A novel approach to auto-calibrating sensorless switched reluctance motor drive

Optimal operation of switched reluctance motor drives is inherently dependent to proper excitation of motor phases with respect to their inductance profile. Based on this fact, utilization of position transducers in the drive is unavoidable. These transducers increase the cost and complexity of the drive. In this paper, a new method for elimination of position sensors in SRM drives is suggested. The proposed method is a marriage of chopping waveform analysis and inductance modeling based on Fourier series. A novel method for online estimation of back emf based on chopping current waveform is introduced. Besides, incremental inductance profile is modeled and automatically calibrated for robust operation. By evaluation of estimated back emf and phase inductance model, position information will be obtained. This approach does not suffer from unpredicted variation of magnetic characteristics since it benefits model auto tuning. The proposed method has implemented successfully on experimental SRM drive based on TMS320F2470 digital signal processor.