CN117962920A - Zero-deflection angle online learning method, front-rear wheel steering angle learning method and system for automatic driving vehicle - Google Patents
Zero-deflection angle online learning method, front-rear wheel steering angle learning method and system for automatic driving vehicle Download PDFInfo
- Publication number
- CN117962920A CN117962920A CN202311369973.8A CN202311369973A CN117962920A CN 117962920 A CN117962920 A CN 117962920A CN 202311369973 A CN202311369973 A CN 202311369973A CN 117962920 A CN117962920 A CN 117962920A
- Authority
- CN
- China
- Prior art keywords
- real
- time
- angle
- zero
- vehicle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000012545 processing Methods 0.000 claims description 18
- 230000004044 response Effects 0.000 claims description 18
- 238000004364 calculation method Methods 0.000 claims description 12
- 238000004891 communication Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 abstract description 4
- 230000008859 change Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 238000005299 abrasion Methods 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Landscapes
- Steering Control In Accordance With Driving Conditions (AREA)
Abstract
The invention provides an automatic driving vehicle zero-deflection angle online learning method, a front-rear wheel steering angle learning method and a system. The zero-deflection angle online learning method of the automatic driving vehicle comprises the following steps: acquiring initial values of zero offset angles of front and rear axles of a vehicle; acquiring vehicle state data, reference track curvature, lateral position deviation and course angle deviation in real time; and responding to the magnitude relation between the real-time vehicle state data and the set value, performing online learning of the front axle zero deflection angle according to the real-time transverse position deviation to obtain the real-time front axle zero deflection angle, and performing online learning of the rear axle zero deflection angle according to the real-time course angle deviation to obtain the real-time rear axle zero deflection angle. The method realizes the online learning of the zero deflection angle of the front axle and the rear axle by utilizing the transverse position deviation and the course angle deviation in the control process of the automatic driving vehicle.
Description
Technical Field
The invention relates to the field of zero-deflection angle learning of vehicle steering, in particular to an automatic driving vehicle zero-deflection angle online learning method, a front-rear wheel steering angle learning method and a system.
Background
The four-wheel steering (four WHEEL STEERING,4 WS) vehicle can realize the same direction or reverse direction independent control of the front wheel and the rear wheel, thereby achieving the purposes of reducing the turning radius and improving the stability of the vehicle body. The four-wheel steering system is realized in mechanical, hydraulic, electric and other modes, and hydraulic steering is adopted for ore-carrying automatic driving vehicles, so that the zero deflection angle of the steering system can be changed along with the abrasion and aging of mechanical structures and the influence of the environment on the sensor.
The steering system abnormality can be judged by the deviation of the steering system when the vehicle is driven by someone, and the normal running of the vehicle is ensured by manual compensation. The automatic driving vehicle has no effective detection means for the change of the zero offset angle of the vehicle, the control precision can be influenced when the change is small, the safety accident can possibly occur when the change is large, and particularly, the zero offset angle of the front axle and the rear axle exists for the four-wheel steering vehicle, so that the real-time and accurate acquisition of the value can be more difficult.
Disclosure of Invention
In order to overcome the defects in the prior art, the invention aims to provide an automatic driving vehicle zero-deflection angle online learning method, a front-rear wheel steering angle learning method and a system.
In order to achieve the above object of the present invention, the present invention provides an online learning method for zero offset angle of an automatic driving vehicle, comprising the steps of:
Acquiring initial values of zero offset angles of front and rear axles of a vehicle;
acquiring vehicle state data, reference track curvature, lateral position deviation and course angle deviation in real time;
And responding to the magnitude relation between the real-time vehicle state data and the set value, performing online learning of the front axle zero deflection angle according to the real-time transverse position deviation to obtain the real-time front axle zero deflection angle, and performing online learning of the rear axle zero deflection angle according to the real-time course angle deviation to obtain the real-time rear axle zero deflection angle.
The method utilizes the transverse position deviation and course angle deviation in the control process of the automatic driving vehicle to realize the on-line learning of the front and rear axle zero deflection angle of the four-wheel steering automatic driving vehicle.
