Data-Driven Optimal Control for Lateral Stability of Vehicle in Straight-Line Driving via Adaptive Dynamic Programming

Xirui Xu,Xiaoyang Gao,Tieshan Li,Tianpeng Huang
DOI: https://doi.org/10.1109/CSIS-IAC60628.2023.10363941
2023-01-01
Abstract:To increase the vehicle's lateral stability when driving straight ahead and to ensure driving safety, a data-driven optimal controller design for vehicle dynamic system is presented in this work. To begin with, the two-degree-of-freedom dynamics model is used by vehicle dynamics model, which is stated in second-order nonlinear form. Then, the proposed algorithm is applied with unknown model information and structure to find the optimal controller utilizing both recorded and current data of the system. In order to carry out this algorithm, a neural network (NN) is built as a model network to recognize the unidentified system. By using adaptive dynamic programming (ADP), the system can preserve the desired heading angle, coordinate of direction, lateral velocity and yaw rate while simultaneously ensuring that the closed-loop system's signals are all uniform ultimately bounded. Finally, the simulation results show the effectiveness of the proposed controller design method.
What problem does this paper attempt to address?