A SELF-TUNING KALMAN FILTER WITH THE COLOURED OBSERVATION NOISE AND INPUT ESTIMATION AND ITS APPLICATION

For discrete linear systems with unknown coloured observation noise and unknown constant input, using the modern time series analysis method¹,based on ARMAX innovation model, this paper presents a new self-tuning Kalman filter with input estimation. As an application example, a novel self-tuning α-β tracking filter with the coloured observation noise and input estimation is proposed, and simulation results show its usefulness.