The automatic driving vehicle zero-deflection angle online learning method comprises the following steps: and responding to the real-time vehicle speed, the real-time curvature and the magnitude relation between the real-time front and rear wheel actual rotation angles and the corresponding set values, carrying out online learning of the front axle zero deflection angle according to the real-time transverse position deviation, and carrying out online learning of the rear axle zero deflection angle according to the real-time course angle deviation.
The automatic driving vehicle zero-deflection angle online learning method comprises the following steps: the front axle zero deflection angle learning step is:
Responding to the real-time vehicle speed being larger than a set vehicle speed value, the real-time curvature absolute value being smaller than a set curvature value and the real-time absolute values of the actual rotation angles of the front and rear wheels being smaller than a set rotation angle value, and iteratively updating the front axle zero offset angle according to the real-time transverse position deviation;
In response to any one of the following conditions: the real-time vehicle speed is not greater than a set vehicle speed value, the real-time curvature absolute value is not less than a set curvature value, and the real-time absolute value of the actual rotation angle of the front wheel and the real-time absolute value of the actual rotation angle of the rear wheel are not less than a set rotation angle value; the front axle zero offset angle at the previous moment is taken as the front axle zero offset angle at the current moment.
The front axle zero deflection angle learned by the alternative proposal has high precision, high calculation speed and high real-time.
The automatic driving vehicle zero-deflection angle online learning method comprises the following steps: the zero offset angle learning step of the rear axle is as follows:
Defining a buffer queue Q with the length of N, storing the real-time course angle deviation into the buffer queue Q in response to the real-time vehicle speed being larger than a set vehicle speed value and the real-time curvature absolute value being smaller than the set absolute value, otherwise, emptying the buffer queue Q;
in response to the elements of the cache queue Q reaching the set number, calculating the average value and standard deviation of the cache queue Q;
and in response to the real-time vehicle speed being greater than the set vehicle speed value, the real-time curvature absolute value being smaller than the set absolute value and the real-time absolute value of the actual rotation angle of the front wheel and the rear wheel being smaller than the set rotation angle absolute value, iteratively updating the zero offset angle of the rear shaft according to the mean value and the standard deviation of the cache queue Q.
The zero offset angle of the rear axle learned by the alternative proposal has high precision, high calculation speed and high real-time.
The automatic driving vehicle zero-deflection angle online learning method comprises the following steps: and when the mean value and the standard deviation of the cache queue Q are calculated, recursively updating the mean value and the standard deviation of the cache queue Q at the current moment according to the mean value and the standard deviation of the cache queue Q at the previous moment.
The alternative scheme improves the calculation speed and reduces the resource consumption of the cpu.
The automatic driving vehicle zero-deflection angle online learning method comprises the following steps: and reporting the fault of the steering system in response to the absolute value of the real-time front axle zero offset angle and/or the rear axle zero offset angle exceeding a set threshold.
This alternative improves the safety of the autonomous vehicle.
The application also provides a front and rear wheel steering angle learning method of the automatic driving vehicle, the real-time front axle zero deflection angle and rear axle zero deflection angle are obtained according to the automatic driving vehicle zero deflection angle online learning method, and the steering control instruction obtained by the path tracking control module in real time is compensated according to the front axle zero deflection angle and the rear axle zero deflection angle, so that the compensated front and rear wheel steering angle is obtained.
The method compensates the path tracking control module instruction by the zero offset angle of the front shaft and the back shaft, and improves the control precision and the robustness of the control algorithm.
The application also provides an automatic driving vehicle zero-deflection angle online learning system, which comprises a data acquisition module, a processing module and a storage module, wherein the data acquisition module acquires vehicle state data, reference track curvature, transverse position deviation and heading angle deviation from a vehicle-mounted sensor and a path tracking control module and transmits the vehicle state data, the reference track curvature, the transverse position deviation and the heading angle deviation to the processing module, the processing module is in communication connection with the storage module, the storage module stores at least one executable instruction, and the executable instruction enables the processing module to execute operations corresponding to the automatic driving vehicle zero-deflection angle online learning method according to the data acquired from the data acquisition module, so that the real-time front axle zero-deflection angle and rear axle zero-deflection angle of the vehicle are obtained.
