Adaptive feedforward assist control method for electric power steering of automobile

Abstract

The application discloses a kind of self-adapting feedforward assist control methods of automobile electric power steering, and main work flow is as follows: electric power steering (EPS) is linear interpolation calculation according to current safe vehicle speed under different vehicle speed under the rack force-target hand force curve that can be calibrated rack force-target hand force curve at current vehicle speed, and combined with EPS mechanical parameters, rack force-target hand force curve at current vehicle speed is calculated to obtain the assist curve, with the assist curve at current vehicle speed and EPS torsion bar torque as the basis, combined with road surface shaking count value, rack end protection activation state, additional request torque of auxiliary driving function, self-adapting adjustment feedforward control related parameters, and then calculate the EPS adaptive feedforward assist torque, so that driver can obtain safe and smooth steering assist under different steering conditions, provide better driving experience for driver.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of automobile technology, and particularly relates to a self-adaptive feedforward assist control method for electric power steering of an automobile. BACKGROUND

[0002] Under the wave of software-defined vehicles, the application layer control software of vehicle chassis dynamics develops rapidly, and the electric power steering machine (EPS) is a key component that affects the lateral dynamics of the chassis. How the electric power steering machine can provide quick and safe and smooth steering assist for the driver under different vehicle driving conditions becomes an important part of improving the vehicle handling stability. The market expects that the electric power steering machine should be able to achieve the expected assist effect under various vehicle driving conditions during the normal driving of the driver, and should meet the functional safety target requirements of the whole vehicle in the non-hand-off steering condition, and should not appear obvious unintended steering, reverse assist, assist torque fluctuation, assist loss, etc.

[0003] At present, the electric power steering machine of some vehicles on the market has poor robustness when driving on a bumpy road or the steering angle is close to the end of the rack, and different degrees of assist motor torque fluctuation, steering wheel shaking, etc. occur, which is mainly caused by the following two reasons: first, the feedforward calibration parameters are only based on the assist curve, and cannot combine the actual situation of the road, so the debugging and optimization of the steering assist parameters are difficult; second, a set of debugging parameters cannot simultaneously compensate for the different phase deviations between the torsion bar torque and the motor assist torque in the dynamic change process. SUMMARY

[0004] To solve the above technical problems, the present application provides a self-adaptive feedforward assist control method for electric power steering of an automobile, comprising the following steps:

[0005] Step 1, the electric power steering system receives the safety vehicle speed signal of the communication input module, the torsion bar torque, the torsion bar torque change gradient, the motor rotor speed, the motor rotor acceleration of the sensor signal acquisition and calculation module, the rack end protection activation state of the rack end protection module, and the additional torque request value of the auxiliary driving function module.

[0006] Step 2, according to the safety vehicle speed signal, the rack force-target torsion bar torque calibration curve under different vehicle speeds is calculated by linear interpolation lookup table, and the rack force-target torsion bar torque relationship curve under the current vehicle speed, i.e. the hand feeling curve under the current vehicle speed, is obtained;

[0007] Step 3, the hand feeling curve under the current vehicle speed is subjected to non-negative gradient amplitude limiting, and the hand feeling curve under the current vehicle speed is subjected to upper and lower safety boundary amplitude limiting;

[0008] Step 4, according to the mechanical conversion coefficient of the steering machine assist motor torque to the rack force and the mechanical conversion coefficient of the torsion bar torque to the rack force, the hand feeling curve after being limited by the safety boundary at the current vehicle speed in step 3 is coordinate transformed to calculate the torsion bar torque-motor assist torque relationship curve at the current vehicle speed, that is, the assist curve at the current vehicle speed;

[0009] Step 5, the assist curve at the current vehicle speed is limited by the upper and lower safety boundaries;

[0010] Step 6, the assist curve after being limited by the safety boundary at the current vehicle speed is redundantly calculated and verified;

[0011] Step 7, according to the motor rotor acceleration signal in step 1, the road surface shaking count value is calculated.

[0012] Step 8, according to the safety vehicle speed signal in step 1, the vehicle running direction is judged, and the assist curve stable gradient limit value is calculated in combination with the road surface shaking count value in step 7;

[0013] Step 9, according to the assist curve stable gradient limit value in step 8, the assist curve after being limited by the safety boundary at the current vehicle speed is gradient limited to obtain the safety assist curve and the stable assist curve.

[0014] Step 10, the additional torque request value in step 1 is superimposed with the torsion bar torque signal to obtain the total requested torsion bar torque;

[0015] Step 11, according to the absolute value of the total requested torsion bar torque in step 10, the safety assist curve and the stable assist curve in step 9 are respectively calculated by linear interpolation, and multiplied by the direction symbol of the total torsion bar torque value to obtain the safety assist torque and the stable assist torque;

[0016] Step 12, according to the rack end protection activation state, the difference between the safety assist torque and the stable assist torque in step 11 is PT1 filtered, when the rack end protection is activated, the alternative filter coefficient under the rack end protection activation state is adopted, otherwise the default filter coefficient is adopted for PT1 filtering calculation, and the filtered output value is superimposed with the stable assist torque and then limited to output to obtain the basic assist torque;

[0017] Step 13, according to the road surface shaking count value in step 7, the dynamic torque compensation parameter and the motor rotor inertia compensation parameter are selected;

[0018] Step 14, according to the rack end protection activation state in step 1, in combination with the dynamic assist torque compensation parameter in step 13, the high assist gradient area dynamic compensation torque calculation and the low assist gradient area dynamic compensation torque calculation are performed;

[0019] Step 15, according to the high power gradient zone dynamic compensation moment and the low power gradient zone dynamic compensation moment calculated in step 14, superimposition is carried out and the maximum safety boundary limiting is carried out, and a dynamic power moment is obtained;

[0020] Step 16, according to the motor rotor acceleration signal and the safety vehicle speed signal in step 1, the motor rotor inertia compensation parameter in step 13 is combined to calculate a motor rotor inertia compensation moment;

[0021] Step 17, the base power moment in step 12, the dynamic power moment in step 15 and the motor rotor inertia compensation moment in step 16 are added to obtain an EPS adaptive feedforward power moment, and the power curve redundancy calculation verification result in step 6 is combined to be output to the motor control module for torque execution.

