CN112158196B

CN112158196B - Automatic parking method and device

Info

Publication number: CN112158196B
Application number: CN202011084782.3A
Authority: CN
Inventors: 吕传龙
Original assignee: Beijing Rockwell Technology Co Ltd
Current assignee: Beijing Rockwell Technology Co Ltd
Priority date: 2020-10-12
Filing date: 2020-10-12
Publication date: 2022-08-12
Anticipated expiration: 2040-10-12
Also published as: CN112158196A

Abstract

The application discloses an automatic parking method and device, and relates to the technical field of electric vehicle control. The method of the present application comprises: generating a target parking track corresponding to the target vehicle according to the target parking space information, the current position information and the surrounding obstacle information corresponding to the target vehicle; generating first parameter adjusting information corresponding to the longitudinal controller and second parameter adjusting information corresponding to the transverse controller according to the target parking track, the first parameter adjuster and the second parameter adjuster; the parameters of the longitudinal controller are optimally adjusted by using the first parameter adjusting information, and the parameters of the transverse controller are optimally adjusted by using the second parameter adjusting information; and generating a target longitudinal acceleration value and a target front wheel deflection angle corresponding to the target vehicle according to the target parking track, the longitudinal controller after the optimization adjustment and the transverse controller after the optimization adjustment, and controlling the target vehicle by using the target longitudinal acceleration value and the target front wheel deflection angle. The method and the device are suitable for the automatic parking process.

Description

Automatic parking method and device

Technical Field

The application relates to the technical field of electric automobile control, in particular to an automatic parking method and device.

Background

With the continuous development of society and the continuous improvement of living standard of people, the demand of people on automobiles is increased day by day. Along with the rapid increase of the automobile holding capacity, the problems brought to the daily life of people by the automobiles are increasingly obvious, wherein the problem of difficult parking is particularly obvious. In order to effectively solve the problem of difficult parking, an automatic parking technology is developed. In the automatic parking process, the automatic parking of the vehicle can be realized without operating a steering wheel by a driver or observing the surrounding environment of the parking space by the driver.

At present, a controller is generally established in advance according to a control algorithm, after an automatic parking track is determined, a control instruction is generated according to the automatic parking track and the controller, and finally the vehicle is controlled to automatically park by using the control instruction. However, the automatic parking control based on the controller has poor control accuracy, so that the situations of collision, scratch and the like of the vehicle in the automatic parking process can be caused, and further personal and economic losses can be caused to a driver.

Disclosure of Invention

The embodiment of the application provides an automatic parking method and device, and mainly aims to improve the control precision of automatic parking control in the automatic parking process.

In order to solve the above technical problem, an embodiment of the present application provides the following technical solutions:

in a first aspect, the present application provides an automatic parking method, comprising:

generating a target parking track corresponding to a target vehicle according to target parking space information, current position information and surrounding obstacle information corresponding to the target vehicle;

generating first parameter adjusting information corresponding to a longitudinal controller and second parameter adjusting information corresponding to a transverse controller according to the target parking track, a first parameter adjuster and a second parameter adjuster, wherein the first parameter adjuster and the second parameter adjuster are obtained based on preset reinforcement learning algorithm training, and the longitudinal controller and the transverse controller are established based on a preset control algorithm;

the first parameter adjusting information is used for carrying out optimization adjustment on the parameters of the longitudinal controller, and the second parameter adjusting information is used for carrying out optimization adjustment on the parameters of the transverse controller;

and generating a target longitudinal acceleration value and a target front wheel deflection angle corresponding to the target vehicle according to the target parking track, the longitudinal controller after the optimization adjustment and the transverse controller after the optimization adjustment, and controlling the target vehicle by using the target longitudinal acceleration value and the target front wheel deflection angle.

Optionally, the generating a target parking track corresponding to the target vehicle according to the target parking space information, the current position information, and the surrounding obstacle information corresponding to the target vehicle includes:

acquiring target parking space information, current position information and surrounding obstacle information corresponding to the target vehicle;

and inputting the target parking space information, the current position information and the surrounding obstacle information into a preset model so that the preset model can output the target parking track and relevant information corresponding to each track point contained in the target parking track, wherein the preset model is a vehicle dynamics model or a vehicle kinematics model.

Optionally, the relevant information corresponding to the track point includes: the reference longitudinal speed value, the reference longitudinal acceleration value, the reference course and the reference transverse position corresponding to the track point; generating first parameter adjustment information corresponding to a longitudinal controller and second parameter adjustment information corresponding to a transverse controller according to the target parking trajectory, the first parameter adjuster and the second parameter adjuster, including:

determining a target track point according to the target parking track and a preset distance value;

acquiring a current longitudinal velocity value, a current longitudinal acceleration value, a current course, a current transverse position and a current course angular velocity corresponding to the target vehicle;

calculating a longitudinal speed deviation and a longitudinal acceleration deviation according to the current longitudinal speed value, the current longitudinal acceleration value, and a reference longitudinal speed value and a reference longitudinal acceleration value corresponding to the target track point;

inputting the longitudinal speed deviation and the longitudinal acceleration deviation as a first observation vector to the first parameter adjuster, so that the first parameter adjuster outputs the first parameter adjustment information;

calculating course deviation and transverse position deviation according to the current course, the current transverse position, and the reference course and the reference transverse position corresponding to the target track point;

and inputting the course deviation, the transverse position deviation, the current longitudinal speed value and the current course angular speed serving as second observation vectors into the second parameter regulator so that the second parameter regulator outputs second parameter regulation information.

Optionally, the related information corresponding to the trace point further includes: a reference longitudinal position corresponding to the trajectory point; generating a target longitudinal acceleration value and a target front wheel deflection angle corresponding to the target vehicle according to the target parking track, the longitudinal controller after the optimization adjustment and the transverse controller after the optimization adjustment, including:

acquiring a current longitudinal position corresponding to the target vehicle;

calculating a longitudinal position deviation according to the current longitudinal position and a reference longitudinal position corresponding to the target track point;

inputting the longitudinal position deviation into the optimally adjusted longitudinal controller so that the optimally adjusted longitudinal controller outputs the target longitudinal acceleration value;

and inputting the course deviation and the transverse position deviation into the transverse controller after the optimization adjustment so as to enable the transverse controller after the optimization adjustment to output the target front wheel slip angle.

