Suboptimal fusion estimation for systems with random delay, packet dropout and uncertain observation

This paper is concerned with the distributed fusion filtering problem for multi-sensor linear discrete-time stochastic systems with one-step random sensor delay, packet dropout and uncertain observation. Three kinds of uncertain phenomena in measurement data are described by three Bernoulli distributed random variables. Based on the innovation approach, the local optimal linear filter, predictor and smoother are given. Further, the distributed suboptimal fusion estimators are given by the covariance intersection fusion algorithm. The computational cost is reduced since the cross-covariance matrix between any two local filters is avoided. Simulation research verifies the effectiveness of the proposed algorithm.