[0022] Compared with the prior art, the present application has the following beneficial effects:

[0023] The present application is based on the rack force-target torsion bar force relationship curve, carries out EPS feedforward power parameter calibration, directly associates the driver's feeling and the actual road working condition, and makes the EPS feedforward power calibration process simpler.

[0024] The present application provides a debugging interface for road roughness detection and rack end protection activation state, thereby improving the robustness of the electric power steering system under rough road and large angle steering working conditions, and making the vehicle driving process safer and smoother.

[0025] The present application can make the electric power steering system provide adaptive feedforward power moment according to different steering working conditions, improve the robustness of the electric power steering system, optimize the debugging process of the steering power parameter, shorten the debugging time and save the project development cost. BRIEF DESCRIPTION OF DRAWINGS

[0026] The present application will be further described in detail below in combination with the drawings and specific embodiments:

[0027] Figure 1 It is the software system architecture diagram of the present application;

[0028] Figure 2 It is the target torsion bar moment minimum safety boundary limiting algorithm principle diagram of the present application;

[0029] Figure 3 It is the power curve redundancy calculation verification algorithm principle diagram of the present application;

[0030] Figure 4 It is the road roughness detection algorithm principle diagram of the present application;

[0031] Figure 5 It is the power curve dynamic gradient limit algorithm principle diagram of the present application. Detailed Implementation

[0032] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can fully understand other advantages and technical effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through different specific embodiments, and the details in this specification can also be applied based on different viewpoints, with various modifications or changes made without departing from the overall design concept of the invention. It should be noted that, unless otherwise specified, the following embodiments and features can be combined with each other. The following exemplary embodiments of the present invention can be implemented in many different forms and should not be construed as being limited to the specific embodiments set forth herein. It should be understood that these embodiments are provided to make the disclosure of the present invention thorough and complete, and to fully convey the technical solutions of these exemplary embodiments to those skilled in the art.

[0033] This invention discloses an adaptive feedforward assist control method for electric power steering in automobiles. Its purpose is to adaptively feedforward control the basic assist output of electric power steering based on signals such as the current safe vehicle speed, torsion bar torque, road vibration count value, and rack end protection activation state, so that the driver can obtain safe and smooth steering assistance under different steering conditions, providing the driver with a better driving experience.

[0034] In this embodiment, the software architecture of the present invention is as follows: Figure 1 As shown, the system includes a communication input module, a sensor signal acquisition and calculation module, a rack end protection module, a road surface unevenness detection module, a hand feel curve interpolation calculation module, an assist curve calculation module, an assist curve safety verification module, a basic assist calculation module, a dynamic assist calculation module, a motor rotor inertia compensation module, an assist torque limiting module, a motor control module, and an auxiliary driving function module. The main workflow is as follows: The electric power steering (EPS) performs linear interpolation calculations on the calibrable rack force-target hand force curves at different vehicle speeds based on the current safe vehicle speed to obtain the rack force-target hand force curve at the current vehicle speed. Combined with the EPS mechanical parameters, it calculates the assist curve at the current vehicle speed. Based on the assist curve at the current vehicle speed and the EPS torsion bar torque, and considering the road vibration count, rack end protection activation status, and additional request torque from the auxiliary driving function, it adaptively adjusts the feedforward control parameters to calculate the EPS adaptive feedforward assist torque.

[0035] Specifically, this embodiment provides an adaptive feedforward assist control method for electric power steering in automobiles, including the following steps:

[0036] Step 1, the electric power steering system receives the safety vehicle speed signal of the communication input module, the sensor signal acquisition and calculation module, the rack end protection module and the auxiliary driving function module in real time.

[0037] Step 2, according to the safety vehicle speed signal, the rack force-target torsion bar torque calibration curve at different vehicle speeds is calculated by linear interpolation lookup table, and the rack force-target torsion bar torque relationship curve at the current vehicle speed, i.e. the current vehicle speed hand feeling curve, is obtained.

[0038] Step 3, the non-negative gradient amplitude limiting of the hand feeling curve at the current vehicle speed is carried out, and the upper and lower safety boundary amplitude limiting of the hand feeling curve at the current vehicle speed is carried out.

[0039] Step 4, according to the mechanical conversion coefficient of the steering machine assist motor torque to the rack force and the mechanical conversion coefficient of the torsion bar torque to the rack force, the hand feeling curve at the current vehicle speed after the safety boundary limiting in step 3 is coordinate transformed, and the torsion bar torque-motor assist torque relationship curve at the current vehicle speed, i.e. the current vehicle speed assist curve, is calculated.

[0040] Step 5, the upper and lower safety boundary amplitude limiting of the assist curve at the current vehicle speed is carried out.

[0041] Step 6, the assist curve at the current vehicle speed after the safety boundary limiting is carried out.

[0042] Step 7, according to the motor rotor acceleration signal in step 1, the road surface jitter count value is calculated.

[0043] Step 8, according to the safety vehicle speed signal in step 1, the vehicle driving direction is judged, and the assist curve stable gradient limit value is calculated combined with the road surface jitter count value in step 7.

