Research on Cooperative Driving Steering Control for Intelligent Vehicles Based on Lateral Deviation Prediction |
Zengke Qin1, Lie Guo1, Longxin Guan1, Jian Wu2, Pingshu Ge3, Xin Liu2 |
1School of Mechanical Engineering, Dalian University of Technology, Dalian, 116024, China 2School of Mechanical and Automotive Engineering, Liaocheng University, Liaocheng, 252000, China 3College of Mechanical and Electronic Engineering, Dalian Minzu University, Dalian, 116600, China |
|
Received: May 20, 2024; Revised: September 1, 2024 Accepted: September 18, 2024. Published online: November 21, 2024. |
|
|
ABSTRACT |
Lane keeping with cooperative control strategy and conflict reduction between human drivers and intelligent driving controller are the key problems in the lane departure prevention (LDP) system. To address these issues, this paper proposes an intelligent vehicle cooperative driving steering control method based on online lateral deviation prediction. In the proposed method, the gate recurrent unit (GRU) model is designed to predict the lateral deviation. The predicted lateral deviation is used as the cooperative driving coefficient in the fuzzy model, which is able to determine the level of assistance controller intervention based on driver intent, thus reducing the potential conflict. The cooperative controller is used to optimize the linear quadratic regulator (LQR) controller with a genetic algorithm (GA) to enhance the lane keeping ability. Finally, after the optimization of the cooperative control strategy, the torque after the superposition of driver and assistance torque acts on the steering column to accomplish cooperative control. Offline training of the GRU model involved collecting driving data from 52 drivers. The proposed strategy was simulated and analyzed, and hardware-in-the-loop experiments were completed to validate it. The strategy not only completes the task of reducing human–machine conflict time but also effectively reduces vehicle lateral deviation. |
Key Words:
Lane departure prediction · Cooperative control · Neural network · Human–machine conflict · Deviation prediction |
|