CN117818753A

CN117818753A - Terminal dynamic self-learning method and system for steering by wire

Info

Publication number: CN117818753A
Application number: CN202311756842.5A
Authority: CN
Inventors: 闵令坤; 张丹; 黄刚; 武建君; 李寅林; 翟继盾; 杨志成; 鞠皋林
Original assignee: Bosch Huayu Steering Systems Co Ltd
Current assignee: Bosch Huayu Steering Systems Co Ltd
Priority date: 2023-12-20
Filing date: 2023-12-20
Publication date: 2024-04-05

Abstract

The invention relates to the technical field of steering systems, in particular to a method and a system for tail end dynamic self-learning of steering by wire control. A method for end-of-line dynamic self-learning for steer-by-wire execution, characterized by: when the vehicle is started and the steering wheel reaches the tail end leftwards or rightwards, recording the current TAS value as a left limit value and a right limit value; when the steering wheel returns to the middle position from the left end and the right end, storing the TAS value recorded currently into a data storage function module; the terminal self-learning module reads the numerical value and compensates the numerical value to the current left and right terminal use values, calculates the angles of the left and right terminals according to the current median value, and uses the calculated angles for each functional module; and if the left end stroke and the right end stroke of the rack exceed the default stroke values, the left end and the right end are safely reset to the initial default values. Compared with the prior art, the method and the system for terminal dynamic self-learning for steering by wire control are provided, so that the position of the steering angle sensor is matched with the actual steering wheel steering angle of the vehicle.

Description

Terminal dynamic self-learning method and system for steering by wire

Technical Field

The invention relates to the technical field of steering systems, in particular to a method and a system for tail end dynamic self-learning of steering by wire control.

Background

With the development of the automobile industry and the improvement of the social requirements on automobile safety, energy conservation and environmental protection, more and more passenger cars are provided with an electric power steering system and an automobile body electronic stabilizing system; the driving auxiliary system related to the driving auxiliary system and the electric control system in the field of chassis are required to accurately turn the angle signals, and the derived functions (soft dead center protection, active centering, automatic parking and the like) are more required to identify the limit position of the turning angle, and are actively limited and accurately controlled; in particular, current steer-by-wire systems, lack of an intermediate shaft relative to conventional steering systems, and precise control of the steering wheel end locking function of the upper steering depends on the angle of the lower steering rack end.

In addition, the right and left maximum limit angle values of the steering wheel of the vehicle are necessary for EPS (electric power steering) to perform various functional controls such as rack end protection, motor lock protection control, and the like.

However, the mechanical manufacturing of the steering system, the assembly of the whole vehicle, the extreme wear loosening and the four-wheel positioning operation all cause deviation, asymmetry and variation of the left and right extreme angles of the steering wheel of the vehicle and the theoretical design value, and have adverse effects on functions and performances related to the extreme angle positions of the steering wheel in EPS control. There is a need to improve the accuracy of the steering wheel limit angle values.

The maximum rotation angle calibration of the existing automobile electric power steering system is judged through the stroke of a gear and a rack. During an offline test (EOL), the left limit angle is determined by rotating the steering wheel leftwards, and the maximum left steering angle is calibrated to be the left limit angle; the right limit angle is determined by turning the steering wheel to the right, and the maximum angle of the right turn is calibrated as the right limit angle. In the existing maximum rotation angle calibration method, recalibration is required for different gear racks. Meanwhile, when the steering structure is maintained and parts are replaced, the calibrated maximum rotation angle also needs to be finely adjusted again due to the difference between the parts.

In the prior art, a manual calibration mode is generally adopted for calibrating the maximum rotation angle of the steering system. With this approach, the first effort costs are high; secondly, calibration is required to be carried out through a certain flow, and the production procedure is complicated; thirdly, the related components such as the gear rack and the like are replaced, and the recalibration and maintenance cost are also required; in addition, the maximum rotation angle calibration method for the existing automobile electric power steering system has higher precision requirements on parts, and is easy to cause insufficient ductility of calibration data.

