JP7235022B2 - steering control device - Google Patents

Description

The present invention relates to steering control devices.

2. Description of the Related Art Conventionally, there has been known a technique for suppressing the influence of noise in a steering control device in which a servo controller calculates a motor output command so that the steering torque follows a target steering torque.

For example, a steering control device disclosed in Patent Document 1 generates a target steering torque that a servo controller follows based on a road reaction force obtained from a target steering torque (Ts ^* ) and a base assist command (Tb ^* ). . Since the steering torque detection value (Ts) on which noise is likely to be superimposed is not used to generate the target steering torque, the influence of noise on the steering torque servo output is suppressed.

Further, in the steering control device disclosed in Patent Document 2, the servo controller (assist controller in Patent Document 2) is composed of a PID controller. The PID controller receives as input the steering torque deviation, which is the difference between the target steering torque and the steering torque, and calculates the proportional, integral, and differential terms of the control torque, in other words, the proportional, integral, and differential control torque components. do.

Japanese Patent No. 6314752 Japanese Patent No. 6252027

For example, in a general PID controller that receives a steering torque deviation as an input, when noise is superimposed on the steering torque signal, the noise is amplified in the calculation of the differential term, that is, the differential control torque component, and appears in the steering torque servo output, causing the motor to operate. to vibrate. Alternatively, the same applies to a configuration provided with a differentiation-preceding PID controller that differentiates the steering torque. In particular, when the motor is vibrated during non-steering or slight steering, the driver is likely to notice the generated sound and mechanical vibrations, resulting in reduced comfort.

SUMMARY OF THE INVENTION The present invention has been created in view of such a point, and its object is to provide a steering control device that suppresses noise appearing in the output of a servo controller that includes calculation of a differential control torque component. .

In the steering control system of the present invention, ^a servo controller (400) calculates an output command (Tb*) for a motor (80) so that a steering torque (Ts) follows a target steering torque (Ts ^* ). The servo controller uses the steering torque deviation (ΔTs), which is the difference between the target steering torque and the steering torque, or the steering torque as the "input to be processed", and calculates the differential control torque component based on the differentiation of the input to be processed. It has a control torque component calculator (50, 505, 60, 605). In the differential control torque component calculator of a general PID controller, the steering torque deviation is the input to be processed, and in the differential control torque component calculator of the differential preemptive PID controller, the steering torque is the input to be processed.

A signal obtained by differentiating the input to be processed and processed by a filter that cuts off high-frequency components equal to or higher than a predetermined cutoff frequency is referred to as a "differential signal to be processed". When the input to be processed is the steering torque deviation, the differential signal to be processed is the steering torque deviation differential signal, and when the input to be processed is the steering torque, the differential signal to be processed is the steering torque differential signal. In the differential control torque component calculation section, "when the change in the input to be processed is relatively large, the degree of contribution of the differential signal to be processed with a relatively high cutoff frequency is relatively high", and "the change in the input to be processed is relatively small, the differential control torque component is calculated such that the contribution of the differential signal to be processed having a relatively low cutoff frequency is relatively high.

In the present invention, by adjusting the degree of contribution of the differential signal to be processed with a different cutoff frequency according to the change in the input to be processed, the noise superimposed on the steering torque signal is Amplification by the derivative term of the PID controller is suppressed. Therefore, it is possible to reduce noise and mechanical vibrations generated by the motor excited by noise, particularly during non-steering and fine steering. Also, during steering, since the responsiveness of the motor output command to the steering torque signal can be ensured, ripple vibration of the steering wheel due to phase lag can be suppressed.

The differential control torque component calculator of the present invention broadly includes two modes. The differential control torque component calculation unit (50, 505) of the first aspect performs weighted addition of differential signals to be processed processed by a plurality of filters having different cutoff frequencies to calculate differential control torque components. .

This differential control torque component calculation unit "relatively increases the degree of contribution of the differential signal to be processed having a relatively high cutoff frequency in a band in which the absolute value of the amplitude of the differential signal to be processed is relatively large", "The degree of contribution of a differential signal to be processed having a relatively low cutoff frequency is made relatively high in a band in which the absolute value of the amplitude of the differential signal to be processed is relatively small."

The differential control torque component calculator (60, 605) of the second aspect processes the signal to be processed by a filter set so that the cutoff frequency becomes higher as the rate of change of the input to be processed increases, and performs differential control. Calculate the torque component.

1 is a schematic configuration diagram of an electric power steering system; FIG. 1 is a schematic configuration diagram of an ECU (steering control device); FIG. FIG. 2 is a block diagram of the servo controller of the first and second embodiments; A map of the torque deadband processor. FIG. 4 is a block diagram of a differential control torque component calculator according to the first embodiment; FIG. 4 is a block diagram showing an equivalent configuration of a pseudo operation unit; (a) First differential dead zone process, (b) Map of second differential dead zone process. 4 is a signal waveform diagram after steering torque deviation and differential differentiation processing; FIG. FIG. 4 is a signal waveform diagram after first LPF processing and first differential dead zone processing; FIG. 4 is a signal waveform diagram after second LPF processing and second differential dead zone processing; FIG. 4 is a waveform diagram of a differential control torque component according to the first embodiment; FIG. 4 is a waveform diagram of a differential control torque component of a comparative example; Time chart 1 showing the behavior of the actual machine. Time chart 2 same as above. FIG. 11 is a block diagram of a differential control torque component calculator according to the second embodiment; (a) A frequency characteristic diagram of a change rate calculator, (b) a map of a filter constant calculator, and (c) a frequency characteristic diagram of a steering torque deviation pseudo differentiation section. The block diagram of the servo controller of 3rd Embodiment. FIG. 11 is a block diagram of a differential control torque component calculator according to the fourth embodiment;

