Framework of integrating trajectory replanning with tracking for self-driving cars

A framework of integrating local trajectory replanning with trajectory tracking is constructed to address the problem of emergency obstacle avoidance for self-driving cars during driving. The framework adds obstacle avoidance trajectory replanning layer above the trajectory tracking layer. The trajectory replanning layer uses behavioral dynamics to replan the local trajectory of obstacle avoidance on the basis of the information of the obstacle and the expected path, and the trajectory tracking layer uses model predictive control to achieve trajectory tracking control. Simulink and CarSim simulations are conducted in the case where moving obstacles exist. Simulation experiment results show that this framework is satisfactory for planning obstacle avoidance trajectory and tracking under emergency conditions.