An intelligent prediction method for production speed of corrugated cardboard production line
HUANG Congqi
JIANG Mian
HUANG Wei
XIE Weiwei
Abstract：The wet-part production speed of corrugated board lines is an important indicator representing the production efficiency.Its intelligent prediction can guide the enterprises to reasonably arrange the production and enhance the control level of the production line,which is of great significance for the efficient and green production of corrugated board.Firstly,data cleaning is performed on multi-type sampled data.The Bessel filters and quartile statistics methods are used to select the data of production speed in the stable interval,and the production parameters corresponding to B-corrugated and BC-corrugated are extracted respectively.Secondly,a BP neural network(GELU as the activation function)and a LightGBM(Light Gradient Boosting Machine)model are established to predict the production speed of the corrugated board line,and the hyperparameters of the two models are optimized by the Bayesian optimization and grid search respectively.The prediction results of the two models are combined by the grey wolf optimization algorithm(GWO).The results show that compared with the BPNN-XGBoost combined model the proposed method can significantly reduce prediction time while maintaining the same prediction accuracy,which greatly facilitates the online prediction of production speed in corrugated board lines.
Keywords：corrugated cardboard production linewet end production speeddata-driven modelingGrey Wolf Optimization algorithm
Publication Date：2025-09-30
Online Publishing Date：2025-10-15(First online date of this platform, not the publication date of the document)
Pages：8( 25-31,38 )