Optionally, after the inputting the heading deviation, the lateral position deviation, the current longitudinal velocity value, and the current heading angular velocity as a second observation vector to the second parameter adjuster, so that the second parameter adjuster outputs the second parameter adjustment information, the method further includes:

inputting the first observation vector and the first parameter adjustment information into the first parameter adjuster so that the first parameter adjuster outputs a first return value;

optimizing and adjusting parameters in the first parameter regulator according to the first return value;

inputting the second observation vector and the second parameter adjustment information into the second parameter adjuster so that the second parameter adjuster outputs a second return value;

and carrying out optimization adjustment on the parameters in the second parameter regulator according to the second return value.

Optionally, before generating the target parking trajectory corresponding to the target vehicle according to the target parking space information, the current position information, and the surrounding obstacle information corresponding to the target vehicle, the method further includes:

establishing a first reinforcement learning model and a second reinforcement learning model based on a preset reinforcement learning algorithm, wherein the preset reinforcement learning algorithm comprises the following steps: any one of a Q Learning algorithm, a DQN algorithm, a PG algorithm, or a DDPG algorithm;

training the first reinforcement learning model based on a preset environment model until a first preset training stop condition is reached to obtain the first parameter regulator;

and training the second reinforcement learning model based on the preset environment model until a second preset training stop condition is reached to obtain the second parameter regulator.

Optionally, before the training of the first reinforcement learning model based on a preset environment model is performed until a first preset training stop condition is reached to obtain the first parameter adjuster, the method further includes:

training the first and second reinforcement learning models using preset training data.

In a second aspect, the present application also provides an automatic parking apparatus, including:

the system comprises a first generating unit, a second generating unit and a control unit, wherein the first generating unit is used for generating a target parking track corresponding to a target vehicle according to target parking space information, current position information and surrounding obstacle information corresponding to the target vehicle;

the second generating unit is used for generating first parameter adjusting information corresponding to a longitudinal controller and second parameter adjusting information corresponding to a transverse controller according to the target parking track, the first parameter adjuster and the second parameter adjuster, wherein the first parameter adjuster and the second parameter adjuster are obtained by training based on a preset reinforcement learning algorithm, and the longitudinal controller and the transverse controller are established based on a preset control algorithm;

the adjusting unit is used for optimizing and adjusting the parameters of the longitudinal controller by using the first parameter adjusting information and optimizing and adjusting the parameters of the transverse controller by using the second parameter adjusting information;

a third generating unit, configured to generate a target longitudinal acceleration value and a target front wheel slip angle corresponding to the target vehicle according to the target parking trajectory, the longitudinal controller after being optimally adjusted, and the lateral controller after being optimally adjusted;

a control unit for controlling the target vehicle using the target longitudinal acceleration value and the target front-wheel slip angle.

Optionally, the first generating unit includes:

the first acquisition module is used for acquiring target parking space information, current position information and surrounding obstacle information corresponding to the target vehicle;

the first input module is used for inputting the target parking space information, the current position information and the surrounding obstacle information into a preset model so that the preset model can output the target parking track and relevant information corresponding to each track point contained in the target parking track, and the preset model is specifically a vehicle dynamics model or a vehicle kinematics model.

Optionally, the relevant information corresponding to the track point includes: the reference longitudinal speed value, the reference longitudinal acceleration value, the reference course and the reference transverse position corresponding to the track point; the second generation unit includes:

the determining module is used for determining a target track point according to the target parking track and a preset distance value;

the second acquisition module is used for acquiring a current longitudinal speed value, a current longitudinal acceleration value, a current course, a current transverse position and a current course angular speed corresponding to the target vehicle;

the first calculation module is used for calculating a longitudinal speed deviation and a longitudinal acceleration deviation according to the current longitudinal speed value, the current longitudinal acceleration value, the reference longitudinal speed value corresponding to the target track point and the reference longitudinal acceleration value;

the second input module is used for inputting the longitudinal speed deviation and the longitudinal acceleration deviation into the first parameter regulator as a first observation vector so that the first parameter regulator can output the first parameter regulation information;

the second calculation module is used for calculating course deviation and transverse position deviation according to the current course, the current transverse position, and the reference course and the reference transverse position corresponding to the target track point;

and the third input module is used for inputting the course deviation, the transverse position deviation, the current longitudinal speed value and the current course angular speed serving as second observation vectors into the second parameter regulator so that the second parameter regulator can output the second parameter regulation information.

Optionally, the related information corresponding to the trace point further includes: a reference longitudinal position corresponding to the trajectory point; the third generation unit includes:

the third acquisition module is used for acquiring the current longitudinal position corresponding to the target vehicle;

the third calculation module is used for calculating the longitudinal position deviation according to the current longitudinal position and the reference longitudinal position corresponding to the target track point;

a fourth input module, configured to input the longitudinal position deviation into the optimally adjusted longitudinal controller, so that the optimally adjusted longitudinal controller outputs the target longitudinal acceleration value;

and the fifth input module is used for inputting the course deviation and the transverse position deviation into the transverse controller after the optimization adjustment so as to enable the transverse controller after the optimization adjustment to output the target front wheel deflection angle.

Optionally, the second generating unit further includes:

a sixth input module, configured to input the heading deviation, the lateral position deviation, the current longitudinal velocity value, and the current heading angular velocity as a second observation vector to the second parameter adjuster, so that after the second parameter adjuster outputs the second parameter adjustment information, the first observation vector and the first parameter adjustment information are input to the first parameter adjuster, so that the first parameter adjuster outputs a first report value;

the first adjusting module is used for optimizing and adjusting the parameters in the first parameter adjuster according to the first return value;

a seventh input module, configured to input the second observation vector and the second parameter adjustment information into the second parameter adjuster, so that the second parameter adjuster outputs a second report value;

and the second adjusting module is used for carrying out optimization adjustment on the parameters in the second parameter regulator according to the second return value.

Optionally, the apparatus further comprises:

the building unit is used for building a first reinforcement learning model and a second reinforcement learning model based on a preset reinforcement learning algorithm before the first generating unit generates a target parking track corresponding to a target vehicle according to target parking space information, current position information and surrounding obstacle information corresponding to the target vehicle, wherein the preset reinforcement learning algorithm is as follows: any one of a Q Learning algorithm, a DQN algorithm, a PG algorithm, or a DDPG algorithm;

the first training unit is used for training the first reinforcement learning model based on a preset environment model until a first preset training stop condition is reached so as to obtain the first parameter regulator;

and the second training unit is used for training the second reinforcement learning model based on the preset environment model until a second preset training stop condition is reached so as to obtain the second parameter regulator.

Optionally, the apparatus further comprises:

and the third training unit is used for training the first reinforcement learning model and the second reinforcement learning model by using preset training data before the first training unit trains the first reinforcement learning model based on a preset environment model until a first preset training stop condition is reached to obtain the first parameter regulator.

