基于PCA-BP的煤矿热电厂辅机故障组合分析法  

Research on Combined Analysis Method of Coal Thermal Power Plant Auxiliary System Fault Based on PCA-BP Neural Network

作　　者：胡晓钦[1] 张晓光[1] 张涛[1] 马云[1] 

机构地区：[1]中国矿业大学机电工程学院,江苏徐州221008

出　　处：《煤矿机械》2010年第4期246-248,共3页Coal Mine Machinery

摘　　要：针对煤矿热电厂部分机组保护功能不完善的现状,研究并提出了一种基于PCA-BP的煤矿热电厂辅机故障组合分析法。该方法用降维处理过后的过程变量来训练和测试BP神经网络,在保证故障诊断精度的同时,减少了神经网络输入神经元,降低了神经网络建模的复杂度,因此PCA-BP组合分析法在热电厂辅机故障诊断运用中能起到很好的效果。For the status that some power units of coal thermal power plant protection is imperfect, study proposed a combined analysis method of coal thermal power plant auxiliary failure based on PCA-BP, which applied the reduced-dimension processing variables in training and testing BP neural network. While ensuring the accuracy of fault diagnosis , reduced the input neurons of neural network, then, reduced the complexity of neural network modeling, thereby, the PCA-BP combined analysis method in the fault diagnosis of thermal power plant auxiliary system can play a good effect.

关 键 词：热电厂辅机 故障诊断 PCA分析法 BP神经网络 

分 类 号：TP306[自动化与计算机技术—计算机系统结构]