A plurality of embodiments of the steering control device of the present invention will be described below with reference to the drawings. An ECU as a "steering control device" is applied to an electric power steering system or a steer-by-wire system of a vehicle, and calculates a motor output command. The following embodiments mainly show examples applied to an electric power steering system. In an electric power steering system, a steering control device outputs an assist torque command to a steering assist motor. The following first to fourth embodiments will be collectively referred to as "this embodiment".

[Configuration of electric power steering system]
As shown in FIG. 1 , the electric power steering system 1 is a system that assists the driver's operation of a steering wheel 91 with the driving torque of a motor 80 . A handle 91 is fixed to one end of the steering shaft 92 , and an intermediate shaft 93 is provided to the other end of the steering shaft 92 . The steering shaft 92 and the intermediate shaft 93 are connected by a torsion bar of a torque sensor 94, and a steering shaft 95 is constructed by these. A torque sensor 94 detects the steering torque Ts based on the twist angle of the torsion bar.

A gear box 96 including a pinion gear 961 and a rack 962 is provided at the end of the intermediate shaft 93 opposite to the torque sensor 94 . When the driver turns the handle 91 , the pinion gear 961 rotates together with the intermediate shaft 93 , and the rack 962 moves left and right along with the rotation of the pinion gear 961 . Tie rods 97 provided at both ends of rack 962 are connected to tires 99 via knuckle arms 98 . The direction of the tire 99 is changed by the tie rod 97 reciprocating left and right and pulling or pushing the knuckle arm 98 .

The motor 80 is, for example, a three-phase AC brushless motor, and outputs assist torque for assisting the steering force of the steering wheel 91 according to the drive voltage Vd output from the ECU 10 . In the case of a three-phase AC motor, the drive voltage Vd means the voltages of the U-phase, V-phase, and W-phase. Rotation of the motor 80 is transmitted to the intermediate shaft 93 via a reduction mechanism 85 composed of a worm gear 86, a worm wheel 87, and the like. Steering of the steering wheel 91 and rotation of the intermediate shaft 93 due to reaction force from the road surface are transmitted to the motor 80 via the reduction mechanism 85 .

The electric power steering system 1 shown in FIG. 1 is of a column-assist type in which the rotation of the motor 80 is transmitted to the steering shaft 95. applicable to In other embodiments, a multiphase AC motor other than three phases or a DC motor with a brush may be used as the steering assist motor.

Here, the entire mechanism from the steering wheel 91 to the tire 99 through which the steering force of the steering wheel 91 is transmitted is referred to as a "steering system mechanism 100". The ECU 10 controls the steering torque Ts generated by the steering system mechanism 100 by controlling the drive torque that the motor 80 outputs to the steering system mechanism 100 . The ECU 10 also acquires a vehicle speed V detected by a vehicle speed sensor 11 provided at a predetermined portion of the vehicle.

The ECU 10 operates with power from an on-vehicle battery (not shown), and calculates an assist torque command Ta ^* based on the steering torque Ts detected by the torque sensor 94, the vehicle speed V detected by the vehicle speed sensor 11, and the like. Then, the ECU 10 applies the driving voltage Vd calculated based on the assist torque command Ta ^* to the motor 80 to cause the steering system mechanism 100 to generate the steering torque Ts.

Various arithmetic processing in the EPS-ECU 15 may be software processing by executing a program stored in advance in a substantial memory device such as a ROM by the CPU, or may be hardware processing by a dedicated electronic circuit. There may be.

[Configuration of ECU]
Next, a specific configuration of the ECU 10 of this embodiment will be described. 2 to 4 show configuration examples common to the first and second embodiments. The ECU 10 of the configuration example shown in FIG. 2 includes an estimated load calculation section 20, a target steering torque calculation section 30, a deviation calculator 39, a servo controller 400, a current feedback ("FB" in the drawing) section 70, and the like.

The estimated load calculation unit 20 includes an adder 21 and an LPF (“low-pass filter”, hereinafter the same) 22 . For example, the adder 21 adds the base assist command Tb ^* and the target steering torque Ts ^* . The LPF 22 extracts components of a predetermined frequency band, for example, 10 Hz or less, from the added torque. The estimated load calculator 20 outputs the frequency component extracted by the LPF 22 as an estimated load Tx.

The target steering torque calculation section 30 of the configuration example of FIG. 2 calculates the target steering torque Ts ^* based on the estimated load Tx and the vehicle speed V. FIG. In other configurations, the target steering torque calculation section 30 may calculate the target steering torque Ts ^* based on the steering angle. A deviation calculator 39 calculates a steering torque deviation ΔTs (=Ts ^* −Ts), which is the difference between the target steering torque Ts ^* and the steering torque Ts.

