JP3721932B2 - Electric power steering device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a steering mechanism having a motor for applying torque to a steering shaft or steering gear of a vehicle, and an electric power steering apparatus having a control device for driving and controlling the motor. The present invention relates to the above-described control device that has a function of learning a position and executes handle return control.
[0002]
[Prior art]
An example of an electric power steering apparatus that has a function of learning a neutral position of steering wheel steering and that performs steering wheel return control is described in, for example, Japanese Patent Application Laid-Open No. 7-132845: Electric power steering with learning function. Things are generally known.
[0003]
FIG. 7 is a block diagram exemplifying a functional configuration of a steering wheel return torque calculation unit having a neutral position learning function implemented by the above-described conventional technology. In the drawing, V _n is a vehicle speed, and τ is a steering wheel. wheel rotation angle steering torque, theta _H is inputted from the steering angle sensor external, theta _O handle neutral position learned illustrates T _R is wheel return torque (command value), respectively.
[0004]
In this way, the above-described conventional technique learns the neutral position of the steering angle under the “condition that the vehicle goes straight” that can be determined from the vehicle speed condition, the steering torque condition, the steering angle condition, etc. The accuracy of the neutral position is gradually increased by the operation, and the steering wheel return control is executed based on the steering angle, so that the steering wheel returning performance particularly at the time of low speed traveling is sufficiently ensured.
[0005]
[Problems to be solved by the invention]
However, in the above prior art, it is possible to gradually increase the accuracy of the neutral position Θ _O of the steering angle over a long time by learning, but at least at the beginning of this learning, the width of the control dead zone Must be set to a very wide range of approximately ± 15 °, and therefore the steering wheel return performance is ensured in the above control dead zone at least until the above-mentioned “condition that the vehicle goes straight” is satisfied a predetermined number of times. There was a problem that could not be done.
[0006]
The power steering function is originally required in many cases during the driving period from when the driver tries to start the vehicle until this "condition that the vehicle goes straight" is satisfied. This problem must be extremely serious in generating the steering wheel return torque.
[0007]
Further, in the above prior art, as shown in the graph of T _R ′ in FIG. 7 in the setting example of the steering wheel return torque, in order to ensure a sufficient steering wheel return performance outside the above control dead zone, this control dead zone. as when the width is large, in the vicinity of the boundary of the control dead zone (outside), no longer has to be rapidly changed the value of T _R '. However, if such a setting is made, the steering wheel return torque rapidly increases immediately after the steering angle escapes from the control dead zone, which causes a problem that a great sense of discomfort occurs in steering. This problem can be solved gradually as the accuracy of the neutral position Θ _O gradually increases, but it must be remarkable especially at the beginning of learning.
[0008]
Further, in the conventional techniques described above, for the neutral position theta _O of accuracy of the steering angle of learning beginning like is very low, left and right symmetry of the steering wheel return torque T _R are exactly the particular learning beginning such guarantee Not. For this reason, when the error of the neutral position Θ _O is large, a left-right asymmetric steering feeling (steering wheel return torque) greatly biased to one side is realized, which is not desirable in terms of operation stability.
[0009]
Further, in the above-described conventional technology, a rotation angle sensor externally attached near the reduction gear is used as the rudder angle sensor. However, when the rudder angle sensor is configured in this way, the number of parts and the material cost are suppressed. This is not possible and causes problems in terms of production costs.
[0010]
The present invention has been made to solve the above-described problems, and an object of the present invention is to provide an electric power steering device capable of always generating an appropriate steering wheel return torque from when the driver tries to start the vehicle. In addition, an electric power steering device that can provide the same steering feeling as that of a hydraulic power steering device is provided at low cost.
[0011]
[Means for Solving the Problems]
In order to solve the above problems, the following means are effective.
That is, the first means is an electric power steering apparatus having a steering mechanism having a motor for applying torque to a steering shaft or steering gear of the vehicle, and a control device for driving and controlling the motor. Neutral position learning means for learning the neutral position of the steering angle θ _H of the steering wheel according to a predetermined statistical operation when the vehicle speed V _n of the vehicle and the steering torque τ exerted on the steering wheel by the driver satisfy a predetermined “straight vehicle condition” When, a calculated and a first handle return control means for calculating the neutral position wheel return torque T _R based on the steering angle theta _H as a reference, the steering wheel return torque T _R based on the rotational angular velocity omega _M of the motor 2 And a second handle return control until at least one neutral position learning by the neutral position learning means is completed. It is to determine the steering wheel return torque T _R by stage.
