基于教与学算法的SRM转矩分配函数优化及控制研究

针对开关磁阻电机(SRM)换相阶段转矩脉动严重问题,提出基于教与学算法的转矩分配函数优化策略。以开通角、关断角、换相重叠角为寻优对象，输出转矩跟随参考转矩能力为评价指标，结合转矩分配函数与直接瞬时转矩控制，构成转矩闭环。为提升系统动态性能，转速环采用分数阶滑模控制方法，将其输出作为转矩环参考转矩。仿真实验结果表明，经教与学算法优化后，转矩可在相绕组换相阶段平滑过渡，且输出转矩脉动得到明显抑制；分数阶滑模较PI控制而言，系统转速迅速进入稳态，且上升过程无超调，有利于提升系统动态特性。

Research on Optimization and Control of SRM Torque Sharing Function Based on Teaching Learning Based Optimization algorithm

Aiming at the serious problem of torque ripple in the commutation phase of switched reluctance motor,a torque sharing function optimization strategy based on teaching learning based optimization algorithm was proposed.With the opening angle,the closing angle and the commutation overlap angle as the optimization objects,the output torque follows the reference torque capacity as the evaluation index,combined with the torque sharing function and the direct instantaneous torque control to form the torque closed loop.In order to improve the dynamic performance of the system,the speed loop adopted the fractional-order sliding mode control method,and its output was used as the torque loop reference torque.The simulation results show that the torque can be smoothly transitioned in the phase winding commutation phase after the teaching learning based optimization algorithm was optimized,and the output torque ripple was obviously suppressed.Compared with PI control,the fractional-order sliding mode quickly enters the steady state,and the ascending process has no overshoot,which is beneficial to improve the dynamic characteristics of the system.