粒子滤波算法综述_胡士强.pdfVIP

第20 卷第4 期 　控　制　与　决　策 2005 年4 月 Vol. 20 No .4 　Control and Decision　 　 Apr . 2005 : 1001-0920(2005) 04-0361-05 1, 2 1 胡士强 , 敬忠良 ( 1. , 200030; 2. , 050054) : 对粒子滤波算法的原理和应用进行综述. 首先针对非线性非高斯 统的状态滤波问题, 阐述粒子滤波的原 理; 然后在分析采样-重要性-重采样算法基础上, 讨论粒子滤波算法存在的主要问题和改进手段; 最后从概率密度函 数的角度出发, 将粒子滤波方法与其他非线性滤波算法进行比较, 阐明了粒子滤波的适应性, 给出了粒子滤波在 一些 研究领域中的应用, 并展望了其未来发展方向. : 粒子滤波; 概率密度; 非线性滤波; 算法 : T P14: A Overview of particle filter algorithm 1, 2 1 H U Shi-qiang , JIN G Zhong-liang ( 1. Department of A tomation, Shanghai Jiaotong Universit y, Shanghai 200030, China; 2. Depart ment of A tomation, Hebei University of Science and Technology , Shijiazh ang 050054, China. Correspondent: HU Shi- , - : @ . . . ) qiang E mail sqh mail sjt ed cn Abstract: The Algorithm and applications related to part icle filter are s rveyed. Aiming at the nonlinear/ non- Ga ssian filter problem , the generic ideas of particle filt er are given, based on the analysis of st andard algorithm of - - , sam pling importance resampling filter the problems of particle filter are disc ssed and some improvement met hods . , - are ill strat ed From view of probabilit y density f nction the com parisons betw een part icle filter and others non linear filter algorithms and applicability are int rod ced, some applications in t he developed areas are reviewed, Finally , f rther research direct ions are pointed o t. : ; ; - ; Key words particle filter probabilit y density non linear filt er algorithm 1引　　言