MPC-Based Path Planning for Lane Changing in Highway Environment

As one of the most basic behaviors of vehicles in the intelligent transportation system, the lane changing in highway environment influences the safety and efficiency of driving. As a result, it has always been paid much attention by researchers. In the process of the lane changing, it is necessary to consider the kinematic constraints and optimization of safety and rapidity. The existed geometrical methods mainly use curves to fit the path without considering the kinematic constraints of the vehicle. In this paper, we study and design the overall frame of generating the path for lane changing using model predictive control. Model predictive control shows its advantage in solving multi-constraint optimization and predicting the future states. Therefore it is used to optimize the rapidity and smoothness of the lane changing process. Software simulation proves that the trajectory planned by the model predictive control has better smoothness than the trajectory planned by the geometrical method. Finally, a lane changing experiment is conducted on the micro vehicle-based hardware simulation platform. The experimental result further verifies the feasibility of the proposed path planning method.