Disclosure of Invention

The invention provides a method and a system for tail end dynamic self-learning of steering by wire control to overcome the defects of the prior art, so as to ensure that the position of a steering angle sensor is matched with the actual steering wheel angle of a vehicle.

To achieve the above object, a method for end-of-line dynamic self-learning for steer-by-wire is designed, which is characterized in that: the specific method comprises the following steps:

(1) The vehicle is started, and in the running process: the speed of the vehicle is more than 3km/h;

(2) When the steering wheel reaches the tail end leftwards or rightwards, and when the TAS angle received by the current steering wheel corner acquisition module corresponding to the left tail end and the right tail end exceeds a default left limit TAS value and a default right limit TAS value, the learning conditions of the left tail end and the right tail end are met, and the current TAS value is recorded as a left limit value and a right limit value;

(3) When the steering wheel returns to the middle position from the left end and the right end, storing the TAS value recorded currently into a data storage function module;

(4) The terminal self-learning module reads the numerical value and compensates the numerical value to the current left and right terminal use values, then calculates the angles of the left and right terminals according to the current median value, and provides the calculated angles for each functional module;

(5) And if the left end stroke and the right end stroke of the rack exceed the default stroke values, the left end and the right end are safely reset to the initial default values.

In the step (4), the self-learning scene of the tail end from the tail end of the learning module to the right is as follows:

1) Right learning scenario one: the middle position learning angle is the right limit angle of the learned tail end, the right limit self-learning angle of the current tail end is the same as the right limit self-learning angle of the current tail end, and the right limit self-learning angle is on one sawtooth edge;

2) Right learning scenario two: the middle position learning angle is the right limit angle of the learned tail end, and the right limit self-learning angle of the current tail end; only the right extreme self-learning angle of the current end is simultaneously on the other sawtooth edge; other angles are on one serrated edge;

3) Right learning scenario three: the middle position learning angle is the right limit angle of the learned tail end, and the right limit self-learning angle of the current tail end; the right extreme self-learning angle of the current end and the right extreme angle of the learned end are simultaneously on the other sawtooth edge; other angles are on one serrated edge;

4) Right learning scenario four: the middle position learning angle is the right limit angle of the learned tail end, and the right limit self-learning angle of the current tail end; the middle position learning angle, one end angle of the current end right extreme self-learning angle and one end angle of the learned end right extreme angle are simultaneously arranged on one sawtooth edge; the current end right limit is from the learned angle and other angles and one end angle of the learned end right limit angle is on the other zigzag edge at the same time.

When the scene is learned to the right, the judgment logic is that

fESStrAngleRightIncreaseflag＝[(wEsStrAngleRightLearnDy-wSAStrAngleOffsetDyNVR<(MAX_TASANGLE/2))；

And (wEsStrAngleRightLearnDy-wSAStrAngleOffsetDyNVR > 0);

and (wConvTASAnge-wSAStrAngleOffsetDyNVR < (-MAX_TASANGLE/2));

wherein wststrengagle offsetdynvr is the median of the rack; wEsStrAngleRightNVR is the right end value of the rack stored in EEProm; wEsStrAngleRightLearnDy is the learned right end value of the rack; wconvtasange is the angle value of the current TAS; fESStrAngle RightIncreateplag is an opening condition learned to the right; max_tsangle is the TAS characteristic maximum, 2016.

When learning scene two to the right, the judgment logic is that

And (wEsStrAngleRightLearnDy-wSAStrAngleOffsetDyNVR < (MAX_TASANGLE/2));

and (wconvtasagle-wsasastrstrangagle offsetdynvr > 0);

and (wConvTASAnge-wSAStrAngleOffsetDyNVR < (MAX_TASANGLE/2));

and (wconvtasagle > wrestraglerightnvr) ];

wherein wststrengagle offsetdynvr is the median of the rack; wEsStrAngleRightNVR is the right end value of the rack stored in EEProm; wEsStrAngleRightLearnDy is the learned right end value of the rack;

wconvtasange is the angle value of the current TAS; fESStrAngle RightIncreateplag is an opening condition learned to the right; max_tsangle is the TAS characteristic maximum, 2016.

