基于R-NN的可控硅励磁系统故障诊断  

Fault Diagnosis of Silicon Controlled Rectifier Excitation System Based on Rough set-Neu-ral Network

作　　者：张江滨[1] 徐显明[2] 

机构地区：[1]西安理工大学 [2]西安理工大学水利水电学院

出　　处：《微计算机信息》2007年第25期186-188,共3页Control & Automation

基　　金：国家自然科学基金重点项目资助(90410019)

摘　　要：基于同步发电机可控硅励磁系统经常发生故障,提出一种基于粗糙集—神经网络(Roughset-NeuralNetwork)相结合的故障诊断方法。以励磁系统中三相桥式可控硅整流回路为核心进行故障诊断研究,对整流回路故障波形的采样数据样本信息进行预处理,通过运用粗糙集理论的知识约简方法形成故障诊断的确定性规则,从而实现故障分类;然后将其结果与故障信息中的输出样本值作为神经网络的输入,实现故障元的定位。通过计算机仿真,结果表明:该方法对三相桥式可控硅整流回路故障诊断简便准确,诊断速度快。Based on the fault of synchronous generators Silicon Controlled Rectifier Excitation System, a fault diagnos is method based on rough set-neural network is presented. Taking three phase bridge SCR（Silicon Controlled Rectifier） of Excitation System as example, The sample-data information of the fault wave is processed in advance and the unnecessary fault signs are simplified via the method of rough set theory to form the correct diagnosis roles and to classify the faults. Then the sampled-data of wave input to a neural network together with the outcome of rough set to locate the fault part. The computer simulation example proves that the method improves the fault diagnosis of three phase bridge SCR speed greatly with high correctness and briefness.

关 键 词：粗糙集 神经网络 可控硅励磁系统 故障诊断 

分 类 号：TP21-27[自动化与计算机技术—检测技术与自动化装置]