A Predictive System for Blast Furnace Cooling Stave Using Heat Transfer Model and Neural Network

Abstract

Blast furnace cooling stave is an essential component for any steel manufacturing industry. It is used for maintaining the inner temperature profile and cooling of the refractory lining. The lifespan of any blast furnace depends on upon furnace cooling system. Therefore, it is important to monitor the internal thermal state of the blast furnace cooling stave. In this paper, a predictive model has been developed using heat transfer model and artificial neural network. Heat transfer model for cooling stave is modeled and analyzed using ANSYS software. A multi-layered neural network is developed to replace the unavoidable noise factor and to fill the deviation between a mathematical model of heat transfer and actual experimental data. All the data used for simulation and analysis purpose are obtained from Rourkela steel plant. The results of this predictive model are compared with the experimental data and result show the predicted value matches well with actual data. Its show the predictive model is effective.