A Compensation Method for Full-Digital Sensorless Control of Permanent Magnet Synchronous Machine based on Model Reference Adaptive System

Abstract

Model reference adaptive system (MRAS) has become a widely used sensorless control strategy in motor drive application. This paper focuses on the sensorless control for the permanent magnet synchronous machine (PMSM) based on MRAS and presents a compensation method that can eliminate steady-state position error caused by full-digital control, which is essential for high-performance control of the high-speed PMSM. The primary MRAS control method is introduced first in this paper. Meanwhile, this article puts forward the problems existing in the traditional digital control process and analyses the cause of the issues. In response to the above problems, an improved MRAS control strategy is proposed, which considers both the existence and the absence of one-beat-delay in the process of modulation. The proposed compensation method can eliminate the steady-state angle error and achieve the accurate control of the machine. Simulation studies verify the validity of the proposed strategy.