Closed Blast Furnace Smelting Process Permeability Prediction Of Intelligent Integrated Modeling Study

The permeability in the Zinc-Lead smelting process of Imperial Smelting Furnace (ISF) is an important parameter that evaluates the production status. The fluctuating permeability often causes unstable production, which leads to a great loss. The accurate prediction of the permeability in the Imperial Smelting Furnace is significant to increase the output of ISF.The Zinc-Lead smelting process is very complicated where many uncertain factors can't be predicted. It is difficult to gain the precise information of measuring, which results in that the general methods are unfit to predict the permeability. The Grey theory Model is a self-adaptable system with the variable parameter and the invariable structure. Being a linear Model, it is not accurate to predict when the relation between the independent variable and the dependent variable is unknown and the data fluctuate seriously. As an intelligent modeling method, the Neural Network has the advantage of good ability to simulate nonlinearly, the self-adaptable and self-learning ability to the complicated and uncertain problem. With slow convergence, the algorithm of training in typical BP model is time consuming, which results in that the Neural Network model can't satisfy the real-time requirement of the system. So, an intelligent integrated model is established based on Grey theory Model and the Neural Network Model. By the means of data updating and multithread technique, the integrated model is self-modified on-line. The software practically runs on the site successfully, which proves that not only the result of the prediction can satisfy the precision which is required to guide the operation on the site, but also the integrated modeling method is practical, effective to the smelting process. The result of the prediction can be used to guide the practical production.The paper is organized as follows. In chapter 1 the existed problem are briefly introduced in the process of ISP Zinc-Lead smelting. In chapter 2 the mechanism knowledge of ISP smelting process with emphasis on the key factors to the permeability and the present control status are studied. In chapter 3 the model for the prediction of Permeability based on the Grey theory is established. In chapter 4 the Neural Network model is proposed. In chapter 5 an integrated model is presented which integrates with the Model-Grey Theory and the Model-Neural Network. In chapter 6 the design and application of the predict system is presented. Chapter 7 concludes the paper and presents the future research.