Design of robust path tracking controller using model predictive control based on steady state input

This paper presents a robust path tracking controller based on a model predictive control (MPC) with steady-state inputs for disturbance compensation. A conventional MPC-based path tracker has a possibility to diverge due to model uncertainty and disturbance. Particularly, the noise of the sensor measurement can cause a deterioration in path tracking performance. A conventional robust controller is used to compensate for the disturbance. However, the constraints for state and inputs are not explicitly reflected in the conventional robust controllers. Therefore, this study focused on the development of the MPC-based robust path tracker. The proposed controller introduces the steady state solution to improve the robustness of MPC. A double lane change scenario is used to evaluate the proposed algorithm by using the co-simulation environment of MATLAB/Simulink and CarSim. Simulation results show that the proposed method is more robust against an increase in disturbance than the sliding mode controller.