Prediction of Strip Width in Roughing-Mill Group Based on PSO-LSSVM

Abstract

Aiming at the low accuracy of width prediction based on mechanism model in Roughing Mills of Hot Rolling Mills (RMHRM), this paper proposes a method of Particle Swarm Optimization (PSO) algorithm to optimize Support Vector Machine (SVM) prediction model parameters, in order to realize the real-time prediction of the strip width change in the rough rolling mill. Firstly, select input variables which are highly correlated with the width of the rough rolled strip, and build the Least Square Support Vector Machine (LSSVM) width prediction model. Secondly, this paper uses the PSO algorithm to optimize the kernel function parameters of LSSVM. Finally, take the 2050-model RMHRM in Shanghai as an example, and the on-site measured sample data is used for training and verification the model. Compared with traditional model prediction results, the prediction method of PSO-LSSVM is more effective.