State Estimation of Multi-sensor System with Observation Delay Based on SCI-AUFK

In order to deal with the fusion estimation problem of the multi-sensor system with observation lag, the local suboptimal recursive Kalman filter is introduced, and the improved covariance cross fusion algorithm is used to propose an improved covariance cross fusion Kalman filter algorithm, that is, fast sequence covariance cross Fusion adaptive unscented Kalman filter algorithm (SCI-AUKF). It can avoid the large computational burden caused by the calculation of cross-covariance, and can deal with the fusion problem of systems with unknown cross-covariance. Compared with the traditional covariance cross-fusion Kalman filter, it has higher robust accuracy. The improved covariance cross-fusion Kalman filter has higher accuracy than each local sensor, and is closer to the accuracy of the Kalman filter weighted by the matrix, so it has Good performance. The simulation example verifies its validity and consistency, and gives a geometric explanation of the accuracy relationship.