Torque Ripple Reduction Of Switched Reluctance Motor Based On Neural Network Sliding Mode Parameter Online Learning

High torque ripple limits the application area of the switched reluctance motor (SRM). To solve this problem, the sliding mode control algorithm is applied to the speed control in SRM. However, the uncertainty of motor parameters significantly impacts the electromagnetic torque of SRM. Therefore, a neural network sliding mode controller (NNSMC) based on parameter online learning is designed in this paper. The internal parameters of SRM are learned online through speed error, resulting in the combined control of the neural network and sliding mode. The Lyapunov stability method is used to prove the stability of the algorithm. The simulation results show that the proposed method can effectively learn the parameters of SRM, reduce torque ripple and improve the operational performance of the motor.