基于元胞遗传算法的移动机器人路径规划  

Path Planning of a Mobile Robot Based on Cellular Genetic Algorithm

作　　者：张屹[1] 刘铮[1] 胡方军[1] 詹腾[1] 丁昌鹏[1] 

机构地区：[1]三峡大学机械与材料学院,湖北宜昌443002

出　　处：《机床与液压》2014年第9期17-20,24,共5页Machine Tool & Hydraulics

基　　金：国家自然科学基金项目(51275274);三峡大学研究生科研创新基金项目(2012CX031)

摘　　要：在移动机器人路径规划问题中,环境建模约束定义难,遗传算法求解易陷入局部收敛,针对上述问题通过建立栅格坐标、栅格序号和栅格状态三者之间的关系,简化了障碍物约束和有效路径判断,同时引入多样性保持较好的元胞遗传算法,使用定长实数编码对生成的路径进行优化。仿真实验表明,由于算法具备较好的隐性迁移机制,保持了解的多样性,提高了算法收敛效率,使移动机器人路径规划问题得到了有效解决。When solving the problem of path planning of a mobile robot,defining constrains in environment modeling was hard, and was easily fallen into the local convergence by traditional genetic algorithm. By aimed at the problem above,and established the re-lationship among grid coordinates,grid number and grid state,the constrains definition of obstacles and effective path judging were simplified,at the same time,the cellular genetic algorithm with better diversity maintaining was also brought in,while optimizing the path by using fixed-length real number encoding. Finally,the simulation results show that the algorithm maintains the better diversity and improves the efficiency of the convergence because of the implicit mechanism of migration of the algorithm,which effectively solves the problem of path planning of the mobile robot.

关 键 词：元胞邻居 遗传算法 移动机器人 环境建模 路径规划 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]