Special Estimation Analysis for the Position of Switched Reluctance Machine using Sinusoidal Inductance Model

Fourier modelings of the inductance of switched reluctance machines have been extensively investigated extensively in the past, but the skills are restricted to the construction of the inductance profile model at a known motor position. This paper gives a presentation of a novel sensorless position estimation method based on online sinusoidal modeling of relative inductance for switched reluctance motors at startup. The background of this method is based on sinusoidal fitting and only very few data points are needed. It is numerically efficient and can provide accurate position information for each phase. There is no special requirement on the inductance profile of the machine, and the magnetic characteristics of the machine are not required. The method is simple and easy to implement. Experimental results have confirmed its validity.