改进自适应遗传算法在移动机器人路径规划中的应用  

Application of improved adaptive genetic algorithm in mobile robot path planning

作　　者：王雷 李明 

机构地区：[1]安徽工程大学机械与汽车工程学院,安徽芜湖241000

出　　处：《南京理工大学学报》2017年第5期627-633,共7页Journal of Nanjing University of Science and Technology

基　　金：国家自然科学基金(51305001);安徽省自然科学基金(1708085ME129);安徽省科技攻关项目(1604A0902183);安徽省高校优秀青年人才支持计划重点项目(gxyq ZD2016125)

摘　　要：为了克服基本遗传算法在求解移动机器人路径规划问题中存在的收敛速度慢、易陷入局部最优等不足,该文提出了一种改进的自适应遗传算法。采用人工势场法对种群进行初始化,设计了自适应交叉和变异概率。同时,采用混合选择方式改善了基本遗传算法收敛速度慢和早熟的现象,提高了算法的进化效率。栅格环境下的仿真实验证明了该文算法在移动机器人路径规划中的可行性和有效性。In order to deal with the problem of the slow convergence speed and local optima of the basic genetic algorithm （ GA） in solving the robot path planning, an improved adaptive genetic algorithm（IGA） is proposed here. An artificial potential field method is employed to create the initial population,and the adaptive crossover probability and the mutation probability are designed. Meanwhile,a hybrid selection method is adopted to improve the convergence speed and the evolutionary efficiency and overcome the premature phenomenon of the basic genetic algorithmobviously. Some experiments under the grid environment verify the feasibility and effectiveness of the improved adaptive genetic algorithm in mobile robot path planning.

关 键 词：自适应遗传算法 路彳圣规划 移动机器人 人工势场 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]