The system has all the advantages of an automatic driving vehicle zero-deflection angle online learning method.
The invention also provides a front and rear wheel steering angle learning system of the automatic driving vehicle, which comprises the automatic driving vehicle zero-deflection angle online learning system, wherein the processing module is connected with the path tracking control module, the path tracking control module acquires the front axle zero-deflection angle and the rear axle zero-deflection angle learned by the processing module, and compensates steering control instructions calculated by the path tracking control module based on the front axle zero-deflection angle and the rear axle zero-deflection angle to obtain the compensated front and rear wheel steering angles.
The system has all the advantages of the front and rear wheel steering angle learning method of the automatic driving vehicle.
The application also provides an automatic driving vehicle, which comprises the front and rear wheel steering angle learning system of the automatic driving vehicle and a vehicle steering control module, wherein the vehicle steering control module acquires the compensated front and rear wheel steering angle obtained by the path tracking control module and directly acts on the vehicle steering, or acquires the steering angles of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel of the vehicle obtained by the path tracking control module and acts on the vehicle steering.
The beneficial effects of the invention are as follows: the invention can accurately and quickly learn the zero deflection angle of the vehicle, has real-time performance, can automatically learn and update along with the change of the zero deflection angle of the vehicle, and ensures the control precision and the robustness of a control algorithm; meanwhile, the steering control instruction calculated by the path tracking control module is compensated according to the learned zero offset angle of the vehicle, so that the reliable and safe operation of the automatic driving vehicle can be ensured even if the zero offset angle is changed, the working efficiency is improved, and the maintenance cost is reduced.
Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.
Drawings
The foregoing and/or additional aspects and advantages of the invention will become apparent and may be better understood from the following description of embodiments taken in conjunction with the accompanying drawings in which:
FIG. 1 is a schematic flow chart of a first embodiment;
FIG. 2 is a schematic flow chart of a second embodiment;
FIG. 3 is a front-rear wheel steering angle reversal schematic diagram of a vehicle supporting the Ackerman steering model;
FIG. 4 is a schematic diagram of front-rear wheel steering angle co-ordinates of a vehicle supporting an Ackerman steering model;
fig. 5 is a functional block diagram of a fourth embodiment.
Detailed Description
Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative only and are not to be construed as limiting the invention.
In the description of the present invention, unless otherwise specified and defined, it should be noted that the terms "mounted," "connected," and "coupled" are to be construed broadly, and may be, for example, mechanical or electrical, or may be in communication with each other between two elements, directly or indirectly through intermediaries, as would be understood by those skilled in the art, in view of the specific meaning of the terms described above.
Example 1
As shown in FIG. 1, the invention provides an automatic driving vehicle zero-deflection angle online learning method, which is particularly suitable for four-wheel steering automatic driving vehicle zero-deflection angle online learning, and comprises the following steps:
When the automatic driving vehicle is electrified or the program is started, the initial value of the zero offset angle of the front axle and the rear axle is read from the file, and the initial value of the zero offset angle of the front axle is recorded as The initial value of the zero offset angle of the rear axle is recorded as/>
Vehicle state data, a reference trajectory curvature ρ, a lateral position deviation ε d, and a heading angle deviation ε θ are acquired in real time.
Specifically, the automatic driving state of the vehicle is obtained first, the subsequent calculation step is completed when the vehicle is in the automatic driving mode, and otherwise, the driving mode is waited for switching. When the vehicle is in an automatic driving state, vehicle state data including the longitudinal speed v of the vehicle, the front axle tire actual turning angle δ f and the rear axle tire actual turning angle δ r are acquired in the present embodiment. The reference track curvature rho, the transverse position deviation epsilon d and the course angle deviation epsilon θ are obtained from a path tracking control module, wherein the path tracking control module calculates the reference track in real time in the running process of the vehicle, the reference track parameters comprise the reference track curvature rho, the transverse position deviation epsilon d is the transverse position difference value between the vehicle locating point and the nearest point of the reference track, and the calculation method is only required by adopting a known technology and is not described in detail herein; the heading angle deviation epsilon θ is the difference value between the heading angle of the vehicle positioning angle and the heading angle of the corresponding point on the reference track, and the transverse position deviation epsilon d and the heading angle deviation epsilon θ are both calculated by the path tracking control module. The path tracking control module is a generic module of the autopilot system, which is not described in detail herein.
