摘要
针对滤波过程中噪声统计特性不准确及非零均值噪声统计特性的情况,本文依据H∞卡尔曼滤波和容积卡尔曼滤波理论,设计了一种H∞模糊自适应容积卡尔曼滤波方法,有效地提高滤波的精度以及对系统未建模动态的鲁棒性。考虑到容积卡尔曼滤波过程中需要已知噪声的先验统计特性的情况,提出了一种模糊自适应方法对系统噪声和测量噪声进行估计,从而提高滤波的稳定性和收敛的快速性。通过仿真实验表明:本文提出的H∞自适应容积卡尔曼滤波能够对噪声特性进行有效的估计,在系统存在参数摄动的情况下具有更高的鲁棒性。
Considering the inaccurate characteristics of noise statistics and the characteristics of non-zero mean noise statistics in the filtering process,in this paper,the H∞fuzzy adaptive cubature Kalman filtering is designed based on the cubature Kalman filtering and H∞filtering theory.Thus,the filter accuracy and robustness of the system unmodeled dynamics are improved effectively.Considering that a priori statistics characteristics of the noise that has been known are necessary during the cubature Kalman filtering process,a fuzzy adaptive method is proposed for estimation of the system noise and measurement noise,so as to improve stability of the filtering and the rapidity of convergence.The simulation results demonstrate that the H∞fuzzy adaptive cubature Kalman filtering is able to effectively estimate the noise characteristics and is more robust under the system perturbations.
作者
刘胜
牛鸿敏
张兰勇
郭晓杰
LIU Sheng;NIU Hongmin;ZHANG Lanyong;GUO Xiaojie(College of Automation,Institute of Automation Engineering Harbin Engineering University,Harbin 150001,China)
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
北大核心
2020年第3期404-410,共7页
Journal of Harbin Engineering University
基金
国家自然科学基金项目(51579047)
黑龙江省自然科学基金项目(QC2017048)
哈尔滨市自然科学基金项目(2016RAQXJ077)
中央高校基础科研业务费(3072019CF407).
关键词
H∞滤波
容积卡尔曼滤波
非线性滤波
模糊规则
自适应算法
噪声统计估计
线性化
鲁棒性
H_∞ filtering
cubature Kalman filtering
nonlinear filtering
fuzzy rules
adaptive algorithm
noise statistics estimator
linearization
robustness