ABSTRACT
In this research, a hierarchical controller based on model predictive control (MPC) for path planning and path following is developed. In the path planning level, obstacles’ information is incorporated in the MPC framework. A new distance-direction based method is proposed for obstacle avoidance, which considers both the distance and moving direction of the obstacle. An alternative distance based method is analyzed for comparison. In the path following level, a bicycle model for MPC is utilized to improve the tracking of the reference from the higher level. Simulation results show that the distance-direction based method can better reflect the threat of the obstacle and generate more reasonable trajectories compared to the alternative distance based method.
