration at the time when the acceleration becomes constant and set as the reference acceleration. The reference acceleration is subtracted from the acceleration to obtain the acceleration deviation amount Δdfm / dt. At this time, the acceleration deviation amount Δdfm / dt is almost zero.

By inputting this acceleration deviation amount Δdfm / dt, the adhesion degree inference unit 5 shows a result of "the adhesion degree is large" by the fuzzy reasoning according to the control rule shown in FIG. The signal S4 is output to the electric motor drive controller 7. The electric motor drive controller 7 first sets the electric motor drive signal S1 based on the torque command signal S5, the electric motor current signal S6, and the rotation speed signal S2, and sets the electric motor drive signal S1 set based on the adhesiveness signal S4 of the electric motor. The size is controlled (corrected) to match the drive.

When the tackiness between the driving wheel driven by the electric motor 2 and the rail becomes low, for example, in the case of rain, the rotation of the electric motor 2 becomes faster and the rotation output by the rotation speed detector 3 is increased. The speed value increases. Then, the acceleration deviation calculating means 4 increases the acceleration obtained by differentiating the rotation speed signal S2. Therefore, a deviation occurs between the reference acceleration and the acceleration, and the acceleration deviation amount Δdfm / dt indicated by the acceleration deviation signal S3 also increases.

Based on the magnitude of the acceleration deviation signal S3, the adhesiveness inference unit 5 infers the adhesiveness with respect to the electric motor 2, and outputs the adhesiveness signal S4 indicating the adhesiveness lower than that at the time of adhesion. The electric motor drive controller 7 controls the adhesion signal S4.
The electric motor drive signal S1 is controlled so as to reduce the rotation speed of the electric motor 2.

When the electric motor 2 starts sticking, the rotation becomes slow, and the acceleration deviation calculating means 4 causes the acceleration deviation amount Δdfm.
The acceleration deviation signal S3 indicating / dt is output as a negative value. The adhesiveness inference unit 5 outputs an inference result that "adhesiveness is a little high", and outputs an adhesiveness signal S4 indicating an adhesiveness higher than when idling occurs to the electric motor drive controller 7. The electric motor drive controller 7 controls the electric motor drive signal S1 so as to increase the rotation speed of the electric motor 2 based on the adhesiveness signal S4. Also in the present embodiment, the same operational effect as that of the first embodiment is obtained.

Example 4. Next, another embodiment of the present invention will be described with reference to FIG. 7, parts corresponding to those in FIG. 5 are designated by the same reference numerals, and detailed description thereof will be omitted. In the fourth embodiment, the motor driving means in the third embodiment is a current detector 8 that detects a motor current, a torque command calculator 6A that generates a torque command, and a current detector 8.
The motor current output by the torque command signal S5 output by the torque command calculator 6A, the rotation speed detector 3a
Rotation speed signals S11 to S14 output by
And an electric motor drive controller 7A which receives the adhesiveness signal S4 output from the adhesiveness inference unit 5A and outputs an electric motor drive signal S1. The electric motor drive controller 7A can drive the electric motor at a variable speed, for example, VVVF.
Use an inverter.

The acceleration deviation calculating means 4A has a function of differentiating the rotational speed to obtain an acceleration and storing the acceleration. Then, the reference acceleration is set from the stored acceleration, and the acceleration deviation amount Δ is calculated from the reference acceleration and the above acceleration.
dfm / dt is obtained, and this acceleration deviation amount Δdfm / dt
The acceleration deviation signals S21 to S24 are supplied to the adhesion degree inference unit 5A as a signal indicating that.

The adhesiveness inference unit 5A estimates the adhesiveness ad_1 by fuzzy inference based on the signals S21 to S24 indicating the supplied acceleration deviation amount Δdfm / dt. An adhesiveness signal S4 indicating the adhesiveness output from the adhesiveness inference unit 5A is supplied to the electric motor drive controller 7A.

Next, the operation will be described. During acceleration or deceleration of the electric vehicle, when the driving wheel and the rail (not shown) are in an adhered state, the rotational speed is accelerated almost constant and the electric vehicle runs. The acceleration deviation calculating means 4A differentiates the rotation speed to obtain the acceleration of each electric motor. Further, the acceleration at the time when the acceleration becomes constant is called from the acceleration storage function to set the reference acceleration. Then, the reference acceleration is subtracted from the acceleration to obtain the acceleration deviation amount Δdfm / dt. At this time, the acceleration deviation amount Δdfm / dt is almost zero.

By inputting this acceleration deviation amount Δdfm / dt, the adhesion degree inference unit 5A shows the result "adhesion degree is large" according to the control rule shown in FIG. 4 by fuzzy reasoning, and the adhesion degree ad_1 indicates the adhesion degree signal. It is output to the motor drive controller 7A as S4. The motor drive controller 7A has a torque command signal S5, a motor current signal S6, and a rotation speed signal S.
First, the electric motor drive signal S1 is set based on S11 to S14, and the electric motor drive signal S1 set based on the adhesiveness signal S4 is controlled (corrected) to a magnitude suitable for driving the electric motor.

