Mechanical Property Prediction of Hot-rolled Strip by Intelligent Correction Network

Based on physical metallurgy and neural network, intelligent prediction models of mechanical property in hot strip mill were developed. A new idea of intelligent correction about mechanical property was proposed. Physical metallurgy models calculated the base value, and the deviation of predicted value with measured value under different technology conditions was obtained by neural network. The simulation indicates that predicted and measured results are in good agreement and the relative error is very low. For 88% yield strength and 98 % tensile strength results, the error is within plusmn3 %, and for 80 % elongation results it is within plusmn6%