To improve the accuracy of the calculation, these parameters are obtained by preprocessing. The present embodiment provides a filtering algorithm for smoothing the above parameters, and an implementation manner is provided herein, where a first-order inertial filter calculation formula is as follows:
Y(i)＝αX(i)+(1-α)Y(i-1)，
Where α e [0,1] is a filter coefficient, X (i) is a current input value, Y (i-1) is a last filter output value, and Y (i) is a current filter output value. The filter is utilized to carry out smooth filtering treatment on the longitudinal speed v, the actual rotation angle delta f of the front axle tire, the actual rotation angle delta r of the rear axle tire, the curvature rho of the reference track, the transverse position deviation epsilon d and the course angle deviation epsilon θ respectively, and the filtering coefficient of each parameter is regulated to achieve an ideal effect.
Then, on-line learning of the front and rear axis zero offset angle delta r is carried out based on the preprocessed data, specifically:
And responding to the magnitude relation between the real-time vehicle state data and the set value, performing online learning of a front axle zero deflection angle delta f according to the real-time transverse position deviation epsilon d to obtain a real-time front axle zero deflection angle delta f, and performing online learning of a rear axle zero deflection angle delta r according to the real-time course angle deviation epsilon θ to obtain a real-time rear axle zero deflection angle delta r.
The learning of the front axle zero deflection angle delta f is mainly to perform on-line learning of the front axle zero deflection angle by utilizing the real-time transverse position deviation epsilon d according to the real-time vehicle speed v, the real-time curvature |rho| and the real-time magnitude relation between the actual rotation angles of the front wheel and the rear wheel and the corresponding set values. Specifically, the learning step of the front axle zero deflection angle delta f is as follows:
In response to the preprocessed real-time vehicle speed v being greater than the set vehicle speed value I v, the real-time curvature absolute value |ρ| being smaller than the set curvature value I ρ, the real-time front wheel actual rotation angle absolute value |δ f |, the real-time rear wheel actual rotation angle absolute value |δ r | being smaller than the set rotation angle value I δ, the front axle zero offset angle delta δ f is iteratively updated according to the real-time lateral position deviation epsilon d, and the update formula is Wherein the superscript i denotes the i-th instant,/>For the front axle zero deflection angle delta f at the ith moment, T is the iterative calculation interval time, beta is the front axle zero deflection angle learning rate, and epsilon di is the transverse position deviation at the ith moment.
In response to any one of the following conditions: the preprocessed real-time vehicle speed v is not more than a set vehicle speed value I v, the real-time curvature absolute value |ρ| is not less than a set curvature value I ρ, the real-time absolute value |delta f | of the actual rotation angle of the front wheel and the real-time absolute value |delta r | of the actual rotation angle of the rear wheel are not less than a set rotation angle value I δ, and the front axle zero offset angle delta f at the last moment is taken as the front axle zero offset angle delta f at the current moment, namely
The learning of the rear axle zero deflection angle delta r is mainly to perform online learning of the rear axle zero deflection angle epsilon θ by utilizing real-time course angle deviation according to the real-time vehicle speed v, real-time curvature |rho| and the real-time magnitude relation between the actual rotation angles of the front wheel and the rear wheel and the corresponding set values. Specifically, the learning step of the zero offset angle delta r of the rear axle is as follows:
Firstly, defining a buffer queue Q with the length of N, and storing real-time course angle deviation epsilon θ into the buffer queue Q in response to the fact that the preprocessed real-time vehicle speed v is larger than a set vehicle speed value I v and the real-time curvature absolute value |rho| is smaller than the set curvature value I ρ. When the real-time vehicle speed v is larger than the set vehicle speed value I v and the real-time curvature absolute value |rho| is smaller than any one of the set curvature value I ρ, the buffer queue Q is emptied.
