蒙特卡洛粒子滤波算法应用研究  

Application Research of Monte Carlo Particle Filter Algorithm

作　　者：韩雨薇 郑安迪 朱俞竹 王璐瑶 董明泽 HAN Yuwei;ZHENG Andi;ZHU Yuzhu;WANG Luyao;DONG Mingze(Liangxin College,China Jiliang University,Hangzhou,Zhejiang Province,310018 China;College of Mechanical and Electrical Engineering,China Jiliang University,Hangzhou,Zhejiang Province,310018 China)

机构地区：[1]中国计量大学量新学院,浙江杭州310018 [2]中国计量大学机电工程学院,浙江杭州310018

出　　处：《科技资讯》2020年第35期12-13,16,共3页Science & Technology Information

摘　　要：该文采用了在移动机器人定位中应用较为广泛的粒子滤波器算法,该算法基于蒙特卡洛方法,将捕捉机器人位姿状态的问题转化为概率问题,该文介绍了贝叶斯滤波理论的推导过程,以及由其发展而来的贝叶斯重要性采样理论,然后简要概括了如何利用重采样方法优化了采样粒子的权值,进而得到较优的粒子权值。该理论在目标追踪及机器人定位等领域中应用较为广泛。This article adopted has been widely applied in mobile robot localization and the particle filter algorithm,the algorithm based on monte carlo method,to capture the state of the robot pose problem into a probability problem,this paper introduced the derivation process of the theory of bayesian filtering,and by the development of bayesian importance sampling theory,and then briefly summarizes how heavy sampling method was used to optimize the weights of sample particle,then get the optimal weights of particles.The theory has been widely used in target tracking and robot positioning.

关 键 词：粒子滤波 蒙特卡洛 机器人定位 贝叶斯滤波理论 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]