关键词: 智能车, 越野环境, 势能场, 概率图, 路径规划

Abstract: Path planning in complex off-road environment is a key technology to realize the autonomous driving of intelligent vehicle. There are obstacles, environmental threats and terrains that affect vehicle movement in off-road environment. In the traditional path planning methods, the shortest path length and time cost are usually taken as the optimization goal, which makes it difficult to plan a feasible and safe driving path in complex off-road environment. An artificial potential field based probabilistic roadmap (APF-PRM) algorithm is proposed to solve the problem. The potential field algorithm is used to model the off-road environment and evaluate the vehicle traffic risk, and then the probabilistic roadmap method is used to conduct the path planning with multi-dimensional traffic cost between path nodes as the goal. Considering the dynamic characteristic of the vehicle, a dynamic curvature smoothing method is used to optimize the vehicle trajectory. Finally, the APF-PRM algorithm is used to conduct the path planning in a simulated off-road environment. The simulated results show that the APF-PRM algorithm utilizes the artificial potential field algorithm to integrate the obstacles, environmental threats and road conditions in the off-road environment in the process of path planning; the probabilistic roadmap method is used to establish a multi-dimensional traffic cost evaluation matrix among the sampling points; and a feasible, safe and efficient path is generated under complex off-road conditions, which provides a multi-objective optimization path planning method for intelligent vehicles.

Key words: intelligentvehicle, off-roadenvironment, potentialfield, probabilisticroadmap, pathplanning