Fast self-tuning weighted measurement fusion Kalman filter for the ARMA signal

Abstract

For the multisensor single channel autoregressive moving average(ARMA) signal with common disturbance measurement noise and sensor bias, when the model parameters, the sensor bias and the noise variances are all unknown, their consistent estimates are obtained by a multistage fused identification method, which includes the recursive extended least squares (RELS) algorithm, correlation method and the Gevers-Wouters algorithm with a dead band. Substituting these estimates into the optimal weighted measurement fusion(WMF) Kalman signal filter, a self-tuning WMF Kalman signal filter with asymptotic global optimality is presented. A fast inversion algorithm of the extended Pei-Radman matrix is presented in order to reduce the computational load. A simulation example verifies the effectiveness of the proposed method.