The Condition Monitor And Diagnostic Of ANSTEEL Blast Furnace Blowers Base On The Neural Network

It is very important to monitor the condition of Machinery. It can help us not only to reduce the maintenance time and the maintenance costs, but also can improve diagnostic accuracy and quality of maintenance. And it can carry considerable economic benefits for us.ANSTEEL Iron Plant was established in July 2000, it is formed by the original Iron Plant and Sintering. Existing eight blast furnaces, the total volume of 20,191 cubic meters, it is largest iron works in Metallurgical industry in China. Blast furnace blower plays an important role in the furnace process, it is directly related to the production of Iron Plant.This thesis mainly study the Iron Plant blast furnace blowers. Basis on the related theory of Machinery vibration diagnosis and the system of integrated artificial neural networks to analyze and diagnose the status and faults of blowers. The system uses advanced hardware to detect the equipment and collect the data. In software, the spectrum analysis to identify the characteristics of blowers fault signal is used by the advanced analytical methods. Energy spectrum was as input of one of sub-neural, and common failures were as the output; the relative amount of vibration and oil temperature was as input of another sub-neural, and common failures were as the output. A comprehensive information is obtained to construct the two sub-neural networks and neural network training and to integrate diagnosis, obtained fault diagnosis system.In this thesis, a conclusion to measure the vibration data of blowers is given out based on the detection sensor. The results show that integrated neural network-based fault diagnosis method of blowers is accurate, practical.