非高斯噪声下的参数自适应高斯混合CQKF算法  

A Parameter Adaptive Gaussian Mixture CQKF Algorithm Under Non-Gaussian Noise

作　　者：孟东[1] 缪玲娟[1] 邵海俊 沈军[1] MENG Dong;MIAO Ling-juan;SHAO Hai-jun;SHEN Jun(School of Automation,Beijing Institute of Technology,Beijing 100081,China)

机构地区：[1]北京理工大学自动化学院,北京100081

出　　处：《北京理工大学学报》2018年第10期1079-1084,共6页Transactions of Beijing Institute of Technology

基　　金：国家自然科学基金资助项目(61153002;61473039)

摘　　要：研究非高斯噪声环境下的高斯混合滤波方法,进行纯方位跟踪系统的目标跟踪.利用改进的参数自适应方法,调整位移参数的大小,从而修正了高斯混合模型,提出了在非高斯噪声下的参数自适应高斯混合CQKF算法;基于非高斯噪声下的离散系统模型,分析了高斯混合CQKF算法中建模过程的局限性,并结合初值优化方法,提出了利用参数自适应方法修正高斯混合滤波模型的方法,从而克服了高斯混合滤波的局限性,提高了滤波精度.仿真实验表明在非高斯噪声下参数自适应高斯混合CQKF算法比原算法有更高的滤波精度.A Gaussian mixture filtering method under non-Gaussian noise environment was studied,and the target tracking of pure azimuth tracking system was carried out.Firstly,a modified parameter adaptive method was used to adjust the size of the displacement parameter,so the Gaussian mixture model could be modified.The parameter adaptive Gaussian mixture CQKF algorithm(PGM-ACQKF)under non-Gaussian noise was proposed.Then based on the discrete system model under non-Gaussian noise,the limitations of the modeling process in the Gaussian mixture CQKF(GM-CQKF)was analyzed.Combining with the initial optimization method,a method to modify the Gaussian mixture model was proposed based on parameter adaptive method.Thus the limitations of GM-CQKF could be overcome and the filtering accuracy could be improved.The simulation results show the effectiveness of the proposed algorithm,which proves that the PGM-ACQKF has higher filtering accuracy than the original algorithm under non-Gaussian noise.

关 键 词：高斯混合模型 容积卡尔曼滤波算法 参数自适应方法 初值优化 

分 类 号：TN927.2[电子电信—通信与信息系统]