Prediction of Rolling Force in the Hot Strip Rolling Using Support Vector Regression with Principal Components Analysis

Abstract

In this paper, a method is proposed to preprocess the data by using principal component analysis (PCA) to obtain the contribution degree of each variable to dependent variables in high-dimensional data, and then select an appropriate number of independent variables as input variables of a prediction model established by support vector regression (SVR) according to the contribution degree, and further predict the rolling force. The data are collected from plants of steel manufacture to ensure the practicability. The variables such as intermediate billet thickness, average strip width, target thickness, head finish inlet temperature and mill speed are selected as the independent variables input to the prediction model to predict the rolling force value. The overall performance of the model is evaluated by mean absolute error (MAE), root mean square error (RMSE), and time required for modeling and prediction. Both prediction accuracy and generalization ability of the proposed model have achieved good results. The proposed model provides a new method and idea for prediction research of rolling force in the hot strip rolling process.