In a third aspect, an embodiment of the present application provides a storage medium including a stored program, where the program is executed to control an apparatus in the storage medium to execute the automatic parking method according to the first aspect.

In a fourth aspect, an embodiment of the present application provides an automatic parking apparatus, including a storage medium; and one or more processors, the storage medium coupled with the processors, the processors configured to execute program instructions stored in the storage medium; the program instructions are executed to execute the automatic parking method according to the first aspect.

By means of the technical scheme, the technical scheme provided by the application at least has the following advantages:

the application provides an automatic parking method and a device, compared with the prior art that after a controller is established according to a control algorithm, automatic parking control is directly carried out based on the controller, the method can generate a target parking track corresponding to a target vehicle according to target parking space information, current position information and surrounding obstacle information corresponding to the target vehicle, then generate first parameter adjusting information corresponding to a longitudinal controller and second parameter adjusting information corresponding to a transverse controller according to the target parking track, a first parameter adjusting information and a second parameter adjusting information, optimize and adjust parameters of the longitudinal controller by using the first parameter adjusting information, optimize and adjust parameters of the transverse controller by using the second parameter adjusting information, and finally optimize and adjust the longitudinal controller and the transverse controller after optimization and adjustment according to the target parking track, and generating a target longitudinal acceleration value and a target front wheel slip angle corresponding to the target vehicle, and controlling the target vehicle by using the target longitudinal acceleration value and the target front wheel slip angle. In the process of each round of automatic parking control, the first parameter adjusting information generated by the first parameter adjuster is used for adjusting the parameters in the longitudinal controller, and the second parameter adjusting information generated by the second parameter adjuster is used for adjusting the parameters in the transverse controller, so that the control precision of the longitudinal controller and the transverse controller can be effectively improved, and the control precision of the automatic parking control can be effectively improved.

The foregoing description is only an overview of the technical solutions of the present application, and the present application can be implemented according to the content of the description in order to make the technical means of the present application more clearly understood, and the following detailed description of the present application is given in order to make the above and other objects, features, and advantages of the present application more clearly understandable.

Drawings

The above and other objects, features and advantages of exemplary embodiments of the present application will become readily apparent from the following detailed description read in conjunction with the accompanying drawings. Several embodiments of the present application are illustrated by way of example and not by way of limitation in the figures of the accompanying drawings and in which like reference numerals refer to similar or corresponding parts and in which:

FIG. 1 is a flow chart illustrating an automatic parking method according to an embodiment of the present application;

FIG. 2 is a flow chart illustrating another method for automatic parking according to an embodiment of the present application;

fig. 3 is a block diagram illustrating components of an automatic parking apparatus according to an embodiment of the present application;

fig. 4 is a block diagram illustrating components of another automatic parking apparatus according to an embodiment of the present application.

Detailed Description

Exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

It is to be noted that, unless otherwise specified, technical terms or scientific terms used herein shall have the ordinary meaning as understood by those skilled in the art to which this application belongs.

An embodiment of the present application provides an automatic parking method, as shown in fig. 1, the method includes:

101. and generating a target parking track corresponding to the target vehicle according to the target parking space information, the current position information and the surrounding obstacle information corresponding to the target vehicle.

The target vehicle is a vehicle needing automatic parking control; the target parking space information corresponding to the target vehicle is information for indicating the position of the parking space to be parked; the current position information corresponding to the target vehicle is information for indicating the current position of the target vehicle; the peripheral obstacle information corresponding to the target vehicle is information used for describing obstacles around the target vehicle and obstacles around the parking space.

In the embodiment of the present invention, the execution subject in each step is an automatic driving controller in the target automobile. In the process of performing the automatic parking control of the current round, the automatic driving controller first needs to generate a target parking trajectory corresponding to the target vehicle according to the target parking space information, the current position information and the surrounding obstacle information corresponding to the target vehicle, so as to perform the automatic parking control based on the target parking trajectory in the following.

102. And generating first parameter adjusting information corresponding to the longitudinal controller and second parameter adjusting information corresponding to the transverse controller according to the target parking track, the first parameter adjuster and the second parameter adjuster.

The longitudinal controller and the transverse controller are specifically established based on a preset control algorithm, and the preset control algorithm may be, but is not limited to: any one of a PID algorithm, an LQR algorithm or an MPC algorithm; the first parameter adjuster and the second parameter adjuster are obtained by training based on a preset reinforcement learning algorithm, and the preset reinforcement learning algorithm can be, but is not limited to: any one of a Q Learning algorithm, a DQN algorithm, a PG algorithm, or a DDPG algorithm.

In the embodiment of the application, after the automatic driving controller generates the target parking track corresponding to the target vehicle according to the target parking space information, the current position information and the surrounding obstacle information corresponding to the target vehicle, first parameter adjustment information corresponding to the longitudinal controller and second parameter adjustment information corresponding to the lateral controller need to be generated according to the target parking track, the first parameter adjuster and the second parameter adjuster.

103. And optimally adjusting the parameters of the longitudinal controller by using the first parameter adjusting information, and optimally adjusting the parameters of the transverse controller by using the second parameter adjusting information.

In the embodiment of the application, after the automatic driving controller generates the first parameter adjustment information corresponding to the longitudinal controller and the second parameter adjustment information corresponding to the transverse controller according to the target parking track, the first parameter adjuster and the second parameter adjuster, the first parameter adjustment information can be used for carrying out optimization adjustment on the parameters of the longitudinal controller, and the second parameter adjustment information can be used for carrying out optimization adjustment on the parameters of the transverse controller, so that the control accuracy of the longitudinal controller and the control accuracy of the transverse controller are improved.

104. And generating a target longitudinal acceleration value and a target front wheel deflection angle corresponding to the target vehicle according to the target parking track, the longitudinal controller after the optimization adjustment and the transverse controller after the optimization adjustment, and controlling the target vehicle by using the target longitudinal acceleration value and the target front wheel deflection angle.

In the embodiment of the application, after the automatic driving controller uses the first parameter adjustment information to optimize and adjust the parameters of the longitudinal controller and uses the second parameter adjustment information to optimize and adjust the parameters of the transverse controller, the automatic driving controller can generate a target longitudinal acceleration value and a target front wheel deflection angle corresponding to the target vehicle according to the target parking track, the longitudinal controller after optimized and adjusted and the transverse controller after optimized and adjusted, and control the target vehicle by using the target longitudinal acceleration value and the target front wheel deflection angle, so that the automatic parking control of the current wheel is completed.

Further, in the embodiment of the present application, the method described in the foregoing 101-104 is repeated until the target vehicle stops at the parking space to be parked, so as to complete the automatic parking of the target vehicle.

