Optimal recursive fusion estimator for asynchronous system

Estimation fusion for multisensor asynchronous sampling system is currently a hot subject of research. Under the assumption with independent and uncorrelated noises, some asynchronous fusion algorithms have been provided. Unfortunately, the general optimal solution with real-time update performance in linear minimum mean square error (LMMSE) hasn´t been given up to now. Simultaneously, it is difficult to design the optimal recursive asynchronous fusion estimator with cross-correlated noises under centralized frame and the research progress is slow. Aiming at this problem, an optimal and general recursive fusion estimator for this kind of multisensor asynchronous sampling system with cross-correlated noises is proposed by use of a universal effective noises decorrelation technology in this paper. The proposed asynchronous fusion estimator not only is optimal in LMMSE, but also can archive the best real-time running performance. Additionally, it can also be even applied to the ldquoout-of-sequencerdquo measurements (OOSM) system. Theoretic analysis and simulation both show the validity and superiority of this fusion estimator.