Distributed resilient fusion filtering for multi-sensor nonlinear singular systems subject to colored measurement noises

Abstract

In this paper, the distributed resilient fusion (DRF) filter is designed for a kind of multi-sensor (MS) nonlinear singular systems with colored measurement noises. The measurement differencing way is used to deal with the colored measurement noises, ensuring that the noises of the measurement output are uncorrelated. During the algorithm implementation, the resilience case that the local filter gain is designed with the certain gain variations is considered, thereby enhancing the system robustness. In this case, our goal is that by using the full-order transformation method, the nonlinear singular system is transformed into an equivalent nonlinear nonsingular one. In addition, the DRF filtering approach is developed in terms of the inverse covariance intersection fusion method, where the local upper bound on the filtering error covariance is deduced and minimized by solving two difference equations and designing the appropriate filter gain, respectively. In the end, the effectiveness of the proposed DRF filtering algorithm is checked by using two numerical examples.