Improvement of disturbance suppression based on disturbance observer

Abstract

This research focuses on improvement of disturbance suppression even under existence of observation noise. Disturbance observer is an effective tool to estimate and compensate disturbance. There is a problem, however, that it cannot accomplish desired disturbance suppression when a large observation noise exists. In this research, a method with which influence of disturbance can be suppressed greatly even with a large observation noise is proposed. The authors focus on an LPF used in velocity calculation. The method is proposed based on the analysis on the characteristics and mechanism of disturbance suppression in disturbance observer with a focus on its cut-off frequencies. In the method, the cut-off frequency of disturbance observer is set higher than that for velocity calculation. Validity of the proposal was confirmed in theoretical analysis, simulation, and experiments