A steering torque deviation ΔTs is input to the servo controller 400 of each embodiment. Further, in the third embodiment, a steering torque Ts is also input as indicated by the dashed line. In the first and second embodiments, the servo controller 400 is controlled by a general PID controller so that the steering torque deviation ΔTs becomes 0, that is, the steering torque Ts follows the target steering torque Ts ^* . Servo control is executed, and a base assist command Tb ^* , which is an output command for the motor 80, is calculated. In the third embodiment, the servo control is executed by the derivative leading PID control that combines the PI calculation based on the steering torque deviation ΔTs and the D calculation based on the steering torque Ts.

In the configuration example of FIG. 2, since the correction torque for the base assist command Tb ^* is not calculated, the base assist command Tb ^* is directly output as the assist torque command Ta ^* . Note that in the configuration in which the correction torque is calculated as shown in FIG. 2 of Patent Document 2 (Japanese Patent No. 6252027), the sum of the base assist command Tb ^* and the correction torque is output as the assist torque command Ta ^* . .

The current feedback unit 70 applies the drive voltage Vd to the motor 80 so that the assist torque corresponding to the assist torque command Ta ^* is applied particularly to the steering shaft 95 on the tire 99 side of the torque sensor 94 . The technique of current feedback control is a well-known technique in the field of motor control, so detailed description thereof will be omitted.

(Servo controller of first and second embodiments)
FIG. 3 shows a configuration example of the servo controller 400 of the first and second embodiments. The servo controller 400 of this configuration example includes a general PID controller 410 and an accumulation calculator 490 . The configuration of the PID controller 410 other than the differential control torque component calculation unit 50 is generally the same as the configuration shown in FIG. The "pseudo-differentiation" in the differential calculation of discrete values corresponds to the calculation of (s/(τs+1) ² ) (where s: Laplace operator, τ: time constant) in terms of the transfer function of a continuous system.

The PID controller 410 calculates proportional, integral, and differential control torque components for each predetermined control cycle by PID control based on the steering torque deviation ΔTs. That is, the PID controller 410 includes a proportional control torque component calculator 430 that calculates the proportional control torque component, an integral control torque component calculator 440 that calculates the integral control torque component, and a differential control torque component that calculates the differential control torque component. It has a calculation unit 50 . In addition, in FIG. 3, the code "50" of the differential control torque component calculation section of the first embodiment is shown as the code of the differential control torque component calculation section.

Proportional control torque component calculator 430 includes delay element 45 , subtractor 463 and proportional gain (Kp) multiplier 473 . Integral control torque component calculator 440 includes delay element 45 , adder 464 and integral gain (Ki) multiplier 474 . The differential control torque component calculator 50 will be described later for each embodiment with reference to FIGS. 5, 15, and the like. A PID component adder 48 adds a PID control torque component for each control cycle and outputs a torque component TM to be processed.

The cumulative processing unit 490 cumulatively processes the processing target torque component TM, which is the sum of the PID control torque components calculated for each control cycle of the PID controller 410, and calculates the current value Tb ^* (n) of the base assist command. . Although accumulation processing is synonymous with integration processing, the term “accumulation” is used here to distinguish it from PID integral control. Depending on the calculation configuration of the servo controller, the cumulative calculation unit may not be provided, and the PID controller may output the torque after the cumulative calculation.

Accumulator 490 includes adder 491 , delay element 492 , torque dead zone processor 493 and limiting operator 494 . The adder 491 adds the previous value Tb ^* (n−1) of the base assist command input via the delay element 492 to the current value of the torque component TM to be processed. As shown in FIG. 4, the torque dead zone processor 493 sets the output Tb*(n) to 0 with a range ⁱⁿ which the absolute value of the input (TM+Tb ^* (n-1)) is equal to or less than the threshold value Td as a dead zone. A limit calculator 494 limits the absolute value of the output of the torque dead zone processor 493 to be equal to or less than the servo limit value.

By the way, when noise is superimposed on the steering torque signal Ts, the noise is amplified by the differential term of the PID controller, appears in the steering torque servo output, and causes the motor to vibrate. Therefore, in the present embodiment, the configuration of the differential control torque component calculator is devised so as to suppress the noise superimposed on the steering torque signal Ts from appearing in the differential control torque component. Hereinafter, the amount input to the differential control torque component computing section 50 will be referred to as "input to be processed". In the first and second embodiments, the steering torque deviation ΔTs is the input to be processed. The differential control torque component calculator 50 calculates a differential control torque component based on differentiation of the steering torque deviation ΔTs, which is an input to be processed.

A signal obtained by differentiating the input to be processed and processed by a filter that cuts off high-frequency components equal to or higher than a predetermined cutoff frequency is referred to as a "differential signal to be processed". The differentiated signal to be processed in the first and second embodiments is the steering torque deviation differentiated signal D(ΔTs). As a technical concept encompassing the first and second embodiments, the differential control torque component calculation unit 50 of the first embodiment and the differential control torque component calculation unit 60 of the second embodiment perform differential control as follows. Calculate the torque component.

(1) When the change in the steering torque deviation ΔTs is relatively large, the degree of contribution of the steering torque deviation differential signal D(ΔTs) having a relatively high cutoff frequency becomes relatively high.

(2) When the change in the steering torque deviation ΔTs is relatively small, the degree of contribution of the steering torque deviation differential signal D(ΔTs) having a relatively low cutoff frequency becomes relatively high.