[0044] Step 9, according to the assist curve stable gradient limit value in step 8, the gradient amplitude limiting of the assist curve at the current vehicle speed after the safety boundary limiting is carried out, and the safety assist curve and the stable assist curve are obtained.

[0045] Step 10, the additional torque request value in step 1 is superimposed with the torsion bar torque signal, and the total requested torsion bar torque is obtained.

[0046] Step 11, according to the absolute value of the total requested torsion bar torque in step 10, the safety assist curve and the stable assist curve in step 9 are respectively calculated by linear interpolation, and multiplied by the direction symbol of the total torsion bar torque value, and the safety assist torque and the stable assist torque are obtained.

[0047] Step 12, according to the rack end protection activation state, the difference between the safety assist torque and the stable assist torque in step 11 is subjected to PT1 filtering, when the rack end protection is activated, the alternative filtering coefficient under the rack end protection activation state is adopted, otherwise the default filtering coefficient is adopted for PT1 filtering calculation, and the filtered output value is superimposed with the stable assist torque and then subjected to amplitude limiting output to obtain the basic assist torque;

[0048] Step 13, according to the road surface jitter count value in step 7, the dynamic torque compensation parameter and the motor rotor inertia compensation parameter are selected;

[0049] Step 14, according to the rack end protection activation state in step 1, in combination with the dynamic assist torque compensation parameter in step 13, the high assist gradient zone dynamic compensation torque calculation and the low assist gradient zone dynamic compensation torque calculation are performed;

[0050] Step 15, according to the high assist gradient zone dynamic compensation torque and the low assist gradient zone dynamic compensation torque calculated in step 14, superposition is performed and maximum safety boundary amplitude limiting is performed to obtain the dynamic assist torque;

[0051] Step 16, according to the motor rotor acceleration signal and the safety vehicle speed signal in step 1, in combination with the motor rotor inertia compensation parameter in step 13, the motor rotor inertia compensation torque is calculated;

[0052] Step 17, the basic assist torque in step 12, the dynamic assist torque in step 15 and the motor rotor inertia compensation torque in step 16 are added to obtain the EPS adaptive feedforward assist torque, and the assist curve redundancy calculation verification result in step 6 is combined to be output to the motor control module for torque execution.

[0053] More specifically, in the step 3, the following steps are implemented:

[0054] Step 31, the maximum target torsion bar torque of the hand feeling curve is subjected to maximum boundary amplitude limiting according to the maximum effective range value of the torsion bar torque sensor;

[0055] Step 32, the ordinate of the last value point of the hand feeling curve is set as the maximum effective range value of the torsion bar torque sensor;

[0056] Step 33, the hand feeling curve is subjected to non-negative gradient amplitude limiting;

[0057] Step 34, the target torsion bar torque is subjected to the lowest safety boundary amplitude limiting according to the following logic:

[0058] When the vehicle speed is less than or equal to the calibrated safety limit vehicle speed V lim , the hand feeling curve is not subjected to the lowest safety target torsion bar torque amplitude limiting;

[0059] When the vehicle speed is greater than the calibrated safety limit speed V lim , the rack force is lower than the first order safety rack force F safe1 , the hand feeling curve within the interval is limited by the first order lowest safety target twist beam torque limit;

[0060] When the vehicle speed is greater than the calibrated safety limit speed V lim , the rack force is higher than the first order safety rack force F safe1 and lower than the second order safety rack force F safe2 , the hand feeling curve within the interval is limited by the first order lowest safety target twist beam torque T safe1 and T safe1 ; safe2 may be calibrated, and T safe1 < T safe2 ;

[0061] When the vehicle speed is greater than the calibrated safety limit speed V lim , the rack force is higher than the second order safety rack force F safe2 , the hand feeling curve within the interval is limited by the second order lowest safety target twist beam torque T safe2 and T safe1 ; safe2 may be calibrated, and T safe1 < T safe2 ;

[0062] The principle of the lowest safety boundary limit algorithm of the hand feeling curve target twist beam torque is shown in Figure 2 , where V lim represents the safety limit speed, F safe1 represents the first order safety rack force, F safe2 represents the second order safety rack force, T safe1 represents the first order lowest safety target twist beam torque, T safe2 represents the second order lowest safety target twist beam torque, and i represents the i-th value point of the hand feeling curve.

[0063] More specifically, the calculation formula of each value point of the assist curve at the current vehicle speed calculated in the step 4 is as follows:

[0064]

[0065] Wherein:

[0066] FRACK Xaxlei is the value of the rack force of the hand feeling curve sampling point, in units of N

[0067] DesTBT Yaxlei is the value of the target twist beam torque of the hand feeling curve sampling point, in units of Nm

[0068] TBT XaxleiThe value of the torsion bar torque for the abscissa of the sampling point of the curve is given in Nm.

[0069] MOT Yaxlei The vertical axis represents the motor assist torque value at the sampling points of the assist curve, in Nm.

[0070] X tbt2rack The mechanical conversion factor for EPS from motor torque to rack force

[0071] X mot2rack The mechanical conversion factor for EPS from torsion bar torque to rack force.

[0072] More specifically, in step 5, the lower safety boundary amplitude of the assist curve at the current vehicle speed is 0, and the upper boundary amplitude is the maximum motor output torque that can be provided.

