Information Fusion Kalman Predictor for Two-Sensor Multichannel ARMA Signal System with Time-Delayed Measurements

Abstract

For the two-sensor multichannel autoregressive moving average (ARMA) signals with time-delayed measurements, an equivalent state space model with time-delayed measurements is obtained. Then a measurement transformation method is presented, which can transform this state space model with measurement delays into the state space model without measurement delays. Furthermore, based on the modern time series analysis method, local Kalman predictors are obtained. Then the covariance intersection (CI) fusion Kalman predictor is presented, which avoids computing the cross-covariance compared with the fused Kalman predictor weighted by matrices. It is proved that its accuracy is higher than each local predictor, and lower than that of the fused Kalman predictor weighted by matrices. The geometric interpretations of the local and fused predictors’ accuracy relation are given based on covariance ellipses. A Monte-Carlo simulation example shows that the CI Kalman fuser has higher accuracy and good performance.