[0016]
The second means is the first means described above, after the neutral position by the neutral position learning means learning has been completed, it is to determine the first wheel return controlling means steering wheel return torque T _R.
The above-described problems can be solved by the above means.
[0017]
[Operation and effect of the invention]
When the accuracy of the neutral position Θ _O of the steering angle is very low, that is, until the neutral position learning by the neutral position learning means is completed at least once, the width of the control dead zone is considerable as shown in FIG. Therefore, the above problem occurs at the beginning of learning. However, if the first means of the present invention is used, in such a case, the steering wheel return torque T _R can be determined based on the motor rotation angular velocity ω _M (or the related value such as the steering velocity ω _H ). when the steering angle theta _H of the steering wheel is located in the control dead zone, i.e., in the learning beginning or the like of the neutral position, approximately it is possible to generate a desired steering wheel return torque T _R.
[0021]
Further, when the second means of the present invention is used, the handle return performance can be sufficiently secured even on a low friction road surface such as a frozen road, and the so-called overshoot phenomenon in which the handle is excessively returned is avoided. It becomes possible.
[0022]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, the present invention will be described based on specific examples. However, the present invention is not limited to the following examples.
FIG. 1 is a hardware configuration diagram of an electric power steering apparatus 100 according to the present embodiment.
[0023]
A steering wheel (steering wheel) 11 is attached to one end of the steering shaft 10, and a pinion shaft 13 supported by a gear box 12 is coupled to the other end. The pinion shaft 13 is engaged with a rack shaft 14 fitted in the gear box 12, and both ends of the rack shaft 14 are connected to steering wheels (not shown) via ball joints or the like, although not shown. Yes. Further, a brushless DC motor M (hereinafter simply referred to as “motor M”) for applying assist torque is connected to the steering shaft 10 via two gears 17.
[0024]
The DC motor M is supplied with motor drive currents iu, iv, and iw for the three phases U, V, and W via the current detector 115 from the drive circuit 113 of the motor control device 110.
Further, the steering shaft 10 is provided with a torque detector 15 for detecting the magnitude and direction (steering torque τ) of the manual steering force applied from the driver to the steering wheel (handle) 11.
[0025]
The motor M has a built-in resolver R / D (rotation angle sensor) for detecting the rotation angle. The CPU 111 outputs the phase n (6 bits of the motor M) output from the resolver R / D. Integer information (0 ≦ n <63), with π / 32 radians as one unit, is input to calculate the rotation angle θ _M (phase: 0 ≦ θ _M <2π) of the motor M. The resolver R / D has a built-in counter (not shown) that counts the rotation angle of one unit, and n increases when the motor M rotates in a predetermined direction and decreases when the motor M rotates in the opposite direction. To do. n is a numerical value that overflows or underflows every time the motor makes one revolution.
[0026]
The motor control device 110 includes a CPU 111, a ROM 112b, a RAM 112a, a drive circuit 113, an input interface (IF) 114, a current detector 115, and the like. The drive circuit 113 includes a battery (not shown), a PWM converter, a PMOS drive circuit, and the like, and supplies electric power to the motor M by changing the drive current to a sine wave by chopper control.
[0027]
The CPU 111 of the motor control device 110 receives output signals (measured values) from the phase sensor _n , the torque sensor 15 used to detect the steering torque τ of the steering wheel, the vehicle speedometer 50 used to calculate the vehicle speed Vn, and the like. ) Is input via the input interface (IF) 114. The CPU 111 determines a torque value (command torque T) to be output from the motor M based on a predetermined torque calculation from these input values, and further, based on the command torque T, each current of the d-axis and the q-axis. The command value (id ^* , iq ^* ) is determined. However, in this embodiment, “id ^* = 0”.
[0028]
FIG. 2 is a block diagram showing a logical configuration of the motor control device 110 that drives and controls the motor M of the present embodiment. The torque current conversion unit 280 in FIG. 2 is a control block that determines the q-axis current command value iq ^* based on the command torque T, and mainly includes an iq ^* -T map (table data). .
[0029]
The command torque calculation unit 200 includes the phase θ _{M of the} motor M calculated by the angle detection unit 310, the steering torque τ of the steering wheel calculated by the steering torque calculation unit 330, and the vehicle speed V _n calculated by the vehicle speed calculation unit 320. Further, based on the steering angle θ _H calculated by the steering angle calculator 500, the value of the desired torque (command torque T) to be output is calculated.
