一种感应电机矢量控制转子电阻在线辨识方案  

Scheme for Motor Rotor Resistance On line Identification in Vector Control of Induction Motor

作　　者：范波[1,2] 刘刚[1] 史光辉 

机构地区：[1]河南科技大学电子信息工程学院,河南洛阳471003 [2]中信重工机械股份有限公司,河南洛阳471003

出　　处：《电气传动》2013年第11期11-16,共6页Electric Drive

基　　金：河南省教育厅自然科学研究项目(2011B470003);河南科技大学校青年基金(2011QN29)

摘　　要：感应电机数学模型具有高阶非线性强耦合的特点,各参数随电机运行状态的变化难以用函数描述,参数辨识是矢量控制等高性能变频调速方案必须面临的问题。提出了一种基于Elman神经网络的感应电机转子电阻参数辨识模型,利用Elman神经网络逼近非线性函数的能力和反馈特性,通过在前馈网络隐含层中增加一个承接层,作为一步延时算子完成记忆目的,使得本模型具有适应时变特征、直接反映系统动态过程特性以及更强计算的能力。仿真及实验结果分析表明,经过充分训练的Elman神经网络具有较高的辨识精度和广泛的适应能力,能够提升感应电机变频调速控制性能。The mathematical model of induction motor has the character of high order, nonlinear and complicate coupling. It is difficult for the motor＇s parameters changing with the work state to describe with a definite function. The parameter＇s identification is a problem confronted by high performance variable frequency speed adjustment system including vector control. The model of induction motor rotor resistance identification based on Elman neural network was proposed. Elman neural network＇s function approximation and unique feedback were used. A context layer to the forward network was added as a delay factor to memorize history state. So this model has the abilities of adapting to time- variable characteristic, reflecting the dynamic characteristics of system directly, and the stronger calculation. The simulation and experimental results analysis show that the Elman neural network with sufficient training in this model has the better accuracy of identification and the extensive adaptability. And this identification method is able to enhance the performance of induction motor＇s variable-frequency speed regulation.

关 键 词：感应电机 矢量控制 转子电阻 ELMAN神经网络 参数辨识 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TM343[自动化与计算机技术—控制科学与工程]