Detection and Modelling of Coloured Noise for Kalman Filter Applications

This paper gives an overview of detecting and modelling of coloured measurement noise for Kalman filter applications. Detection and modelling are important as the Kalman filter is based on the assumption of white (measurement) noise. Modelling is done by using autoregressive techniques, resulting in a shaping filter which can then be incorporated into the Kalman filter model. Modelling is performed in the time and frequency domain. The analysis is performed for single-channel systems.