自适应强跟踪容积卡尔曼滤波算法  

An adaptive strong tracking cubature Kalman filter

作　　者：赵利强[1] 罗达灿[1] 王建林[1] 于涛[1] 

机构地区：[1]北京化工大学信息科学与技术学院,北京100029

出　　处：《北京化工大学学报（自然科学版）》2013年第3期98-103,共6页Journal of Beijing University of Chemical Technology(Natural Science Edition)

基　　金：中央高校基本科研业务费(ZY1110)

摘　　要：提出了一种自适应强跟踪容积卡尔曼滤波算法(ASTSCKF),该算法在平方根容积卡尔曼滤波算法(SCKF)步骤中引入强跟踪滤波器(STF),通过渐消因子在线修正一步预测误差协方差矩阵,强迫输出残差序列正交,使得算法具有应对系统状态突变等不确定因素的能力,增强了算法的鲁棒性;结合改进渐消记忆时变噪声统计估计器,对噪声方差阵进行实时在线估计,有效解决了SCKF算法由于噪声统计不准确、未知或时变性带来的滤波发散问题,使其具有应对噪声变化的自适应能力。仿真实验结果表明:ASTSCKF算法在系统状态发生突变并且噪声变化的情况下,能够表现出良好的滤波性能,较SCKF算法有更强的鲁棒性以及噪声变化的自适应性。An improved square-root cubature Kalman filter（SCKF）-based strong tracking filter and noise statistic estimator has been proposed.By introducing a strong tracking filter（STF） into the SCKF and modifying the predicted error covariance with a fading factor,the residual sequence was forced to be orthogonal.The resulting adaptive strong tracking cubature Kalman filter（ASTSCKF） has the capability to deal with uncertainty factors such as sudden changes in system states and thus the robustness of the algorithm was enhanced.Combining the improved fading memory time-varying noise statistic estimator and estimating on-line the noise covariance matrix in real-time,the algorithm can effectively overcome the filtering divergence problems caused by unknown,uncertain or time-dependent noise statistics.Therefore,it also has noise-changing adaptive capacity.The simulated experimental results indicate that ASTSCKF can still present good filtering performance when system states suddenly change and noise changes occur.ASTSCKF has greater robustness and noise-changing adaptive capacity than SCKF.

关 键 词：容积卡尔曼滤波算法 强跟踪滤波器 渐消因子 噪声统计估计器 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]