针对粒子滤波中重采样导致粒子多样性减弱造成的滤波精度下降问题,给出了一种基于模糊支持度采样的改进粒子滤波算法;该算法在重采样过程后,首先根据MCMC(Markov Chain Monte Carlo)原理抽取候选粒子,然后依据重采样粒子和候选粒子自身数据中的蕴含信息,并结合模糊理论构建支持度函数和支持度矩阵,以充分地提取数据中的有效信息,在增强粒子多样性的同时实现其对于粒子的优选;最后仿真结果表明,该算法可有效地提高对于系统状态的估计精度。
By the analysis that the degeneracy of particles diversity in the course of re-sampling causes the descent of filtering precision,an improved particle filtering algorithm based on fuzzy support degree sampling is proposed.After the re-sampling process,the new algorithm firstly extracts candidate particles based on Markov Chain Monte Carlo principle,and then,according to the information implicated from the re-sampling and candidate particles and fuzzy theory establishes support degree function and matrix to make fully use of effective information.Thereby the increasing of the diversity of particles and the optimizing selection of particles are realized.Finally,simulation results show the method can effectively improve state estimation precision.