激光技术, 2006, 30 (5): 0471, 网络出版: 2010-06-03
基于神经网络的粉末冶金材料激光焊接工艺优化
Optimizing laser welding parameters of powder metallurgical material based on artificial neural network
材料 激光焊接 金刚石钻头 粉末冶金 人工神经网络 气孔缺陷 material laser welding diamond core driller powder metallurgy artificial neural networks porosity defect
摘要
针对薄壁金刚石钻头的激光焊接应用,采用德国LSM240全自动激光焊接系统进行单面焊接试验,建立了粉末冶金材料激光焊接工艺优化的误差反向传播人工神经网络模型,应用该模型研究了激光焊接工艺参数对气孔率和焊缝强度等焊接质量因素的影响,并对薄壁金刚石钻头激光焊接进行了工艺参数优化处理,获得了无气孔缺陷的优质焊接接头。结果表明,气孔率同激光功率、焊接速度之间具有幂函数关系;焊缝强度同激光功率、焊接速度之间具有高斯函数关系。
Abstract
The model for optimizing laser welding parameters of powder metallurgical material on back propagation(BP)artificial neural network was presented for laser welding application of thin-wall diamond core driller. The effects of laser welding parameters on porosity and welding strength were investigated with the model on the basis of one-side laser welding experiment by LSM240 auto-welding system. Laser welding parameters of thin-wall diamond core driller were optimized with the model,high quality laser welding seam with porosity defect are obtained. The results showed that the mathematic relation of the porosity with laser power and welding speed was power function and the mathematic relation of the welding strength with laser power and welding speed was Gaussian function.
唐霞辉, 秦应雄, 钟如涛, 周金鑫, 李正佳. 基于神经网络的粉末冶金材料激光焊接工艺优化[J]. 激光技术, 2006, 30(5): 0471. TANG Xia-hui, QIN Ying-xiong, ZHONG Ru-tao, ZHOU Jin-xin, LI Zheng-jia. Optimizing laser welding parameters of powder metallurgical material based on artificial neural network[J]. Laser Technology, 2006, 30(5): 0471.