Compared with the prior art that automatic parking control is directly performed based on a controller after the controller is established according to a control algorithm, the automatic parking control method can generate first parameter adjustment information corresponding to a longitudinal controller and second parameter adjustment information corresponding to a transverse controller according to a target parking track, a first parameter adjuster and a second parameter adjuster after the target parking track corresponding to a target vehicle is generated according to target parking space information, current position information and surrounding obstacle information corresponding to the target vehicle, optimize and adjust parameters of the longitudinal controller by using the first parameter adjustment information and optimize and adjust parameters of the transverse controller by using the second parameter adjustment information, and finally optimize and adjust the longitudinal controller and the transverse controller after optimization and adjustment according to the target parking track, and generating a target longitudinal acceleration value and a target front wheel slip angle corresponding to the target vehicle, and controlling the target vehicle by using the target longitudinal acceleration value and the target front wheel slip angle. In the process of each round of automatic parking control, the first parameter adjusting information generated by the first parameter adjuster is used for adjusting the parameters in the longitudinal controller, and the second parameter adjusting information generated by the second parameter adjuster is used for adjusting the parameters in the transverse controller, so that the control precision of the longitudinal controller and the transverse controller can be effectively improved, and the control precision of the automatic parking control can be effectively improved.

For the purpose of more detailed description, the present application provides another automatic parking method, specifically as shown in fig. 2, including:

201. and establishing a first reinforcement learning model and a second reinforcement learning model, and training the first reinforcement learning model and the second reinforcement learning model to obtain a first parameter regulator and a second parameter regulator.

In the embodiment of the present application, a first reinforcement learning model and a second reinforcement learning model need to be established in advance, and the first reinforcement learning model and the second reinforcement learning model need to be trained, so as to obtain a first parameter adjuster and a second parameter adjuster.

Specifically, in this step, a first reinforcement learning model and a second reinforcement learning model may be established in the following manner; training the first reinforcement learning model and the second reinforcement learning model to obtain a first parameter adjuster and a second parameter adjuster:

(1) establishing a first reinforcement learning model and a second reinforcement learning model based on a preset reinforcement learning algorithm, wherein the preset reinforcement learning algorithm can be but is not limited to: any one of a Q Learning algorithm, a DQN algorithm, a PG algorithm, or a DDPG algorithm.

(2) And training the first reinforcement learning model based on a preset environment model until a first preset training stop condition is reached to obtain a first parameter regulator.

The first preset training stop condition may be, but is not limited to: the first reinforcement learning model converges, and the accumulated training times reaches the preset training times or the accumulated training time reaches the preset training time.

In the embodiment of the present application, after the first reinforcement learning model is established based on the preset reinforcement learning algorithm, the first reinforcement learning model can be trained based on the preset environment model, namely, the preset environment model is controlled to output a longitudinal observation vector, the longitudinal observation vector is input into the first reinforcement learning model, so that the first reinforcement learning model outputs longitudinal parameter adjustment information, and then the longitudinal parameter adjustment information is input into the preset environment model, so as to output a longitudinal return value by the preset environment model, finally, carrying out optimization adjustment on parameters in the first reinforcement learning model according to the longitudinal return value to complete the training of the current round, repeating the steps until a first preset training stop condition is reached, stopping the training of the first reinforcement learning model, and determining the first reinforcement learning model as the first parameter adjuster, wherein the longitudinal observation vector may include, but is not limited to: longitudinal velocity bias and longitudinal acceleration bias, etc.

(3) And training the second reinforcement learning model based on the preset environment model until a second preset training stop condition is reached to obtain a second parameter regulator.

Wherein, the second preset training stop condition may be, but is not limited to: the second reinforcement learning model converges, and the accumulated training times reaches the preset training times or the accumulated training time reaches the preset training time.

In the embodiment of the present application, after the second reinforcement learning model is established based on the preset reinforcement learning algorithm, the second reinforcement learning model can be trained based on the preset environment model, namely, the preset environment model is controlled to output the transverse observation vector, the transverse observation vector is input into the second reinforcement learning model, so that the second reinforcement learning model outputs the transverse parameter adjustment information, and then the transverse parameter adjustment information is input into the preset environment model, so as to output a transverse return value by the preset environment model, finally, carrying out optimization adjustment on parameters in the second reinforcement learning model according to the transverse return value to complete the training of the current round, repeating the steps until reaching a second preset training stop condition, stopping the training of the second reinforcement learning model, and determining a second reinforcement learning model as a second parameter modifier, wherein the lateral observation vector may include, but is not limited to: course deviation, transverse position deviation, current longitudinal speed value, current course angular speed and the like.

Further, in this embodiment of the present application, before training a first reinforcement learning model and a second reinforcement learning model based on a preset environment model, preset training data may also be used to train the first reinforcement learning model and the second reinforcement learning model, so as to reduce training time for training the first reinforcement learning model and the second reinforcement learning model based on the preset environment model, where the preset training data is priori knowledge training data pre-stored by a worker, the preset training data includes multiple sets of longitudinal training data and multiple sets of transverse training data, and each set of longitudinal training data includes: the current moment longitudinal observation vector, the longitudinal parameter adjustment information corresponding to the current moment longitudinal observation vector, the longitudinal report value corresponding to the current moment longitudinal observation vector and the next moment longitudinal observation vector, wherein each group of transverse training data comprises: the current time transverse observation vector, the transverse parameter adjustment information corresponding to the current time transverse observation vector, the transverse report value corresponding to the current time transverse observation vector and the next time transverse observation vector.

202. And generating a target parking track corresponding to the target vehicle according to the target parking space information, the current position information and the surrounding obstacle information corresponding to the target vehicle.

In the embodiment of the invention, in the process of the current round of automatic parking control, the automatic driving controller firstly needs to generate the target parking track corresponding to the target vehicle according to the target parking space information, the current position information and the surrounding obstacle information corresponding to the target vehicle. The following describes in detail how the automatic driving controller generates a target parking trajectory corresponding to the target vehicle according to the target parking space information, the current position information, and the surrounding obstacle information corresponding to the target vehicle.

(1) And acquiring target parking space information, current position information and surrounding obstacle information corresponding to the target vehicle.

In the embodiment of the application, the automatic driving controller needs to acquire target parking space information, current position information and surrounding obstacle information corresponding to a target vehicle.

Specifically, in this step, the autopilot controller may obtain target parking space information, current position information, and surrounding obstacle information corresponding to the target vehicle through the high-precision map, the preset positioning sensor, the preset ultrasonic radar, and the preset look-around camera, but is not limited thereto.

