Self-tuning multisensor measurement fusion Kalman filter

Abstract

For the multisensor system with unknown noise statistics, and with the measurement matrices having the same factor, based on the weighted least squares (WLS) method, a weighted fusion measurement equation is obtained, and it together with the state equation to constitute a equivalent weighted measurement fusion system. Based on the on-line identification of the moving average (MA) innovation model parameters for weighted measurement fusion system, using the modern time series analysis method, a self-tuning weighted measurement fusion Kalman filter is presented. It is proved that it converges to globally optimal measurement fusion Kalman filter with known noise statistics, so that it has asymptotic global optimality. A simulation example for a tracking system with 4 sensors shows its effectiveness.