The fusion estimation problem is investigated for a class of uncertain stochastic systems with stochastic sensor gain degradation. The sensor gain degradation is described by random variable whose probability is assumed to be known. The model’s uncertainty is described by stochastic parameter perturbations considered in the system matrix. A kind of local unbiased estimator structure is proposed, and an optimization problem which sets the local filter gains and the finite horizon estimation error to be the decision variables and the cost function, respectively, is established. Then for the given scalar fusion weights, obtaining the closed form of the optimal consensus filter gains is a challenging problem, so a set of sub-optimal local filter gains are computed based on minimizing an upper bound of the cost function. Finally, simulation example is given to illustrate the effectiveness of the proposed approach.

sensor gain degradation|model’s uncertainty|local unbiased estimator|scalar fusion weights|decentralized fusion estimator