Interactive Trajectory Planning of Autonomous Vehicle Based on State Lattice and Potential Field

In order to solve the problem of trajectory planning in complex traffic scenarios, such as falling into local optimization and generating trajectory machinery, it will lead to low acceptance degree of users. We propose an interactive trajectory planning framework based on the combination of state lattice and potential field. The lattice search method is used to discretization the driving environment and reduce the solution space. At the same time, by constructing an interactive road environment potential field, considering changes in traffic participants' intentions, the trajectory is updated in a timely manner, and the trajectory is optimized based on the scene. This article conducts real-world experiments in various challenging and complex scenarios, and the experimental results verify that the trajectory planner can generate safe, comfortable, and flexible trajectories with high timeliness.