Research on Automatic Driving Motion Control Based on Double Estimator Reinforcement Learning Combined with Forward Predictive Control

被引：0

作者：

Du G. ^{[1
,2
]}

Zou Y. ^{[1
]}

Zhang X. ^{[1
]}

Sun W. ^{[1
]}

机构：

[1] School of Mechanical Engineering, Beijing Institute of Technology, Beijing

[2] Institute of Dynamic System and Control, ETH Zurich, Zurich

来源：

Qiche Gongcheng/Automotive Engineering | 2024年 / 46卷 / 04期

关键词：

autonomous vehicle; double estimator reinforcement learning algorithm; forward predictive control method; motion control optimization;

D O I：

10.19562/j.chinasae.qcgc.2024.04.002

中图分类号：

学科分类号：

摘要：

Motion control research is an important part to achieve the goal of autonomous driving. To solve the problem of suboptimal control sequence due to the limitation of single-step decision in traditional reinforcement learning algorithm,a motion control framework based on the combination of double estimator reinforcement learning algorithm and forward predictive control method(DEQL-FPC)is proposed. In this framework,double estimators are introduced to solve the problem of action overestimation of traditional reinforcement learning methods and improve the speed of optimization. The forward predictive multi-step decision making method is designed to replace the single step decision making of traditional reinforcement learning so as to effectively improve the performance of global control strategies. Through virtual driving environment simulation,the superiority of the control framework applied in path tracking and safe obstacle avoidance of autonomous vehicles is proved,and the accuracy,safety,rapidity and comfort of motion control are guaranteed. © 2024 SAE-China. All rights reserved.

引用

页码：564 / 576

页数：12

共 20 条

[1] HAYDARI A, YILMAZ Y., Deep reinforcement learning for intelligent transportation systems:a survey, IEEE Transactions on Intelligent Transportation Systems, 23, 1, pp. 11-32, (2020)
[2] LI K Q, DAI Y F, LI S B, Et al., State-of-the-art and technical trends of intelligent and connected vehicles, Journal of Automotive Safety and Energy, 8, 1, pp. 1-14, (2017)
[3] JIN L S, HAN G D, XIE X Y, Et al., Review of autonomous driving decision-making research based on reinforcement learning, Automotive Engineering, 45, 4, pp. 527-540, (2023)
[4] HAN G, FU W, WANG W, Et al., The lateral tracking control for the intelligent vehicle based on adaptive PID neural network[J], Sensors, 17, 6, (2017)
[5] PAN S J, LI Y L, LI Z X, Et al., Path following method of intelligent vehicles based on improved pure tracking, Automotive Engineering, 45, 1, pp. 1-8, (2023)
[6] HU J, ZHONG X K, CHEN R N, Et al., Path tracking control of intelligent vehicles based on fuzzy LQR, Automotive Engineering, 44, 1, pp. 17-25, (2022)
[7] TANG L, YAN F, ZOU B, Et al., An improved kinematic model predictive control for high-speed path tracking of autonomous vehicles[J], IEEE Access, 8, pp. 51400-51413, (2020)
[8] YANG B, ZHANG H H, JIANG Z S., Simulation of vehicle obstacle avoidance path tracking control based on model predictive control, Intelligent Computer and Applications, 10, 12, pp. 99-103, (2020)
[9] LI J, WAN W X, HAO S Q, Et al., Research on model predictive control of autonomous vehicle path tracking under complex road condition, Automotive Engineering, 44, 5, pp. 664-674, (2022)
[10] GAO F, HAN Y, DANG D., Balancing accuracy and efficiency: fast motion planning based on nonlinear model predictive control [C], Proceedings of the 2021 5th CAA International Conference on Vehicular Control and Intelligence, pp. 3074-3089, (2021)

← 1 2 →