带乘性噪声广义系统的观测噪声最优估计算法  

Optimal Estimation of Measurement Noise for Singular Systems with Multiplicative Noise

作　　者：褚东升[1] 于宇 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）》2009年第5期1115-1118,1124,共5页Periodical of Ocean University of China

基　　金：国家高技术研究发展计划项目(2006AA09Z115);国家自然科学基金项目(60704023)资助

摘　　要：研究带乘性噪声广义系统的观测噪声最优估计问题。在假设系统正则的情况下,针对乘性噪声为一般随机矩阵即各观测通道乘性噪声同时刻相关的情况,通过受限等价变换和状态扩维的方法,给出了线性最小方差意义下的观测噪声最优滤波算法和最优平滑估计算法。数字仿真结果表明了算法的有效性。This paper provides a filtering algorithm and a smoothing algorithm of measurement noise in singular systems with multiplicative noise, both of which are optimal in the sense of linear minimum-variance. Assuming the singular systems are regular, transform them into equivalent nonsingular systems by restricted equivalent transformation and state augmentation. The whole transformation changes most parameters except for the measurement noise. Hence, the filtering algorithm, as well as the smoothing one, can be acquired in the nonsingular systems. Simulations are illustrated to show the validity of the algorithms.

关 键 词：广义系统 乘性噪声 线性最小方差估计 观测噪声估计器 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]