When learning scene three to the right, the judgment logic is that

fESStrAngleRightIncreaseflag＝[(wEsStrAngleRightLearnDy-wSAStrAngleOffsetDyNVR<(-MAX_TASANGLE/2))；

And (wConvTASAnge-wSAStrAngleOffsetDyNVR < (-MAX_TASANGLE/2));

and (wconvtasagle > wussstralglichtlearndy) ];

In the step (4), the self-learning scene of the tail end from the tail end of the learning module to the left is as follows:

1) Left learning scenario one: the middle position learning angle is the left limit angle of the learned tail end, the left limit self-learning angle of the current tail end is the same as the right limit self-learning angle of the current tail end, and the current tail end is on one sawtooth edge;

2) Learning scene two to the left: the middle position learning angle is the left limit angle of the learned tail end, and the left limit self-learning angle of the current tail end; only the current end left limit self-learning angle is on the other sawtooth edge at the same time; other angles are on one serrated edge;

3) Learning scene three to the left: the middle position learning angle is the left limit angle of the learned tail end, and the left limit self-learning angle of the current tail end; the left extreme self-learning angle of the current tail end and the left extreme angle of the learned tail end are simultaneously on the other sawtooth edge; other angles are on one serrated edge;

4) Learning scene four to the left: the middle position learning angle is the left limit angle of the learned tail end, and the left limit self-learning angle of the current tail end; the middle position learning angle, one end angle of the current end left limit self-learning angle and one end angle of the learned end left limit angle are simultaneously arranged on one sawtooth edge; the current end left limit self-learns the angle and other angles and the one end angle of the learned end left limit angle are on the other zigzag edge at the same time.

When the scene is learned leftwards, the judgment logic is that

fESStrAngleLeftIncreaseflag＝(wEsStrAngleLeftLearnDy-wSAStrAngleOffsetDyNVR>(MAX_TASANGLE/2)；

And (wEsStrAngleLeftLearnDy-wSAStrAngleOffsetDyNVR < 0);

and (wConvTASAnge-wSAStrAngleOffsetDyNVR > (MAX_TASANG/2);

wherein wststrengagle offsetdynvr is the median of the rack; wEsStrAngle LeftNVR is stored at the left end value of the EEProm rack; wEsStrAngleLeftLearnDy is the left end value of the learned rack; wconvtasange is the angle value of the current TAS; fESStrAngle LeftIncreateslag is a leftward learned starting condition; max_tsangle is the TAS characteristic maximum, 2016.

When learning scene two to the left, the judgment logic is that

fESStrAngleLeftIncreaseflag＝[wEsStrAngleLeftLearnDy-wSteeringAngleOffsetDyNVR>(-MAX_TASANGLE/2)；

And wEsStrAngleLeftLearnDy-wSteeringAngleOffsetDyNVR <0;

and wConvTASAnge-wSteringAngleOffsetDyNVR > (-MAX_TASANG/2);

and wConvTASAnge-wSteeringAngleOffsetDyNVR <0;

and (wconvtasagle < wconstregleleflearndy);

When learning scene three to the left, the judgment logic is that

fESStrAngleLeftIncreaseflag＝(wEsStrAngleLeftLearnDy-wSAStrAngleOffsetDyNVR>(MAX_TASANGLE/2))；

And (wConvTASAnge-wSAStrAngleOffsetDyNVR > (MAX_TASANGLE/2));

and (wconvtasagle < wconstregleleflearndy);

