Path tracking control of unmanned ground vehicles considering the signal time delay
Guoxing Bai,Yu Meng,Li Liu,Guodong Wang,Jianxiu Huang,Haoqing Zheng
DOI: https://doi.org/10.1177/09544070231163348
2023-03-30
Abstract:Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, Ahead of Print. Research on path tracking control of unmanned ground vehicles based on control methods such as model predictive control (MPC) is gradually prosperous. However, the problem caused by the signal time delay has not been taken seriously. It is found that the iterative mechanism of MPC is beneficial for solving this problem, and proposed a solution based on this mechanism. The principle of this solution is to use the control inputs of previous control periods as the inputs of the iteration periods of the prediction model corresponding to the signal time delay. So that the start moment of the iteration period where the first value of the control input sequence of the current control period is located can correspond to the moment when the control input arrives at the vehicle precisely. Based on this solution, a nonlinear MPC (NMPC) controller is designed and verified by co-simulation with MATLAB and Carsim. The simulation results show that the proposed controller can complete the path tracking control in the system with the signal time delay which the traditional NMPC controller and Stanley controller fail, and has high accuracy and real-time performance. In all simulation results, the absolute value of the displacement error does not exceed 0.2622 m, the absolute value of the heading error does not exceed 0.0909 rad, and the solution cost in each control period does not exceed 13 ms. Furthermore, it is also confirmed that the performance of this solution is superior to other solutions dealing with signal time delay.
engineering, mechanical,transportation science & technology