Global Trajectory Planning for Autonomous Race-Car with Vehicle Dynamics Model-Based Optimization

The trajectory of a race car has a significant impact on the lap time. Therefore, optimizing the racing trajectory is a crucial aspect of winning a race. This paper presents a global optimization method based on a vehicle dynamics model to address this issue. A global method considers all parts of the track and can achieve a better lap time than methods that only take one part of track into account. In this study, a more complex vehicle dynamics model is used to represent the actual racing vehicle at high velocities or at its handling limits. By incorporating constraints in critical dynamics states, this method could also be altered to simulate different vehicles, improve lap time, and generate a safe trajectory when the model does not match the real track condition well.