协同进化粒子群滚动优化的机器人路径规划  

Rolling Path Planning of Mobile Robot Based on Co-evolutionary Particle Swarm Optimization

作　　者：陆萍[1] 董虎胜[1,2] 钟宝江[2] 

机构地区：[1]苏州经贸职业技术学院信息系,江苏苏州215009 [2]苏州大学计算机科学与技术学院,江苏苏州215009

出　　处：《计算机测量与控制》2013年第11期3128-3130,3147,共4页Computer Measurement &Control

基　　金：国家自然科学基金(61075040);江苏省省属高校自然科学研究重大项目(10KJA520047);江苏省自然科学基金(BK2012645);江苏省高校"青蓝工程"优秀青年教师项目

摘　　要：提出了一种在静态未知环境下,采用协同进化粒子群滚动优化的机器人路径规划方法;该方法首先对协同进化粒子群算法进行了改进,在多子群协同进化中引入群体质心与优胜劣汰的进化策略,提高了种群的搜索能力;在机器人向目标点前进中遇到障碍时,采用协同粒子群优化算法进行避障,机器人前进路径不断动态修改,直至到达目标点;在对典型多障碍物环境、狭缝及凹型障碍物等各种复杂环境的测试中,采用该方法始终能够规划出有效的避障路径,与标准粒子群等算法相比,改进后算法在避障处理中具有更快的收敛速度与更优的搜索精度,规划的路径更有效。A rolling path planning method of mobile robot based on co evolutionary particle swarm optimization is presented, which works in a static environment where the global information is unknown. The coevolutionary particle swarm optimization algorithm is im proved by adopting population center and the strategy of survival of the fittest. As a result, the population can gain better searching capacity. When the way is blocked during the robot moving towards the target, coevolutionary particle swarm optimization is used to avoid the obsta cles. So the path for the robot is modified dynamically until the goal is found. Simulation results show that the algorithm can obtain available path in typical multiobstacles environment, even in more complex situation with the confirmation of slip or concave can it work well. Com pared to PSO and CPSO, better path can be acquired based on the improved CPSO algorithm with faster convergence and more precision.

关 键 词：粒子群优化 协同进化 路径规划 移动机器人 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]