A hierarchical prediction method based on hybrid-kernel GWO-SVM for metal tube bending wrinkling detection

Metal bending tube is widely used in industry while its forming defects extremely affect the quality. Among all defects, bending-inside wrinkling caused by non-uniform compressive stress a zero-tolerated defect, particularly when for transportation. However, current detection approach, suffering from lack of insight into mechanism, normally posteriori. To obtain priori condition certain go-to-bend tube, we put forward metal hierarchical prediction method based on hybrid-kernel gray wolf optimizer (GWO) support vector machine (SVM). Three typical kernel combinations are utilized GWO-SVM model. verify proposed method, aluminum alloy series tubes tested. By constructing 12 designations tubes’ finite element simulation case base, model trained through three GWO-SVMs, respectively. The results compared with traditional SVM and GWO-SVM, which show that has best performance hierarchically predicting wrinkling. Analysis predicted shows relative wall thickness less than 0.015, very likely to occur any radius within range. On contrary, there tendency wrinkle. At same time, smaller R/D, higher hierarchy This lays foundation prevention.