A Distributed Robust Fusion Filter Design with Randomly Occurring Parameter Uncertainty and Sensor Resolutions

This paper discusses the distributed fusion (DF) filtering problem for a class of multi-sensor (MS) nonlinear systems undergo randomly occurring parameter uncertainty (ROPU) and sensor resolutions. Among them, the ROPU is considered in the state model. which can be depicted by using the norm-bounded uncertainty and Bernoulli distributed random variable. In addition, the sensor resolution that can not be ignored in the process of algorithm design is a basic parameter of the sensor. In this case, the goal of the presented DF filtering issue is that by exploiting the covariance intersection fusion rule, the DF filter is designed. Meanwhile, the local upper bound concerning on the filtering error covariance is presented by using the matrix inequalities and inductive approach, and this upper bound is minimized in terms of choosing the proper filter gain. A numerical simulation is performed to test the effectiveness of the presented fusion method.