And in response to the elements of the cache queue Q reaching the set number N, calculating the average value and standard deviation of the cache queue Q.
In particular, the method comprises the steps of,Wherein Q j is the element in queue Q, N is the number of queue elements, wherein/>To just reach the average value of the buffer queue Q when the set number N,/>In order to just reach the standard deviation of the cache queue Q when the set number N is reached, the average value and the standard deviation of the cache queue Q are not calculated when the elements of the cache queue Q do not reach the set number N.
The method that the last part presses the latest element is adopted for storing the elements of the cache queue Q, when the number of the elements reaches N, the last part presses the latest element when new elements exist, and the elements overflow the first part of the cache queue Q.
When the number of elements of the cache queue Q reaches N, in order to accelerate calculation and reduce cpu resource consumption, the subsequent mean value and standard deviation are calculated by adopting a recurrence formula, where Q N is the element pressed into the cache queue Q at last, and Q o is the element overflowed from the cache queue Q, and the calculation formula is as follows:
wherein/> Is the average value of the buffer queue Q at the ith moment,/>Is the standard deviation of the buffer queue Q at the ith moment,/>Is the average value of the buffer queue Q at the i-1 time,/>Is the standard deviation of the buffer queue Q at the i-1 th moment, if/>And/>When the element number of the cache queue Q does not exist and reaches N, adopting/>And/>The mean value and standard deviation of the buffer queue Q at the ith moment are calculated according to the calculation formula. In response to the preprocessed real-time vehicle speed v being greater than the set vehicle speed value I v, the real-time curvature absolute value |ρ| being smaller than the set curvature value I ρ, the real-time front wheel actual rotation angle absolute value |δ f |, and the real-time rear wheel actual rotation angle absolute value |δ r | being smaller than the set rotation angle value I δ, the rear axle zero offset angle Δδ r is iteratively updated according to the mean value and standard deviation of the real-time buffer queue Q, and the update formula is: /(I)Wherein/>For the rear axle zero offset angle delta r,/>, at time iFor the rear axle rotation angle control instruction at the ith moment, the rear axle rotation angle control instruction is directly obtained from a path tracking control module or obtained by preprocessing after being obtained from the path tracking control module, gamma is a rear axle zero deflection angle learning rate parameter, influences the convergence speed of the rear axle zero deflection angle delta r together with the standard deviation of the course angle deviation epsilon θ, and is calculated as the standard deviation/>And if the convergence speed is larger, the convergence speed is slow, otherwise, the convergence speed is fast.
And responding to the fact that the absolute value of the real-time front axle zero deflection angle delta f and/or the real-time rear axle zero deflection angle delta r exceeds a set threshold value, reporting the fault of the steering system, and reminding maintenance personnel of overhauling the steering system or performing safety braking control.
The last learned front axle zero offset angle delta f and rear axle zero offset angle delta r are saved in a file to be used as initial values when the next starting is performed.
Example two
As shown in fig. 2, the present invention further provides a front-rear wheel steering angle learning method of an automatic driving vehicle, where the automatic driving vehicle zero-offset angle online learning method according to the first embodiment obtains a real-time front axle zero-offset angle Δδ f and a rear axle zero-offset angle Δδ r, and compensates a steering control instruction obtained by a path tracking control module in real time according to the front axle zero-offset angle Δδ f and the rear axle zero-offset angle Δδ r, so as to obtain a compensated front-rear wheel steering angle.
The compensation formula is: Wherein/> For the compensated front wheel steering angle at the i-th momentFor the compensated rear wheel steering angle at the i-th time/>For the i-th moment front axle rotation angle control commandIs the rear axle rotation angle control command/>, at the i-th momentBoth are obtained directly from the path tracking control module or obtained by preprocessing after being obtained from the path tracking control module.
