摘要
In order to improve the control precision of strip coiling temperature for hot strip mill,the BP neural network was combined with mathematical model to calculate convective heat-transfer coefficient of laminar flow cooling.The off-line calculated results indicate that the standard deviation of coiling temperature prediction is reduced by 22.84 % with the convective heat-transfer coefficient calculated by BP neural network.The prospects of this method for online application are bright.This method is more helpful to increasing the control precision of coiling temperature for hot strip steel.
In order to improve the control precision of strip coiling temperature for hot strip mill,the BP neural network was combined with mathematical model to calculate convective heat-transfer coefficient of laminar flow cooling.The off-line calculated results indicate that the standard deviation of coiling temperature prediction is reduced by 22.84 % with the convective heat-transfer coefficient calculated by BP neural network.The prospects of this method for online application are bright.This method is more helpful to increasing the control precision of coiling temperature for hot strip steel.
基金
Item Sponsored by National Natural Science Foundation of China(50104004)