[0073] More specifically, in step 6, the principle of the assist curve redundancy calculation and verification algorithm is as follows: Figure 3 As shown, the redundancy calculation and verification results of the assist curve are divided into the following 5 types:

[0074] 1) When the ordinate of the first point of the feel curve used for security verification is RevDesTBT Yaxlei The deviation from 0 Nm exceeds the set threshold Tol m Or the x-coordinate of the first point of the feel curve used for security verification, RevFRACK Xaxlei The deviation from 0 Nm exceeds the set threshold Tol n If the error occurs, the verification failure result will be output, and the fault code Err will be output as 1.

[0075] 2) When the ordinate of the last point of the feel curve used for security verification is RevDesTBT Yaxlei With the maximum effective range value TBT of the torsion bar torque sensor max The deviation exceeds the set threshold Tol m And when the ordinate of the last point of the boost curve is RevMOT Yaxlei With the maximum motor output torque (MOT) that can be provided max The deviation exceeds the set threshold Tol m If the error occurs, the verification failure result will be output, and the fault code Err will be output as 2.

[0076] 3) When the ordinate of any point in the feel curve used for security verification is RevDesTBT Yaxlei The minimum safety boundary limiting condition in step 3 is triggered, and the threshold Tol is exceeded below the minimum safety boundary. m If the error occurs, the verification failure result will be output, and the fault code Err 4 will be output.

[0077] 4) the ordinate RevDesTBT of any one of the value points of the feel curve used for safety check Yaxlei the ordinate DesTBT of the corresponding point of the safety feel curve Yaxlei deviation exceeds the set threshold Tol m , or the abscissa RevFRACK of any one of the value points of the feel curve used for safety check Xaxlei the abscissa FRACK of the corresponding point of the safety feel curve Xaxlei deviation exceeds the set threshold Tol n , then the check failure result is output, and the fault code Err is output 8;

[0078] 5) in other cases, the check pass result is output, and the fault code 0 is output.

[0079] More specifically, in step 7, the road surface unevenness detection algorithm principle is as shown in Figure 4 The method for calculating the road surface jitter count value through the rotor acceleration signal is as follows:

[0080] The road surface jitter counter enters the initialization state, and the road surface jitter timing value and the road surface jitter count value are set to 0;

[0081] When it is detected that the motor rotor acceleration changes in a direction by more than the road surface jitter calibration gradient value, the rotor acceleration state and direction are recorded, and the road surface jitter timing is started;

[0082] When it is detected that the motor rotor acceleration changes in the opposite direction of the recorded rotor acceleration state by more than the road surface jitter calibration gradient value, the rotor acceleration state and direction are recorded, the road surface jitter count value is increased by 1, and the road surface jitter timing value is set to 0;

[0083] When the road surface jitter timing value exceeds the road surface jitter detection maximum interval time, the road surface jitter counter returns to the initialization state, and the road surface jitter timing value and the road surface jitter count value are set to 0;

[0084] More specifically, in step 8, the assist curve dynamic gradient limit value algorithm principle is as shown in Figure 5 The calculation method of the assist curve stable gradient limit value is as follows:

[0085] When the vehicle travel direction signal is forward travel, and the road surface jitter count value Count cr is less than the set threshold Count lim : the assist curve stable gradient limit value is equal to the assist curve stable gradient limit value Grad spd at the current vehicle speed when the vehicle travels forward.

[0086] When the vehicle travel direction signal is forward travel, and the road surface jitter count value Count crEqual to or greater than the set threshold value Count lim Equal to or greater than the set threshold value Count spd Equal to or greater than the set threshold value Count cr The smaller one of the two.

[0087] When the vehicle driving direction signal is reverse driving, and the road surface jolt count value Count cr is less than the set threshold value Count lim : the assist curve stable gradient limit value is equal to the assist curve stable gradient limit value Grad revspd at the current vehicle speed when the vehicle is reverse driving.

[0088] When the vehicle driving direction signal is reverse driving, and the road surface jolt count value Count cr is greater than or equal to the set threshold value Count lim : the assist curve stable gradient limit value is equal to the assist curve stable gradient limit value Grad revspd at the current vehicle speed when the vehicle is reverse driving. cr The smaller one of the two.

[0089] More specifically, in step 9, according to the assist curve safety gradient calibration limit value and the assist curve stable gradient limit value, the safety-limited assist curve at the current vehicle speed is gradient-limited in turn to obtain the safety assist curve and the stable assist curve, and the safety gradient calibration limit value is greater than the assist curve stable gradient limit value.

[0090] More specifically, in step 10, the specific calculation method of the total requested torsion bar torque is as follows: the maximum boundary limiting of the torsion bar torque is performed according to the maximum effective range value of the torsion bar torque sensor; the limited torsion bar torque is superimposed with the additional torque request value to obtain the total requested torsion bar torque.

[0091] More specifically, in step 11, the linear interpolation method of the stable assist curve is as follows:

[0092] If the absolute value of the total requested torsion bar torque is less than or equal to the abscissa of the value point with the largest abscissa in the stable assist curve, the absolute value of the total requested torsion bar torque is used to perform internal linear interpolation calculation on the stable assist curve to obtain the stable assist torque.

[0093] If the absolute value of the total request torque is greater than the abscissa of the maximum value point of the stable assist curve, the absolute value of the total request torque is used to perform external linear interpolation calculation on the stable assist curve to obtain the stable assist torque, and the ordinate of the maximum value point of the safety assist curve is used to limit the maximum boundary of the stable assist torque.