The drive-by-wire vehicle supporting the bicycle model instruction can directly compensate the front axle rotation angleCompensated front axle angle/>The command acts on the vehicle.
According to the compensated front axle angle when for a vehicle supporting only the ackerman steering modelCompensated front axle angle/>The calculated rotation angle of four wheels is as follows:
In response to the compensated front-rear wheel steering reversal, the steering angles of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel are:
The obtained corners of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel are acted on the vehicle to realize automatic driving control, as shown in fig. 3.
In response to the compensated front and rear wheel rotation angles being in the same direction, the rotation angles of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel are as follows:
the obtained corners of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel act on the vehicle to realize automatic driving control, as shown in figure 4, wherein L is the axle distance of the front axle and the rear axle of the vehicle, D is the wheel distance, and/ > To compensate for the front wheel angle,/>To compensate for the rear wheel steering angle.
The steps are repeatedly executed, so that the converged front axle zero deflection angle delta f and rear axle zero deflection angle delta r can be obtained through online learning in the automatic driving and driving process of the vehicle, and the automatic learning and updating are carried out along with the change of the vehicle zero deflection angle, so that the control precision and the robustness of a control algorithm are ensured.
Example III
The invention also provides an automatic driving vehicle zero-deflection angle online learning system, which comprises a zero-deflection angle self-learning module, wherein the zero-deflection angle self-learning module comprises a data acquisition module, a processing module and a storage module, the data acquisition module acquires vehicle state data, a reference track curvature rho, a transverse position deviation epsilon d and a course angle deviation epsilon θ from a vehicle-mounted sensor and a path tracking control module and transmits the vehicle state data, the reference track curvature rho, the transverse position deviation epsilon d and the course angle deviation epsilon θ to the processing module, the processing module is in communication connection with the storage module, and the storage module stores at least one executable instruction, and the executable instruction enables the processing module to execute operations corresponding to the automatic driving vehicle zero-deflection angle online learning method according to the data acquired by the data acquisition module, so as to acquire a real-time front axle zero-deflection angle delta f and a rear axle zero-deflection angle delta r.
In this embodiment, the vehicle state data includes the longitudinal speed v of the vehicle, the front axle tire actual turning angle δ f and the rear axle tire actual turning angle δ r.
Example IV
The invention also provides a front and rear wheel steering angle learning system of the automatic driving vehicle, as shown in fig. 5, comprising the automatic driving vehicle zero-deflection angle online learning system in the third embodiment, wherein a processing module is connected with a path tracking control module, the path tracking control module acquires the front axle zero-deflection angle delta f and the rear axle zero-deflection angle delta r learned by the processing module, and compensates steering control instructions calculated by the path tracking control module based on the front axle zero-deflection angle delta f and the rear axle zero-deflection angle delta r to obtain the compensated front and rear wheel steering angles.
The drive-by-wire vehicle supporting the bicycle model instruction can directly compensate the front axle rotation angleCompensated front axle angle/>The command acts on the vehicle. For a vehicle supporting only the ackerman steering model, the path tracking control module calculates the corners of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel of the vehicle according to the front-rear wheel corner learning method of the automatic driving vehicle in the second embodiment, and applies the obtained corners of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel to the vehicle to realize automatic driving control.
Example five
The invention also provides an automatic driving vehicle, which comprises a front and rear wheel steering angle learning system of the automatic driving vehicle and a vehicle steering control module, wherein the vehicle steering control module obtains the compensated front and rear wheel steering angle obtained by the path tracking control module and directly acts on the vehicle steering, or obtains the steering angles of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel of the vehicle obtained by the path tracking control module and acts on the vehicle steering.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
While embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that: many changes, modifications, substitutions and variations may be made to the embodiments without departing from the spirit and principles of the invention, the scope of which is defined by the claims and their equivalents.
