Dynamic speed trajectory generation and tracking control for autonomous driving of intelligent high-speed trains combining with deep learning and backstepping control methods

The development of autonomous transportation systems has received increasing attention over the last decades. Different from existing automatic train control systems, the decision-making capability in the autonomous driving system enables a train to adapt to the complicated and dynamic circumstances...

Full description

Saved in:
Bibliographic Details
Published in:Engineering applications of artificial intelligence Vol. 115; p. 105230
Main Authors: Wang, Xi, Li, Shukai, Cao, Yuan, Xin, Tianpeng, Yang, Lixing
Format: Journal Article
Language:English
Published: Elsevier Ltd 01.10.2022
Subjects:
Autonomous driving
Backstepping control
Deep learning
High-speed trains
Tracking control
Deep learning
Backstepping control
High-speed trains
Autonomous driving
Tracking control
ISSN:0952-1976
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The development of autonomous transportation systems has received increasing attention over the last decades. Different from existing automatic train control systems, the decision-making capability in the autonomous driving system enables a train to adapt to the complicated and dynamic circumstances. This paper in particular focuses on the decision-making problem for the autonomous driving of intelligent high-speed trains, and proposes a novel decision-making framework by combining the deep learning and backstepping control methods. By exploiting the deep learning methods, a speed trajectory generator is trained with the actual driving data, and dynamically calculates the reference speed trajectory according to the real-time driving condition. Then, a backstepping controller is designed to track the reference speed trajectory such that the separation, cohesion and alignment requirements for the autonomous driving of high-speed trains are achieved. Simulation experiments are implemented to illustrate the effectiveness of our methods.
ISSN:0952-1976
DOI:10.1016/j.engappai.2022.105230