基于人工神经网络的次同步谐振监测方法  

A method based on ANN for real-time monitoring of sub-synchronous resonance

作　　者：董青迅[1] 李兴源[1] 张淼[2] 穆子龙[1] 顾威[1] 

机构地区：[1]四川大学电气信息学院,四川成都610065 [2]重庆市电力公司沙坪坝供电局,重庆400030

出　　处：《电力系统保护与控制》2011年第9期21-25,共5页Power System Protection and Control

基　　金：国家科技支撑计划项目(2008BAA13B01);四川电网公司科技项目~~

摘　　要：提出一种可实时监测次同步谐振的方法,用最容易取得的电流量作为次同步谐振的监测量。通过构造特殊的多层前馈神经网络,建立相应的次同步谐振监测电路,利用其强大的记忆功能和非线性映射特性,得到与采样电流对应的扭振转矩的表现,通过观察转矩的变化趋势,判断次同步谐振发生与否。给出人工神经网络训练样本的形成方法、训练算法和具体的操作步骤。以IEEE第一标准测试系统模型作为仿真算例,试验结果表明所提出的基于人工神经网络的次同步谐振监测方法的有效性。A new method based on artificial neural network for real-time monitoring of sub-synchronous resonance is proposed.The current which is easy to get is selected as the sub-synchronous resonance monitoring amount.By constructing a special multi-layer feed forward neural network,a corresponding SSR measurement circuit is built.Using the strong memory function and the nonlinear mapping behavior,the performance of torque corresponding to the sample current is got.And then whether sub-synchronous resonance occurs or not can be identified by observing the change of the torque.In addition,the forming method of ANN training samples,the algorithm and the procedure for training the measurement circuit are given.IEEE first benchmark system model for sub-synchronous resonance is selected as the simulation example.The simulation results illustrate the effectiveness of the presented approach.

关 键 词：次同步谐振 人工神经网络 监测 电流 转矩 

分 类 号：TM935.2[电气工程—电力电子与电力传动]