Claims (10)
1. An online learning method of zero offset angle of an automatic driving vehicle is characterized by comprising the following steps:
Acquiring initial values of zero offset angles of front and rear axles of a vehicle;
acquiring vehicle state data, reference track curvature, lateral position deviation and course angle deviation in real time;
And responding to the magnitude relation between the real-time vehicle state data and the set value, performing online learning of the front axle zero deflection angle according to the real-time transverse position deviation to obtain the real-time front axle zero deflection angle, and performing online learning of the rear axle zero deflection angle according to the real-time course angle deviation to obtain the real-time rear axle zero deflection angle.
2. The method for online learning zero offset angle of an autonomous vehicle according to claim 1, wherein the online learning of zero offset angle of the front axle is performed according to real-time lateral position deviation in response to real-time vehicle speed, real-time curvature, real-time magnitude relation between actual rotation angles of the front and rear wheels and respective corresponding set values, and the online learning of zero offset angle of the rear axle is performed according to real-time heading angle deviation.
3. The method for learning zero offset angle on line of an automatically driven vehicle according to claim 1 or 2, wherein the front axle zero offset angle learning step is:
Responding to the real-time vehicle speed being larger than a set vehicle speed value, the real-time curvature absolute value being smaller than a set curvature value and the real-time absolute values of the actual rotation angles of the front and rear wheels being smaller than a set rotation angle value, and iteratively updating the front axle zero offset angle according to the real-time transverse position deviation;
In response to any one of the following conditions: the real-time vehicle speed is not greater than the set vehicle speed value, the real-time curvature absolute value is not less than the set curvature value, the real-time absolute value of the actual rotation angle of the front wheel and the real-time absolute value of the actual rotation angle of the rear wheel are not less than the set rotation angle value, and the front axle zero offset angle at the last moment is taken as the front axle zero offset angle at the current moment.
4. The method for learning zero offset angle on line of an automatically driven vehicle according to claim 1 or 2, wherein the step of learning zero offset angle of the rear axle is:
Defining a buffer queue Q with the length of N, storing the real-time course angle deviation into the buffer queue Q in response to the real-time vehicle speed being larger than a set vehicle speed value and the real-time curvature absolute value being smaller than the set absolute value, otherwise, emptying the buffer queue Q;
calculating the mean value and standard deviation of the cache queue Q in response to the set number of elements of the cache queue Q;
And in response to the real-time vehicle speed being greater than the set vehicle speed value, the real-time curvature absolute value being smaller than the set absolute value and the real-time absolute values of the actual rotation angles of the front and rear wheels being smaller than the set rotation angle absolute value, iteratively updating the zero offset angle of the rear axle according to the mean value and standard deviation of the cache queue Q.
5. The method for online learning zero offset angle of an automatically driven vehicle according to claim 4, wherein when the mean value and standard deviation of the cache queue Q are calculated, the mean value and standard deviation of the cache queue Q at the current time are recursively updated according to the mean value and standard deviation of the cache queue Q at the previous time.
6. The method of on-line learning zero offset angle of an autonomous vehicle of claim 1, wherein a steering system fault is reported in response to an absolute value of the real-time front axle zero offset angle and/or rear axle zero offset angle exceeding a set threshold.
7. The method for learning the front and rear wheel steering angles of the automatic driving vehicle is characterized in that the method for learning the front and rear wheel steering angles on line of the automatic driving vehicle with zero deflection is used for obtaining real-time front axle zero deflection angles and rear axle zero deflection angles according to any one of claims 1-6, and the steering control instructions obtained by the real-time calculation of the path tracking control module are compensated according to the front axle zero deflection angles and the rear axle zero deflection angles, so that the compensated front and rear wheel steering angles are obtained.
8. The automatic driving vehicle zero-offset online learning system is characterized by comprising a data acquisition module, a processing module and a storage module, wherein the data acquisition module acquires vehicle state data, reference track curvature, transverse position deviation and heading angle deviation from a vehicle-mounted sensor and a path tracking control module and transmits the vehicle state data, the reference track curvature, the transverse position deviation and the heading angle deviation to the processing module, the processing module is in communication connection with the storage module, the storage module stores at least one executable instruction, and the executable instruction enables the processing module to execute the operation corresponding to the automatic driving vehicle zero-offset online learning method according to the data acquired from the data acquisition module to acquire the real-time front axle zero-offset and rear axle zero-offset of the vehicle.
