Fuzzy Mathematical Medols Study In BF Wall Temperature Monitor Control And Trouble Diagnoses

Metallurgical industry is the basic industry of national economy and Blast Furnace (BF) iron-making is the upper main procedure of metallurgical industry. To improve the production procedure and lower the cost of iron-making is significant for the development of metallurgical industry and economical use of resources. BF wall temperature monitor control and diagnoses can safeguard stable production, BF equipment longevity and safety of workers. So foreign steel corporations stress great emphasis on it in their development of Expert Systems. The instantaneity, instability and unpredictability of BF wall temperatures have stopped it from effective monitor control solely by Classical Cybernetics. To achieve this, it is needed to construct a model based on data analysis and fuzzy logic.Nowadays, fuzzy mathematics and fuzzy logic have been widely applied to engineering. In this dissertation, first, we built the Pattern Recognition logic diagnoses model of BF wall temperature and temperature-changing. Second, we built the fuzzy-synthetical functions, computed the weight coefficient by Analytical Hierarchy Process and worked out the evaluation of temperature alarm level. Third, we constructed transition and relieving threshold functions and logic consequence process, then built the alarm level transition-relieving model which is based on Effectual-time. At last, a model of holistic trouble diagnoses and four sub-models of trouble diagnoses were built by the interrelated data and new parameters which were given by the third model.The temperature data collected by Expert System from BF No.7 of Handan Steel Co. is used to evaluate the models. And two different temperature curves are selected to test the third models. In the same condition, the alarm level transition-relieving model have basically fulfilled the targets of enlarging dangerous signs and filtering of small fluctuation.