Torque Ripple Reduction in Switched Reluctance Motor Drives Using B-Spline Neural Networks

A switched reluctance motor torque ripple reduction scheme using a B-spline neural network (BSNN) is presented. Closed-loop torque control can be implemented using an on-line torque estimator. Due to the local weight updating algorithm used for the BSNN, an appropriate phase current profile for torque ripple reduction can be obtained on-line in real time. It has good dynamic performance with respect to changes in torque demand. The scheme does not require high-bandwidth current controllers. Simulation and experimental results demonstrate the validity of the scheme