基于BP神经网络的板材多点成型中的回弹预测
首发时间:2026-01-09
摘要:通过正交试验设计和极差分析方法,利用数值模拟软件对不同工况下板材对压成型的回弹进行了模拟,得出影响板材对压成型的回弹因素从大到小为设计半径、板材厚度、冲头排列间距和板材屈服强度,并找到了最优水平组合。采用BP神经网络预测并验证对压成型板材的回弹量,预测结果与有限元仿真值接近,具有一定准确性和适用性。
For information in English, please click here
Springback Prediction in Sheet Metal Forming
Abstract:Through orthogonal experimental design and range analysis methods, numerical simulation software was used to simulate the springback of sheet metal under different working conditions during compression molding. The factors affecting the springback of sheet metal during compression molding, from large to small, were determined as design radius, sheet thickness, punch arrangement spacing, and sheet yield strength, and the optimal horizontal combination was found. Using BP neural network to predict and verify the springback of pressed sheet metal, the predicted results are close to the finite element simulation values, with certain accuracy and applicability.
Keywords: Steel structure processing Multi-point molding Neural network Rebound prediction
基金:
引用

No.****
同行评议
共计0人参与
勘误表
基于BP神经网络的板材多点成型中的回弹预测
评论
全部评论