粒子群法在非完整移动操作机运动规划中的应用  

Motion planning for nonholonomic mobile manipulators using particle swarm optimization

作　　者：付宜利[1] 闫庆辉[1] 马玉林[1] 

机构地区：[1]哈尔滨工业大学现代生产技术中心,哈尔滨150001

出　　处：《哈尔滨工业大学学报》2008年第5期735-739,共5页Journal of Harbin Institute of Technology

基　　金：黑龙江省留学归国基金资助项目(20010701018)

摘　　要：研究了给定末端点路径条件下非完整轮式移动操作机运动规划问题,提出了基于粒子群法的运动规划方法.将移动操作机的末端点路径离散化.移动操作机从初始位形出发,利用粒子群法来搜索下一个最优位形,直至达到末端点路径终点.选择平台的控制输入矢量作为粒子,使得非完整约束条件自然被满足.在障碍环境下,将具有启发信息的势函数引入粒子的适应度函数,有效地实现了系统的无碰运动.通过计算机仿真验证了该算法的有效性.The motion planning of nonholonomic wheeled mobile manipulators （WMMs） along given end-effector （EE） paths was studied. A motion planning approach using the particle swarm optimization （PSO） was presented. The EE path was divided into n discrete points. The WMM start from an initial configuration, and then the next optimal configuration of the WMM was found using PSO, until the WMM reach the end of the EE path. The control input vector of the platform was chosen as the particle, and the nonholonomic constrains can be naturally met. In the presence of obstacles, a potential function was introduced into the fitness function ; the free motion of the WMM was effectively realized. Simulation results were given to demonstrate the effectiveness of the proposed method.

关 键 词：轮式移动操作机 非完整约束 粒子群优化算法 运动规划 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]