Also, for example, in rainy weather, the electric motor 2a
When the adhesiveness between the driving wheel and the rail driven by the motor becomes low, the rotation of the electric motor 2a becomes faster and the rotation speed detector 3
The rotation speed value output by a increases. Then, in the acceleration deviation calculating means 4A, the acceleration obtained by differentiating the rotation speed signal S11 becomes large. Therefore, a deviation occurs between the reference acceleration and the acceleration, and the acceleration deviation amount Δdfm / dt indicated by the acceleration deviation signal S21 also increases.

Based on the magnitude of the acceleration deviation signal S21, the adhesiveness inferring section 5A infers the adhesiveness with respect to the electric motor 2a, and the adhesiveness signal S4 showing the adhesiveness lower than that at the time of adhesiveness.
Is output. The electric motor drive controller 7A controls the electric motor drive signal S1 so as to reduce the rotation speed of the electric motor 2a based on the adhesiveness signal S4. When the electric motor 2a starts sticking, the rotation becomes slower, and the acceleration deviation calculating means 4A outputs a negative acceleration deviation signal S21 indicating the acceleration deviation amount Δdfm / dt. The adhesiveness inference unit 5A outputs an inference result that "adhesiveness is a little higher", and outputs a signal S4 indicating an adhesiveness higher than when idling occurs to the electric motor drive controller 7A. The electric motor drive controller 7A controls the electric motor drive signal S1 so as to increase the rotation speed of the electric motor 2a by the adhesiveness signal S4. Also in the present embodiment, the same operational effect as that of the first embodiment is obtained.

Example 5. Next, another embodiment of the present invention will be described with reference to FIG. 8, parts corresponding to those in FIG. 1 are designated by the same reference numerals, and detailed description thereof will be omitted. In the fifth embodiment, the electric motor driving means includes a current detector 8 that detects the electric motor current, and a torque command calculation unit 6B that generates a torque command based on the adhesion signal S4.
The torque command signal S5 output from the torque command calculation unit 6B,
A motor drive controller 7 that outputs a drive signal S1 based on the motor current and the rotation speed output from the rotation speed detector 3.
It is configured with B.

The adhesion degree inferring section 5B is supplied with the acceleration deviation signal S3 from the acceleration deviation calculating means 4B to obtain the adhesion degree signal S4 by fuzzy inference, and supplies this adhesion degree signal S4 to the torque command calculating section 6B. The operation of fuzzy reasoning in the stickiness reasoning unit 5B is the same as the example of FIG. In the torque command calculation unit 6B, based on the adhesion degree signal S4, the torque command signal S5 is hardly corrected during adhesion, and the torque command signal S5 is corrected during idling so as to suppress idling. The motor drive controller 7B receives the corrected torque command signal S
5, the electric motor drive signal S1 is controlled. Example 5
Also in this, the same effect as the third embodiment is obtained.

Example 6. Next, another embodiment of the present invention will be described with reference to FIG. 9, parts corresponding to those in FIG. 5 are designated by the same reference numerals, and detailed description thereof will be omitted. In the sixth embodiment, the electric motor drive means 1 includes a current detector 8 that detects an electric motor current, and a torque command calculation unit 6C that generates a torque command based on the adhesiveness signal S4.
And a motor current signal S6 output from the current detector 8, a torque command signal S5 output from the torque command calculator 6C, and a rotation speed signal S1 output from the rotation speed detectors 3a to 3d.
1 to S14 and outputs a motor drive signal S1 and supplies the motor 2 with a motor drive controller 7C.

The adhesion degree inferring section 5C is supplied with the acceleration deviation signals S21 to S24 from the acceleration deviation calculating means 4C to obtain the adhesion degree signal S4 by fuzzy reasoning, and supplies this adhesion degree signal S4 to the torque command calculating section 6C. . The operation of fuzzy inference in the adhesiveness inference unit 5C is the same as the example of FIG. The torque command calculation unit 6C hardly corrects the torque command signal S5 at the time of adhesion based on the adhesion degree signal S4,
When idling, the torque command signal S5 is corrected so as to suppress idling. The electric motor drive controller 7C controls the electric motor drive signal S1 based on the corrected torque command signal S5. Also in the sixth embodiment, the same operational effect as that of the second embodiment is obtained.

Example 7. In the first to sixth embodiments described above, the acceleration deviation calculating means 4 subtracts the reference acceleration from the acceleration and outputs the deviation amount as a signal representing the acceleration deviation. However, in the seventh embodiment, the acceleration is divided by the reference acceleration. The same effect can be obtained by using the deviation rate.