(2) And inputting the target parking space information, the current position information and the surrounding obstacle information into a preset model so that the preset model can output the target parking track and relevant information corresponding to each track point contained in the target parking track.

The preset model is specifically a vehicle dynamics model or a vehicle kinematics model; the relevant information corresponding to any one track point included in the target parking track may include, but is not limited to: the track point corresponds to a reference longitudinal speed value, a reference longitudinal acceleration value, a reference heading, a reference transverse position, a reference longitudinal position and the like.

In the embodiment of the application, after obtaining the target parking space information, the current position information and the surrounding obstacle information corresponding to the target vehicle, the automatic driving controller may input the target parking space information, the current position information and the surrounding obstacle information corresponding to the target vehicle into a preset model (a vehicle dynamics model or a vehicle kinematics model), so that the preset model outputs the target parking track and relevant information corresponding to each track point included in the target parking track.

203. And generating first parameter adjusting information corresponding to the longitudinal controller and second parameter adjusting information corresponding to the transverse controller according to the target parking track, the first parameter adjuster and the second parameter adjuster.

In the embodiment of the present invention, after the automatic driving controller generates the target parking trajectory corresponding to the target vehicle according to the target parking space information, the current position information, and the surrounding obstacle information corresponding to the target vehicle, the automatic driving controller needs to generate the first parameter adjustment information corresponding to the longitudinal controller and the second parameter adjustment information corresponding to the lateral controller according to the target parking trajectory, the first parameter adjuster, and the second parameter adjuster. The following will describe in detail how the automatic driving controller generates the first parameter adjustment information corresponding to the longitudinal controller and the second parameter adjustment information corresponding to the lateral controller according to the target parking trajectory, the first parameter adjuster, and the second parameter adjuster.

(1) And determining a target track point according to the target parking track and the preset distance value.

The preset distance value may be, but is not limited to: 0.1m, 0.3m, 0.5 m.

In the embodiment of the application, after the automatic driving controller generates the target parking track corresponding to the target vehicle according to the target parking space information, the current position information and the surrounding obstacle information corresponding to the target vehicle, the target track point can be determined according to the target parking track and the preset distance value, namely the track point which is in the target parking track and is away from the starting point of the target parking track by the preset distance value is determined as the target track point.

Further, in the embodiment of the present invention, after the automatic driving controller inputs the target parking space information, the current position information, and the surrounding obstacle information corresponding to the target vehicle into the preset model, the preset model outputs the target parking trajectory and the relevant information corresponding to each trajectory point included in the target parking trajectory, so that after the automatic driving controller determines the target trajectory point according to the target parking trajectory and the preset distance value, the relevant information corresponding to the target trajectory point can be obtained: and the reference longitudinal speed value, the reference longitudinal acceleration value, the reference course, the reference transverse position, the reference longitudinal position and the like corresponding to the target track point.

(2) And acquiring a current longitudinal velocity value, a current longitudinal acceleration value, a current course, a current transverse position and a current course angular velocity corresponding to the target vehicle.

In the embodiment of the application, after the automatic driving controller determines the target track point according to the target parking track and the preset distance value, the current longitudinal velocity value, the current longitudinal acceleration value, the current heading, the current transverse position and the current heading angular velocity corresponding to the target vehicle need to be acquired.

(3) And calculating the longitudinal speed deviation and the longitudinal acceleration deviation according to the current longitudinal speed value, the current longitudinal acceleration value, the reference longitudinal speed value corresponding to the target track point and the reference longitudinal acceleration value.

In the embodiment of the application, after obtaining the current longitudinal velocity value and the current longitudinal acceleration value corresponding to the target vehicle, the automatic driving controller can calculate the longitudinal velocity deviation and the longitudinal acceleration deviation according to the current longitudinal velocity value corresponding to the target vehicle, the current longitudinal acceleration value, the reference longitudinal velocity value corresponding to the target track point and the reference longitudinal acceleration value, namely calculate the difference value between the reference longitudinal velocity value corresponding to the target track point and the current longitudinal velocity value corresponding to the target vehicle, and determine the difference value as the longitudinal velocity deviation; and calculating the difference value between the reference longitudinal acceleration value corresponding to the target track point and the current longitudinal acceleration value corresponding to the target vehicle, and determining the difference value as the longitudinal acceleration deviation.

(4) The longitudinal velocity deviation and the longitudinal acceleration deviation are input as a first observation vector to the first parameter adjuster, so that the first parameter adjuster outputs first parameter adjustment information.

In the embodiment of the application, after the longitudinal speed deviation and the longitudinal acceleration deviation are obtained through calculation, the automatic driving controller can input the longitudinal speed deviation and the longitudinal acceleration deviation into the first parameter regulator as the first observation vector, and the first parameter regulator can output first parameter regulation information corresponding to the longitudinal controller.

(5) And calculating course deviation and transverse position deviation according to the current course, the current transverse position, the reference course corresponding to the target track point and the reference transverse position.

In the embodiment of the application, after the automatic driving controller obtains the current course and the current transverse position corresponding to the target vehicle, the automatic driving controller can calculate course deviation and transverse position deviation according to the current course, the current transverse position corresponding to the target vehicle and the reference course and the reference transverse position corresponding to the target track point, namely calculate the difference value between the reference course corresponding to the target track point and the current course corresponding to the target vehicle, and determine the difference value as course deviation; and calculating the difference value between the reference transverse position corresponding to the target track point and the current transverse position corresponding to the target vehicle, and determining the difference value as the transverse position deviation.

(6) And inputting the course deviation, the transverse position deviation, the current longitudinal speed value and the current course angular speed as second observation vectors into a second parameter regulator so that the second parameter regulator outputs second parameter regulation information.

In the embodiment of the application, after the automatic driving controller calculates and obtains the heading deviation and the lateral position deviation, the heading deviation and the lateral position deviation can be input into the second parameter regulator as the second observation vector, and the second parameter regulator can output second parameter regulation information corresponding to the lateral controller.

Further, in the embodiment of the present application, after obtaining first parameter adjustment information corresponding to the longitudinal controller and second parameter adjustment information corresponding to the lateral controller, the automatic driving controller may input the first observation vector and the first parameter adjustment information into the first parameter adjuster, so that the first parameter adjuster outputs a first return value, and then performs optimization adjustment on parameters in the first parameter adjuster according to the first return value; inputting the second observation vector and the second parameter adjustment information into a second parameter adjuster so that the second parameter adjuster outputs a second return value, and then carrying out optimization adjustment on parameters in the second parameter adjuster according to the second return value, thereby further improving the accuracy of outputting the first parameter adjustment information by the first parameter adjuster in the next round of automatic parking control process; and the accuracy of outputting the second parameter adjusting information by the second parameter adjuster in the next round of automatic parking control process can be further improved.

