基于改进人工势场法的自动驾驶路径规划方法  

Automatic Driving Path Planning Based on Improved Artificial Potential Field Method

作　　者：马庆禄[1] 黄光浩 MA Qing-lu;HUANG Guang-hao(School of Transportation,Chongqing Jiaotong University,Chongqing 400074,China)

机构地区：[1]重庆交通大学交通运输学院,重庆400074

出　　处：《计算机仿真》2022年第8期160-165,共6页Computer Simulation

基　　金：国家社会科学基金国家应急管理体系建设研究专项(20VYJ023);重庆市科技局技术预见与制度创新专项(cstc2019jsyjyzysbA0058)。

摘　　要：针对传统人工势场法在路径规划时存在目标不可达和陷入局部极小值导致停滞不前的问题,提出一种改进传统人工势场的路径规划方法,旨在利用距离阈值和斥力调节因子对引力场函数和斥力场函数在避障过程中的影响范围进行分段研究。实验以百度阿波罗自动驾驶小车为动力学原型,并对小车行驶路线和障碍物的校园地图进行抽象符号化处理,利用MATLAB进行仿真的结果显示,改进的自动驾驶车辆的路径规划算法成功解决了传统势场函数的目标不可达问题,路径能够顺利抵达目标位置,并且改进后的路径规划较传统势场在评价值上提高了64%,具有较好的路径规划优势。Aiming at the problems of target unreachable and falling into local minima in path planning of traditional artificial potential field method, an improved path planning method of traditional artificial potential field is proposed, which aims to study the influence range of gravitational field function and repulsive field function in the process of obstacle avoidance by using distance threshold and repulsion adjustment factor. The experiment took Baidu Apollo autonomous car as the dynamic prototype, and carried out abstract symbolic processing on the campus map of the car’s driving route and obstacles. The simulation results based on MATLAB show that the improved path planning algorithm of autonomous vehicles successfully solves the target unreachable problem of the traditional potential field function, and the path can reach the target position smoothly, And the improved path planning is 64% higher than the traditional potential field in the evaluation value, which has better path planning advantages.

关 键 词：人工势场 路径规划 自动驾驶 

分 类 号：TB24[一般工业技术—工程设计测绘]