A system employing the terminal dynamic self-learning method of any one of the above claims, characterized in that: the system comprises:

upper steering road feel module: the system comprises a steering wheel angle acquisition module, a lower steering execution module and a lower steering execution module, wherein the steering wheel angle acquisition module is mainly used for calculating the angle of an upper steering road feel module, and is transmitted to the lower steering execution module through a private CAN (controller area network), and the angle storage activation condition is used for terminal learning in a terminal dynamic self-learning function;

the whole vehicle signal processing system comprises: the system comprises a vehicle speed acquisition module, a lower steering execution module and a lower steering execution module, wherein the vehicle speed acquisition module is mainly used for calculating a vehicle speed signal and transmitting the vehicle speed signal to the lower steering execution module through a CAN (controller area network), and is used for one of terminal angle self-learning activation conditions in a terminal dynamic self-learning function;

the lower steering execution module: the system comprises a signal processing function module, a tail end self-learning function module, a TAS angle acquisition module, a fault processing module, a data storage function module and other wire control angle related function modules; the signal processing functional module is used for storing terminal angle self-learning activation condition bars and learning values, and related signals comprise a vehicle speed signal of a vehicle CAN signal and a steering wheel corner signal of a private CAN.

Compared with the prior art, the invention provides a method and a system for tail end dynamic self-learning of steering by wire control, so as to ensure that the position of a steering angle sensor is matched with the actual steering wheel angle of a vehicle.

Drawings

FIG. 1 is a schematic diagram of a dual redundant power-assisted steering system.

FIG. 2 is a schematic diagram of the end-point dynamic self-learning system according to the present invention.

FIG. 3 is a dynamic end-to-end self-learning flow chart of the power-assisted steering system of the present invention.

FIG. 4 is a schematic diagram of a dynamic end self-learning "right" learning scenario for the power-assisted steering system of the present invention.

Fig. 5 is an enlarged view of the scene in fig. 4.

Fig. 6 is an enlarged schematic view of the second scenario in fig. 4.

Fig. 7 is an enlarged schematic view of the scene three in fig. 4.

Fig. 8 is a schematic diagram of the learning process of the "right" learning of the present invention.

FIG. 9 is a schematic diagram of a dynamic end self-learning "left" learning scenario for a power-assisted steering system of the present invention.

Fig. 10 is an enlarged view of the scene of fig. 9.

Fig. 11 is an enlarged schematic view of the second scene in fig. 9.

Fig. 12 is an enlarged schematic view of the scene three in fig. 9.

Fig. 13 is a schematic diagram of the learning process of the "left" learning of the present invention.

Fig. 14 is a schematic diagram of the end angle learning value compensation.

Fig. 15 is a schematic diagram of an end safety reset scenario.

Detailed Description

The invention is further described below with reference to the accompanying drawings.

As shown in fig. 1, a schematic structural diagram of a dual-redundancy drive-by-wire power steering system provided by an embodiment of the present invention is provided, the system mainly comprises an upper steering feel system and a lower steering executing system, the upper steering feel module mainly comprises a steering wheel 1, an upper steering TAS torque sensor 2, an upper steering column assembly 3, an upper steering feel dual-redundancy ECU10, and an upper steering turbine assembly 11; the lower steering execution module mainly comprises a pull rod assembly 4, a lower steering turbine assembly 5, a rack assembly 6, a lower steering execution double-redundancy ECU7, a motor rotor position sensor 8 and a lower steering TAS (total internal rotation) angle sensor 9.

The upper steering road feel module and the lower steering execution module are in communication interaction with each other through a private CAN13, and the upper system module and the lower system module are in communication interaction with the whole vehicle through a CAN bus 12. The steering wheel 1 is used for driving by a driver, and the input end of the upper steering column assembly 3 is connected with the steering wheel 1; the upper steering TAS torque sensor 2 detects the steering torque and the angle of a driver, the upper steering road feel dual-redundancy ECU10 is used for acquiring and processing the steering torque, and transmitting the steering torque to the lower steering execution dual-redundancy ECU7 through the private CAN13 for judging conditions in the tail end self-learning process; the road sensing motor is arranged on the upper steering column assembly 3 and provides road sensing moment for a driver through a speed reducing mechanism.

