Abstract
This paper proposes an adaptive flux observer for a sensorless switched reluctance motor. The observer adaptive gains are designed using the Lyapunov theory to guarantee both the accuracy and stability of the sensorless control of a switched reluctance motor. A nonlinear inductance model is developed based on a finite element analysis data and used in the estimation algorithms for rotor position and speed. The adaptive flux observer estimates the rotor position at low, medium, and high speeds. A low-frequency ramp method is proposed to excite the switched reluctance motor during standstill where the voltage and current signals are unobservable. The proposed hybrid method is characterized by simplicity, accuracy, ease of implementation, and low real-time computation burden. Therefore, the sensorless control technique depends only on active phase measurements without extra hardware and memory storage for real-time implementation. Complete sensorless control of a three-phase 6/4-pole switched reluctance motor drive system is carried out using Matlab/Simulink. Also, it is implemented experimentally in real-time using the digital signal processor-DS1102 control board. The simulation and experimental results of the proposed sensorless scheme demonstrate the accurate estimation of both the speed and rotor position during the transient and steady states.