基于模糊神经网络的供热管网故障损坏程度诊断分析

发布时间：2018-03-17 10:47

本文选题：供热管网　切入点：故障诊断　出处：《河北工程大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着经济的快速发展,城镇化集中供热规模不断增加,随之而来的是供热故障的发生。伴随计算机技术的不断进步,为了提高供热系统的经济效益和社会效益,利用智能化手段对集中供热系统进行实时监控和管理是现代发展的趋势。本文尝试用模糊神经网络诊断供热管网故障损坏程度,主要做了以下几方面的研究工作:本文总结供热管网故障诊断常用的智能方法,以及供热系统国内外发展现状、供热事故研究现状、供热管网故障诊断研究和进展。概述BP神经网络和模糊逻辑系统的基本理论知识。对邯郸市热力公司供热系统情况进行统计,通过举例供热管网故障案例证明预测供热系统故障的重要性,为供热管网运行提出建议。分析供热管网故障原因并提出应对故障的措施。采用BP神经网络为模型对供热管网进行诊断,运用MATLAB软件实现了模型的训练和仿真,结果证明了BP神经网络可以用于故障诊断,但是也发现了BP神经网路存在很多劣势。为了避免模型的缺点,本文决定将BP神经网络和模糊逻辑系统结合在一起用于供热管网故障诊断分析。运用隶属度函数将样本数据模糊化,根据模糊规则和模糊推理结合BP神经网络形成模糊神经网络结构,从而对供热管网进行诊断。以邯郸市热力管网故障损坏程度为例,输入因素为竣工时间、投运时间、管道管径,输出因素为故障损坏程度。运用MATLAB程序进行训练和仿真,仿真结果表明模糊神经网络比BP神经网络收敛速度快、准确率高,模糊神经网可以用在供热管网的故障损坏程度诊断。
[Abstract]:With the rapid development of economy, the urbanization of central heating scale increasing, followed by heating failure. With the continuous development of computer technology, in order to improve the heating system of the economic and social benefits, the use of intelligent means of modern trends in the development of central heating system for real-time monitoring and management. This paper attempts to use the fuzzy neural network fault diagnosis for damage, mainly do the following research work: This paper summarizes the commonly used methods of intelligent heating pipe network fault diagnosis, and the heating system at home and abroad, the research status of heating accidents, heating pipe network fault diagnosis research and progress. An overview of BP neural network and fuzzy logic system of the basic theory of knowledge. The statistics of the Thermotics Inc of heating system in Handan City, through the example of heating pipe network fault prediction system for heat proof case Fault importance, put forward the proposal for the heating network operation. The causes of malfunction of heat pipe network and put forward corresponding measures of failure. By using BP neural network to diagnose the heating network model, using the MATLAB software to realize the training and simulation model, results show that BP neural network can be used for fault diagnosis, but also found the BP nerve the Internet has many disadvantages. In order to avoid the shortcomings of this model, the BP neural network and fuzzy logic system are combined for analysis of heating network fault diagnosis. Using the membership function of the fuzzy sample data, according to the fuzzy rules and fuzzy inference BP neural network combined with fuzzy neural network structure, thus the diagnosis of heating network. To the extent of the damage fault heat pipe network Handan city as an example, the input factors for the completion time, operation time, pipe diameter, output factors for fault. The MATLAB program is used for training and simulation. The simulation results show that the convergence speed of fuzzy neural network is faster than that of BP neural network, and the accuracy rate is high. Fuzzy neural network can be used to diagnose the degree of fault damage in heating network.

【学位授予单位】：河北工程大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TU995.3

【参考文献】