Search

During the slab sizing press (SP) process, the pressing force corresponds to the slab profile, which guides the production schedule design and the final profile control. To accomplish the prediction of pressing force for SP, an improved ensemble method based on chaotic Harris hawks optimizer (CHHO) and stacking is proposed. A mechanistic knowledge is introduced for feature selection that enhances the rationality of input features. Subsequently, 11 machine learning models are compared and 5 of them are selected as candidate learners for the stacking method. Based on the candidates learners, 8 stacking strategies are constructed, which the stacked model with extratree regressor, gradient boosted decision trees, and kernel ridge regression (KRR) as base‐learners and KRR as meta‐learner performs the best. The R2, MAE, mean square error, mean squared log error, and mean absolute percentage error for the test dataset are 0.9912, 0.0856, 0.0167, 0.0005, and 2.00%, respectively, and 95% of the prediction errors are less than 0.15 MN. Then, sensitivity analysis and predictive analysis based on Shapley Additive Explanations are performed to demonstrate the good alignment of the proposed model with physical reality. Furthermore, to cope with the complexity and uncertainty of the production process, the proposed CHHO‐stacking model and the kernel density estimation method are integrated to model the prediction interval. [ABSTRACT FROM AUTHOR]

The accuracy of flatness actuator efficiency (FAE) is a prerequisite to achieve high precision automatic flatness control. This study proposes an accurate acquisition approach of FAE based on the combination of simulation modeling and data‐driven modeling for flatness control of cold‐rolled strip. First, a 3D rolling feature space model is constructed according to the change interval of rolling process parameters based on rolling conditions. A 3D finite element simulation model is subsequently established to determine the prior value of actuator efficiency at each node in the feature space model. Then, to improve the accuracy of prior value, an online optimization model is developed by means of data‐driven mechanism. The sorting algorithm and the central limit theorem are introduced to filter the measured process data and calculate the variable weighting. The processed data are imported into the model to improve the fitting model. Finally, a linear output prediction model based on the trend extrapolation method is proposed to realize the accurate acquisition of FAE under any rolling conditions. The data analysis of the production process shows that the model is adaptive and accurate, which can realize high precision flatness control. [ABSTRACT FROM AUTHOR]