Proposal of high performance methods for linear Kalman filter with disturbance estimation and application to motion control systems

Abstract

This study aims to realize fast disturbance estimation with a Kalman filter (KF). This paper presents a methodology of fast estimation and rejection for some types disturbances. This methodology attempts to estimate the state of mechatronics systems that rejected the influences of disturbance and noise by using KF. Most of known disturbance estimation method is the disturbance observer (DOB). However, the conventional DOB can not estimate the disturbance that is contaminated of noises. Generally, KF is known as state estimation algorithm that can consider influences of various noise. In addition, we had done some previous studies to design a linear KF with disturbance estimation. We obtained a fact that the disturbance estimation speed by the KF is slow from these studies results. Therefore, this paper proposes a new design method of KF with disturbance estimation which includes a design parameter to speed up disturbance estimation, where we have never concerned in this paper with KF design method except linear systems. Finally, we show the design method on the proposed method and usefulness via the simulation results relative to the position control for an simple cart system.