Conduction Angle Adaptive Control Base on NSGA-II Optimize BP Neural Network

Aiming at the large torque ripple problem of the permanent magnet assisted switched reluctance motor, an adaptive control for the turn-on angle and turn-off angle is proposed in this paper. Firstly, for the problem that the parameters of BP neural network are difficult to determine, the NSGA-II multi-objective genetic algorithm is used to optimize its parameters so that the network prediction output is more accurate. Then, select the appropriate angle datas for various working conditions and use it as the training output data of the BP neural network prediction model. Finally, the simulation results show that: under different speed and load conditions, the control strategy proposed in this paper realizes the angle adaptive adjustment, reduces the current amplitude and torque ripple.