9. The front-rear wheel steering angle learning system of the automatic driving vehicle is characterized by comprising the automatic driving vehicle zero-deflection angle online learning system according to claim 7, wherein the processing module is connected with the path tracking control module, the path tracking control module obtains the front-axis zero-deflection angle and the rear-axis zero-deflection angle learned by the processing module, and compensates steering control instructions calculated by the path tracking control module based on the front-axis zero-deflection angle and the rear-axis zero-deflection angle to obtain the compensated front-rear wheel steering angle.
10. An automatic driving vehicle, comprising the front and rear wheel steering angle learning system of the automatic driving vehicle according to claim 9 and a vehicle steering control module, wherein the vehicle steering control module obtains the compensated front and rear wheel steering angle obtained by the path tracking control module and directly acts on the vehicle steering, or obtains the steering angles of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel of the vehicle obtained by the path tracking control module and acts on the vehicle steering.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311369973.8A CN117962920A (en) | 2023-10-20 | 2023-10-20 | Zero-deflection angle online learning method, front-rear wheel steering angle learning method and system for automatic driving vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311369973.8A CN117962920A (en) | 2023-10-20 | 2023-10-20 | Zero-deflection angle online learning method, front-rear wheel steering angle learning method and system for automatic driving vehicle |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117962920A true CN117962920A (en) | 2024-05-03 |
Family
ID=90846865
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311369973.8A Pending CN117962920A (en) | 2023-10-20 | 2023-10-20 | Zero-deflection angle online learning method, front-rear wheel steering angle learning method and system for automatic driving vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117962920A (en) |
-
2023
- 2023-10-20 CN CN202311369973.8A patent/CN117962920A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3932761A1 (en) | Vehicle abnormal lane change control method, device and system | |
| US10077072B2 (en) | Vehicle steering device and vehicle steering control method with reduced lateral speed | |
| CN107963126B (en) | Large-curvature automatic driving steering control method for multi-axle steering vehicle | |
| CN111007722B (en) | Transverse robust fault-tolerant control system and method for four-wheel steering automatic driving automobile | |
| CN103057436B (en) | Yawing moment control method of individual driven electromobile based on multi-agent | |
| CN112014122B (en) | Wheel corner measuring device and operation method | |
| CN111891125A (en) | Lane departure active deviation correction method based on torque control | |
| US20180304918A1 (en) | Automatic driving system | |
| CN115435816B (en) | Method, system, medium and equipment for calibrating internal and external parameters of AGV (automatic guided vehicle) with double steering wheels on line | |
| CN103121466B (en) | Arc linear motor power-assisted steering system and road feel control method thereof | |
| CN115715263A (en) | Steering control method and device | |
| CN116560371A (en) | Self-adaptive model predictive control-based automatic driving vehicle path tracking method | |
| CN117962920A (en) | Zero-deflection angle online learning method, front-rear wheel steering angle learning method and system for automatic driving vehicle | |
| CN115993089B (en) | PL-ICP-based online four-steering-wheel AGV internal and external parameter calibration method | |
| CN113415340B (en) | Parameter setting method for steering control of Ackerman-like steering mechanism | |
| CN102717726A (en) | Electronic differential control method and electrically driven mine car using same | |
| CN115848162A (en) | Control method for differential steering of six-wheel independent electrically-driven unmanned vehicle | |
| CN110949499B (en) | Unmanned driving corner compensation system of commercial vehicle and control method thereof | |
| CN112162553B (en) | Automatic alignment control system and method for cotton picker | |
| CN110723200B (en) | Steering centering and intermediate position control system and control method thereof | |
| JP3168758B2 (en) | Detection yaw rate correction device | |
| CN112092800A (en) | Steering wheel corner optimization method and system during automatic parking | |
| CN117519133B (en) | Unmanned cotton picker track tracking control method | |
| JPH072130A (en) | Method for controlling rear wheel steering device | |
| EP4183601A1 (en) | A method for controlling axle load distribution of a vehicle |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |