Guaranteed Cost Robust Centralized Fusion Kalman Predictor for Systems with Moving Average Colored Measurement Noise, Missing Measurements and Uncertain Noise Variances

The paper is concerned with the guaranteed cost robust centralized fusion (CF) prediction problem for systems with moving average colored measurement noise, missing measurements and uncertain noise variances. The considered system is converted into one only with uncertain noise variances by model transformation method. The method includes rewriting moving average model of state space form, augmenting method and fictitious noise technology. By parameterizing the perturbations of uncertain noise variances, two classes of guaranteed cost robust CF Kalman predictors are presented based on the minimax robust estimation principle. Two problems can be converted into optimisation problems with constraints, and they can be solved by Lagrange multiplier method and linear programme method. The proposed guaranteed cost predictors can concurrently give maximal lower bound and minimal upper bound of accuracy deviations. The proof of the guaranteed cost robustness is proved by the Lyapunov equation approach. A simulation example used to UPS shows the correctness and effectiveness of the proposed results.