CN-115320638-B - Transverse control method and device based on vehicle dynamic prediction model

CN115320638BCN 115320638 BCN115320638 BCN 115320638BCN-115320638-B

Abstract

Inputting a first steering wheel turning angle in a pre-acquired target time window into a vehicle dynamic prediction model, and convolving the vehicle dynamic prediction model based on the input first steering wheel turning angle in the target time window and a unit impulse response of the vehicle dynamic prediction model to obtain a predicted yaw rate; based on the predicted yaw rate and a pre-established cost function, a predicted steering wheel angle is obtained. The invention improves the accuracy of predicting the steering wheel angle, reduces the adjustment of the actuator while ensuring the precision, thereby improving the durability of the actuator and reducing the Total Cost of Ownership (TCO), and in addition, the invention iterates the optimal constraint condition on the algorithm through a large amount of operation data, thereby ensuring more accurate automatic driving behavior.

Inventors

ZHOU PEI
ZHANG SHENGQI
ZHANG HANQING
ZHU WANGZHEN

Assignees

嬴彻星创智能科技(上海)有限公司

Dates

Publication Date: 20260508
Application Date: 20220831

Claims (9)

1. A lateral control method based on a dynamic prediction model of a vehicle, comprising: Inputting a first steering wheel turning angle in a pre-acquired target time window into a vehicle dynamic prediction model, wherein the vehicle dynamic prediction model carries out convolution on the basis of the input first steering wheel turning angle in the target time window and a unit impulse response of the vehicle dynamic prediction model to obtain a predicted yaw rate; obtaining a predicted steering wheel angle based on the predicted yaw rate and a pre-established cost function to control the vehicle in accordance with the predicted steering wheel angle; the obtaining the predicted steering wheel angle based on the predicted yaw rate and a pre-established cost function includes: Obtaining a plurality of costs and second steering wheel angles corresponding to the costs according to the predicted yaw rate and by combining a pre-established cost function; and selecting the corresponding second steering wheel angle as the predicted steering wheel angle based on the minimum cost and the preset constraint condition.
2. The lateral control method based on a dynamic predictive model of a vehicle according to claim 1, wherein the obtaining a plurality of costs and the second steering wheel angle corresponding to each cost according to the predicted yaw rate and in combination with a pre-established cost function includes: Obtaining a yaw rate deviation error according to the predicted yaw rate and the pre-obtained estimated yaw rate; Acquiring a course angle, a side deflection angle and a transverse position of the same track point corresponding to the first steering wheel angle in a target time window according to the first steering wheel angle corresponding to the predicted yaw angle speed; Obtaining a course angle deviation error according to the yaw rate deviation error, the course angle and the side deviation angle; Obtaining a transverse position offset error according to the yaw rate offset error, the course angle offset error and the transverse position; According to the yaw rate offset error, the course angle offset error and the transverse position offset error, and combining And obtaining a plurality of costs and a second steering wheel angle corresponding to each cost by a pre-established cost function.
3. The lateral control method based on a vehicle dynamic prediction model according to claim 2, characterized by comprising, before the deriving a yaw-rate offset error from the predicted yaw-rate and a pre-acquired estimated yaw-rate: acquiring track points in a target time window; Obtaining the curvature of each track point according to the track points; And obtaining a presumed yaw rate based on the curvature of each track point and the vehicle speed corresponding to the track point.
4. The lateral control method based on a dynamic predictive model of a vehicle according to claim 1, wherein selecting a corresponding second steering wheel angle based on a minimum cost and a preset constraint condition, to obtain a predicted steering wheel angle, comprises: Determining a lateral position error constraint and a steering wheel rotational speed constraint based on the target travel region; Selecting a corresponding second steering wheel corner according to the transverse position error constraint and the steering wheel rotating speed constraint; And selecting the second steering wheel angle corresponding to the minimum cost as the predicted steering wheel angle according to the cost of the selected second steering wheel angle.
5. The lateral control method based on a vehicle dynamic prediction model according to claim 1, characterized by comprising, before inputting a first steering wheel angle within a target time window acquired in advance into the vehicle dynamic prediction model: acquiring a first steering wheel angle and a first yaw rate in the target time window; Normalizing the first steering wheel angle in the target time window; after the predicted yaw rate is obtained, comprising: and updating the vehicle dynamic prediction model according to the predicted yaw rate, the first yaw rate in the target time window and the normalized first steering wheel corner.
6. The lateral control method based on the vehicle dynamic prediction model according to claim 4, further comprising, after acquiring the first steering wheel angle and the first yaw rate within the target time window: obtaining a corresponding steering wheel angle average value and a corresponding yaw rate average value according to the first steering wheel angle and the first yaw rate in the target time window; respectively reducing each first steering wheel angle in the target time window according to the steering wheel angle average value; and respectively reducing each first yaw rate in the target time window according to the yaw rate average value.
7. A lateral control device based on a dynamic prediction model of a vehicle, comprising: the yaw rate prediction module is used for inputting a first steering wheel turning angle in a target time window obtained in advance into a vehicle dynamic prediction model, and the vehicle dynamic prediction model is used for carrying out convolution on the basis of the input first steering wheel turning angle in the target time window and a unit impulse response of the vehicle dynamic prediction model to obtain a predicted yaw rate; An angle prediction module that obtains a predicted steering wheel angle based on the predicted yaw rate and a pre-established cost function to control a vehicle in accordance with the predicted steering wheel angle; the angle prediction module comprises: The steering angle acquisition unit is used for acquiring a plurality of costs and second steering wheel steering angles corresponding to the costs according to the predicted yaw rate and in combination with a pre-established cost function; and the turning angle prediction unit is used for selecting a corresponding second steering wheel turning angle as a predicted steering wheel turning angle based on the minimum cost and a preset constraint condition.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the lateral control method based on a dynamic predictive model of a vehicle according to any one of claims 1 to 6 when the program is executed by the processor.
9. A non-transitory computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the lateral control method based on a dynamic predictive model of a vehicle according to any one of claims 1 to 6.