[0094] More specifically, in step 12, the specific calculation formula of the basic assist torque is:

[0095] M pt1 = (M safe -M steady -M pt1last )*X pt1 +M pt1last

[0096] M basic =M steady +M pt1

[0097] Wherein:

[0098] X pt1 is the PT1 filter coefficient

[0099] M safe is the safety assist torque, unit Nm

[0100] M steady is the stable assist torque, unit Nm

[0101] M pt1last is the output value of the PT1 filter value in the last operation period, and the initial value is 0, unit Nm

[0102] M pt1 is the PT1 filter value, unit Nm

[0103] M basic is the basic assist torque, unit Nm

[0104] More specifically, in step 13, when the road surface jitter count value exceeds the set value, the dynamic assist torque compensation replacement parameter is used; when the road surface jitter count value does not exceed the set value, the safety vehicle speed and motor rotor speed linear interpolation calculation is used to obtain the low assist gradient zone and high assist gradient zone default parameters. When the road surface jitter count value exceeds the set value, the safety vehicle speed linear interpolation calculation is used to obtain the motor rotor inertia compensation replacement parameter; when the road surface jitter count value does not exceed the set value, the safety vehicle speed linear interpolation calculation is used to obtain the motor rotor inertia compensation default parameter.

[0105] More specifically, in step 14, the calculation methods of the high assist gradient zone dynamic compensation torque and the low assist gradient zone dynamic compensation torque are as follows:

[0106] When the rack end protection trigger is activated and the road shock count value exceeds the set value: the high assistance gradient zone compensation torque is 0, and the low assistance gradient zone compensation torque is calculated as follows:

[0107] M LOffset = X Chattersub * T grad

[0108] Wherein:

[0109] M LOffset is the low assistance gradient zone compensation torque, unit Nm

[0110] X Chattersub is the dynamic assistance torque compensation substitute parameter

[0111] T grad is the torsion bar torque change gradient, unit Nm / ms

[0112] When the rack end protection trigger is activated and the road shock count value does not exceed the set value: the high assistance gradient zone compensation torque is calculated as follows:

[0113] M HOffset = X Hintpol * T grad * X AssistGrad

[0114] Wherein:

[0115] M HOffset is the high assistance gradient zone compensation torque, unit Nm

[0116] X Hintpol is the high assistance gradient zone parameter linearly interpolated from the safe vehicle speed and the motor rotor speed

[0117] T grad is the torsion bar torque change gradient, unit Nm / ms

[0118] X AssistGrad is the assistance curve gradient, unit Nm / Nm

[0119] The low assistance gradient zone compensation torque is calculated as follows:

[0120] M LOffset = X Endstopsub * T grad

[0121] Wherein:

[0122] M LOffset is the low assistance gradient zone compensation torque, unit Nm

[0123] X EndstopsubLinear interpolation of rack end protection alternative parameter for safe vehicle speed

[0124] T grad Torsion bar torque change gradient, unit Nm / ms

[0125] When rack end protection is not triggered and road surface bump count exceeds a set value: high assist gradient zone compensation torque is calculated as follows:

[0126] M Hoffset = X Chattersub * T grad * X AssistGrad

[0127] Where:

[0128] M HOffset High assist gradient zone compensation torque, unit Nm

[0129] X Chattersub Dynamic assist torque compensation alternative parameter

[0130] T grad Torsion bar torque change gradient, unit Nm / ms

[0131] X AssistGrad Assist curve gradient, unit Nm / Nm

[0132] Low assist gradient zone compensation torque is calculated as follows:

[0133] M LOffset = X Chattersub * T grad

[0134] Where:

[0135] M LOffset Low assist gradient zone compensation torque, unit Nm

[0136] X Chattersub Dynamic assist torque compensation alternative parameter

[0137] T grad Torsion bar torque change gradient, unit Nm / ms

[0138] When rack end protection is not triggered and road surface bump count does not exceed a set value: high assist gradient zone compensation torque is calculated as follows:

[0139] M HOffset = X Hintpol * T grad * X AssistGrad

[0140] Where:

[0141] M HOffset For compensation torque in the high-assist gradient region, the unit is Nm

[0142] X Hintpol High assist gradient region parameters obtained by linear interpolation of safe vehicle speed and motor rotor speed

[0143] T grad This represents the gradient of the torsion bar torque, in Nm / ms.

[0144] X AssistGrad To aid the gradient of the curve, the unit is Nm / Nm

[0145] The compensation torque in the low boost gradient region is calculated using the following formula:

[0146] M Loffset =X Lintpol *T grad

[0147] in:

[0148] M LOffset Compensation torque for low boost gradient region, unit Nm

[0149] X Lintpol Low assist gradient region parameters obtained by linear interpolation of safe vehicle speed and motor rotor speed

[0150] T grad This represents the gradient of the torsion bar torque, in Nm / ms.

[0151] More specifically, in step 15, the dynamic assist torque is calculated using the following formula:

[0152] M dynamic =M HOffset +M LOffset

[0153] in:

[0154] M dynamic Dynamic assist torque, unit Nm

[0155] M HOffset For compensation torque in the high-assist gradient region, the unit is Nm

[0156] M LOffset Compensation torque for low boost gradient region, unit Nm

[0157] For dynamic assist torque M dynamic Output after applying the maximum safety boundary limit.

[0158] More specifically, in step 16, the dynamic assist torque is calculated using the following formula:

[0159] M InertiaComp =ARotAcc X RotFactor

[0160] Wherein:

[0161] M InertiaComp is the motor rotor inertia compensation motor torque, unit Nm

[0162] A RotAcc is the motor rotor acceleration, unit 1 / S 2

[0163] X RotFactopr is the motor rotor inertia compensation parameter, unit kg*m 2

[0164] More specifically, in step 17, the EPS adaptive feedforward assist torque calculation formula executed by the steering assist motor is as follows:

[0165]

[0166] Wherein:

[0167] M out is the EPS adaptive feedforward assist torque executed by the steering assist motor, unit Nm

[0168] M basic is the basic assist torque, unit Nm

[0169] M dynamic is the dynamic assist torque, unit Nm

[0170] M InertiaComp is the motor rotor inertia compensation motor torque, unit Nm

[0171] Err is the fault code output by the assist curve safety verification module

[0172] The present application has been described in detail through specific embodiments and examples, but these do not constitute a limitation on the present application. Those skilled in the art can also make many modifications and improvements without departing from the principles of the present application, and these should also be considered as falling within the scope of protection of the present application.

