CN-121989938-A - Method and related equipment for determining braking track of automatic driving vehicle

CN121989938ACN 121989938 ACN121989938 ACN 121989938ACN-121989938-A

Abstract

The invention belongs to the technical field of automatic driving, and provides a method and related equipment for determining braking track of an automatic driving vehicle, wherein the method comprises the steps of acquiring state information of the automatic driving vehicle and braking indication information of a current road section from one or more data acquisition devices associated with the automatic driving vehicle; the method comprises the steps of determining a braking mode of an automatic driving vehicle according to state information and braking instruction information, constructing a plurality of candidate braking tracks of the automatic driving vehicle according to the braking mode and the state information, respectively inputting the plurality of candidate braking tracks into a braking track evaluation model for evaluating the candidate braking tracks to obtain a grading value of each candidate braking track, determining the candidate braking track with the highest grading value as the braking track of the automatic driving vehicle, and controlling the automatic driving vehicle to brake according to the braking tracks. The invention can improve the accuracy of the braking track of the automatic driving vehicle.

Inventors

LIU XIUYANG
HU HONGFEI
ZHU TIAN
PENG ZHICHUAN
ZHOU YANHUI
ZHANG ZHITENG
LIU GUANGWEI
ZHU ZEMIN
ZHANG YONG
CHEN YANNING

Assignees

长沙中车智驭新能源科技有限公司

Dates

Publication Date: 20260508
Application Date: 20241104

Claims (10)

1. A method for determining a braking trajectory of an autonomous vehicle, comprising: Acquiring, from one or more data acquisition devices associated with an autonomous vehicle, status information of the autonomous vehicle and brake indication information of a current road segment, the brake indication information including traffic signs for indicating a speed limit of the autonomous vehicle and location coordinates of the traffic signs; Determining a braking mode of the automatic driving vehicle according to the state information and the braking indication information, wherein the braking mode is braking deceleration or braking parking; Constructing a plurality of candidate braking tracks of the automatic driving vehicle according to the braking mode and the state information, wherein the braking time corresponding to different candidate braking tracks in the plurality of candidate braking tracks is different from each other; The method comprises the steps of respectively inputting a plurality of candidate braking tracks into a braking track evaluation model for evaluating the candidate braking tracks, obtaining a grading value of each candidate braking track, and determining the candidate braking track with the highest grading value as the braking track of the automatic driving vehicle, wherein the braking track evaluation model evaluates the candidate braking tracks from three dimensions of speed, acceleration and obstacle distance of the automatic driving vehicle; And controlling the automatic driving vehicle to brake according to the braking track.
2. The method of determining a braking trajectory of an autonomous vehicle according to claim 1, wherein the status information includes vehicle coordinates, linear velocity, and acceleration; the traffic sign is a parking sign or a deceleration sign.
3. The method of determining a braking trajectory of an autonomous vehicle according to claim 2, wherein said determining a braking mode of the autonomous vehicle based on the state information and the braking instruction information comprises: Calculating a distance between the vehicle coordinates and the position coordinates; When the distance is within a preset distance interval and the traffic sign is a parking sign, determining that the braking mode is braking and parking; And when the distance is within a preset distance interval and the traffic sign is a deceleration sign, determining that the braking mode is braking deceleration.
4. The method of determining a braking trajectory of an autonomous vehicle according to claim 3, wherein the expression of the candidate braking trajectory is: S(t)＝A 5 ·t 5 +A 4 ·t 4 +A 3 ·t 3 +A 2 ·t 2 +A 1 t+A 0 wherein S (t) represents the candidate braking trajectory, wherein a 0 ,A 1 ,A 2 ,A 3 ,A 4 ,A 5 represents a polynomial coefficient, S represents vehicle coordinates of the autonomous vehicle, and t represents braking time by solving the following set of equations: the result a 0 ,A 1 ,A 2 ,A 3 ,A 4 ,A 5 ,S 1 represents the position of the autonomous vehicle after braking, Represents the linear velocity of the autonomous vehicle after braking, Indicating the acceleration of the autonomous vehicle after braking, S 0 indicating the position of the autonomous vehicle before braking, Representing the linear velocity of the autonomous vehicle before braking, Indicating the acceleration of the autonomous vehicle before braking, t 0 indicating the time before braking of the autonomous vehicle, and t 1 indicating the time after braking of the autonomous vehicle.
5. The method of claim 4, wherein the brake trajectory evaluation model is a deep reinforcement learning model, and wherein the reward function of the deep reinforcement learning model is r=r 1 +R 2 +R 3 R dist =1+S(8) Wherein R represents a reward value, R 1 represents a reward function for evaluating the candidate braking track from a speed dimension, R 2 represents a reward function for evaluating the candidate braking track from an acceleration dimension, R 3 represents a reward function for evaluating the candidate braking track from an obstacle distance dimension, W speed represents a weight of the speed reward function R speed , W dist represents a weight of the distance reward function R dist , length represents a total length of planning of the candidate braking track, j t represents a longitudinal planned acceleration of the candidate braking track at each planned time t, v ref_t represents a reference maximum speed of the candidate braking track at each planned time t, each planned time interval is 100ms until t=8, Representing a planned speed at each planned time t in the candidate braking trajectory, S (8) representing a longitudinal position of the candidate braking trajectory at t=8 seconds, F c representing a collision risk function, d representing an obstacle distance, σ representing a standard deviation.
6. The method of claim 5, wherein training the brake trajectory evaluation model has a loss function of: Wherein L i (θ) represents a loss function, which is a mean square error between the target value network output value y i and the current value network output value Q (S, A; θ), θ represents weights of all the neural network layers, R (S) represents rewards under the state information S, γ represents discount factors, S v represents the state information of the autonomous vehicle at the next braking moment, A ′ represents the braking mode of the autonomous vehicle at the next braking moment, θ - represents the neural network parameters of the target value network, and max A′ Q(S ′ ,A ′ ;θ - ) represents the maximum state action value when the target value network selects the state S ′ .
7. The method of determining a braking trajectory of an autonomous vehicle according to claim 6, wherein said controlling the autonomous vehicle to brake according to the braking trajectory comprises: Generating a vehicle control signal indicative of the braking trajectory and inputting the vehicle control signal into a control system of the autonomous vehicle; A control system responsive to the vehicle control signal brakes the autonomous vehicle.
8. An automatic driving vehicle braking trajectory determination device, characterized by comprising: an information acquisition module for acquiring, from one or more data acquisition devices associated with an autonomous vehicle, status information of the autonomous vehicle and brake indication information of a current road segment, the brake indication information including a traffic sign for indicating a speed limit of the autonomous vehicle and position coordinates of the traffic sign; the braking mode determining module is used for determining a braking mode of the automatic driving vehicle according to the state information and the braking indication information, wherein the braking mode is braking deceleration or braking parking; The track generation module is used for constructing a plurality of candidate braking tracks of the automatic driving vehicle according to the braking mode and the state information, wherein the braking time corresponding to different candidate braking tracks in the plurality of candidate braking tracks is different; The track evaluation module is used for respectively inputting the plurality of candidate braking tracks into a braking track evaluation model for evaluating the candidate braking tracks, obtaining a grading value of each candidate braking track, and determining the candidate braking track with the highest grading value as the braking track of the automatic driving vehicle, wherein the braking track evaluation model evaluates the candidate braking tracks from three dimensions of the speed, the acceleration and the obstacle distance of the automatic driving vehicle; And the braking module is used for controlling the automatic driving vehicle to brake according to the braking track.
9. A terminal device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the method of determining the braking trajectory of an autonomous vehicle as claimed in any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, implements the autonomous vehicle braking trajectory determination method according to any one of claims 1 to 7.

