模糊遗传粒子滤波算法研究  

Research on Fuzzy Genetic Particle Filtering Algorithm

作　　者：刘淑波 张园 初俊博 史新鹏 邓加川 LIU Shubo;ZHANG Yuan;CHU Junbo;SHI Xinpeng;Deng Jiachuan(Department of Basic,Dalian Naval Academy,Dalian 116018)

机构地区：[1]海军大连舰艇学院基础部,大连116018

出　　处：《舰船电子工程》2020年第2期18-20,53,共4页Ship Electronic Engineering

基　　金：海军大连舰艇学院科研发展基金“智能粒子滤波算法及其应用研究”资助。

摘　　要：论文针对粒子滤波算法的退化及重采样使得样本枯竭问题,将智能优化算法--模糊遗传算法融入基本的粒子滤波算法,对其进行改进,提出了模糊遗传粒子滤波(FGA-PF)算法。由对非线性机动目标跟踪问题仿真结果得出,该算法与基本的PF算法相比,大幅提高跟踪精度;与基于重要密度函数选择的改进粒子滤波算法--UPF算法相比,提高跟踪精度伴随着大幅缩减计算时间。该算法为粒子滤波的其它智能优化改进提供了工程支持。Aiming at the problem of the sample exhaustion caused by the degradation and resampling of the particle filtering al gorithm,the fuzzy genetic particle filtering algorithm is developed in this paper,which integrates the fuzzy genetic algorithm of the intelligent optimization algorithm into the basic particle filtering algorithm.The simulation results of nonlinear maneuvering target tracking problem show that,compared with the basic PF algorithm,tracking accuracy is significantly improved using this algorithm.Compared with the UPF algorithm,which is improved based on important density function selection of PF algorithm,the tracking ac curacy is improved,and time is reduced significantly.This algorithm provides engineering support for other intelligent optimization improvements of PF algorithm.

关 键 词：粒子滤波算法 退化 重采样 模糊遗传算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]