The intelligent production of sheet metal is a comprehensive technology involving control science, computer science, and sheet metal production. Intelligent rolling is an important development in the production process of sheet metal. In this paper, a new symmetrical four-roller bending (SFRB) process is introduced, which consists of feeding, pre-bending, reverse roll bending, second bending, and forward roll bending and unloading. A control strategy is proposed for the process, including on-line monitoring of curvature, on-line identification of the springback law, on-line prediction of final reduction, and control strategy. A convenient and reliable on-line curvature monitoring method is proposed. The quantitative relationship between the reduction and the curvature, in the form of a quadratic function, was established by physical experiments and numerical simulation, and the online identification of the springback law was realized. An on-line prediction method of the final reduction is proposed, and the determination principle of the reduction of three pre-bending processes is given. Finally, the control strategy of the SFRB process was verified by physical experiments. The relative error of the curvature radius of the final formed parts can be controlled within 0.8%. This research provides new insights into intelligent rolling.
Roll-bending technology has a high degree of flexibility and does not require special molds. However, based on the existing plastic mechanics theory and finite element simulation, it is difficult to accurately analyze the complex spatial relationship of profile roll forming. Therefore, a fixed-curvature prediction model is constructed based on XGBoost (extreme gradient boosting), and the coupling effect of the process parameters and material performance parameters on the roll-forming process is explored. Combined with a Bayesian optimization algorithm, the hyperparameters of the fixed-curvature prediction model are optimized. In addition, based on the prediction result of the fixed curvature, a variable-curvature prediction model is established using the conditional random field (CRF). To further improve the prediction accuracy, an error compensation network is added after the result of the CRF in order to map the discrete sequence to the continuous sequence. The experimental results show that the mean square error (MSE), mean absolute error (MAE), and mean absolute percentage error (MAPE) predicted by the models above are much smaller than other methods, which verifies the superiority of the prediction models.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.