论文部分内容阅读
本文研究了具有丢失观测的多传感器线性离散随机不确定系统的最优线性估计问题,其中不同的传感器具有不同的丢失率.首先将乘性噪声转化为加性噪声,然后基于矩阵满秩分解和加权最小二乘理论,提出了具有较小计算负担的加权观测融合估计算法.分析了加权观测融合估计算法的稳态特性,给出了稳态存在的一个充分条件.所提出的加权观测融合估值器与集中式融合估值器具有相同的精度,即具有全局最优性.仿真研究验证了算法的有效性.
In this paper, we study the optimal linear estimation problem for discrete multisensor systems with missing observations, in which different sensors have different loss rates.Firstly, the multiplicative noise is transformed into additive noise, and then based on the matrix full rank decomposition and Weighted least squares theory, a weighted observation fusion estimation algorithm with a small computational burden is proposed.The steady state characteristics of the weighted observation fusion estimation algorithm are analyzed and a sufficient condition for existence of steady state is given. The proposed weighted observation fusion estimation The device has the same precision as the centralized fusion estimator, ie, it has the global optimality. Simulation studies verify the effectiveness of the algorithm.