Claims

1. A self-adaptive feedforward assist control method for electric power assisted steering of an automobile, characterized by, The method comprises the following steps: Step 1, the electric power steering system receives the safety vehicle speed signal of the communication input module in real time, the sensor signal acquisition and calculation module of the torsion bar torque, the torsion bar torque change gradient, the motor rotor speed, the motor rotor acceleration, the rack end protection module of the rack end protection activation state, the additional torque request value of the auxiliary driving function module; Step 2, according to the safety vehicle speed signal, the rack force-target torsion bar torque calibration curve at different vehicle speeds is calculated by linear interpolation lookup table, and the rack force-target torsion bar torque relationship curve at the current vehicle speed, that is, the current vehicle speed hand feeling curve is obtained; Step 3, the hand feeling curve at the current vehicle speed is subjected to non-negative gradient limiting, and the hand feeling curve at the current vehicle speed is subjected to upper and lower safety boundary limiting; Step 4, according to the mechanical conversion coefficient of the steering machine assist motor torque to the rack force and the mechanical conversion coefficient of the torsion bar torque to the rack force, the hand feeling curve at the current vehicle speed subjected to the safety boundary limiting in step 3 is subjected to coordinate transformation, and the torsion bar torque-motor assist torque relationship curve at the current vehicle speed, that is, the current vehicle speed assist curve is calculated; Step 5, the assist curve at the current vehicle speed is subjected to upper and lower safety boundary limiting; Step 6, the assist curve at the current vehicle speed subjected to the safety boundary limiting is subjected to redundancy calculation verification; Step 7, according to the motor rotor acceleration signal in step 1, the road surface jitter count value is calculated; Step 8, according to the safety vehicle speed signal in step 1, the vehicle driving direction is judged, and the assist curve stability gradient limit value is calculated in combination with the road surface jitter count value in step 7; Step 9, according to the assist curve stability gradient limit value in step 8, the assist curve subjected to the safety boundary limiting at the current vehicle speed is subjected to gradient limiting, and the safety assist curve and the stable assist curve are obtained; Step 10, the additional torque request value in step 1 is superimposed with the torsion bar torque signal to obtain the total requested torsion bar torque; Step 11, according to the absolute value of the total requested torsion bar torque in step 10, the safety assist curve and the stable assist curve in step 9 are respectively subjected to linear interpolation calculation, and are multiplied by the direction symbol of the total torsion bar torque value to obtain the safety assist torque and the stable assist torque; Step 12, according to the rack end protection activation state, the difference between the safety assist torque and the stable assist torque in step 11 is subjected to PT1 filtering, when the rack end protection is activated, the alternative filtering coefficient under the rack end protection activation state is adopted, otherwise the default filtering coefficient is adopted for PT1 filtering calculation, and the filtered output value is superimposed with the stable assist torque and subjected to limiting output to obtain the basic assist torque; Step 13, according to the road surface jitter count value in step 7, the dynamic torque compensation parameter and the motor rotor inertia compensation parameter are selected; Step 14, according to the rack end protection activation state in step 1, in combination with the dynamic assist torque compensation parameter in step 13, the high assist gradient area dynamic compensation torque calculation and the low assist gradient area dynamic compensation torque calculation are performed; Step 15, according to the high boost gradient zone dynamic compensation torque and low boost gradient zone dynamic compensation torque calculated in step 14, superimpose and do the maximum safety boundary limiting, get dynamic boost torque; Step 16, according to the motor rotor acceleration signal and the safe vehicle speed signal in step 1, combined with the motor rotor inertia compensation parameter in step 13, the motor rotor inertia compensation torque is calculated; Step 17, the basic boost torque in step 12, the dynamic boost torque in step 15 and the motor rotor inertia compensation torque in step 16 are added to get the EPS adaptive feedforward boost torque, and the boost curve redundancy calculation check result in step 6 is combined to output to the motor control module for torque execution.

2. The adaptive feedforward assist control method for electric power assisted steering of an automobile according to claim 1, characterized by, In the step 3, the following steps are implemented: Step 31, according to the maximum effective range value of the torsion bar torque sensor, the maximum target torsion bar torque of the hand feeling curve is limited to the maximum boundary; Step 32, the ordinate of the last value point of the hand feeling curve is set to the maximum effective range value of the torsion bar torque sensor; Step 33, the hand feeling curve is limited to non-negative gradient; Step 34, the target torsion bar torque is limited to the minimum safety boundary according to the following logic: When the vehicle speed is less than or equal to the calibrated safe limit speed V lim , the minimum safety target torsion bar torque limit of the hand feeling curve is not limited. When the vehicle speed is greater than the calibrated safety limit speed V lim , the rack force is not lower than the first-order safety rack force F safe1 The lowest safety target torsion bar torque limit is performed on the hand feeling curve in the interval. When the vehicle speed is greater than the calibrated safe limit speed V lim The rack force is higher than the first-order safe rack force F safe1 And lower than the second-order safe rack force F safe2 The first-order minimum safe target torsion bar torque T safe1 Amplitude limiting; When the vehicle speed is greater than the calibrated safe limit speed V lim The rack force is higher than the second-order safe rack force F safe2 The hand feeling curve within the interval is subjected to the second-order lowest safe target torsion rod torque T safe2 Amplitude limiting.

3. The adaptive feedforward assist control method for electric power assisted steering of an automobile according to claim 1, characterized by, In the step 4, the calculation formula of each value point of the boost curve at the current vehicle speed is as follows: where: FRACK Xaxlei is the value of the rack force in N for the hand feel curve sampling point. DesTBT Yaxlei DesTBT is the longitudinal coordinate of the sampling point of the hand feeling curve, the value of the target torsion bar torque, unit Nm; TBT Xaxlei TbT is the value of the lateral rod torque for the sampling point of the force curve, in Nm. MOT Yaxlei MOT is the value of the motor assist torque in Nm for the longitudinal coordinate of the assist curve sampling point. X tbt2rack Mechanical conversion factor from motor torque to rack force for EPS; X mot2rack Mechanical conversion factor from torsion bar torque to rack force for EPS.

4. The adaptive feedforward assist control method for electric power assisted steering of an automobile according to claim 1, characterized by, In step 5, the lower safety boundary amplitude of the boost curve at the current vehicle speed is 0, and the upper boundary amplitude is the maximum motor output torque that can be provided.

5. The adaptive feedforward assist control method for electric power assisted steering of an automobile according to claim 1, wherein In step 7, the method for calculating the road surface jitter count value through the rotor acceleration signal is as follows: The road surface jitter counter enters the initialization state, and the road surface jitter timing value and the road surface jitter count value are set to 0; When the motor rotor acceleration is detected to change in a certain direction beyond the road surface jitter calibration gradient value, the rotor acceleration state and direction are recorded, and the road surface jitter timing is started; When the motor rotor acceleration is detected to change in the opposite direction of the last recorded rotor acceleration state beyond the road surface jitter calibration gradient value, the rotor acceleration state and direction are recorded, and the road surface jitter count value is increased by 1, and the road surface jitter timing value is set to 0; When the road surface jitter timing value exceeds the maximum interval time of road surface jitter detection, the road surface jitter counter returns to the initialization state, and the road surface jitter timing value and the road surface jitter count value are set to 0.

6. The adaptive feedforward assist control method for electric power assisted steering of an automobile according to claim 1, wherein In step 8, the calculation method of the boost curve stable gradient limit value is as follows: When the vehicle travel direction signal is forward travel, and the road surface roughness count value Count cr is less than the set threshold value Count lim : the boost curve stable gradient limit value is equal to the boost curve stable gradient limit value Grad spd at the current vehicle speed when the vehicle is traveling forward. When the vehicle travel direction signal is forward travel, and the road surface roughness count value Count cr is greater than or equal to a set threshold value Count lim : the assist curve stable gradient limit value is equal to the smaller of the assist curve stable gradient limit value Grad spd at the current vehicle speed when the vehicle is traveling forward and the current vehicle speed when the road surface roughness is too large cr . When the vehicle travel direction signal is reverse travel, and the road surface roughness count value Count cr is less than the set threshold value Count lim : the boost curve stable gradient limit value is equal to the boost curve stable gradient limit value Grad revspd at the current vehicle speed when the vehicle is traveling in reverse. When the vehicle travel direction signal is reverse travel, and the road surface roughness count value Count cr is greater than or equal to a set threshold value Count lim : the assist curve stable gradient limit value is equal to the smaller of the assist curve stable gradient limit value Grad revspd at the current vehicle speed when the vehicle is traveling in reverse and the road surface roughness is too large cr .

7. The adaptive feedforward assist control method for electric power assisted steering of an automobile according to claim 1, wherein In step 9, according to the boost curve safety gradient calibration limit value and the boost curve stable gradient limit value, the boost curve after being limited by the safety boundary is sequentially gradient-limited at the current vehicle speed to obtain the safety boost curve and the stable boost curve, and the safety gradient calibration limit value is greater than the boost curve stable gradient limit value.

8. The adaptive feedforward assist control method for electric power assisted steering of an automobile according to claim 1, wherein In step 10, the specific calculation method of the total requested torsion bar torque is as follows: according to the maximum effective range value of the torsion bar torque sensor, the maximum boundary of the torsion bar torque is limited; the limited torsion bar torque and the additional torque request value are superimposed to obtain the total requested torsion bar torque.

9. The adaptive feedforward assist control method for electric power assisted steering of an automobile according to claim 1, wherein In step 11, the linear interpolation method of the stable boost curve is as follows: If the absolute value of the total request torsion bar torque is less than or equal to the abscissa of the maximum value point of the abscissa of the stable boost curve, the absolute value of the total request torsion bar torque is used to perform internal linear interpolation calculation on the stable boost curve to obtain the stable boost torque; If the absolute value of the total request torsion bar torque is greater than the abscissa of the maximum value point of the abscissa of the stable boost curve, the absolute value of the total request torsion bar torque is used to perform external linear interpolation calculation on the stable boost curve to obtain the stable boost torque, and the ordinate of the maximum value point of the ordinate of the safety boost curve is used to limit the maximum boundary of the stable boost torque.

10. The adaptive feedforward assist control method for electric power assisted steering of an automobile according to claim 1, characterized by, In step 12, the specific calculation formula of the basic boost torque is: M pt1 = (M safe - M steady - M pt1last )* X pt1 + M pt1last M basic = M steady + M pt1 Wherein: X pt1 is the PT1 filter coefficient; M safe Safety assist torque, in Nm; M steady Stable assist torque, unit Nm; M pt1last PT1 is the output value of the PT1 filter value in the last operation period, the initial value of which is 0, in units of Nm; M pt1 PT1 is the filter value in Nm; M basic Base assist torque, in Nm.

