一种基于BP神经网络的异步电动机转子速度辨识

A Novel Speed Identification Scheme Based on the BP Artificial Neural Network

针对异步电动机直接转矩控制系统,基于三相异步电机方程提出一种新型的采用人工神经网络辨识电机转速的方法。利用BP 神经网络强大的函数逼近能力,对电机转速进行在线辨识。理论推导证明了其可行性,仿真和试验结果表明了该方法具有较好的静动态性能。

The paper presents a novel speed identification scheme based on the BP Neural Network for the DTC motor drive system. BP Neural Network＇ s powerful ability toapproximate any unknown, non-linear functions made it useful in the online identification of the motor speed.Simulation shows that the scheme provides the system with good performance both static and dynamic.