The lower steering power-assisted motor is arranged on the lower steering turbine assembly 5, provides power-assisted torque for a driver through a speed reducing mechanism, and acts on the rack assembly 6 and the pull rod assembly 4 through the lower steering turbine assembly 5 to realize steering power assistance; the lower steering TAS steering angle sensor 9 is configured to detect an angle value of the steering shaft to determine a lower steering angle median value; the motor rotor position sensor 8 is used for detecting the rotor position rotation angle of the lower steering power-assisted motor, and feeds back the detected rotation angle of the lower steering power-assisted motor to the lower steering execution dual redundancy ECU7 for calculating the lower steering angle.

Fig. 2 is a schematic diagram of a terminal dynamic self-learning functional system according to the present invention, which includes three functional systems:

1. upper steering road feel module: the system comprises a steering wheel corner acquisition module, mainly a function of calculating the angle of an upper steering road feel module, and the angle is transmitted to a lower steering execution module through a private CAN, and the angle is used for storing and activating conditions for end learning in an end dynamic self-learning function.

2. The whole vehicle signal processing system comprises: the system comprises a vehicle speed acquisition module which is mainly used for calculating a vehicle speed signal and transmitting the vehicle speed signal to a lower steering execution module through a CAN, wherein the vehicle speed acquisition module is used for one of terminal angle self-learning activation conditions in a terminal dynamic self-learning function.

3. The lower steering execution module: the system comprises a signal processing function module, a tail end self-learning function module, a TAS angle acquisition module, a fault processing module, a data storage function module (EEPROM) and other wire control angle related function modules; the signal processing functional module is used for storing terminal angle self-learning activation condition bars and learning values, and related signals comprise a vehicle speed signal of a vehicle CAN signal and a steering wheel corner signal of a private CAN.

The TAS angle acquisition module is used for providing an extreme learning angle of the tail end; the end self-learning function stores an end learning value with a self-learning Xi Moduan (left/right) angle exceeding a default limit angle when the steering wheel is turned to a target angle, and provides the learning value to the associated function module. The data storage functional module (EEPROM) is used as a function of storing the end (left/right) self-learning value, can still keep the learning value after the EPS is powered on and powered off, and meets the non-volatility of the end self-learning value. The fault handling module stores faults and performs related safety measures when unexpected faults are encountered during learning.

As shown in fig. 3, a dynamic end self-learning flow chart of the power-assisted steering system of the present invention includes the following steps:

(6) The vehicle is started, and in the running process: the speed of the vehicle is more than 3km/h;

(7) When the steering wheel reaches the tail end in the left direction or in the right direction, when the TAS angle received by the current steering wheel corner acquisition module corresponding to the left and right tail ends exceeds the default left and right limit TAS values, the left and right tail end learning conditions are met, and the current TAS values are recorded as the left and right limit values;

(8) When the steering wheel returns to the middle position from the left end and the right end, storing the TAS value recorded currently into a data storage functional module (EEPROM);

(9) The terminal self-learning module reads the numerical value and compensates the numerical value to the current left and right terminal use values, then calculates the angles of the left and right terminals according to the current median value, and provides the calculated angles for each functional module;

(10) And if the left end stroke and the right end stroke of the rack exceed the default stroke values, the left end and the right end are safely reset to the initial default values.

Due to the nature of the TAS angle sensor itself, cyclic saw tooth angle data is provided. In use, four types of scenes are typically encountered, as shown in fig. 4, the right limit, i.e., the "right" learning scene is as follows:

as shown in fig. 5, "right" learning scenario one: the median learn angle, the learned end right limit angle, the current end right limit self-learn angle, and on one zigzag edge.