In the first embodiment, the differential control torque component is calculated by weighted addition of the steering torque deviation differential signal D (ΔTs) processed by a plurality of filters with different cutoff frequencies. The weighted addition also includes adjustment of the weight by the differential gain corresponding to each torque deviation differential signal D(ΔTs).

In the second embodiment, the differential control torque component is calculated by processing the steering torque deviation differential signal D (ΔTs) with a filter set so that the cutoff frequency increases as the rate of change of the steering torque deviation ΔTs increases. .

Next, configuration examples of the differential control torque component computing section 50 of the first embodiment and the differential control torque component computing section 60 of the second embodiment will be described in order. Also, the effects of the first embodiment will be mainly described.

(Differential control torque component calculator of the first embodiment)
The configuration of the differential control torque component calculator of the first embodiment will be described with reference to FIGS. 5 to 7. FIG. As shown in FIG. 5, the differential control torque component calculator 50 of the first embodiment includes first and second differential calculators and a weighted adder 58 provided in parallel. Note that the configuration of the modified example indicated by broken lines in FIG. 5 will be supplemented later, and the basic example will be explained by ignoring the broken lines.

The first differential operation section includes a first pseudo differential section 511 , a first differential dead band processor 541 , a delay element 551 , a subtractor 561 and a first differential gain (Kd1) multiplier 571 . Similarly, the second differential operation section includes a second pseudo differential section 512 , a second differential dead band processor 542 , a delay element 552 , a subtractor 562 and a second differential gain (Kd2) multiplier 572 . FIG. 6 shows a configuration equivalent to the first and second pseudo differential units 511 and 512. As shown in FIG. FIG. 7 shows a map example of the first and second differential dead zone processors 541 and 542. FIG.

The first pseudo-differentiation section 511 in the first differential operation section is equivalent to a configuration in which the differential differentiation processing section 52 and the first LPF 531 are combined, as shown in FIG. A differential differentiation processing unit 52 calculates a differential differential D(ΔTs) of the steering torque deviation ΔTs. The first LPF 531 passes the first steering torque deviation differential signal D(ΔTs)1_0 in which high-frequency components higher than a relatively high first cutoff frequency fco1 (eg, 350 Hz) are cut off. "_0" at the end represents the signal before dead zone processing.

The first differential dead band processor 541 defines a small amplitude region in which the absolute value of the amplitude of the input steering torque deviation differential signal D(ΔTs)1_0 is relatively small as a dead band, and a large amplitude region in which the absolute value of the amplitude is relatively large. In the region, the steering torque deviation differential signal D(ΔTs)1_# corresponding to the input is output. "_#" at the end represents the signal after dead zone processing.

In the example shown in FIG. 7A, the first differential dead band processor 541 outputs 0 in a small amplitude region where the absolute value of the amplitude of the input signal is less than the small threshold α, and the absolute value of the amplitude of the input signal is Outputs the same value as the input in a large amplitude region greater than the large threshold β. In addition, the first differential dead band processor 541 changes the absolute value of the output as the absolute value of the amplitude of the input signal increases in the transition region where the absolute value of the amplitude of the input signal is between the small threshold value α and the large threshold value β. Gradually increase from 0 to a large threshold β.

Subtractor 561 calculates the difference between the current value of steering torque deviation differential signal D(ΔTs)1_# after dead zone processing output from first differential dead zone processor 541 and the previous value inputted via delay element 551. calculate. A first differential gain multiplier 571 outputs a value obtained by multiplying the output of the subtractor 561 by a first differential gain Kd1 reflecting the degree of contribution. Note that the delay element 551 and the subtractor 561 are an example of an arithmetic configuration and are not essential. Therefore, the first differential gain multiplier 571 may be interpreted as multiplying the output of the first differential dead band processor 541 by the first differential gain Kd1.

Descriptions of the second differential operation section that overlap with those of the first differential operation section will be omitted. The second pseudo-differentiation section 512 is equivalent to a configuration in which the differential differentiation processing section 52 and the second LPF 532 are combined. The second LPF 532 passes the second steering torque deviation differential signal D(ΔTs)2_0 in which high-frequency components equal to or higher than a relatively low second cutoff frequency fco2 (for example, 150 Hz) are blocked. That is, of the differential differential D (ΔTs) of the steering torque deviation ΔTs, the first pseudo-differential section 511 outputs a differential signal containing more high-frequency components, and the second pseudo-differential section 512 outputs a differential signal containing less high-frequency components. Output a signal.

The second differential dead band processor 542 treats a large amplitude region in which the absolute value of the amplitude of the input steering torque deviation differential signal D(ΔTs) 2_0 is relatively large as a dead band, and a small amplitude region in which the absolute value of the amplitude is relatively small. In the region, the steering torque deviation differential signal D(ΔTs)2_# corresponding to the input is output.

In the example shown in FIG. 7(b), the second differential dead band processor 542 outputs the same value as the input in a small amplitude region where the absolute value of the amplitude of the input signal is less than the small threshold α, and the amplitude of the input signal 0 is output in a large amplitude region whose absolute value is greater than the large threshold value β. In addition, the second differential dead band processor 542 changes the absolute value of the output as the absolute value of the amplitude of the input signal increases in the transition region where the absolute value of the amplitude of the input signal is between the small threshold value α and the large threshold value β. Gradually decrease from the small threshold α to 0.

