摘要
针对传统粒子群算法(PSO)在机器人路径规划时易陷入早熟、收敛速度较慢和搜索精度较低的问题,提出了一种麻雀优化粒子群算法(SSA-PSO)来解决机器人路径规划问题。将麻雀算法(SSA)的警惕机制与PSO算法的种群结合,对PSO算法中的惯性权重因子、学习因子进行优化,防止算法在快速收敛的同时出现早熟收敛。基于Matlab平台对SSA-PSO算法在栅格地图上进行了仿真实验,并在同一地图上与其他改进粒子群算法的适应度、路径长度、平均规划时间进行了比较。仿真结果表明,相较于PSO算法,SSA-PSO算法在文中3种地图上的路径长度分别缩短了2 m,3.071 m和10 m,性能分别提升了7%,10%和14%。与其他改进粒子群算法相比,SSA-PSO算法的复杂度相对较低,规划的路径最短,路径规划时间也最短。
Aiming at the problems that traditional particle swarm optimization(PSO)is prone to fall into premature,slow convergence and low search accuracy when robot path planning,this paper proposed a sparrow search algorithm and particle swarm optimization(SSA-PSO)algorithm to solve the robot path planning problem.The vigilance mechanism of sparrow search algorithm(SSA)was combined with the population of PSO algorithm to optimize the inertia weight factor and learning factor of PSO algorithm,so as to prevent premature convergence of the algorithm while fast convergence occurs.Finally,the SSA-PSO algorithm was simulated on the grid map based on Matlab platform.In addition,it was compared with other improved particle swarm optimization algorithms on the same map in terms of fitness value,path length and average planning time.The results show that the SSA-PSO algorithm in solving the robot path planning problem has reduced the path length by 2 m,3.071 m and 10 m on maps of different sizes compared with the PSO algorithm,and its performance has improved by 7%,10%and 14%respectively.The SSA-PSO algorithm has relatively low complexity,the shortest path distance and the shortest path planning time.
作者
刘伯威
董小瑞
张志文
屠畅
LIU Bowei;DONG Xiaorui;ZHANG Zhiwen;TU Chang(School of Energy and Power Engineering,North University of China,Taiyuan 030051,China)
出处
《中北大学学报（自然科学版）》
CAS
2023年第4期374-380,共7页
Journal of North University of China(Natural Science Edition)
基金
山西省青年科技研究基金(201901D211208)。
关键词
规划时间
路径规划
适应度
麻雀优化粒子群算法
planning time
path planning
fitness
sparrow optimization particle swarm optimization algorithm
