Prediction Model And Online Self-learning System For Slab Re-heating Furnaces

Reheating furnace is an indispensable equipment of steel production process. The main function is to heat billet evenly to comply with the temperature requirements of rolling process. Furnace power consumption is large50%of energy on steel-rolling production, and20%of steel production, which affects directly product cost and power consumption. Therefore, reheating furnace optimization control can not only improve product quality and heating efficiency, but also reduce cost and power consumption.Prediction model is a key technology of the optimization control, which affects control accuracy directly. According to the characteristic and problems of steel production process, the paper proposed the corresponding solution of the prediction model and self-learning system. Main work is done in the paper as follows.(1) Based on the fundamental laws of thermodynamics, fully considered the heat transfer relationship in reheating furnace, the prediction model is established, which describes the multi-dimensional billet temperature distribution.(2) Because of the complexity of radiation and convection heat transfer, there are some inaccurate parameters in prediction model, which can reduce the model accuracy. In order to solve the problem, based on the incomplete information, using analysis of the model sensitivity, the self-learning system is designed.(3) Designed the reheating furnace control system platform with C#and Oracle software.