As for the precision rolling process, a thickness prediction model was constructed for precision rolling exit by introducing a time domain convolutional network algorithm. The feature information of time-series data of the precision rolling process was extracted by using this time-domain convolutional network model, and the prediction performance of the precision rolling exit thickness was improved by optimizing the structure and parameters of the model. The simulation results of the actual steel dataset show that the proposed time-domain convolutional network algorithm, compared to traditional methods, has significant advantages in evaluation indicators, such as root mean square error, average absolute percentage error, and coefficient of determination, which can provide critical information for decision of on-site engineers.

strip steel  /  hot rolling  /  thickness prediction  /  time-domain convolutional network  /  precision rolling process  /  time-series data  /  feature extraction  /  root mean square error