204. And optimally adjusting the parameters of the longitudinal controller by using the first parameter adjusting information, and optimally adjusting the parameters of the transverse controller by using the second parameter adjusting information.

In step 204, the parameters of the longitudinal controller are optimally adjusted by using the first parameter adjustment information, and the parameters of the lateral controller are optimally adjusted by using the second parameter adjustment information, which may refer to the description of the corresponding part in fig. 1, and the embodiments of the present application will not be described again here.

205. And generating a target longitudinal acceleration value and a target front wheel deflection angle corresponding to the target vehicle according to the target parking track, the longitudinal controller after the optimization adjustment and the transverse controller after the optimization adjustment.

In the embodiment of the invention, after the automatic driving controller uses the first parameter adjustment information to optimize and adjust the parameters of the longitudinal controller and uses the second parameter adjustment information to optimize and adjust the parameters of the transverse controller, the automatic driving controller can generate the target longitudinal acceleration value and the target front wheel deflection angle corresponding to the target vehicle according to the target parking track, the optimized and adjusted longitudinal controller and the optimized and adjusted transverse controller. The following describes in detail how the automatic driving controller generates a target longitudinal acceleration value and a target front wheel slip angle corresponding to the target vehicle according to the target parking trajectory, the optimally adjusted longitudinal controller, and the optimally adjusted lateral controller.

(1) And acquiring the current longitudinal position corresponding to the target vehicle.

In the embodiment of the application, after the automatic driving controller uses the first parameter adjustment information to perform the optimal adjustment on the parameters of the longitudinal controller and uses the second parameter adjustment information to perform the optimal adjustment on the parameters of the lateral controller, the automatic driving controller needs to acquire the current longitudinal position corresponding to the target vehicle.

(2) And calculating the longitudinal position deviation according to the current longitudinal position and the reference longitudinal position corresponding to the target track point.

In this embodiment, after obtaining the reference longitudinal position corresponding to the target track point in step 203(1) and obtaining the current longitudinal position corresponding to the target vehicle in step 205(1), the automatic driving controller may calculate a longitudinal position deviation according to the current longitudinal position corresponding to the target vehicle and the reference longitudinal position corresponding to the target track point, that is, calculate a difference value between the reference longitudinal position corresponding to the target track point and the current longitudinal position corresponding to the target vehicle, and determine the difference value as the longitudinal position deviation.

(3) And inputting the longitudinal position deviation into the optimally adjusted longitudinal controller so that the optimally adjusted longitudinal controller outputs a target longitudinal acceleration value.

In the embodiment of the application, after the automatic driving controller calculates and obtains the longitudinal position deviation, the longitudinal position deviation can be input into the longitudinal controller after the optimization and adjustment, and the longitudinal controller after the optimization and adjustment can output the target longitudinal acceleration value.

(4) And inputting the course deviation and the transverse position deviation into the transverse controller after the optimization adjustment so as to output a target front wheel deflection angle by the transverse controller after the optimization adjustment.

In the embodiment of the present application, after obtaining the target longitudinal acceleration value, the automatic driving controller may input the heading deviation and the lateral position deviation calculated in step 203(5) into the optimally adjusted lateral controller, and the optimally adjusted lateral controller may output the target front wheel slip angle.

206. The target vehicle is controlled using the target longitudinal acceleration value and the target front-wheel slip angle.

In step 206, the target vehicle is controlled by using the target longitudinal acceleration value and the target front-wheel slip angle, reference may be made to the description of the corresponding part in fig. 1, and details of the embodiment of the present application will not be repeated here.

In order to achieve the above object, according to another aspect of the present application, an embodiment of the present application further provides a storage medium including a stored program, where the program is executed to control a device in which the storage medium is located to execute the automatic parking method.

In order to achieve the above object, according to another aspect of the present application, an embodiment of the present application further provides an automatic parking device, which includes a storage medium; and one or more processors, the storage medium coupled with the processors, the processors configured to execute program instructions stored in the storage medium; when the program instructions are operated, the automatic parking method is executed.

Further, as an implementation of the method shown in fig. 1 and 2, another embodiment of the present application further provides an automatic parking apparatus. The embodiment of the apparatus corresponds to the embodiment of the method, and for convenience of reading, details in the embodiment of the apparatus are not repeated one by one, but it should be clear that the apparatus in the embodiment can correspondingly implement all the contents in the embodiment of the method. The device is applied to improving the control precision of automatic parking control in the automatic parking process, and particularly comprises the following components in percentage by weight as shown in fig. 3:

the first generating unit 31 is configured to generate a target parking track corresponding to a target vehicle according to target parking space information, current position information, and surrounding obstacle information corresponding to the target vehicle;

a second generating unit 32, configured to generate, according to the target parking trajectory, the first parameter adjuster and the second parameter adjuster, first parameter adjustment information corresponding to a longitudinal controller and second parameter adjustment information corresponding to a lateral controller, where the first parameter adjuster and the second parameter adjuster are obtained by training based on a preset reinforcement learning algorithm, and the longitudinal controller and the lateral controller are established based on a preset control algorithm;

an adjusting unit 33, configured to perform optimal adjustment on the parameters of the longitudinal controller by using the first parameter adjustment information, and perform optimal adjustment on the parameters of the lateral controller by using the second parameter adjustment information;

a third generating unit 34, configured to generate a target longitudinal acceleration value and a target front wheel slip angle corresponding to the target vehicle according to the target parking trajectory, the longitudinal controller after being optimally adjusted, and the lateral controller after being optimally adjusted;

a control unit 35 for controlling the target vehicle using the target longitudinal acceleration value and the target front-wheel slip angle.

Further, as shown in fig. 4, the first generation unit 31 includes:

a first obtaining module 3101, configured to obtain target parking space information, current position information, and surrounding obstacle information corresponding to the target vehicle;

the first input module 3102 is configured to input the target parking space information, the current position information, and the surrounding obstacle information into a preset model, so that the preset model outputs the target parking trajectory and relevant information corresponding to each trajectory point included in the target parking trajectory, where the preset model is specifically a vehicle dynamics model or a vehicle kinematics model.

