Integration Of Self-tuning Observation Decoupled Wiener State Estimators And Its Application

From the 1970s onwards, a new subject - multisensor information fusion has been developed rapidly, and was applied widely to modern C3I systems and many weapons platform. Information fusion as an interdisciplinary integrated information processing theory, involves systems theory, information theory, control theory, artificial intelligence and computer communications, and many other fields and disciplines. As a new technology for multisource date processing, multisensor information fusion can synthesize multisource date from one object, generate more accurate estimation than single source.This paper, for the multisensor system, using the weighted least squares (WLS) criterion, two weighted measurement fusion equation are obtained which accompanies the state equation to constitute two equivalent weighted measurement fusion systems. Based on this, using the modern time series analysis method, based on the autoregressive moving average (ARMA) innovation models, two kinds of the optimal weighted measurement fusion decoupled Wiener state estimators (filter, predictor and smoother) are presented and verified to have completely functional equivalence to the centralized measurement fusion decoupled Wiener state estimators. For the multisensor systems with unknown noise statistics, based on the identification of moving average (MA) innovation models and the solutions of the matrix equations for correlation functions, two kinds of self-tuning weighted measurement fusion decoupled Wiener state estimators are presented. For each kind of them, four kinds of self-tuning weighted measurement fusion algorithms are presented. Based on self-tuning weighted measurement fusion decoupled Wiener state estimators, the multisensor multichannels ARMA signals self-tuning weighted measurement fusion decoupled Wiener estimators are also presented. The convergences of the estimators are proved by the dynamic error system analysis (DESA) method. That is to say, it is strictly proved that the self-tuning fuser converges to the corresponding optimal weighted measurement fusion decoupled Wiener fuser in a realization or with probability one, so that it has asymptotic global optimality. The many simulation examples for the tracking system show their effectiveness.