Non-Linear Regression of Rolling Torque for Medium and Heavy Plate Based on Genetic Programming

Rolling torque calculation for the hot rolling of medium and heavy plates is studied.Analyzing and dealing with the field data of rolling in the processes of medium and heavy plate rolling applying the technique of GP,means of determination are obtained for computational model and correlation parameters about non-linear regression of the rolling torque.With genetic programming method those nearly correlative factors are found with changing rolling torque taking these factors as arguments.Then it will generate functional expressions in order to embody rules of change in rolling torques and setup the non-linear regression model for the rolling torque.Then algorithm design and programming testing can be done.Test results showed that the average relative error between the model calculation value and measured value for rolling torque was not more than 6.23%.The model has been used in LISCO for 2?800?mm medium plate rolling in practice.