[0062]

According to the first aspect of the present invention, there is provided a control device for an electric vehicle, which is a rotational speed detector for detecting a rotational speed of an electric motor, and information on an acceleration deviation of the electric motor based on an output of the rotational speed detector. , A tackiness reasoning section that performs fuzzy reasoning based on the output of the computing element to output a tackiness degree, and an electric motor that drives the electric motor based on the tackiness degree that the tackiness degree reasoning section outputs. Since it is equipped with a driving means, it does not require slipping slip detection using a slipping slip detector, and fuzzy reasoning can be applied as an adhesion reasoning unit, so it is always optimal for running conditions of the wheels and rails. There is an effect that can be performed. Further, by inputting the information regarding the acceleration deviation as the input of the stickiness inference, there is an effect that it is possible to perform a highly accurate control capable of widely handling from a small idling to an idling that accelerates sharply. Furthermore, since only one variable is configured as an inference input, the control rule can be easily constructed.

According to a second aspect of the present invention, there is provided a control device for an electric vehicle, wherein a rotational speed detector for detecting rotational speeds of a plurality of electric motors and accelerations of the electric motors based on outputs of the rotational speed detectors. An arithmetic unit for calculating information on deviation, an adhesiveness inference unit for performing fuzzy inference based on the output of the arithmetic unit to output an adhesiveness, and a plurality of units based on the adhesiveness output by the adhesiveness inference unit Since the electric motor drive means for driving the electric motor is provided, the same effect as in claim 1 can be obtained even when there are a plurality of electric motors. Also, by inputting the information about the acceleration deviation into the adhesiveness inference unit,
Since the degree of adhesion for each axis can be inferred regardless of the variation in the idling speed when the idling occurs, it is possible to achieve optimum drive control even for a plurality of electric motors. Further, since it is configured so that only one variable is used as the inference input, the control rule can be easily constructed.

An electric vehicle controller according to a third aspect of the present invention is the electric vehicle controller according to the first or second aspect, wherein the electric motor drive means outputs a torque command signal for driving the electric motor. Since the electric motor is driven based on the torque command signal and the adhesiveness, the effect corresponding to claim 1 or claim 2 is achieved.

An electric vehicle controller according to a fourth aspect of the present invention is the electric vehicle controller according to the first or second aspect, wherein the electric motor driving means outputs a torque command signal for driving the electric motor. Since the torque command calculation unit is configured to output the torque command signal based on the degree of adhesion, the torque command calculation unit achieves the effects of claim 1 and claim 2, respectively.

[Brief description of drawings]

FIG. 1 is a block diagram showing a basic configuration of a control device for an electric vehicle according to an embodiment of the present invention.

FIG. 2 is a table showing fuzzy control rules used in the present invention.

FIG. 3 is a deviation amount Δdfm / d used in the present invention.
It is a figure which shows the membership function R1 of t.

FIG. 4 is a diagram showing a membership function RR of the adhesiveness ad_1 used in the present invention.

FIG. 5 is a block diagram showing a control device for an electric vehicle according to a second embodiment of the present invention.

FIG. 6 is a block diagram showing a control device for an electric vehicle according to a third embodiment of the present invention.

FIG. 7 is a block diagram showing a control device for an electric vehicle according to a fourth embodiment of the present invention.

FIG. 8 is a block diagram showing a control device for an electric vehicle according to a fifth embodiment of the present invention.

FIG. 9 is a block diagram showing a control device for an electric vehicle according to a sixth embodiment of the present invention.

FIG. 10 is a block diagram showing a conventional electric vehicle control method.

[Explanation of symbols]

1 electric motor drive means, 2, 2a-2d electric motors, 3, 3
a to 3d Rotational speed detector 4,4A to 4C Acceleration deviation calculating means 5,5A to 5C Adhesion degree inference unit 6,6A
6C Torque command calculation unit, 7, 7A to 7C Electric motor drive controller, 8 Current amount detector, 9 Reference speed calculation unit.

Claims (4)

[Claims] 1. A rotation speed detector for detecting a rotation speed of an electric motor, an arithmetic unit for calculating information on an acceleration deviation of the electric motor based on an output of the rotation speed detector, and an output for the arithmetic unit. A tackiness reasoning unit that performs fuzzy reasoning to output a tackiness degree, and an electric motor driving unit that drives the electric motor based on the tackiness degree output by the tackiness reasoning unit are included in the electric vehicle. Control device. 2. A rotation speed detector for detecting rotation speeds of a plurality of electric motors, an arithmetic unit for calculating information on acceleration deviations of the electric motors based on outputs of the rotational speed detectors, and the arithmetic unit. A fuzzy inference based on the output of the adhesiveness inference unit to output the adhesiveness, and an electric motor drive unit that collectively drives the plurality of electric motors based on the adhesiveness output from the adhesiveness inference unit. A control device for an electric vehicle, which is characterized in that 3. The electric motor drive means includes a torque command calculation unit that outputs a torque command signal for driving the electric motor, and drives the electric motor based on the torque command signal and the adhesiveness. The control device for an electric vehicle according to claim 1 or 2. 4. The electric motor drive means includes a torque command calculation unit that outputs a torque command signal for driving the electric motor, and the torque command calculation unit outputs the torque command signal based on the adhesiveness. The control device for an electric vehicle according to claim 1 or 2, wherein: