基于RBF网络的开关磁阻电动机自适应滑模控制  

Adaptive Sliding Mode Control for Switched Reluctance Motors Based on RBF Neural Network

作　　者：周素莹[1] 林辉[1] 

机构地区：[1]西北工业大学,陕西西安710072

出　　处：《微特电机》2011年第7期55-57,76,共4页Small & Special Electrical Machines

摘　　要：结合滑模控制和神经网络各自的优点,对开关磁阻电动机提出了一种基于RBF神经网络的开关磁阻电动机自适应滑模控制方案。设计了基于反馈线性化的滑模变结构控制器,通过RBF神经网络的在线学习实时估计系统参数变化,减小系统参数变化对控制效果的影响,实现了开关磁阻电动机的自适应滑模控制。理论分析证明了所设计自适应神经滑模控制器的稳定性和可行性,仿真结果证明了所设计控制器可行性。With the combination of the merits of sliding-mode control and neural network, a neural-network-based adaptive sliding-mode control scheme for switched reluctance was proposed. First, a sliding-mode controller based on the feedback-linearzing was designed, then an RBF neural network was used to estimate structure parameter variations in realtime. Thus, the effects of structure parameter variations on control performance are reduced. Theoretical analysis and simulation resuhs prove that the proposed strategy is steady and feasible.

关 键 词：开关磁阻电动机 自适应滑模控制 神经网络 反馈线性化 

分 类 号：TM352[电气工程—电机]