Sensorless Control for Switched Reluctance Motor Based on Quadrature Feature Position Observer at Medium and High Speeds

The conventional linear prediction has discontinuous estimated speed and position, and the error of flux linkage reduces the estimation accuracy. To solve the above problems, this article proposes a quadrature feature position observer for switched reluctance motor (SRM) at medium and high speeds, which includes harmonic decouple network (HDN), multiple second-order generalized integrator (MSOGI) and quadrature signal generator. This method uses HDN and MSOGI to filter the nonfundamental frequency of the feature position signal. Then a quadrature signal generator to transform the feature position signals into a set of quadrature signals, and the rotor position and speed are estimated by using the phase locked loop. Compared with the linear prediction method, proposed method has higher estimation accuracy, and there are no discontinuous estimated speed and position. In addition, this method does not need to measure the 3-D magnetic characteristics, requires less data storage space, the implementation is simpler, and has a wider application range. Eventually, the effectiveness of this method is verified on the 750 W 12/8 SRM setup, experimental results show that proposed method can improve the steady state and transient performance of sensorless control system.