Description

Method and related equipment for determining braking track of automatic driving vehicle Technical Field The invention belongs to the technical field of automatic driving, and particularly relates to a method for determining braking track of an automatic driving vehicle and related equipment. Background In recent years, rapid development of automatic driving technology has driven progress in intelligent transportation systems, particularly in terms of vehicle braking technology. The braking system is a critical safety component of an autonomous vehicle, ensuring that the vehicle is able to respond effectively to driving demands in various driving environments. Currently, braking techniques for autonomous vehicles rely primarily on sensor data and historical driving patterns to generate a braking trajectory. The prior art generally utilizes a plurality of sensors (e.g., lidar, cameras, and ultrasonic sensors) to collect environmental data and analyze historical driving behavior through advanced algorithms to predict and calculate a vehicle's braking trajectory under certain conditions. The design of these algorithms is initially to build a driving model through analysis of historical data for automatic application when similar scenarios are encountered. However, this data analysis method based on historical features has significant drawbacks, resulting in an inaccurate generated braking trajectory. In particular, the prior art often neglects the impact of real-time dynamic factors and environmental changes on braking decisions. Although historical data can provide a certain reference, due to the complexity and variability of the traffic environment, single dependency on historical features is difficult to cope with various emergencies in real-time situations. For example, in the case of sudden braking or emergency obstacle avoidance, the historical data may not cover all possible scenarios, resulting in a lag or inadaptation in braking decisions, thereby affecting traffic safety. In addition, existing brake control algorithms often fail to adequately account for physical characteristics of the vehicle (e.g., acceleration) and environmental factors (e.g., obstacle distance), which further results in reduced accuracy of the braking trajectory. Therefore, in order to improve the safety and reliability of an autonomous vehicle in complex and dynamic environments, a more optimal method for determining the braking trajectory of an autonomous vehicle is needed. Disclosure of Invention The invention solves the technical problem of providing a method and related equipment for determining the braking track of an automatic driving vehicle so as to improve the accuracy of the braking track of the automatic driving vehicle. In a first aspect, the present invention provides a method of determining a braking trajectory of an autonomous vehicle, the method comprising the steps of: Acquiring, from one or more data acquisition devices associated with the autonomous vehicle, status information of the autonomous vehicle and brake indication information of a current road segment, the brake indication information including traffic signs and position coordinates of the traffic signs for indicating a speed limit of the autonomous vehicle; Determining a braking mode of the automatic driving vehicle according to the state information and the braking indication information, wherein the braking mode is braking deceleration or braking parking; Constructing a plurality of candidate braking tracks of the automatic driving vehicle according to the braking mode and the state information, wherein the braking time corresponding to different candidate braking tracks in the plurality of candidate braking tracks is different from each other; Respectively inputting a plurality of candidate braking tracks into a braking track evaluation model for evaluating the candidate braking tracks to obtain a grading value of each candidate braking track, and determining the candidate braking track with the highest grading value as the braking track of the automatic driving vehicle, wherein the braking track evaluation model evaluates the candidate braking track from three dimensions of the speed, the acceleration and the obstacle distance of the automatic driving vehicle; And controlling the automatic driving vehicle to brake according to the braking track. Optionally, the status information includes vehicle coordinates, linear velocity, and acceleration; The traffic sign is a parking sign or a deceleration sign. Optionally, determining the braking mode of the automatic driving vehicle according to the state information and the braking indication information includes: Calculating the distance between the coordinates of the vehicle and the coordinates of the position; When the distance is within a preset distance interval and the traffic sign is a parking sign, determining that the braking mode is braking and parking; And when the distance is within the preset