Real-time Trajectory optimization for Autonomous Vehicle Racing using Sequential Linearization

摘要

This paper presents a real-time Model Predictive Controller (MPC) for racing trajectory optimization. The vehicle must respect its dynamic limitations and the track boundaries while simultaneously minimizing lap times. Due to the track boundary constraints, the feasible set of the optimization problem is non-convex. This paper presents the method of sequential linearization to solve this non-convex optimization problem.