Thus, the first and second substrates are bonded together,

and (wEsStrAngleRightLearnDy-wSAStrAngleOffsetDyNVR > 0);

and (wConvTASAnge-wSAStrAngleOffsetDyNVR < (-MAX_TASANGLE/2)).

As shown in fig. 6, "right" learning scenario two: the middle position learning angle is the right limit angle of the learned tail end, and the right limit self-learning angle of the current tail end; only the right extreme self-learning angle of the current end is simultaneously on the other sawtooth edge; other angles are on one serrated edge.

Thus, the first and second substrates are bonded together,

and (wEsStrAngleRightLearnDy-wSAStrAngleOffsetDyNVR < (MAX_TASANGLE/2));

and (wconvtasagle-wsasastrstrangagle offsetdynvr > 0);

and (wConvTASAnge-wSAStrAngleOffsetDyNVR < (MAX_TASANGLE/2));

and (wconvtasagle > wrestraglerightnvr) ].

As shown in fig. 7, "right" learning scenario three: the middle position learning angle is the right limit angle of the learned tail end, and the right limit self-learning angle of the current tail end; the right extreme self-learning angle of the current end and the right extreme angle of the learned end are simultaneously on the other sawtooth edge; other angles are on one serrated edge.

Thus, the first and second substrates are bonded together,

and (wConvTASAnge-wSAStrAngleOffsetDyNVR < (-MAX_TASANGLE/2));

and (wconvtasagle > wusstrestrighleglearndy) ].

"Right" learning scenario four: the middle position learning angle is the right limit angle of the learned tail end, and the right limit self-learning angle of the current tail end; the middle position learning angle, one end angle of the current end right extreme self-learning angle and one end angle of the learned end right extreme angle are simultaneously arranged on one sawtooth edge; the current end right limit is from the learned angle and other angles and one end angle of the learned end right limit angle is on the other zigzag edge at the same time.

During running, the vehicle speed is more than 3Km/h, when the steering wheel is driven near the right end and the learning opening condition fESStrAngle RightIncreatesflag is more than 1, the current TAS value wConvTASAnge is the newly learned end angle value wEsStrAngle RightLearnDy=wConvTASAnge to be learned and recorded.

When the steering wheel is returned to the vicinity of the center position, the learned data wEsStrAngleRightLearnDy of the learning record is stored in a data storage functional module (EEPROM), and the value read from the data storage functional module (EEPROM) is wEsStrAngleRightNVR. The entire rightward learning process is shown in fig. 8.

As shown in fig. 9, the learning scenario for the right limit, i.e. "left", is as follows:

as shown in fig. 10, "left" learning scenario one: the median learn angle, the learned end left limit angle, the current end left limit self-learn angle, and on one zigzag edge.

Thus, the first and second substrates are bonded together,

and (wEsStrAngleLeftLearnDy-wSAStrAngleOffsetDyNVR < 0);

and (wConvTASAnge-wSAStrAngleOffsetDyNVR > (MAX_TASANG/2).

As shown in fig. 11, "left" learning scenario two: the middle position learning angle is the left limit angle of the learned tail end, and the left limit self-learning angle of the current tail end; only the current end left limit self-learning angle is on the other sawtooth edge at the same time; other angles are on one serrated edge.

Thus, the first and second substrates are bonded together,

and wEsStrAngleLeftLearnDy-wSteeringAngleOffsetDyNVR <0;

and wConvTASAnge-wSteringAngleOffsetDyNVR > (-MAX_TASANG/2);

and wConvTASAnge-wSteeringAngleOffsetDyNVR <0;

and (wconvtasagle < wconstregleleflearndy).

As shown in fig. 12, "left" learning scenario three: the middle position learning angle is the left limit angle of the learned tail end, and the left limit self-learning angle of the current tail end; the left extreme self-learning angle of the current tail end and the left extreme angle of the learned tail end are simultaneously on the other sawtooth edge; other angles are on one serrated edge.

