A Novel Time Series based Seq2Seq Model for Temperature Prediction in Firing Furnace Process

Abstract

It is important to maintain the temperature uniformly in the calcining process at the manufacturing site. However, it is not easy to maintain a constant temperature. To predict the temperature, it is necessary to predict the life of the heating element in the calcining process. The breakdown time of the process should be reduced by replacing the heating element at an appropriate time before the life of the heating element is reached. In this paper, to predict the temperature of the firing furnace used in the manufacturing process, the temperature of the firing furnace was predicted using the Sequence to Sequence (Seq2Seq) model, which is often used for machine translation. The implemented model proved to have very high accuracy.