Anti-Collision Path Planning and Tracking of Autonomous Vehicle Based on Optimized Artificial Potential Field and Discrete LQR Algorithm

This paper introduces an enhanced APF method to address challenges in automatic lane changing and collision avoidance for autonomous vehicles, targeting issues of infeasible target points, local optimization, inadequate safety margins, and instability when using DLQR. By integrating a distance adjus...

Full description

Saved in:
Bibliographic Details
Main Authors: Chaoxia Zhang, Zhihao Chen, Xingjiao Li, Ting Zhao
Format: Article
Language:English
Published: MDPI AG 2024-11-01
Series:World Electric Vehicle Journal
Subjects:
Online Access:https://www.mdpi.com/2032-6653/15/11/522
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850227674529660928
author Chaoxia Zhang
Zhihao Chen
Xingjiao Li
Ting Zhao
author_facet Chaoxia Zhang
Zhihao Chen
Xingjiao Li
Ting Zhao
author_sort Chaoxia Zhang
collection DOAJ
description This paper introduces an enhanced APF method to address challenges in automatic lane changing and collision avoidance for autonomous vehicles, targeting issues of infeasible target points, local optimization, inadequate safety margins, and instability when using DLQR. By integrating a distance adjustment factor, this research aims to rectify traditional APF limitations. A safety distance model and a sub-target virtual potential field are established to facilitate collision-free path generation for autonomous vehicles. A path tracking system is designed, combining feed-forward control with DLQR. Linearization and discretization of the vehicle’s dynamic state space model, with constraint variables set to minimize control-command costs, aligns with DLQR objectives. The aim is precise steering angle determination for path tracking, negating lateral errors due to external disturbances. A Simulink–CarSim co-simulation platform is utilized for obstacle and speed scenarios, validating the autonomous vehicle’s dynamic hazard avoidance, lane changing, and overtaking capabilities. The refined APF method enhances path safety, smoothness, and stability. Experimental data across three speeds reveal reasonable steering angle and lateral deflection angle variations. The controller ensures stable reference path tracking at 40, 50, and 60 km/h around various obstacles, verifying the controller’s effectiveness and driving stability. Comparative analysis of visual trajectories pre-optimization and post-optimization highlights improvements. Vehicle roll and sideslip angle peaks, roll-angle fluctuation, and front/rear wheel steering vertical support forces are compared with traditional LQR, validating the optimized controller’s enhancement of vehicle performance. Simulation results using MATLAB/Simulink and CarSim demonstrate that the optimized controller reduces steering angles by 5 to 10°, decreases sideslip angles by 3 to 5°, and increases vertical support forces from 1000 to 1450 N, showcasing our algorithm’s superior obstacle avoidance and lane-changing capabilities under dynamic conditions.
format Article
id doaj-art-d0233bdf28f0409b897bb1cb00615e17
institution OA Journals
issn 2032-6653
language English
publishDate 2024-11-01
publisher MDPI AG
record_format Article
series World Electric Vehicle Journal
spelling doaj-art-d0233bdf28f0409b897bb1cb00615e172025-08-20T02:04:45ZengMDPI AGWorld Electric Vehicle Journal2032-66532024-11-01151152210.3390/wevj15110522Anti-Collision Path Planning and Tracking of Autonomous Vehicle Based on Optimized Artificial Potential Field and Discrete LQR AlgorithmChaoxia Zhang0Zhihao Chen1Xingjiao Li2Ting Zhao3School of Mechanicalan Electrical Engineering and Automation, Foshan University, Foshan 528225, ChinaSchool of Mechanicalan Electrical Engineering and Automation, Foshan University, Foshan 528225, ChinaCenter of Teacher’s Teaching Development, Guangdong University of Education, Guangzhou 510303, ChinaSchool of Mechanicalan Electrical Engineering and Automation, Foshan University, Foshan 528225, ChinaThis paper introduces an enhanced APF method to address challenges in automatic lane changing and