带乘性噪声系统状态滤波的自适应算法  

Adaptive State Filtering Algorithm for System with Multiplicative Noise

作　　者：褚东升[1] 宁云磊[1] 张玲[1] 

机构地区：[1]中国海洋大学工程学院,山东青岛266100

出　　处：《中国海洋大学学报（自然科学版）》2011年第12期109-113,117,共6页Periodical of Ocean University of China

基　　金：国家自然科学基金项目(60704023);山东省自然科学基金项目(ZR2010DQ003)资助

摘　　要：带乘性噪声系统由于其广泛的适用性,一直成为研究的热点。针对带乘性噪声系统状态最优估计的自适应算法进行研究,探讨在噪声服从平稳正态分布情况下,对未知动态噪声方差阵与观测噪声方差阵的辨识问题。在证明带乘性噪声系统新息在稳态时和线性系统新息有着相似稳定特性的前提下,通过对线性系统辨识方法的改进,完成对带乘性噪声系统噪声方差阵的辨识,并利用新息特性对该方法进行进一步改进,以提高辨识精度;最后通过仿真验证该方法的有效性。The system with multiplicative noise has always been the global research hot spot for its wide applicability.This paper aims at finding an adaptive algorithm for the system with multiplicative noise to identify the covariance of the measurement noise and dynamic noise with Gaussian distribution.Since the characteristic of the innovation of the system with multiplicative noise is similar to that of the linear system during the steady state,the conventional identification method is revised to solve the state estimation problem of the nonlinear system.Furthermore this paper proposes further improvement to enhance the accuracy based on the characteristic of the innovation.Finally,the simulation results validate the method.

关 键 词：乘性噪声 状态滤波 自适应 新息协方差 

分 类 号：TP31[自动化与计算机技术—计算机软件与理论]