This paper is concerned with information fusion estimation problems for multi-sensor networked systems with random packet losses. Based on a recent developed compensation strategy of packet losses that the predictor of lost observation is used as the observation when a packet is lost, centralized fusion estimators (CFEs), including the filter, predictor and smoother, in the linear unbiased minimum variance (LUMV) sense are first designed by completing square method. Then, local optimal estimators are designed for each sensor subsystem. Estimation error cross-covariance matrices between any two local estimators are derived. Based on local estimators and cross-covariance matrices, distributed fusion estimators (DFEs) are presented by using the matrix-weighted fusion estimation algorithm in the LUMV sense. Compared with the existing results with zero-input and hold-input compensations, the proposed algorithms with prediction compensations can obviously improve the estimation accuracy. Two simulation examples show their effectiveness.
Financed by the National Centre for Research and Development under grant No. SP/I/1/77065/10 by the strategic scientific research and experimental development program:
SYNAT - “Interdisciplinary System for Interactive Scientific and Scientific-Technical Information”.