基于模糊Petri网的高炉故障诊断系统  

Blast Furnace Fault Diagnosis System Based on Fuzzy Neural Petri Net

作　　者：刘静[1] 潘炼[1] 

机构地区：[1]武汉科技大学信息科学与工程学院冶金工业过程系统科学湖北省重点实验室,武汉430081

出　　处：《计算机测量与控制》2013年第5期1157-1159,1162,共4页Computer Measurement &Control

基　　金：武汉科技大学绿色制造和节能减排科技研究中心开放基金(A1024);湖北省冶金过程系统科学重点实验室(武汉科技大学)(Y201112)

摘　　要：采用模糊神经Petri网专家系统来对高炉进行故障诊断,结合神经网络自适应、自学习和模糊控制逻辑推理强、能够表达模糊语义的优点,建立了一种模糊神经Petri网故障诊断模型;针对高炉故障特点对先兆预测分析原理进行了分析研究,提出了自己的设计方法;在分析模糊Petri网及神经网络的优点的基础上,构建了基于模糊神经Petri网的专家诊断系统,通过测试风量,风压,喷吹量,炉顶温度,炉身温度,透气性指数得到诊断结果,使得诊断推理与网络学习更为完善,并运用VC++6.0中MFC功能设计了一套与前面诊断系统相对应的仿真程序,证明了该诊断模型的可行性。taking use of a kind of expert system based on fuzzy neural Petri net for fault diagnosis. Considering that neural network has advantage of adaptive, sell--learning and fuzzy control of high logical reasoning and that can express the fuzzy semantic, we also establish a kind of diagnostic model based on fuzzy neural Petri net. For the blast furnace has such characteristic that we take use of it into research on aura forecast analysis principle. Considering the advantage of both fuzzy Petri net and neural network, we construct a set of expert diagnosed system based on fuzzy neural Petri net. We take use of MFC of VC＋＋6.0 to design a set of simulation program corresponds to previous diagnostic system. To do so, we can prove the feasibility of the diagnostic model.

关 键 词：模糊神经Petri网络 专家系统 故障诊断 

分 类 号：TP302[自动化与计算机技术—计算机系统结构]