For example, the angle detection unit 310 calculates the rotation angle (phase) θ _M of the motor M by the following equation (1) based on the output value n of the resolver R / D.
[Expression 1]
θ _M = 2πn / 64 (n is an integer. 0 ≦ n <63) (1)
[0030]
Command torque calculation section 200 mainly calculates an assist torque calculation unit 210 for calculating the assist torque T _A, the inertia compensation torque calculation unit 220 for calculating the inertia compensation torque T _K, the damper torque T _D Damper a torque calculating unit 230, and a steering wheel return torque calculation unit 240 or the like calculates the steering wheel return torque T _R. The command torque calculator 200 calculates the command torque T according to the following equation (2).
[Expression 2]
_{T = T A + T K +} T D + T R ... (2)
[0031]
FIG. 3 is a block diagram illustrating a functional configuration of the steering wheel return torque calculation unit 240 before completion of the initial learning of the neutral position learning unit 241 of the present embodiment. The neutral position learning unit 241 starts learning the neutral position (zero point of the steering angle) Θ _{O after} the ignition key of the vehicle is switched to the ON state. In this learning, the steering angle θ _H of the steering wheel is sampled when the vehicle speed V _n of the vehicle and the steering torque τ exerted on the steering wheel by the driver satisfy a predetermined “straight vehicle traveling condition”. And executed by a predetermined statistical operation. Therefore, as long as the number of samples does not reach the predetermined number, the neutral position Θ _O of the steering wheel is never output from the neutral position learning unit 241.
[0032]
The steering wheel return torque T _R (command value) is determined based on the steering speed ω _H until the steering wheel neutral position Θ _{O as} a learning result is output at least once (first time). The steering speed ω _H is calculated by the steering speed estimation unit 242 based on the steering torque τ and the motor rotation angle θ _M. This steering speed ω _H is the sum of the rotational angular velocity of the torsion bar of the torque sensor 15 calculated from the amount of change in the steering torque τ and the rotational angular velocity of the pinion shaft 13 calculated from the amount of change in the motor rotational angle θ _M. Calculated.
[0033]
Further, thereafter, until at least the first cycle is completed in the neutral position learning wheel return torque T _R are determined in accordance with the following equation (3) to (5).
[Equation 3]
T _R = G1 · T _R ″ (3)
[Expression 4]
G1 = g1 (V _n ) (4)
[Equation 5]
T _R ″ = f 1 (ω _H ) (5)
Here, the functions g1 and f1 are defined in the form of, for example, the data map 243 and the data map 244 in FIG.
[0034]
Thus, until at least the first cycle is completed in the neutral position learning, it is determined based on the command value T _R of the steering wheel return torque to the steering speed omega _H, conventional large taken compelled not control Even in the dead zone, a desired handle return torque can be output from the motor M.
[0035]
Further, after at least the first one of the neutral position learning is completed, more preferably, the accuracy of the steering wheel neutral position Θ _O learned by the neutral position learning unit 241 is sufficiently improved, and the subsequent control is performed. The steering wheel is difficult to rotate due to self-aligning torque (that is, the motor M is difficult to rotate due to the road surface reaction force), and sufficient steering wheel return performance can be exhibited even on a low friction road surface such as a frozen road. Then, steering wheel return control based on the steering angle (steering wheel rotation angle) θ _H is performed.
According to the control based on the steering angle θ _H , it is possible to sufficiently secure the handle returning performance on the low friction road that is difficult to compensate by the control based on the steering speed ω _H , and at the same time, the steering wheel returns excessively. The so-called overshoot phenomenon can also be avoided.
[0036]
That is, more specifically, such control of the steering wheel return torque based on the steering angle θ _H can be realized by, for example, the steering wheel return torque calculation unit 240 in FIG.
FIG. 4 is a block diagram illustrating a functional configuration of the steering wheel return torque calculation unit 240 after completion of learning by the neutral position learning unit 241 of the present embodiment.
[0037]
As shown in FIG. 4, after the number of steering wheel steering angle θ _H samples reaches a predetermined number, the steering wheel neutral position Θ _O is output from the neutral position learning unit 241.
Then, steering wheel return torque T _R can be determined according to the following equation (6) to (9).
