基于粒子群优化的移动机器人MCL全局定位算法  

MCL Global Localization Algorithm for Mobile Robot Based on Particle Swarm Optimization

作　　者：龚建铭 范阳 周建辉 GONG Jian-ming;FAN Yang;ZHO Jian-hui(Shunde Graduate School,Beijing University of Science and Technology,Foshan Guangdong 528399,China;School of Mechanical Engineering,University of Science and Technology Beijing,Beijing 100083,China;Department of Mechanical Engineering,Qian’an College of North China University of Technology,Qian’an Hebei 064400,China)

机构地区：[1]北京科技大学顺德研究生院,广东佛山528399 [2]北京科技大学机械工程学院,北京100083 [3]华北理工大学迁安学院机械工程系,河北迁安064400

出　　处：《湖北第二师范学院学报》2022年第8期38-44,共7页Journal of Hubei University of Education

摘　　要：针对室内移动机器人运行时可能发生碰撞、打滑和被人为搬运等情况而导致定位失效的问题,提出了一种基于粒子群优化的移动机器人MCL(Monte Carlo Localization)全局定位算法。基于MCL算法框架,通过定义粒子有效数目的方式来识别测量更新后定位算法的准确性,并利用粒子群优化算法将粒子集的位置向观测概率高的方向调整,调整后粒子集的姿态调整依靠重要性采样完成粒子集的更新过程。在Matlab下的仿真和实验结果表明改进后的MCL全局定位算法在位姿估计失效后恢复定位的性能明显优于MCL算法,可为MCL算法应用于机器人全局定位提供依据。Aiming at the problem of localization failure of indoor mobile robot caused by collision,slipping and manual handling during operation,a MCL(Monte Carlo localization)global localization algorithm for mobile robots based on particle swarm optimization(PSO)is proposed.Based on the framework of MCL algorithm,the accuracy of localization algorithm after measurement updating is identified by defining the effective number of particles,and the position of particle set is adjusted to the direction with high observation probability by using particle swarm optimization algorithm,and the attitude adjustment of adjusted particle set relies on importance sampling to complete the updating process of particle set.The simulation and experimental results in MATLAB show that the performance of the improved MCL algorithm is obviously better than that of MCL algorithm after the failure of pose estimation,which can be used to guide the improved MCL algorithm.

关 键 词：移动机器人 粒子群优化 MCL 全局定位 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]