Simultaneous Stability and Path Following Control for 4WIS4WID Autonomous Vehicles Based on Computationally Efficient Offset Free MPC
Linhe Ge,Yang Zhao,Shouren Zhong,Zitong Shan,Fangwu Ma,Zhiwu Han,Konghui Guo
DOI: https://doi.org/10.1007/s12555-022-0326-y
IF: 2.964
2023-01-01
International Journal of Control Automation and Systems
Abstract:The steady-state error problem of autonomous vehicle MPC-based motion control has not been effectively solved for a long time. This problem is more serious for lateral and longitudinal coupling control problems of vehicles with over-actuated configurations. Based on our newly designed general offset free MPC (OF-MPC) solver and the TMeasy tire model, a steady-state error free control strategy for simultaneous stability and path following control of four-wheel steering and four-wheel drive vehicles is proposed. OF-MPC uses the disturbances term to describe the model mismatch and external disturbances, then uses the Kalman filter to observe the disturbances, and finally considers the disturbances in the optimization stage to realize the control without steady-state error. Realtime simulation results show that OF-MPC can solve model mismatch and external disturbances problems, and the steady-state error free control is realized. The simulation results of the double lane change maneuver show that the OF-MPC dynamic control performance is also better than the traditional MPC (TRA-MPC), which is more obvious when the vehicle is at the stability boundary and under various constant or time-varying disturbances. Regardless of the dimensions and complex constraints of this problem, real-time performance is still guaranteed, thanks to the proposed OF-MPC. When the horizon length is 100, the average time consumption is only about 15 milliseconds.