Online Calibration of Sensorless Position Estimation for Switched Reluctance Motors With Parametric Uncertainties

This article proposes an online calibration strategy of sensorless position estimation for switched reluctance motors (SRMs) with parametric uncertainties. For conventional sensorless SRM drives, the voltage drops on power devices and the winding resistance changing with the current and temperature are neglected, which affects the accuracy of the estimated position. In this article, an inductance calculation scheme is proposed to improve the real-time position estimation with these parametric uncertainties. First, the calculation errors of the flux linkage and inductance are both investigated and the error correction of the estimated rotor position is analyzed in detail. Then, the inductance values of the current overlapped phases are calculated. The dc component of the phase inductance is online adjusted, where all three phase inductances can be easily obtained accordingly. Finally, the real-time rotor position is acquired by the coordinate transformation of the calculated three-phase inductance during the current overlapped region. Compared to conventional SRM sensorless drives, the error of the estimated position due to parametric uncertainties can be significantly reduced without pulse injection, extra hardware, and prestored flux linkage and inductance profiles. The simulation and experiments are carried out on three-phase and four-phase SRMs to validate the effectiveness of the proposed scheme.