基于转矩估计与磁链补偿的SRM转矩脉动控制

由于开关磁阻电机强非线性、强耦合等特点,导致传统磁链控制过程中转矩脉动过大。针对该问题,提出了一种基于SRM转矩特性神经网络的瞬时转矩估计与磁链前馈补偿相结合的控制策略。利用神经网络构建了SRM的瞬时转矩估计器,在该网络结构中设计了能够体现SRM转矩变化规律的激励函数,对神经网络的输入进行预处理,通过自适应学习率训练,实现对瞬时转矩的实时估计。根据转矩估计得到的转矩偏差求得磁链偏差,在磁链模型基础上实现对磁链的前馈补偿,通过磁链滞环控制配合下实现对SRM转矩脉动的抑制。仿真实验表明,基于瞬时转矩估计和磁链前馈补偿的控制方案相较于传统控制策略可以有效地抑制转矩脉动,改善了系统的动态性能。

Torque Ripple Suppression Control of Switched Reluctance Motor Based on Torque Estimation and Flux Linkage Compensation

Due to the strong nonlinearity,strong coupling and other characteristics of the switched reluctance motor,the torque ripple in the traditional flux linkage control process is too large.Aiming at this problem,a control strategy based on the combination of instantaneous torque estimation of SRM torque characteristic neural network and flux feedforward compensation is proposed.The instantaneous torque estimator of SRM is constructed by neural network.In the network structure,an excitation function that can reflect the law of SRM torque is designed,the input of the neural network is preprocessed,the adaptive learning rate training is implemented to achieve Real-time estimation of instantaneous torque.According to the torque deviation obtained by the torque estimation,the flux linkage deviation is obtained.Based on the flux linkage model,the feedforward compensation of the flux linkage is realized,and the flux linkage hysteresis control is used to suppress the SRM torque ripple.Simulation shows that the control scheme based on instantaneous torque estimation and flux-linkage feedforward compensation can effectively suppress torque ripple and improve the dynamic performance of the system compared with traditional control strategies.