Position Sensorless Torque Ripple Control of Switched Reluctance Motor Drive using B-Spline Neural Network

This paper presents a method for the rotor position estimation of a switched reluctance motor (SRM) based on B-spline neural network to develop a position sensor-less drive. The neural network estimates the rotor position by measuring the phase voltage and current to calculate the flux-linkage. The neural-network can be trained online or by using the flux-linkage characteristics of the SRM. The efficacy of the model is validated by employing it to minimize the torque ripple in a 8/6 SRM by simulation in MATLAB/Simulink environment.