Description

Transverse control method and device based on vehicle dynamic prediction model Technical Field The invention relates to the technical field of automatic driving, in particular to a transverse control method and device based on a dynamic prediction model of a vehicle. Background With the popularity of autonomous vehicles, the autonomous vehicles may be used as taxis or public transportation vehicles, and passengers need to input destinations when using the autonomous vehicles, which generate a travel route based on a current location and the destination, and travel according to the generated travel route. At present, a planning control system is mostly adopted to generate a driving route, and a control module is utilized to track the driving route. The planning control system is used as a sufficient requirement for safe driving of the vehicle, and provides five great values of safe driving, accurate control, comfortable riding, economical use and good durability for customers. However, the existence of some mutually restricted indicators in the control module, such as pursuing control accuracy, inevitably results in more frequent adjustment of the controlled object, thereby reducing the durability of the actuator. Disclosure of Invention The invention provides a transverse control method and a device based on a dynamic prediction model of a vehicle, which are used for solving the defect that the durability of an actuator is reduced when the accuracy is ensured due to mutual restriction among indexes in the prior art, and reducing the adjustment of the actuator on the premise of ensuring a safety boundary so as to improve the durability of the actuator. The invention provides a transverse control method based on a vehicle dynamic prediction model, which comprises the steps of inputting a first steering wheel turning angle in a target time window obtained in advance into the vehicle dynamic prediction model, wherein the vehicle dynamic prediction model carries out convolution on the basis of the first steering wheel turning angle in the input target time window and a unit impulse response of the vehicle dynamic prediction model to obtain a predicted yaw rate, and obtains the predicted steering wheel turning angle based on the predicted yaw rate and a pre-established cost function so as to control a vehicle according to the predicted steering wheel turning angle. The transverse control method based on the vehicle dynamic prediction model comprises the steps of obtaining a plurality of costs and second steering wheel corners corresponding to the costs according to the predicted yaw rate and combining the pre-established cost function, and selecting the corresponding second steering wheel corners as predicted steering wheel corners based on the minimum cost and a preset constraint condition. The transverse control method based on the vehicle dynamic prediction model comprises the steps of obtaining a plurality of costs and second steering wheel angles corresponding to each cost according to the predicted yaw rate and a preset cost function, obtaining a yaw rate deviation error according to the predicted yaw rate and the preset estimated yaw rate, obtaining a heading angle, a side deviation angle and a transverse position of the same track point corresponding to the first steering wheel angle in a target time window according to the first steering wheel angle corresponding to the predicted yaw rate, obtaining a heading angle deviation error according to the transverse deviation error, the heading angle deviation error and the side deviation angle, obtaining a transverse position deviation error according to the yaw rate deviation error, the heading angle deviation error and the transverse position, and obtaining a plurality of costs and second steering wheel angles corresponding to each cost according to the yaw rate deviation error, the heading angle deviation error and the transverse position deviation error and a preset cost function. The transverse control method based on the vehicle dynamic prediction model comprises the steps of obtaining track points in a target time window, obtaining curvature corresponding to each track point according to the track points, and obtaining the estimated yaw rate based on the curvature of each track point and the vehicle speed corresponding to the track point before obtaining the yaw rate deviation error according to the predicted yaw rate and the pre-obtained estimated yaw rate. The transverse control method based on the vehicle dynamic prediction model comprises the steps of determining transverse position error constraint and steering wheel rotating speed constraint based on a target driving area, selecting the corresponding second steering wheel rotating angle according to the transverse position error constraint and the steering wheel rotating speed constraint, and selecting the second steering wheel rotating angle corresponding to the minimum cost according to the cost