[Formula 6]
T _R = G1, G2, T _R ′ (6)
[Expression 7]
T _R '= f2 (Θ) (7)
[Equation 8]
Θ = θ _H −Θ _O (8)
[Equation 9]
G2 = g2 (τ) (when Θ> 0),
G2 = g2 (−τ) (when Θ <0) (9)
Here, the functions f2 and g2 are defined in the form of, for example, the data map 245 and the data map 246 (when Θ> 0) in FIG.
[0038]
Here, the function g2 implements the steering wheel return torque limiting means of the present invention, and the gain G2 is set to 1 when the driver performs the steering wheel return operation slowly (or weakly) or in the hand-off state. It is configured like this. Therefore, the steering state discriminating means and the steering wheel return current limiting means of the present invention that can be realized using the method of FIG. 4, for example, the steering judgment map 246 of FIG. 4 (in FIG. 4, “Θ> 0” According to the case, for example, the steering wheel return torque T _R is also increased when the driver performs the steering wheel return operation slowly (or weakly) from the state in which the steering wheel is cut during low-speed driving, and also in the hand-off state. Since this is effective, it is possible to realize a steering feeling substantially similar to that of a hydraulic power steering apparatus.
[0039]
FIG. 5B illustrates a modified example (steering determination map 246b) of the steering determination map 246 of the steering wheel return torque calculation unit 240 after completion of learning according to the present embodiment. This graph shows the form of a function g3 which is a modification of the function g2 when “Θ> 0”. As described above, when the absolute value | τ | of τ is large, the magnitude of G2 is adjusted so that the sum of the assist torque T _A and the handle return torque T _R does not become too large when the handle is returned. A method is also conceivable. According to such a method, it is possible to more reliably prevent a phenomenon (overshoot) that the motor M applies too much torque to the steering system (steering mechanism) when the handle is returned.
[0040]
Further, such adjustment may be performed on the side of the assist torque calculation unit 210 for calculating the assist torque T _A (Figure 1). For example, according to a method such as inputting the steering angle Θ whose neutral position after learning has been adjusted to the assist torque calculation unit 210 side, such adjustment can be performed in substantially the same manner as the above method. Is possible.
[0041]
FIG. 6 is a flowchart of a control program 500a for realizing the steering angle calculation unit 500 (steering angle estimation means) of FIG. In the present control program 500a, first, at step 520, the rotation angle (phase) θ _M of the motor M calculated by the angle detection unit 310 (FIG. 1) according to the equation (1) is used, and the phase θ _M A change amount D (= θ _M −θ0) from the previous control cycle is obtained. Here, θ0 is the value of the phase θ _M in the previous control period.
[0042]
Next, in step 530, the magnitude of the change amount D is determined. If greater than π, the process proceeds to step 540, and if not, the process proceeds to step 550. In step 540, the value of the change amount D is decreased by 2π (during underflow).
In step 550, the magnitude of the change amount D is determined. If smaller than −π, the process proceeds to step 560, and if not, the process proceeds to step 570. In step 560, the value of the change amount D is increased by 2π (when overflowing).
[0043]
In step 570, the value of the motor rotational angle theta _m, based on the change amount D obtained in the above process, that is, calculated according to the following equation (10).
[Expression 10]
θ _m = θ _m + D (10)
[0044]
In step 580, the rotation angle θ _H of the handle (pinion shaft 13) is calculated according to the following equation (11). Here, R is the reduction ratio between the motor and the handle (the gear ratio between the two gears (17) in FIG. 1).
[Expression 11]
θ _H = θ _m / R (11)
In step 590, the current phase θ _{M of the} motor M (in the current control cycle) is stored in the retreat area θ0.
[0045]
In such overflow / underflow determination, the maximum value ω _Hmax of the rotational angular velocity of the steering wheel (that is, the steering speed ω _H ) and the control execution cycle t0 in which the program 500a is periodically executed are expressed by the following equation (12). It is established when satisfying.
[Expression 12]
π> R · ω _Hmax · t0 (12)
[0046]
Although the maximum value ω _Hmax of the steering speed ω _H slightly depends on the configuration of the steering mechanism, this is usually the maximum value of the steering speed that can be operated by a human, and is usually about 20 [rad / sec]. It is. Therefore, as can be seen from the equation (12), by configuring the system so that the value of the product R · t0 of the reduction ratio R and the control execution period t0 is surely below π / ω _Hmax , That is, the steering angle estimating means can be realized, that is, the program 500a can be operated effectively.
