摘要
采用多维BP神经网络方法进行锅炉故障诊断建模,其中BP神经网络采用多维结构,输入层采用模糊数学方法对运行参数进行量化,通过参数之间和参数与故障类之间的关联关系,建立了多维BP神经网络模型。以锅炉管泄漏为例,进行了故障仿真试验,试验结果表明此方法能有效、快速地进行锅炉故障诊断。
By using the multi-dimensional BP neural network method,a model for fault diagnosis of boilers has been esablished.In this model,the BP neural network adopts multi-dimensional structure,and the operating parameters are quantified by using the fuzzy mathematic method in the input layer.Through interrelated relationship among the parameters,as well as between parameters and fault types,a model of multi-dimensional BP neural network has been constructed.Taking pipe leakage in a boiler as example,an emulation test of said fault has been carried out.Results of test show that the said method can effectively and rapidly diagnose the boiler fault.
出处
《热力发电》
CAS
北大核心
2011年第4期77-80,共4页
Thermal Power Generation
关键词
BP神经网络
锅炉
故障诊断
建模
仿真
BP neural network
boiler
fault diagnosis
modelling
emulation
