摘要
利用正常蒸汽管网的实测数据作为神经网络学习样本,通过神经网络的学习不断调整通过传热理论建立的正常管网传热模型中的温度分布系数,由此建立一个逼近实际蒸汽管网模型的传热模型,并以此作为蒸汽管网的诊断系统的诊断标准.再结合数据采集系统和对管网实时运行状况模拟和故障报警系统,组成蒸汽管网的智能监测系统.最后,用VC编程实现了该系统.
A model very close to realistic vapor pipe-net is established for heat transfer. It is used as the standard of the diagnosis system of the pipe-net. This model is established by using the measured data of the normal vapor pipe-net as the studying specimen of a neural net and by the continual studying of this neural net to turn the temperature factor included in the model of the pipe-net without fault that is established by the theory of heat transfer. The intelligence supervision system of vapor pipe-net consists of the system of fault diagnosis, the system of data acquisition, the simulation system of the real time running of the pipe-net, and the fault warning system. The author accomplished this system by the programming of visual C++.
出处
《中国计量学院学报》
2005年第2期127-130,共4页
Journal of China Jiliang University
关键词
蒸汽管网
神经网络
智能诊断系统
vapor pipe-net, neural net, intelligence diagnosis system
