Real-time Motion Control with Iterative Optimization and Robustness Analysis for Autonomous Driving

In this paper, a real-time motion controller with iterative optimization and robustness analysis is proposed for an autonomous vehicle. The motion controller is divided into lateral and longitudinal controllers. In this study, this study only considers the lateral controller in which the design is based on linear quadratic regulator (LQR) algorithm and a look-up table is applied to accelerate the computation for varying longitudinal velocity. A complete robustness analysis using Monte Carlo method is carried out to test the motion controller's robustness on parameter uncertainties. Additionally, an iterative control simulation is proposed to adapt the control maneuver command such that the controller's tracking performance is improved. Simulation results are provided to show the effectiveness of the proposed motion control strategy.