应用激光, 2010, 30 (6): 479, 网络出版: 2011-03-24
基于BP神经网络的光纤激光焊接工艺参数优化及性能预测
Optimization of Fiber Laser Welding Process Viarables and Performance Prediction Based on BP Neural Network
光纤激光焊接 BP神经网络 工艺优化 焊接质量 fiber laser welding Back Propagation(BP)neural network process viarables optimization welding quality
摘要
采用连续光纤激光器对304不锈钢薄板进行了焊接工艺研究;为了提高激光焊接质量,引入BP神经网络对激光焊接工艺参数进行了优化,建立了焊接质量指标与焊接工艺参数之间的神经网络预测模型,并利用神经网络模型选择了较优的工艺参数。实验结果表明采用该工艺参数进行激光焊接可获得成形良好、无缺陷的焊缝。神经网络的性能预测指标与实际值间的偏差小于5%,可用于激光焊接工艺设计。
Abstract
Welding process viarables of 304 stainless steel sheet welded by continuous fiber laser was researched in this paper. In order to improve the welding quality, Back Propagation(BP)neural network was used to optimize process parameters. BP neural network model was developed to express the relationship between welding parameters and weld quality index,and the effect of optimization were analyzed. The results show that the select process parameters for laser welding obtain good shaping, outweld defects in welded joints. The deviation between the performance prediction index of neural network and the actual value was less than 5%, so BP neural network can be used for laser welding process design.
*郭亮, 王少华, 张庆茂, 徐鹏嵩, 庞振华. 基于BP神经网络的光纤激光焊接工艺参数优化及性能预测[J]. 应用激光, 2010, 30(6): 479. Guo Liang, Wang Shaohua, Zhang Qingmao, Xu Pengsong, Pang Zhenhua. Optimization of Fiber Laser Welding Process Viarables and Performance Prediction Based on BP Neural Network[J]. APPLIED LASER, 2010, 30(6): 479.