基于人工神经网络的电压电流互补模型转子磁链估计

Rotor Flux Estimation Using Voltage-current Model Based on Artificial Neural Networks

为了准确对感应电机磁链进行实时估计,提出了一种基于人工神经网络(ANN)电压电流互补模型的估计方案。该方案利用基于ANN的电压与电流模型之间的磁链估计差值在线对电机参数进行更新,以减小因电机参数变化所引起的模型偏差,并最终通过电机频率范围决策转子磁链。实验结果表明,该方案对磁链估计有效,实时控制效果好。

In order to estimate the induction motors online flux accurately,a estimation method using voltage-current model based on artificial neural networks（ANN） was proposed.Online updating the induction motors parameters were implemented by the error of flux estimation between the voltage model and current model based on ANN,thereby,reduced the model error because of parameters change.Furthermore,rotor flux estimation was selected by motor frequency.Experimental results show that this method obtained the high precision rotor flux and high-performance.