Further, as shown in fig. 4, the relevant information corresponding to the track point includes: the reference longitudinal speed value, the reference longitudinal acceleration value, the reference course and the reference transverse position corresponding to the track point; the second generation unit 32 includes:

the determining module 3201 is used for determining a target track point according to the target parking track and a preset distance value;

a second obtaining module 3202, configured to obtain a current longitudinal velocity value, a current longitudinal acceleration value, a current heading, a current lateral position, and a current heading angular velocity corresponding to the target vehicle;

a first calculating module 3203, configured to calculate a longitudinal velocity deviation and a longitudinal acceleration deviation according to the current longitudinal velocity value, the current longitudinal acceleration value, and a reference longitudinal velocity value and a reference longitudinal acceleration value corresponding to the target track point;

a second input module 3204, configured to input the longitudinal speed deviation and the longitudinal acceleration deviation as a first observation vector to the first parameter adjuster, so that the first parameter adjuster outputs the first parameter adjustment information;

the second calculation module 3205 is configured to calculate a heading deviation and a lateral position deviation according to the current heading, the current lateral position, and a reference heading and a reference lateral position corresponding to the target track point;

a third input module 3206, configured to input the heading deviation, the lateral position deviation, the current longitudinal velocity value, and the current heading angular velocity as a second observation vector to the second parameter adjuster, so that the second parameter adjuster outputs the second parameter adjustment information.

Further, as shown in fig. 4, the related information corresponding to the track point further includes: a reference longitudinal position corresponding to the trajectory point; the third generation unit 34 includes:

a third obtaining module 3401, configured to obtain a current longitudinal position corresponding to the target vehicle;

a third calculating module 3402, configured to calculate a longitudinal position deviation according to the current longitudinal position and a reference longitudinal position corresponding to the target track point;

a fourth input module 3403, configured to input the longitudinal position deviation into the optimally adjusted longitudinal controller, so that the optimally adjusted longitudinal controller outputs the target longitudinal acceleration value;

a fifth input module 3404, configured to input the heading deviation and the lateral position deviation into the optimally adjusted lateral controller, so that the optimally adjusted lateral controller outputs the target front-wheel slip angle.

Further, as shown in fig. 4, the second generating unit 32 further includes:

a sixth input module 3207, configured to, after the third input module 326 inputs the heading bias, the lateral position bias, the current longitudinal velocity value, and the current heading angular velocity as a second observation vector to the second parameter adjuster, so that the second parameter adjuster outputs the second parameter adjustment information, input the first observation vector and the first parameter adjustment information to the first parameter adjuster, so that the first parameter adjuster outputs a first return value;

a first adjusting module 3208, configured to perform an optimal adjustment on a parameter in the first parameter adjuster according to the first reported back value;

a seventh input module 3209, configured to input the second observation vector and the second parameter adjustment information into the second parameter adjuster, so that the second parameter adjuster outputs a second report value;

a second adjusting module 3210, configured to perform optimal adjustment on the parameter in the second parameter adjuster according to the second return value.

Further, as shown in fig. 4, the apparatus further includes:

the establishing unit 36 is configured to establish a first reinforcement learning model and a second reinforcement learning model based on a preset reinforcement learning algorithm before the first generating unit 31 generates the target parking trajectory corresponding to the target vehicle according to the target parking space information, the current position information, and the surrounding obstacle information corresponding to the target vehicle, where the preset reinforcement learning algorithm is: any one of a Q Learning algorithm, a DQN algorithm, a PG algorithm, or a DDPG algorithm;

a first training unit 37, configured to train the first reinforcement learning model based on a preset environment model until a first preset training stop condition is reached, so as to obtain the first parameter adjuster;

a second training unit 38, configured to train the second reinforcement learning model based on the preset environment model until a second preset training stop condition is reached, so as to obtain the second parameter adjuster.

Further, as shown in fig. 4, the apparatus further includes:

a third training unit 39, configured to train the first reinforcement learning model and the second reinforcement learning model using preset training data before the first training unit 37 trains the first reinforcement learning model based on a preset environment model until a first preset training stop condition is reached to obtain the first parameter adjuster.

The embodiment of the application provides an automatic parking method and a device, compared with the prior art that after a controller is established according to a control algorithm, automatic parking control is directly carried out based on the controller, the embodiment of the application can generate first parameter adjusting information corresponding to a longitudinal controller and second parameter adjusting information corresponding to a transverse controller according to a target parking track, a first parameter adjuster and a second parameter adjuster after the target parking track corresponding to a target vehicle is generated according to target parking space information, current position information and surrounding obstacle information corresponding to the target vehicle, optimize and adjust parameters of the longitudinal controller by using the first parameter adjusting information, optimize and adjust parameters of the transverse controller by using the second parameter adjusting information, and finally optimize and adjust the longitudinal controller and the transverse controller after optimization and adjustment according to the target parking track, and generating a target longitudinal acceleration value and a target front wheel slip angle corresponding to the target vehicle, and controlling the target vehicle by using the target longitudinal acceleration value and the target front wheel slip angle. In the process of each round of automatic parking control, the first parameter adjusting information generated by the first parameter adjuster is used for adjusting the parameters in the longitudinal controller, and the second parameter adjusting information generated by the second parameter adjuster is used for adjusting the parameters in the transverse controller, so that the control precision of the longitudinal controller and the transverse controller can be effectively improved, and the control precision of the automatic parking control can be effectively improved.

The automatic parking device comprises a processor and a memory, wherein the first generation unit, the second generation unit, the adjustment unit, the third generation unit, the control unit and the like are stored in the memory as program units, and the processor executes the program units stored in the memory to realize corresponding functions.

The processor comprises a kernel, and the kernel calls the corresponding program unit from the memory. One or more than one kernel can be set, and the control precision of automatic parking control is improved in the automatic parking process by adjusting kernel parameters.

The embodiment of the application provides a storage medium, which comprises a stored program, wherein when the program runs, a device where the storage medium is located is controlled to execute the automatic parking method.

The storage medium may include volatile memory in a computer readable medium, Random Access Memory (RAM) and/or nonvolatile memory such as Read Only Memory (ROM) or flash memory (flash RAM), and the memory includes at least one memory chip.

The embodiment of the application also provides an automatic parking device, which comprises a storage medium; and one or more processors, the storage medium coupled with the processors, the processors configured to execute program instructions stored in the storage medium; when the program instructions are operated, the automatic parking method is executed.

