变载荷感应电动机的神经网络建模研究

根据采用晶闸管三相调压器控制变载荷电动机运行的特点,阐明了采用神经网络方法对该系统建模的必要性.基于带有回归单元的Elman神经网络,对变载荷三相异步电动机的晶闸管三相调压器系统进行了建模.采用一种带惯性项的动态反向传播学习算法,克服了通常的BP算法振荡和收敛速度慢的弱点,使变载荷电动机系统跟随负载变化对电动机实现调压控制.对Elman神经网络的结构运用方法,以及惯性项的动态反向传播学习算法做了较详细的介绍,对由晶闸管三相调压器构成的拖动系统建模所选向量参数进行了说明.实例表明,利用该方法迭代后的学习结果更容易将误差减小至期望值.

An ANN Model Test for Induction Motor Dynamic Loads

The modeling of an induction motor that is powered by an SCR voltage regulator is very complex in traditional modeling methods because the load on the motor is constantly changing.But it can be simplified by using a dynamic recurrent ANN(Artificial Neural Network),namely Elman ANN,based on the operation characteristics of the system.A dynamic back-propagation(BP) learning method with an inertia item is used for ANN training,which has advantages over traditional BP method plagued by algorithm resonance and slow convergence rate.The trained ANN makes it possible to regulate the voltage across the motor with varied load.The ANN configuration,the modeling procedure and the dynamic BP learning method were described.The selection methods of the parameters for system modeling were also discussed. The simulation results show that the learning results obtained by the iterative process have a relatively smaller error that falls within expected values.