The adaptive tracking of slowly varying processes with coloured-noise disturbances

Abstract

An adaptive tracking technique is developed to estimate the output of a slowly varying process which is corrupted by a coloured noise disturbance. There are two estimators in the adaptive tracker. The primary estimator and the vernier estimator is simply a low-pass filter or its equivalent. The vernier estimator is self-tuning and is adaptive to the varying disturbances. Since the process contains unknown parameters, the usual Kalman filter can not be used. A conventional low-pass filter might be employed but it would give a significant phase lag in this situation. The adaptive tracker has a small phase lag and has particular advantages over the conventional low-pass filter when the frequency of the coloured noise disturbance is varying.