High-gain observers are an important technique for state estimation in nonlinear systems. This technique utilizes a high gain in order to achieve fast reconstruction of the system states. While the high gain enables fast recovery of the states and good disturbance rejection properties, it can also act to amplify measurement noise. The trade-off between robustness to measurement noise and fast recovery of states is an important and well known problem within the study of high-gain observers.One method to attenuate the effects of measurement noise is to use a low-pass filter in conjunction with the high-gain observer. The focus of this work will be an analysis of this technique. Four different configurations will be considered. The first three configurations place the filter before the observer and study the outcome of making the filter order lower than that of the plant, the same as that of the plant, or higher than that of the plant. The fourth configuration places the filter after the observer. Simulations of all four configurations and comparisons between them will be presented. The low-pass filter is shown to be a simple and effective means of reducing the effects of measurement noise on systems that utilize high- gain observers.
Read
