On-Line Compensation of Resolver Periodic Error for PMSM Drives

Accurate rotor position plays a critical role in field-oriented control of permanent magnet synchronous machines (PMSMs). However, the resolver periodic error has posed great challenges to the control process. It can be interpreted that the periodic position error will lead to a distinction between the stator currents fed back to the current controller and the real stator currents, generating current harmonics and torque ripples in PMSMs. This article proposes a novel method for on-line compensation of resolver periodic error based on model reference adaptive system and phase-locked loop. The principle of the novel method is elaborated, in which a reference model and an adaptive model are constructed. The reference current vector representing the fundamental current is generated by the reference model. The adaptive current vector reflecting the harmonic current is generated by the adaptive model. The cross product of reference and adaptive current vectors is implemented to obtain the information of the position error. The output of the adaptive model is adapted to the output of the reference model. The periodic error is compensated in the adaptive process. The control system of the PMSM, including the compensation method, is introduced in this article. The feasibility and reliability of the proposed on-line compensation method is verified by simulation and experiments.

I. Introduction

Recently, the demand for high-performance permanent magnet synchronous machines (PMSMs) has increased rapidly in industrial applications, such as aerospace [1], railway traction [2], and electric vehicles [3], [4]. The field-oriented control (FOC) method of PMSMs has been applied extensively because of its excellent performance, in which the accurate rotor position is required [5]. Due to the robustness, reliability and capability to suppress the common-mode noise, resolvers have been widely used to give position information through a resolver-to-digital (R/D) converter, especially in harsh environments [6], [7].