基于神经网络的异步机直接转矩控制及仿真

Direct self-control and simulation of asynchronous machines based on neural network

在直接转矩控制的基础之上,采用神经网络对电机模型进行辩识．仿真的结果表明,用此方法解决了直接转矩控制中模型难于精确辩识的问题,使控制具有良好的动、静态性能指标．

Based on direct self-control, neural networks were used to identify the asynchronous machine model. This approach can overcome the difficulty of the model＇s identification. The simulation results show that this approach has very good dynamic and static properties.