Fuzzy Identification Prediction And Control To B.F Ironmaking Process

Blast Furnace (BF) ironmaking process is highly complicated, whose operating mechanism is characteristic of nonlinearity, time lag, high dimension, big noise and distribution parameter etc. What's more, BF is an open system with heat transport and chemical reaction coupling. Study on BF ironmaking process with intelligent cybernetics and realization of intelligent control to it are the frontier in the field of metallurgic development.In present dissertation, No.1 BF (750m~3) at Laiwu Iron and Steel Group Co. and No.6 BF(380m~3) at Linfen Iron and Steel Co., which are the representation of medium-sized BF in China, were selected out as the studying objective.The complexity of BF ironmaking process was analyzed in detail in the first three chapters of this article, based on Fuzzy Mathematics and knowledge of ironmaking experts. The following conclusions were drawn: Not only many unconventionality problems (including prediction and diagnosis of abnormal BF state, diagnosis of BF equipment etc) exist in BF ironmaking process, but much fuzzy information is encountered in analysis BF ironmaking process (for example, identification, prediction and control of [Si]), which makes Simple-inference disabled, Membership Functions in Fuzzy Mathematics must be used to solved those problems. Many kinds of Membership Functions were constructed, for example, the relationship between hot metal temperature and it's confidence coefficients was confirmed with the method of three-dimensional membership function, and the amount of rules can be reduced through three-dimensional membership function. Nonparametric similar degree method was used in the fuzzy cluster of [Si], and the fuzzy similarity relation matrix of [Si] was presented, with the fuzzy entropy as the evaluation criterion.With the development of intelligent cybernetics, new and more efficient prediction and diagnosis models of abnormal BF state should be established on intelligent cybernetics, and it is the object of ironmaking process. Those models were presented in the fourth chapter of this paper. Some key parameters of BF state were analyzed in detail, based on fuzzy mathematics. And then, Fuzzy prediction model and fuzzy neural net diagnostic model of BF state were presented. Not only the internal logical-inference, but quantitative calculation of conformity and variation trend of BF state was given with those two fuzzy models, and they are operational models.Study on prediction and control models of temperature in BF is the most difficult problem in automation of BF ironmaking process and practical production. The exact...