Thus, the first and second substrates are bonded together,

and (wConvTASAnge-wSAStrAngleOffsetDyNVR > (MAX_TASANGLE/2));

and (wconvtasagle < wconstregleleflearndy).

Learning scene four "left": the middle position learning angle is the left limit angle of the learned tail end, and the left limit self-learning angle of the current tail end; the middle position learning angle, one end angle of the current end left limit self-learning angle and one end angle of the learned end left limit angle are simultaneously arranged on one sawtooth edge; the current end left limit self-learns the angle and other angles and the one end angle of the learned end left limit angle are on the other zigzag edge at the same time.

During running, the vehicle speed is more than 3Km/h, when the steering wheel is driven near the left end and the learning opening condition fESStrAngle LeftIncreateflag= 1, the current TAS value wConvTASAnge is the newly learned end angle wEsStrAngle LeftLearnDy = wConvTASAnge data record.

When the steering wheel is returned to the vicinity of the center position, the learned record value wEsStrAngleLeftLearnDy is stored in a data storage function module (EEPROM), and the read value is wEsStrAngleLeftNVR. The entire rightward learning process is shown in fig. 13.

As shown in fig. 14, according to the self-learning value compensation strategy of the last-bit dynamic self-learning of the power-assisted steering system, the steering wheel is restored to the dynamic left and right end values of the data storage functional module (EEPROM), the left and right end values currently used are updated in time after being restored, and in order to avoid the instantaneous change of the angle caused by the instantaneous updating of the end-bit values, the original current last-bit learning value is compensated by a low-pass filter to gradually compensate to the actual left and right end values of the steering wheel, so that the whole process of dynamic self-learning of the end angle is completed.

Thus, the compensation strategy is determined according to the following formula: d (D) _n ＝p*Δ+(1-p)*D _n-1 。

Wherein D is _n The current end angle compensation value; d (D) _n-1 The last end angle compensation value; delta is the difference between the compensation target end angle and the actual end angle; p is the compensation coefficient.

The dynamic tail end self-learning safe reset of the drive-by-wire power-assisted steering system is used for avoiding the problem that the support mechanism of the tail end rack is deviated due to assembly or abrasion, so that the tail end is learned leftwards and rightwards, and a special situation, namely, the tail end is learned towards one side continuously, and the numerical value is updated continuously; the other direction will never be learned and the value will remain unchanged, resulting in a continuously increasing learning stroke and a loss of relative accuracy of the end angle related function.

The current method is to reset the original position values of the left and right ends by adopting a safe resetting method after the stroke between the left and right ends after learning exceeds a preset stroke, and then to perform self-learning of the left and right ends. As shown in fig. 15, an end safety reset scenario is schematically shown:

1) Before the rack moves: s1 is the current end stroke.

2) After the rack moves: s2 is the current end stroke.

3) Rack movement S3: when S2> S1+DeltaS, the left and right S4 distances are reset to the tail ends by taking the middle position as the center; ΔS is generally set to 5 to 10mm.

Claims

1. A method for end-of-line dynamic self-learning for steer-by-wire execution, characterized by: the specific method comprises the following steps:

the vehicle is started, and in the running process: the speed of the vehicle is more than 3km/h;

when the steering wheel reaches the tail end leftwards or rightwards, and when the TAS angle received by the current steering wheel corner acquisition module corresponding to the left tail end and the right tail end exceeds a default left limit TAS value and a default right limit TAS value, the learning conditions of the left tail end and the right tail end are met, and the current TAS value is recorded as a left limit value and a right limit value;

when the steering wheel returns to the middle position from the left end and the right end, storing the TAS value recorded currently into a data storage function module;

the terminal self-learning module reads the numerical value and compensates the numerical value to the current left and right terminal use values, then calculates the angles of the left and right terminals according to the current median value, and provides the calculated angles for each functional module;

and if the left end stroke and the right end stroke of the rack exceed the default stroke values, the left end and the right end are safely reset to the initial default values.

