Predicting Temperatures Inside a Steel Slab Reheating Furnace Using Neural Networks

The steel industry, characterized by complex and costly production processes, stands to gain significantly from the integration of intelligent systems for automation. This study details the development of a data-driven computer system, leveraging artificial neural networks, specifically for predicting temperatures in slab reheating furnaces. Recurrent Artificial Neural Networks (RNNs) have been extensively studied for their ability to make predictions based on historical sequences with temporal dependencies, which is ideal for monitoring industrial process variables. This research evaluates the performance of several predictive neural models, including Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and Temporal Convolutional Network (TCN). It also delves into data preprocessing and hyperparameter tuning to enhance model accuracy. Prediction accuracy was assessed using metrics such as Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). The models were tested across various prediction horizons to determine their effectiveness in delivering accurate forecasts several steps ahead.