Abstract: Aiming at the inefficiency and premature of intelligent optimization algorithms solving the path planning problem of Automated Guided Vehicle(AGV),an improved Grey Wolf Optimization (GWO) algorithm was proposed for solving AGV path planning problems in complex environment,in which path fine-tuning operator and neighborhood mutation operator were introduced to improve the exploitation capability of GWO.A new initialization algorithm was further introduced to improve the quality of initial population,and an improved path segment and obstacle intersection judgment approach was employed to enhance the improved GWO.A new obstacle avoidance operator was proposed to improve the efficiency of path segments avoiding obstacles,and an AGV path planning prototype platform with multiple intelligent algorithms was developed by MATLAB GUI tools to verify the improved GWO compared with traditional genetic algorithms with single and multi-population strategy under different complex static environments.

Key words: grey wolf optimization algorithm, path fine-tuning operator, neighborhood mutation operator, genetic algorithms, initialization method

摘要： 针对智能优化算法求解自动导引小车路径规划问题效率低且易早熟的缺陷,提出一种用于求解复杂环境下自动导引小车路径规划问题的改进灰狼优化算法。算法引入路径微调算子和邻域变异算子来提高灰狼优化算法的局部开发能力,又引入新的初始解生成算法提高初始种群的质量;采用改进的路径片段与障碍物相交判断算法来提高算法的运行效率,再采用新的避障算子来提高路径片段避开障碍物的效率。基于MATLAB GUI开发工具开发了带有多种智能优化算法的自动导引小车路径规划仿真原型平台,并与单种群遗传算法、多种群遗传算法和改进遗传算法进行对比,验证了改进灰狼优化算法求解自动导引小车路径规划的有效性。

关键词: 灰狼优化算法, 路径微调算子, 邻域变异算子, 遗传算法, 初始解生成算法