2. A method for end-of-line dynamic self-learning of steer-by-wire as claimed in claim 1, wherein: in the step (4), the self-learning scene of the tail end from the tail end of the learning module to the right is as follows:

3. A method for end-of-line dynamic self-learning of steer-by-wire as claimed in claim 2, wherein: when the scene is learned to the right, the judgment logic is that

fESStrAngleRightIncreaseflag=[(wEsStrAngleRightLearnDy-wSAStrAngleOffsetDyNVR<(MAX_TASANGLE/2))；

And (wEsStrAngleRightLearnDy-wSAStrAngleOffsetDyNVR > 0);

and (wConvTASAnge-wSAStrAngleOffsetDyNVR < (-MAX_TASANGLE/2));

4. A method for end-of-line dynamic self-learning of steer-by-wire as claimed in claim 2, wherein: when learning scene two to the right, the judgment logic is that

And (wEsStrAngleRightLearnDy-wSAStrAngleOffsetDyNVR < (MAX_TASANGLE/2));

and (wconvtasagle-wsasastrstrangagle offsetdynvr > 0);

and (wConvTASAnge-wSAStrAngleOffsetDyNVR < (MAX_TASANGLE/2));

and (wconvtasagle > wrestraglerightnvr) ];

5. A method for end-of-line dynamic self-learning of steer-by-wire as claimed in claim 2, wherein: when learning scene three to the right, the judgment logic is that

fESStrAngleRightIncreaseflag=[(wEsStrAngleRightLearnDy-wSAStrAngleOffsetDyNVR<(-MAX_TASANGLE/2))；

And (wConvTASAnge-wSAStrAngleOffsetDyNVR < (-MAX_TASANGLE/2));

and (wconvtasagle > wussstralglichtlearndy) ];

6. A method for end-of-line dynamic self-learning of steer-by-wire as claimed in claim 1, wherein: in the step (4), the self-learning scene of the tail end from the tail end of the learning module to the left is as follows:

7. A method for end-of-line dynamic self-learning of steer-by-wire as claimed in claim 6, wherein: when the scene is learned leftwards, the judgment logic is that

fESStrAngleLeftIncreaseflag=(wEsStrAngleLeftLearnDy-wSAStrAngleOffsetDyNVR>(MAX_TASANGLE/2)；

And (wEsStrAngleLeftLearnDy-wSAStrAngleOffsetDyNVR < 0);

and (wConvTASAnge-wSAStrAngleOffsetDyNVR > (MAX_TASANG/2);

8. A method for end-of-line dynamic self-learning of steer-by-wire as claimed in claim 6, wherein: when learning scene two to the left, the judgment logic is that

fESStrAngleLeftIncreaseflag=[wEsStrAngleLeftLearnDy-wSteeringAngleOffsetDyNVR>(-MAX_TASANGLE/2)；

And wEsStrAngleLeftLearnDy-wSteeringAngleOffsetDyNVR <0;

and wConvTASAnge-wSteringAngleOffsetDyNVR > (-MAX_TASANG/2);

and wConvTASAnge-wSteeringAngleOffsetDyNVR <0;

and (wconvtasagle < wconstregleleflearndy);

9. A method for end-of-line dynamic self-learning of steer-by-wire as claimed in claim 6, wherein: when learning scene three to the left, the judgment logic is that

fESStrAngleLeftIncreaseflag=(wEsStrAngleLeftLearnDy-wSAStrAngleOffsetDyNVR>(MAX_TASANGLE/2))；

And (wConvTASAnge-wSAStrAngleOffsetDyNVR > (MAX_TASANGLE/2));

and (wconvtasagle < wconstregleleflearndy);

10. A system employing the terminal dynamic self-learning method of any one of the above claims, characterized in that: the system comprises: