A novel prediction scheme for hot rolled strip thickness based on extreme learning machine

Abstract

In order to predict the hot-rolled strip thickness, two extreme learning machine (ELM)-based thickness modeling algorithms based on clustering and differential evolution algorithm are proposed in this paper. These two kinds of modeling methods are used to predict the thickness, and the experimental results are compared with the standard ELM. The final results show that the two models proposed in this paper are better than the standard ELM model, and these two kinds of modeling methods can be selected according to different production conditions.