Motor Load Balancing with Roll Force Prediction for a Cold-Rolling Setup with Neural Networks

The use of machine learning algorithms to improve productivity and quality maximize efficiency in the steel industry has recently become a major trend. In this paper, we propose an algorithm that automates setup cold-rolling process maximizes by predicting roll forces motor loads with multi-layer perceptron networks addition balancing increase production speed. proposed method first constructs multilayer models all available information from components, hot-rolling process, process. Then, variables related normal part set-up are adjusted balance among rolling stands. To validate method, used data set 70,533 instances 128 types steels 78 variables, extracted actual manufacturing was found be superior physical prediction model currently for setups regard accuracy, load balancing,