11. The adaptive feedforward assist control method for electric power assisted steering of an automobile according to claim 1, characterized by, In step 13, when the road surface jitter count value exceeds the set value, the dynamic boost torque compensation substitute parameter is used; when the road surface jitter count value does not exceed the set value, the safety vehicle speed and the motor rotor speed linear interpolation calculation is used to obtain the low boost gradient zone and the high boost gradient zone default parameter; When the road surface jitter count value exceeds the set value, the safety vehicle speed linear interpolation calculation is used to obtain the motor rotor inertia compensation substitute parameter; when the road surface jitter count value does not exceed the set value, the safety vehicle speed linear interpolation calculation is used to obtain the motor rotor inertia compensation default parameter.

12. The adaptive feedforward assist control method for electric power assisted steering of an automobile according to claim 1, wherein In step 14, the calculation methods of the high boost gradient zone dynamic compensation torque and the low boost gradient zone dynamic compensation torque are as follows: When the rack end protection trigger is activated and the road surface jitter count value exceeds the set value: the high boost gradient zone compensation torque is 0, and the low boost gradient zone compensation torque is calculated according to the following formula: M Loffsst = X Chattersub T grad Wherein: M LOffset Mcomp, low Compensating torque for low-assist gradient region, unit Nm; X Chattersub is a dynamic assist torque compensation substitute parameter; T grad for the torsion bar torque change gradient, in Nm / ms; When the rack end protection trigger is activated and the road surface jitter count value does not exceed the set value: the high boost gradient zone compensation torque is calculated according to the following formula: M HOffset = X Hintpol T grad X AssistGrad Wherein: M HOffset Mcompensate, high assist gradient zone compensation torque, unit Nm; X Hintpol a high-assist gradient region parameter linearly interpolated from the safe vehicle speed and the motor rotor speed; T grad for the torsion bar torque change gradient, in Nm / ms; X AssistGrad For the assist force curve gradient, units Nm / Nm; The low boost gradient zone compensation torque is calculated according to the following formula: M Loffset = X Endstopsub T grad Wherein: M LOffset Mcomp, low Compensating torque for low assist gradient region, unit Nm; X Endstopsub Linear interpolation of rack end protection alternative parameters for safe vehicle speed T grad for the torsion bar torque change gradient, in Nm / ms; When the rack end protection is not activated and the road surface jitter count value exceeds the set value: the high boost gradient zone compensation torque is calculated according to the following formula: M Hoffsst = X Chattersub T grad X AssistGrad Wherein: M HOffset Mcompensate, high assist gradient zone compensation torque, unit Nm; X Chattersub is a dynamic assist torque compensation substitute parameter; T grad for the torsion bar torque change gradient, in Nm / ms; X AssistGrad For the assist curve gradient, units Nm / Nm; The low boost gradient zone compensation torque is calculated according to the following formula: M LOffset = X Chattersub T grad Wherein: M LOffset Mcomp, low Compensating torque for low assist gradient region, unit Nm; X Chattersub is a dynamic assist torque compensation substitute parameter; T grad for the torsion bar torque change gradient, in Nm / ms; When the rack end protection is not activated and the road surface jitter count value does not exceed the set value: the high boost gradient zone compensation torque is calculated according to the following formula: M HOffset = X Hintpol T grad X AssistGrad Wherein: M HOffset Mcompensate, high assist gradient zone compensation torque, unit Nm; X Hintpol High boost gradient region parameters linearly interpolated for safe vehicle speed and motor rotor speed; T grad for the torsion bar torque change gradient, in Nm / ms; X AssistGrad For the assist curve gradient, units Nm / Nm; The low boost gradient zone compensation torque is calculated according to the following formula: M LOffset = X Lintpol T grad Wherein: M LOffset Mcompensate, low assist gradient zone, unit Nm; X Lintpol Low-assist gradient zone parameters linearly interpolated for the safe vehicle speed and motor rotor speed T grad For torsion bar torque change gradient, unit Nm / ms.

13. The adaptive feedforward assist control method for electric power assisted steering of an automobile according to claim 1, wherein In step 15, the calculation formula of the dynamic boost torque is as follows: M dynamic = M HOffset + M LOffset Wherein: M dynamic M is the dynamic assistance torque, in Nm; M HOffset Mcompensate, high assist gradient zone compensation torque, unit Nm; M LOffset Mcompensate, low assist gradient zone, unit Nm; On dynamic assist torque M dynamic Output after maximum safety margin limiting 14. The adaptive feedforward assist control method for electric power assisted steering of an automobile according to claim 1, wherein, In step 16, the calculation formula of the dynamic boost torque is as follows: M InertiaComp = A RotAcc X RotFactor Wherein: M InertiaComp Motor torque for motor rotor inertia compensation, unit Nm; A RotAcc for the motor rotor acceleration, units 1 / S 2 ; X RotFactor Jm 0.0001 for motor rotor inertia compensation parameter, unit kg*m 2 .

15. The adaptive feedforward assist control method for electric power assisted steering of an automobile according to claim 1, wherein, In step 17, the calculation formula of the EPS adaptive front feed boost torque executed by the steering boost motor is as follows: Wherein: M out EPS adaptive feed-forward assist torque for steering assist motor, unit Nm; M basic Mbase is the base torque, in Nm; M dynamic M is the dynamic assistance torque, in Nm; M InertiaComp Motor torque for motor rotor inertia compensation, unit Nm; Err is the fault code output by the boost curve safety check module.

Images