基于分层PSO的智能车辆超车路径规划

Overtaking Lane Hierarchical Planning for Intelligent Vehicles Based on Multi-Objective Particle Swarm Optimization

针对自动驾驶智能车辆安全超车问题,提出一种基于分层优化算法的路径规划方案。该方案采用三层结构获得可行、安全的优化路径。首先,通过分析驾驶员超车操作流程,综合路况环境、车辆结构等约束条件,构建超车换道模型,进而采用三角分解法构建智能车辆的可运行空间。其次,在可运行空间内,采用Dijkstras算法生成一条无碰撞路径。最后,利用超车换道模型构造优化目标函数和约束条件,通过多目标粒子群优化的方式获得最优路径,并采用Pareto优先规则提高优化算法的更新速度。利用CarSim8.10构造车辆动力和运动学模型,基于Simulink仿真平台开展实验验证。仿真结果表明,该方案能够保证智能车辆实现安全、有效地超车。该方案具有良好的可扩展性能,对于道路拥塞、路况不明朗等复杂行车条件下的安全超车提供一定的方法参考和借鉴。

In order to improve the driving safety of intelligent vehicles,a hierarchical path planning method is proposed in this paper.This method uses a three-level structure to obtain a feasible and safe optimized path in overtaking vehicles.Firstly,by analyzing the laws of overtaking behaviors and integrating the constraints such as road conditions and vehicle structure,the overtaking lane changing model is constructed and the trigonometric decomposition method is used to construct the operational space of the intelligent vehicle.Second,the Dijkstras algorithm is used to find the collision-free path.Finally,an objective function and constraint condition are constructed by the overtaking lane changing model.A multi-objective particle swarm optimization method with accelerated updating method is proposed based on the Pareto priority rule.The optimal path is obtained by optimizing the objective function under constraints.Using MATLAB,a model-based path tracking control Simulink simulation platform,where the vehicle dynamics and kinematics model is constructed by CarSim8.10,is built.The simulation results show that the improved path planning model achieves the overtaking behavior of intelligent vehicles safely and effectively in an optimal way.The program has good scalability and provides certain methodological reference for safety overtaking under complicated traffic conditions such as road congestion and unclear road conditions.