Prediction of Strip Temperature on the Run Out Table Of A Hot Strip Mill Using A Neural Network

摘要

The aim of this paper has been to improve the temperature calculation used to temperature control of strip on the Run Out Table (ROT) cooling section. The control of the coiling temperature and the cooling patterns are important because of the physical transitions in steel. This paper proposes a novel neural model for describing temperature distribution of the strip in ROT. The proposed model takes into account variation of therm physical properties of strip with temperature. The cooling process is represented by the heat equation in Lagrange coordinates which is approximated by a set of ordinary differential equations. Differential equations that describe the mathematical model are solved by numerical methods, and make data for training the neural network model. Tests with various parameters in NN model show the level of accuracy of the model.