Variable Gain Super Twisting Algorithm을 활용한 영구자석 동기전동기의 위치 추종기설계

Abstract

This paper presents a variable-gain super-twisting sliding observer for a permanent magnet synchronous motor's position and speed. The observer algorithm ensures global, finite-time convergence to the desired sliding surface when the matched perturbations/uncertainties are Lipschitz-continuous functions of time that are bounded, together with their derivatives, known functions. The proposed algorithm is similar to the first sliding mode but has many advantages over the first sliding mode in that the gain value of the controller changes in real-time, and chattering is reduced.