Subtractor 562 calculates the difference between the current value of steering torque deviation differential signal D(ΔTs)2_# after dead band processing output from second differential dead band processor 542 and the previous value input via delay element 552. calculate. A second differential gain multiplier 572 outputs a value obtained by multiplying the output of the subtractor 562 by a second differential gain Kd2 reflecting the degree of contribution. As above, the second differential gain multiplier 572 may be interpreted as multiplying the output of the second differential deadband processor 542 by the second differential gain Kd2.

The weighted adder 58 multiplies the output of the first differential dead band processor 541 by the first differential gain Kd1, which is the "first differential control torque component", and the output of the second differential dead band processor 542 to the second differential A "second differential control torque component" which is a value multiplied by the gain Kd2 is added to calculate a differential control torque component.

Alternatively, as in the modification shown by the dashed line, first and second differential gain multipliers 571 and 572 are arranged before the first and second pseudo differential units 511 and 512, and the steering torque deviation ΔTs is calculated by the first and second Pseudo differentiation and dead zone processing may be performed on the values multiplied by the differential gains Kd1 and Kd2. In the modified example, the weighted adder 58 is "the output of the first differential dead band processor 541 that receives the steering torque deviation differential signal D(ΔTs)1_0 based on the steering torque deviation ΔTs multiplied by the first differential gain Kd1". and "the output of the second differential dead zone processor 542 that receives as input the steering torque deviation differential signal D(ΔTs)2_0 based on the steering torque deviation ΔTs multiplied by the second differential gain Kd2" for differential control. Calculate the torque component.

To summarize the above, the differential control torque component calculator 50 of the first embodiment converts the steering torque deviation differential signals D(ΔTs)1 and D(ΔTs)2 processed by a plurality of filters having different cutoff frequencies. Weighted addition is performed to calculate the differential control torque component. The differential control torque component calculation unit 50 calculates the degree of contribution of the steering torque deviation differential signal D(ΔTs)1 having a relatively high cutoff frequency, and calculates the relative absolute value of the amplitude of the steering torque deviation differential signal D(ΔTs)1. relatively high in a large band. Further, the differential control torque component calculation unit 50 calculates the degree of contribution of the steering torque deviation differential signal D(ΔTs)2 having a relatively low cutoff frequency, and the absolute value of the amplitude of the steering torque deviation differential signal D(ΔTs)2 as Make it relatively high in a relatively small band.

Here, in the examples shown in FIGS. 7A and 7B, the small threshold value α and the large threshold value β of the first and second differential dead zone processors 541 and 542, that is, the breakpoints of the input/output characteristics are common values. . As a result, the steering torque deviation differential signal D(ΔTs) in the band cut by the dead zone in the first differential dead zone processor 541 can be supplemented with the dead zone passage output of the second differential dead zone processor 542 . However, not limited to this, the small threshold value or the large threshold value of the first and second differential dead zone processors 541 and 542 may be set to different values.

Next, the process of calculating the differential control torque component by the differential control torque component calculating section 50 will be described with reference to FIGS. 8 to 12. FIG. That is, the discrete computation model shown in FIG. 5 is schematically expressed as a differential element of normal PID control, and the image of signal waveforms for each process is shown in FIGS. 8 to 12. FIG. Although illustration is omitted, the horizontal axis in FIGS. 8 to 12 is time. Also, (a) and (b) arranged vertically in FIGS. 8, 9, and 10 are treated as a set of related diagrams. [Nm] and [Nm/s] on the vertical axis of each figure are not meant as numerical units, but are for specifying dimensions of physical quantities.

FIG. 8(a) shows the steering torque deviation ΔTs superimposed with noise, and FIG. 8(b) shows the differential differential D(ΔTs) of the steering torque deviation. The thin solid line is the envelope of the differential differential signal D(ΔTs) of the steering torque deviation. The differential differential signal D(ΔTs) is sent to the first LPF process and the second LPF process, respectively. In the first LPF processing, high frequency components above a relatively high first cutoff frequency fco1 (eg, 350 Hz) are cut off. In the second LPF processing, high frequency components above a relatively low second cutoff frequency fco2 (for example, 150 Hz) are cut off.

As shown in FIG. 9A, much noise remains in the signal D(ΔTs)1_0 after the first LPF processing. However, as seen in (*1), the phase lag with respect to the envelope curve of the differential differential signal D(ΔTs) is small and the responsiveness is excellent. On the other hand, as shown in FIG. 10(a), the signal D(ΔTs)2_0 after the second LPF processing has little noise, but as seen in (*2), the envelope of the differential differential signal D(ΔTs) The phase lag with respect to the line is large.

In the first differential dead zone processing shown in FIG. 9B, the amplitude of the input signal in the small amplitude region is set to 0 to reduce noise. As a result, a signal D(ΔTs)1_# in which only the input signals in the transition region and the large amplitude region have passed through the dead band is output. In the second differential dead zone processing shown in FIG. 10(b), the amplitude of the input signal in the large amplitude region is set to zero. As a result, a signal D(ΔTs)2_# in which only the input signal in the small amplitude region and the reduced input signal in the transition region have passed through the dead band is output.

As shown in FIG. 11, the value obtained by multiplying the signal D(ΔTs)1_# after the first differential dead band processing by the first differential gain Kd1, and the signal D(ΔTs)2_# after the second differential dead band processing are combined with the second A value multiplied by the differential gain Kd2 is added and output as a differential control torque component. This differential control torque component has reduced noise and a reduced phase delay with respect to the envelope of the differential differential signal D(ΔTs).