More specifically, for example, by setting R = 5 and t0 = 1/100 [sec], it is possible to reliably satisfy such a condition.
[0047]
Further, the output value n (6-bit information: 0 to 63) of the resolver R / D (rotation angle sensor) represents only the phase of the motor M (0 ≦ θ _M <2π). If the information amount of the output value n is expanded to about 3 to 5 bits, for example, on the upper bit side, the motor rotation angle θ _m calculated by the program 500a can be easily determined, that is, the determination process such as overflow can be omitted. It can also be obtained directly from the hardware (resolver R / D) .
[0049]
Further, the predetermined “vehicle straight running condition” may be dynamically changed according to a learning history such as the number of learnings or a learning result, for example. As an example thereof, for example, a method of gradually increasing the boundary value of the vehicle speed used for the determination of “straight condition of vehicle” from about 30 km / hr to about 50 km / hr as the number of learning increases. Is mentioned. For example, according to such means, it is possible to shorten the time until the completion of the initial learning and simultaneously improve the learning accuracy according to the learning history or the like.
[Brief description of the drawings]
FIG. 1 is a hardware configuration diagram of an electric power steering apparatus 100 according to an embodiment of the present invention.
FIG. 2 is a block diagram showing a logical configuration of a motor control device 110 that drives and controls a motor M of an electric power steering device 100 according to an embodiment of the present invention.
FIG. 3 is a block diagram showing a functional configuration of a steering wheel return torque calculation unit 240 before completion of learning of the neutral position learning unit 241 according to the embodiment of the present invention.
FIG. 4 is a block diagram showing a functional configuration of a handle return torque calculation unit 240 after completion of learning by a neutral position learning unit 241 according to the embodiment of the present invention.
FIGS. 5A and 5B are graphs illustrating a steering determination map 246 of the steering wheel return torque calculation unit 240 after learning according to the embodiment of the present invention.
FIG. 6 is a flowchart of a control program 500a for realizing a steering angle calculation unit (steering angle estimation means) 500 according to an embodiment of the present invention.
FIG. 7 is a block diagram showing a functional configuration of a steering wheel return torque calculation unit having a neutral position learning function according to the prior art.
[Explanation of symbols]
M: Motor R / D: Resolver (Motor M rotation angle sensor)
n: Resolver output signal θ _M : Motor rotation angle (phase: 0 ≦ θ _M <2π)
θ _m ... Motor rotation angle θ _H ... Steering wheel rotation angle ω _H ... Steering speed Θ _O ... Steering wheel neutral position T _R ... Steering wheel return torque (command value)
T ... Command torque τ ... Steering torque V _n ... Vehicle speed 10 ... Steering shaft 11 ... Steering wheel (handle)
DESCRIPTION OF SYMBOLS 100 ... Electric power steering apparatus 110 ... Motor control apparatus 111 ... CPU
200 ... Command torque calculation unit 240 ... Steering wheel return torque computing unit 241 ... Neutral position learning unit 246 ... Steering determination map (corresponding to a steering state determining unit and a steering wheel return current limiting unit. In FIG. Ex.)
500 ... Steering angle calculation unit (steering angle estimating means)
500a ... Control program for realizing a steering angle calculation unit (steering angle estimation means)

Claims (2)

In an electric power steering apparatus comprising: a steering mechanism including a motor that applies torque to a steering shaft or steering gear of a vehicle; and a control device that drives and controls the motor.
When the vehicle speed V _n of the vehicle and the steering torque τ exerted on the steering wheel by the driver satisfy a predetermined “straight vehicle condition”, the neutral position of the steering angle θ _H of the steering wheel is learned according to a predetermined statistical operation. Neutral position learning means,
Based on the said neutral position, on the basis of the steering angle theta _H, a first handle return control means for calculating a torque T _R back handle,
Based on the rotational angular velocity omega _M of the motor, and a second handle return control means for calculating the steering wheel return torque T _R,
At least until the said neutral position by the neutral position learning means learning is completed once, the electric power steering apparatus characterized by determining the steering wheel return torque T _R by said second wheel return controlling means. The neutral position after learning of the neutral position is completed by the learning means, the electric power steering apparatus according to claim 1, wherein the determining by the steering wheel return torque T _R to return the first handle control means.

Images