带不确定方差乘性和加性噪声系统鲁棒加权 融合稳态Kalman估值器

本文研究带不确定方差乘性和加性噪声和带状态相依及噪声相依乘性噪声的多传感器系统鲁棒加权融合估计问题.通过引入虚拟噪声补偿乘性噪声的不确定性,将原系统化为带确定参数和不确定加性噪声方差的系统,进而利用Lyapunov方程方法提出在统一框架下的按对角阵加权融合极大极小鲁棒稳态Kalman估值器(预报器、滤波器和平滑器),其中基于预报器设计滤波器和平滑器,并给出每个融合器的实际估值误差方差的最小上界.证明了融合器的鲁棒精度高于每个局部估值器的鲁棒精度.应用于不间断电源(uninterruptible power system,UPS)系统鲁棒融合滤波的仿真例子说明了所提结果的正确性和有效性.

Robust weighted fusion steady-state Kalman estimators for systems with uncertain-variance multiplicative and additive noises

The robust weighted fusion estimation problem is studied for the multi-sensor systems with uncertain-variance multiplicativeand additive noises and with state-dependent and noise-dependent multiplicative noises in this paper. By introducing the fictitiousnoises to compensate the uncertainties of multiplicative noises, the original system is converted into one with deterministic parametersand uncertain additive noise variances. By the Lyapunov equation approach, the unified mini-max robust fusion steady-state Kalmanestimators (predictor, filter and smoother) weighted by diagonal matrices are presented, where the filter and smoother are designed basedon the predictor, and the minimal upper bound of actual estimation error variances of each fuser is given. It is proved that the robust accuraciesof fusers are higher than that of each local estimator. A simulation example applied to robust fusion filtering of uninterruptiblepower system (UPS) shows the correctness and effectiveness of the proposed results