Here, in the comparative example shown in FIG. 12, a value obtained by multiplying the signal D(ΔTs)1_0 after the first LPF processing by the first differential gain Kd1 is output as the differential control torque component. In other words, the comparative example does not include the dead zone processor and the second differential operation section. In this case, the noise of the steering torque signal Ts remarkably appears in the differential control torque component, and the motor is vibrated to generate sound and vibration. In particular, when the motor is vibrated during non-steering or slight steering, the driver is likely to notice the generated sound and mechanical vibrations, resulting in reduced comfort.

On the other hand, in the first embodiment, the steering torque deviation differential signal processed by two filters with different cutoff frequencies is subjected to dead zone processing and the addition of differential gain to adjust the degree of contribution. This suppresses the noise superimposed on the steering torque signal from being amplified by the differential term of the PID controller.

Therefore, it is possible to reduce noise and mechanical vibrations generated by the motor excited by noise, particularly during non-steering and fine steering. Also, during steering, since the responsiveness of the motor output command to the steering torque signal can be ensured, ripple vibration of the steering wheel due to phase lag can be suppressed.

Next, with reference to FIGS. 13 and 14, the behavior of the steering torque Ts, the differential control torque component, and the base assist command Tb ^* when the steering wheel is slightly steered from the neutral position while the vehicle is stopped, that is, when the steering wheel is slightly steered will be described. do. A plurality of diagrams arranged vertically in FIGS. 13 and 14 are treated as a set of related diagrams.

FIG. 13 shows, from the top, the steering torque Ts, the differential control torque component of the comparative example, and the base assist command Tb ^* . Similar to FIG. 12, this comparative example does not include a dead zone processor and a second differential operation section. In the comparative example, noise of about ±2 Nm of the differential control torque component is always superimposed on the base assist command Tb ^* . In particular, noise with a large amplitude is seen after X seconds. Such noise excites the motor at high frequencies and appears as sound and vibration.

FIG. 14 shows from top to bottom the first differential control torque component, the second differential control torque component, and the base assist command Tb ^* according to the first embodiment. The steering torque Ts shown in the upper part of FIG. 13 is input to the first and second differential calculation units. By passing through the first differential dead zone processing, the first differential control torque component (D(ΔTs)1_#×Kd1) hardly appears except for noise after X seconds during fine steering.

On the other hand, since the second cutoff frequency fco2 for the second LPF process is set lower than the first cutoff frequency fco1 for the first LPF process, the second differential control torque component (D(ΔTs)2_#×Kd2) has reduced high-frequency noise. Since the input signal in the small amplitude region passes through the second differential dead zone processing, the second differential control torque component with less noise can be obtained. As a result, since the base assist command Tb ^* with little noise is output, the torque that excites the motor at high frequency is weak and does not appear as sound or vibration.

(Differential control torque component calculator of the second embodiment)
Next, with reference to FIGS. 15 and 16, the configuration of the differential control torque component calculator of the second embodiment will be described. The differential control torque component calculation unit 60 of the second embodiment uses a filter that is set so that the cutoff frequency increases as the rate of change in the steering torque deviation ΔTs, which is the input to be processed, increases. ΔTs) to calculate the differential control torque component. In this example, the relatively high cutoff frequency is 350 Hz and the relatively low cutoff frequency is 50 Hz.

As shown in FIG. 15 , the differential control torque component calculator 60 has a change rate calculator 61 , a filter constant calculator 62 , a steering torque deviation pseudo differentiator 63 , and a differential gain processor 64 . A change rate calculator 61 calculates a change rate of the steering torque deviation ΔTs. FIG. 16(a) shows the output characteristics of the steering torque deviation change rate Rc(ΔTs) when the steering torque deviation ΔTs is input. In the pseudo-differentiation used in the rate-of-change calculation, filtering is performed by a one-stage LPF after differential differentiation is calculated.

The filter constant calculation unit 62 calculates a filter constant such that the cutoff frequency of the LPF used to calculate the steering torque deviation differential signal D(ΔTs) increases as the rate of change of the steering torque deviation ΔTs increases. FIG. 16(b) shows the relationship between the absolute value |Rc(ΔTs)| of the steering torque deviation change rate and the filter constant. When the absolute value |Rc(ΔTs)| of the steering torque deviation change rate is close to 0, the filter constant is set to a value corresponding to the cutoff frequency of 50 Hz.

When the absolute value |Rc(ΔTs)| of the rate of change of steering torque deviation is greater than the critical value Fh, the filter constant is set to a constant value corresponding to the cutoff frequency of 350 Hz. of the steering torque deviation change rate |Rc(ΔTs)|

A steering torque deviation pseudo differentiating section 63 as a "processed pseudo differentiating section" calculates a steering torque deviation differential signal D(ΔTs) based on a filter constant. FIG. 16(c) shows the output characteristic of the steering torque deviation pseudo differentiating section 63 when the steering torque deviation ΔTs is input. In pseudo-differentiation of the steering torque deviation ΔTs, filter processing is performed by a two-stage LPF after differential differentiation is calculated. As a result, when the absolute value |Rc(ΔTs)| of the rate of change of the steering torque deviation is greater than the critical value Fh, the frequency characteristic shown by LPF_H is obtained, and the absolute value |Rc(ΔTs)| The closer it gets, the closer it gets to the frequency characteristic indicated by LPF_L, and the noise in the high frequency band is suppressed.

