Zero-Sequence Voltage Error Elimination-Based Offline VSI Nonlinearity Identification for PMSM Drives

Abstract

The offline inverter nonlinearity identification is a useful method to deal with the inverter nonlinearity effect in machine drive. However, the conventional methods can only be achieved at particular initial rotor position, which reduces the application universality. Meanwhile, it is revealed in this paper that the zero-sequence voltage error exists in the zero-axis voltage, which reduces the identification accuracy and has rarely been investigated. To solve the above problems, an offline inverter nonlinearity identification scheme applicable at arbitrary initial rotor positions is proposed, where the zero-sequence voltage error is fully considered. In this paper, the characteristics of zero-sequence voltage error at different rotor position cases is fully investigated by the theoretical analysis for the first time. The offline identification approach at different rotor position cases is then introduced, with the zero-sequence voltage error eliminated. Finally, the second order Newton interpolation polynomial (NIP) based fitting strategy is proposed to reduce the calculation complexity. The proposed method is validated on a 2.2kW permanent magnet synchronous motor (PMSM) platform.