基于级联神经网络的电机转速和负载转矩辨识

Speed and Torque Estimation for Motor Using Cascade Networks

合理选择电机容量具有重要意义,而电机容量可根据电机转速和负载转矩确定。此处提出利用级联BP神经网络估计电机转速和负载转矩的方法,将两个神经网络级联,第一级网络输入为电机定子电压,输出为定子电流和转速;第二级网络输入为第一级网络的输出,输出为负载转矩,形成了级联BP神经网络模型,实现了异步电机转速和负载转矩的同时估计。实验结果表明,该方法能较准确地估计出异步电机的转速和负载转矩。

It is important to choose right specification for induction motor,which is determined by speed and load torque estimator.This paper proposes speed and load torque estimation scheme based on the cascade BP neural net-work.The in-puts of the first level are the stator voltages,and the outputs are the stator currents and speed.The inputs of the second level are the outputs of the first level,and the output is load torque.The speed and load torque of induction motor are both estimated in the cascade configuration.The experimental results are presented to validate the proposed approach.