In the example shown in FIG. 15, the differential gain processing unit 64 multiplies the steering torque deviation differential signal D(ΔTs) output by the steering torque deviation pseudo differentiating unit 63 by the differential gain Kd to calculate the differential control torque component. Alternatively, the differential gain processing section may be configured to multiply the steering torque deviation ΔTs input to the steering torque deviation pseudo differentiating section 63 by the differential gain Kd.

As described above, the differential control torque component calculator 60 of the second embodiment uses a filter whose cutoff frequency is variably set, and uses the steering torque deviation differential signal D (ΔTs) according to the steering torque deviation change rate Rc (ΔTs). ) to calculate the differential control torque component. As a result, the same effects as in the first embodiment can be obtained. Since a plurality of filters are not required as compared with the first embodiment, the configuration is simplified.

As described above, in the first and second embodiments, the input to be processed by the differential control torque component calculators 50 and 60 is the steering torque deviation ΔTs. The key concepts of the first and second embodiments are expressed as follows. "Steering control that outputs a differential control torque component with a wide band when the change in the steering torque deviation ΔTs is large, and outputs a differential control torque component with a narrower band when the change in the steering torque deviation ΔTs is small compared to when the change is large. Apparatus.” Therefore, not only the configurations illustrated in the first and second embodiments, but also other configurations according to this concept may be employed.

In contrast, in the third and fourth embodiments described below, the input to be processed by the differential control torque component calculators 505 and 605 is the steering torque Ts. The differential control torque component calculator calculates a differential control torque component based on differentiation of the steering torque Ts, which is the input to be processed. Further, the differentiated signal to be processed in the third and fourth embodiments is the steering torque differentiated signal D(Ts). The differential control torque component calculation units 505 and 605 of the third and fourth embodiments calculate the differential control torque component as follows.

(1) When the change in the steering torque Ts is relatively large, the degree of contribution of the steering torque differential signal D(Ts) having a relatively high cutoff frequency is relatively high.

(2) When the change in the steering torque Ts is relatively small, the degree of contribution of the steering torque differential signal D(Ts) having a relatively low cutoff frequency is relatively high.

The key concept of this embodiment, which includes the first to fourth embodiments, is expressed as follows. "When the change in the input to be processed by the differential control torque component calculation section is large, the differential control torque component with a wide band is output, and when the change in the input to be processed by the differential control torque component calculation section is small, the band A steering control device that outputs a differential control torque component that narrows the

(Third and fourth embodiments)
Next, with reference to FIG. 17, the configuration of the third embodiment will be described. The servo controller 400 of the third embodiment includes a derivative-first type PID controller 415 instead of the general PID controller 410 of the first embodiment shown in FIG. Differential precedence type PID control is a type of improved PID control, and in order to prevent the output from becoming an extremely large value due to differentiation when the target value changes abruptly, the input for differentiation is not the steering torque deviation ΔTs. , and the steering torque Ts, which is a controlled variable.

The configuration itself of the differential control torque component calculation unit 505 in the differential leading type PID controller 415 is substantially the same as the differential control torque component calculation unit 50 shown in FIG. However, instead of the steering torque deviation ΔTs, the steering torque Ts is input to the first and second pseudo differentiating sections 511 and 512 . Also, the first and second differential gain multipliers 591 and 592 multiply the differential gains -Kd1 and -Kd2 whose signs are inverted.

The first and second differential dead band processors 541 and 542 perform dead band processing on the input steering torque differential signals D(Ts)1_0 and D(Ts)2_0 to obtain the steering torque differential signals D(Ts)1_# and D( Ts) Output 2_#. That is, in the third embodiment, "deviation" is deleted from the terms such as "differential signal" in the first embodiment, and (ΔTs) in the symbols is replaced with (Ts). As for the characteristics of the first and second differential dead zone processors 541 and 542, the characteristics similar to those shown in FIGS. 7A and 7B can be used as they are.

Note that the proportional control torque component calculation unit 430 and the integral control torque component calculation unit 440 in the differential precedence type PID controller 415 perform PI calculation based on the steering torque deviation ΔTs, like the general PID controller 410 . Also, the 400 accumulation calculation unit 490 of the servo controller is the same as that of FIG.

Also in the configuration of the third embodiment, when noise is superimposed on the steering torque Ts, the noise is amplified by the differential term. However, based on the key concept of the present embodiment, the degree of contribution of the steering torque differential signal with a different cutoff frequency is adjusted according to the change in the steering torque Ts, which is the input to be processed by the differential control torque component calculator. Thus, noise appearing in the output of the servo controller 400 can be suppressed.

FIG. 18 shows a differential control torque component calculator 605 of the fourth embodiment corresponding to the second embodiment. The differential control torque component calculator 605 can be replaced with the differential control torque component calculator 505 in FIG. In the fourth embodiment, the input to be processed by the differential control torque component calculation unit 605 is changed from the steering torque deviation ΔTs to the steering torque Ts, and the differential gain processing unit 64 multiplies by the differential gain (−Kd) whose sign is inverted. . The reference numeral "63" of the steering torque pseudo-differentiation section as the "processed pseudo-differentiation section" shares the reference numeral of the steering torque deviation pseudo-differentiation section of the second embodiment. In the fourth embodiment, noise appearing in the output of the servo controller 400 can be similarly suppressed with only such a slight difference in configuration.

