Lateral spread prediction based on hybrid CNN-LSTM model for hot strip finishing mill

•A new deep learning network is built to predict lateral spread in hot strip mill.•The Pearson correlation analysis is used to select factors that affect lateral spread.•Using Convolutional Neural Network (CNN) to extract key features from data.•Using Long Short-Term Memory (LSTM) Network to learn more information from the features for lateral spread prediction.•The CNN + LSTM model achieves highest prediction accuracy among algorithms. During the hot strip rolling process, the material undergoes thermoplastic deformation. And the product’s quality is significantly impacted by the width variation of materials. Traditionally, the lateral spread is calculated based on empirical formulas derived from the manufacturing process. However, there exists a significant discrepancy between the empirical value and actual value. Based on the rolling parameters, a hybrid neural network model integrating Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) Network was developed to enhance the prediction accuracy of the lateral spread in this paper. Pearson correlation analysis was used to analyze the relationship between rolling parameters and the lateral spread. Parameters with high correlation were selected as input variables for the model. The performance of the proposed model is better than other three models, achieving a lower Root Mean Square Error (RMSE) value of 1.0816 mm and a higher prediction accuracy of 87.2596 %. The hybrid model provides a new insight for the improvement of the strip width deviation feedback system, which can improve the economic efficiency.