New method for selecting adaptive Kalman filter fading factor

A new algorithm of adaptively adjusting the fading factor for filter without divergence is presented. First the reason of Kalman filter divergence and the principles of divergence restraining schemes are discussed in detail, then the physical meaning of observation noise covariance matrix with an exponential weighting factor is analyzed, and a lemma is proved. The advantage of the new method is that fading factor adjusting simple and precise. Simulation results show the effectivencess of the new method.