The embodiment of the application provides equipment, the equipment comprises a processor, a memory and a program which is stored on the memory and can run on the processor, and the following steps are realized when the processor executes the program:

Further, the generating a target parking track corresponding to the target vehicle according to the target parking space information, the current position information, and the surrounding obstacle information corresponding to the target vehicle includes:

Further, the relevant information corresponding to the track point includes: the reference longitudinal speed value, the reference longitudinal acceleration value, the reference course and the reference transverse position corresponding to the track point; generating first parameter adjustment information corresponding to a longitudinal controller and second parameter adjustment information corresponding to a transverse controller according to the target parking trajectory, the first parameter adjuster and the second parameter adjuster, including:

and inputting the course deviation, the transverse position deviation, the current longitudinal speed value and the current course angular speed serving as second observation vectors into the second parameter regulator so that the second parameter regulator can output second parameter regulation information.

Further, the related information corresponding to the trace points further includes: a reference longitudinal position corresponding to the trajectory point; generating a target longitudinal acceleration value and a target front wheel deflection angle corresponding to the target vehicle according to the target parking track, the longitudinal controller after the optimization adjustment and the transverse controller after the optimization adjustment, including:

acquiring a current longitudinal position corresponding to the target vehicle;

Further, after the inputting the heading deviation, the lateral position deviation, the current longitudinal speed value and the current heading angular speed as a second observation vector into the second parameter adjuster so that the second parameter adjuster outputs the second parameter adjustment information, the method further includes:

Further, before generating a target parking track corresponding to the target vehicle according to the target parking space information, the current position information, and the surrounding obstacle information corresponding to the target vehicle, the method further includes:

Further, before the training the first reinforcement learning model based on a preset environment model until a first preset training stop condition is reached to obtain the first parameter adjuster, the method further includes:

The present application further provides a computer program product adapted to perform program code for initializing the following method steps when executed on a data processing device: generating a target parking track corresponding to a target vehicle according to target parking space information, current position information and surrounding obstacle information corresponding to the target vehicle; generating first parameter adjusting information corresponding to a longitudinal controller and second parameter adjusting information corresponding to a transverse controller according to the target parking track, a first parameter adjuster and a second parameter adjuster, wherein the first parameter adjuster and the second parameter adjuster are obtained based on preset reinforcement learning algorithm training, and the longitudinal controller and the transverse controller are established based on a preset control algorithm; the first parameter adjusting information is used for carrying out optimization adjustment on the parameters of the longitudinal controller, and the second parameter adjusting information is used for carrying out optimization adjustment on the parameters of the transverse controller; and generating a target longitudinal acceleration value and a target front wheel deflection angle corresponding to the target vehicle according to the target parking track, the longitudinal controller after the optimization adjustment and the transverse controller after the optimization adjustment, and controlling the target vehicle by using the target longitudinal acceleration value and the target front wheel deflection angle.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). The memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in the process, method, article, or apparatus that comprises the element.

The above are merely examples of the present application and are not intended to limit the present application. Various modifications and changes may occur to those skilled in the art to which the present application pertains. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. An automatic parking method, comprising:

generating a target longitudinal acceleration value and a target front wheel deflection angle corresponding to the target vehicle according to the target parking track, the longitudinal controller after the optimization adjustment and the transverse controller after the optimization adjustment, and controlling the target vehicle by using the target longitudinal acceleration value and the target front wheel deflection angle;

the generating of the target parking track corresponding to the target vehicle according to the target parking space information, the current position information and the surrounding obstacle information corresponding to the target vehicle includes:

inputting the target parking space information, the current position information and the surrounding obstacle information into a preset model so that the preset model can output the target parking track and relevant information corresponding to each track point contained in the target parking track, wherein the preset model is a vehicle dynamics model or a vehicle kinematics model;

wherein, the relevant information corresponding to the track point comprises: the reference longitudinal speed value, the reference longitudinal acceleration value, the reference course and the reference transverse position corresponding to the track point;

generating first parameter adjustment information corresponding to a longitudinal controller and second parameter adjustment information corresponding to a transverse controller according to the target parking trajectory, the first parameter adjuster and the second parameter adjuster, including:

2. The method according to claim 1, wherein the related information corresponding to the track point further comprises: a reference longitudinal position corresponding to the trajectory point; generating a target longitudinal acceleration value and a target front wheel deflection angle corresponding to the target vehicle according to the target parking track, the longitudinal controller after the optimization adjustment and the transverse controller after the optimization adjustment, including:

acquiring a current longitudinal position corresponding to the target vehicle;

3. The method of claim 1, wherein after said inputting said heading bias, said lateral position bias, said current longitudinal velocity value, and said current heading angular velocity as a second observation vector into said second parametric modulator, such that said second parametric modulator outputs said second parametric modulation information, the method further comprises:

4. The method of claim 1, wherein before generating the target parking trajectory corresponding to the target vehicle according to the target parking space information, the current position information, and the surrounding obstacle information corresponding to the target vehicle, the method further comprises:

5. The method of claim 4, wherein before the training the first reinforcement learning model based on a preset environment model until a first preset training stop condition is reached to obtain the first parameter adjuster, the method further comprises:

6. An automatic parking device, comprising:

a second generating unit, configured to generate first parameter adjustment information corresponding to a longitudinal controller and second parameter adjustment information corresponding to a lateral controller according to the target parking trajectory, the first parameter adjuster, and the second parameter adjuster, where the first parameter adjuster and the second parameter adjuster are obtained by training based on a preset reinforcement learning algorithm, and the longitudinal controller and the lateral controller are established based on a preset control algorithm;

a control unit for controlling the target vehicle using the target longitudinal acceleration value and the target front-wheel slip angle;

the first generation unit includes:

the first input module is used for inputting the target parking space information, the current position information and the surrounding obstacle information into a preset model so that the preset model can output the target parking track and relevant information corresponding to each track point contained in the target parking track, and the preset model is specifically a vehicle dynamics model or a vehicle kinematics model;

the second generation unit includes:

the first calculation module is used for calculating a longitudinal speed deviation and a longitudinal acceleration deviation according to the current longitudinal speed value, the current longitudinal acceleration value, and a reference longitudinal speed value and a reference longitudinal acceleration value corresponding to the target track point;

7. The apparatus according to claim 6, wherein the related information corresponding to the track point further comprises: a reference longitudinal position corresponding to the trajectory point; the third generation unit includes:

8. The apparatus of claim 6, wherein the second generating unit further comprises:

9. The apparatus of claim 6, further comprising:

10. The apparatus of claim 9, further comprising:

11. A storage medium characterized by comprising a stored program, wherein a device in which the storage medium is located is controlled to execute the automatic parking method according to any one of claims 1 to 5 when the program is executed.

12. An automatic parking apparatus, characterized in that the apparatus comprises a storage medium; and one or more processors, the storage medium coupled with the processors, the processors configured to execute program instructions stored in the storage medium; the program instructions when executed perform the method for automatic parking according to any one of claims 1 to 5.