Research On The Application Of Intelligent Control On BOF Endpoint Dynamic Control Modeling

The aim of endpoint control of EOF (Basic Oxygen Furnace) is to secure hitting both the element contents and temperature targets by adjusting the amounts of oxygen and cooling materials when blowing process ceases. But the process of steel-making is so complex and the environment is so harsh that it is very difficult to measure the process signals continuously, therefore, we can't apply normal methods of process control on the blowing process. So accurate prediction of the endpoint's element contents and temperature becomes very important. It is the basis of endpoint control. At present, the most successful and widely applicable method in such field is the combination of static control and dynamic control based on the measured information of sublance.Strictly, the content of other elements, such as phosphor and manganese, should also be included in endpoint control, but the nuclear task should be the control of carbon content and temperature. Furthermore, according to the current level of equipment and technology of our country, it is more realistic to consider only such two targets in this stage. The most astringent fact is that the process of steel-making is a synthesis of multi-substance, multi-phase, high temperature and attended by many physical chemistry reactions. The research on the mechanism of reactions is not very clear. And there are severe nonlinear and coupling relation between the inputs and outputs. So the effects of normal models are always not ideal. But what's worthy to be felicitated is that the development of modern artificial intelligence provides a new way to solve such question.So after summarizing the former work of other people, this paper mainly discusses the dynamic control models of carbon content and temperature based on the sublance's information by applying intelligent control theory. The chief work can be included in such three aspects:1. The paper briefly introduces the basic concepts of endpoint control of BOF, systematically looks back the development and status quo of computer's endpoint control of BOF, and particularly discusses the construction methods of static control model and dynamic control model in detail.2. During the past years, Neural Network has made breakthrough both in the theory, implement method and many application fields. Especially, in the nonlinear modeling field, Neural Network displays unparalleled predominance. Comparedwith BP NN, RBF NN has more quick convergence velocity and can effectively avoid the problem of local minimum, but the selection of centers of RBF has become the bottleneck that retards the broader application of RBF NN. This paper adopts ROLS to select the centers of RBF NN, and constructs an effective dynamic control model of carbon content.3. Fuzzy model is an application of fuzzy set and fuzzy logic in the modeling process of system. It provides a new way for the modeling of complex and uncertain system. LM fuzzy model has a strong ability in describing the uncertainty of system and has a more brief structure than other fuzzy model, so this paper adopts LM fuzzy model to approximate the temperature's complex changing law during the ultimate stage. The simulating results shows that the method is effective and completely may be used to practical BOF process.