collision avoidance for autonomous vehicles, targeting issues of infeasible target points, local optimization, inadequate safety margins, and instability when using DLQR. By integrating a distance adjustment factor, this research aims to rectify traditional APF limitations. A safety distance model and a sub-target virtual potential field are established to facilitate collision-free path generation for autonomous vehicles. A path tracking system is designed, combining feed-forward control with DLQR. Linearization and discretization of the vehicle’s dynamic state space model, with constraint variables set to minimize control-command costs, aligns with DLQR objectives. The aim is precise steering angle determination for path tracking, negating lateral errors due to external disturbances. A Simulink–CarSim co-simulation platform is utilized for obstacle and speed scenarios, validating the autonomous vehicle’s dynamic hazard avoidance, lane changing, and overtaking capabilities. The refined APF method enhances path safety, smoothness, and stability. Experimental data across three speeds reveal reasonable steering angle and lateral deflection angle variations. The controller ensures stable reference path tracking at 40, 50, and 60 km/h around various obstacles, verifying the controller’s effectiveness and driving stability. Comparative analysis of visual trajectories pre-optimization and post-optimization highlights improvements. Vehicle roll and sideslip angle peaks, roll-angle fluctuation, and front/rear wheel steering vertical support forces are compared with traditional LQR, validating the optimized controller’s enhancement of vehicle performance. Simulation results using MATLAB/Simulink and CarSim demonstrate that the optimized controller reduces steering angles by 5 to 10°, decreases sideslip angles by 3 to 5°, and increases vertical support forces from 1000 to 1450 N, showcasing our algorithm’s superior obstacle avoidance and lane-changing capabilities under dynamic conditions.https://www.mdpi.com/2032-6653/15/11/522path planningartificial potential field algorithmscollision avoidancepath trackinglinear quadratic optimal controller
spellingShingle Chaoxia Zhang
Zhihao Chen
Xingjiao Li
Ting Zhao
Anti-Collision Path Planning and Tracking of Autonomous Vehicle Based on Optimized Artificial Potential Field and Discrete LQR Algorithm
World Electric Vehicle Journal
path planning
artificial potential field algorithms
collision avoidance
path tracking
linear quadratic optimal controller
title Anti-Collision Path Planning and Tracking of Autonomous Vehicle Based on Optimized Artificial Potential Field and Discrete LQR Algorithm
title_full Anti-Collision Path Planning and Tracking of Autonomous Vehicle Based on Optimized Artificial Potential Field and Discrete LQR Algorithm
title_fullStr Anti-Collision Path Planning and Tracking of Autonomous Vehicle Based on Optimized Artificial Potential Field and Discrete LQR Algorithm
title_full_unstemmed Anti-Collision Path Planning and Tracking of Autonomous Vehicle Based on Optimized Artificial Potential Field and Discrete LQR Algorithm
title_short Anti-Collision Path Planning and Tracking of Autonomous Vehicle Based on Optimized Artificial Potential Field and Discrete LQR Algorithm
title_sort anti collision path planning and tracking of autonomous vehicle based on optimized artificial potential field and discrete lqr algorithm
topic path planning
artificial potential field algorithms
collision avoidance
path tracking
linear quadratic optimal controller
url https://www.mdpi.com/2032-6653/15/11/522
work_keys_str_mv AT chaoxiazhang anticollisionpathplanningandtrackingofautonomousvehiclebasedonoptimizedartificialpotentialfieldanddiscretelqralgorithm
AT zhihaochen anticollisionpathplanningandtrackingofautonomousvehiclebasedonoptimizedartificialpotentialfieldanddiscretelqralgorithm
AT xingjiaoli anticollisionpathplanningandtrackingofautonomousvehiclebasedonoptimizedartificialpotentialfieldanddiscretelqralgorithm
AT tingzhao anticollisionpathplanningandtrackingofautonomousvehiclebasedonoptimizedartificialpotentialfieldanddiscretelqralgorithm