基于柔性神经网络的SRM直接瞬时转矩控制  

Direct Instantaneous Torque Control of SRM Based on Flexible Neural Network

作　　者：党选举[1] 袁小唐 胡景佳 

机构地区：[1]桂林电子科技大学电子工程与自动化学院,广西桂林541004

出　　处：《电气传动》2014年第5期62-66,共5页Electric Drive

基　　金：国家自然基金项目(60964001;61263013);广西教育厅重大项目(201101ZD007);广西信息科学实验中心重大项目(20130110)

摘　　要：开关磁阻电机(SRM)有着广阔的应用前景,而低速运行时的高转矩脉动这一缺陷限制了其在汽车等领域内的应用。针对SRM强非线性和高度耦合性的特点,借鉴传统直接瞬时转矩控制(DITC)策略,提出了基于所构造的柔性神经网络(FNN)的SRM直接瞬时转矩控制策略。其中改进的控制策略外环采用不完全微分模糊PID对速度进行调节,内环采用以转矩误差的平方为性能指标函数的FNN自适应PID对转矩进行调节。在Matlab/Simulink环境下,仿真结果表明改进的DITC策略能够更有效地降低SRM的低速转矩脉动。Switched reluctance motor have broad application prospect. However, high torque ripple is a serious drawback at low speed, and limits its widespread application on the automobile industry and other fields. SRM direct instantaneous torque control strategy based on constructed flexible neural network was proposed, which refers to the ideas of traditional direct instantaneous torque control to solve the strong nonlinear and high coupling of the SRM. In the improved control strategy, the outer-loop incomplete derivative fuzzy-PID is used to adjust the motor speed, and the inner-loop flexible neural network self-adaptive PID is employed to regulate the motor torque by using the square of the torque error as performance index function. In the Matlab/Simulink environment, simulation results prove that the improved DITC strategy proposed reduces the torque ripple of the SRM more effective at low speed, comparing with the traditional DITC strategy.

关 键 词：开关磁阻电机 直接瞬时转矩控制 模糊PID 柔性神经网络 

分 类 号：TM352[电气工程—电机]