Accurate Position Sensorless Control for Switched Reluctance Machines Based on Data Fusion with Kalman Algorithm

In this paper, a novel position sensorless control for switched reluctance machines based on data fusion with Kalman algorithm is proposed. First, the unsaturated inductance of the detection phase is obtained by high-frequency pulse injection and then calculated to obtain rotor position information. At the same time, the inductance of the other two phases is calculated and converted to unsaturated inductance by simple equations and then calculated to obtain rotor position information similarly. Data fusion is performed using the Kalman algorithm to assign different weights to the rotor position information obtained by the two methods to obtain the optimal rotor position information. The simulation results prove that the proposed method not only doesn't need additional measurement hardware and complex data tables but also significantly reduces the rotor position estimation error compared to the conventional method.