Multi-sensor optimal information fusion Kalman filter for discrete multichannel ARMA signals

A new multi-sensor optimal information fusion criterion weighted by covariance is presented in the linear minimum variance sense. Based on this optimal fusion criterion, using the measurement white noise filters, a general multi-sensor optimal information fusion distributed Kalman filter is given for the discrete multichannel ARMA (autoregressive moving average) signals with correlated noises. It has a two-layer fusion structure with fault tolerant and robust properties. When all local sensor subsystems are faultless, the precision of the fusion filter is lower than that of the centralized filter. When some sensors are fault, the fusion filter has better reliability. The precision of the fusion filter is higher than that of any local sensor subsystem. Applying it into a double-channel system with three sensors shows its effectiveness.