基于BP神经网络的定子绕组智能观测器的研究  

Research on the Stator Resistance Intelligent Observer Based on BP Neutral Network

作　　者：郑慧芳[1] 高长举[1] 刘英环[1] 袁丽肖[1] 

机构地区：[1]河北工业大学电气与自动化学院,天津300130

出　　处：《河北工业大学学报》2006年第2期100-104,共5页Journal of Hebei University of Technology

摘　　要：在直接转矩控制系统中,异步电动机定子电阻在低速时受温度影响而变化,从而影响了控制系统的低速性能.介绍、分析了人工神经网络的特性及其算法,设计了一种基于BP神经网络的定子绕组智能观测器来估计运行中定子电阻的变化,并借助MATLAB辅助工具进行仿真.仿真结果表明该神经网络观测器检测误差在工程允许的范围之内.In direct torque control system, the stator resistance ofasynchronism electro motor, is variable due to the effect of temperature. Consequently, it influence the low speed performance of the system. In order to solve this problem, this paper introduce and analyse the characteristic and arithmetic of ANN, then designs a stator resistance intelligent observer based on BP neutral net＇~ork and simulink by means of Matlab. With the result of simulation, it has been proved that the neural network resistance observer enjoys a accury.

关 键 词：直接转矩控制 算法 BP神经网络 智能观测器 仿真 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]