(Other embodiments)
The steering control device of the present invention is applicable not only to the electric power steering system but also to the reaction force control device of the steer-by-wire system in which the steering wheel and the steered wheels are mechanically separated, as disclosed in FIG. 11 of Patent Document 2. may be In this case, in the servo controller, the steering torque Ts detected by the reaction force detection device follows the target steering torque Ts ^* calculated according to the load and steering angle calculated by the tire steering device. controlled.

The derivative leading type PID control shown in the third and fourth embodiments is an example of servo control other than general PID control. The present invention can be applied to other servo control configurations as long as the control configuration has a processing path for differentiating the steering torque including noise.

The present invention is not limited to such embodiments, and can be embodied in various forms without departing from the scope of the invention.

The controller and techniques described in this disclosure may be implemented by a dedicated computer provided by configuring a processor and memory programmed to perform one or more functions embodied by the computer program. may be Alternatively, the controller and techniques described in this disclosure may be implemented by a dedicated computer provided by configuring the processor with one or more dedicated hardware logic circuits. Alternatively, the controller and techniques described in this disclosure can be implemented by a combination of a processor and memory programmed to perform one or more functions and a processor configured by one or more hardware logic circuits. It may also be implemented by one or more dedicated computers configured. The computer program may also be stored as computer-executable instructions on a computer-readable non-transitional tangible recording medium.

10 ... steering control device,
400 Servo controller,
50, 505, 60, 605... Differential control torque component calculator,
80... Motor.

Claims (5)

A steering control device for calculating an output command (Tb ^* ) for a motor (80) by a servo controller (400) so that a steering torque (Ts) follows a target steering torque (Ts ^* ),
The servo controller uses a steering torque deviation (ΔTs), which is a difference between a target steering torque and a steering torque, or a steering torque as an input to be processed, and differentiates the differential control torque component based on the differentiation of the input to be processed. having a control torque component calculator (50, 505, 60, 605),
When the input to be processed is differentiated and a signal processed by a filter that cuts off high-frequency components equal to or higher than a predetermined cutoff frequency is used as the differentiated signal to be processed,
The differential control torque component calculation unit is
When the change in the input to be processed is relatively large, the contribution of the differentiated signal to be processed having a relatively high cutoff frequency becomes relatively high,
A steering control device that calculates the differential control torque component such that when the change in the input to be processed is relatively small, the contribution of the differential signal to be processed having a relatively low cutoff frequency becomes relatively high. The differential control torque component calculator (50, 505)
weighted addition of the differential signal to be processed processed by a plurality of filters having different cutoff frequencies to calculate the differential control torque component,
making the degree of contribution of the differential signal to be processed having a relatively high cutoff frequency relatively high in a band in which the absolute value of the amplitude of the differential signal to be processed is relatively large;
2. The steering control device according to claim 1, wherein the contribution of the differential signal to be processed having a relatively low cutoff frequency is made relatively high in a band in which the absolute value of the amplitude of the differential signal to be processed is relatively small. The differential control torque component calculation unit is
The differential signal to be processed from which high-frequency components equal to or higher than the first cutoff frequency (fco1) are cut off is input, and a small amplitude region in which the absolute value of the amplitude of the differential signal to be processed is relatively small is defined as a dead zone, and the absolute value of the amplitude is a first differential dead band processor (541) for outputting the differential signal to be processed according to the input in a large amplitude region with relatively large values;
The differential signal to be processed in which high-frequency components equal to or higher than a second cutoff frequency (fco2) lower than the first cutoff frequency are cut off is input, the large amplitude range is set as a dead zone, and the small amplitude range is controlled according to the input. a second differential dead zone processor (542) for outputting the differential signal to be processed;
The first differential signal to be processed based on the input to be processed multiplied by the first differential gain or the value obtained by multiplying the output of the first differential dead band processor by the first differential gain (Kd1) The differential to be processed based on the output of the differential dead band processor and the value obtained by multiplying the output of the second differential dead band processor by a second differential gain (Kd2), or the input to be processed multiplied by the second differential gain. a weighted adder (58) for calculating the differential control torque component by adding the output of the second differential dead band processor to which a signal is input;
3. A steering control device according to claim 2, comprising: The differential control torque component calculator (60, 605)
2. The steering control device according to claim 1, wherein the differential control torque component is calculated by processing the differential signal to be processed by a filter set so that the cutoff frequency increases as the rate of change of the input to be processed increases. . The differential control torque component calculation unit is
a change rate calculation unit (61) for calculating a change rate of the input to be processed;
a filter constant calculation unit (62) for calculating a filter constant such that the cutoff frequency used for calculation of the differential signal to be processed increases as the rate of change of the input to be processed increases;
a processing target pseudo-differentiation unit (63) that calculates the processing target differential signal based on the filter constant;
a differential gain processing unit (64) that multiplies a differential gain (Kd) to the differential signal to be processed output by the pseudo-differential unit to be processed or the input to be processed that is input to the pseudo-differential unit to be processed;
5. The steering control device according to claim 4, comprising:

Images