基于神经网络与遗传算法的铝合金激光摆动焊工艺参数优化  

Optimal Design for Laser Oscillating Welding Process Parameter Based on Artificial Neural Networks and Genetic Algorithm for Aluminum Alloy

作　　者：梁超 张熊[1] 米高阳 王春明[1] LIANG Chao;ZHANG Xiong;MI Gaoyang;WANG Chunming(State Key Laboratory of Material Processing and Die&Mould Technology,School of Materials Science and Engineering,Huazhong University of Science and Technology,Wuhan 430074,China)

机构地区：[1]华中科技大学材料科学与工程学院材料成型与模具技术国家重点实验室,湖北武汉430074

出　　处：《电焊机》2022年第8期43-49,64,共8页Electric Welding Machine

基　　金：国家自然科学基金(52075199、52105338)。

摘　　要：铝合金焊接时容易产生气孔,严重影响焊缝的力学性能。激光摆动焊接工艺可以显著降低铝合金焊接过程中的气孔率,提高接头拉伸强度,但其工艺参数繁多且互相影响,很难直接对工艺参数进行优化。因此设计了18组正交试验,通过极差分析研究了不同工艺参数对气孔率的影响程度,并通过正交优化设计对工艺参数进行优化。通过不同的学习算法建立BP神经网络,结果表明使用BR算法的模型均方误差最小,预测性能最好。采用遗传算法结合BP神经网络对焊缝性能和焊接效率进行多目标优化,获得的焊缝拉伸强度相比正交优化所得焊缝提升了3.02%,焊接效率提升了18.3%。BP神经网络-遗传算法组合模型可在保证焊接性能的同时提高焊接效率。Aluminum alloys tend to produce pores during welding, which critically affects the mechanical properties of the weld. The laser oscillating welding can significantly reduce the porosity of aluminum alloy welding. However, it is very difficult to figure out and optimize the welding parameters since they are in a complex relationship. To study and figure out the influence of different parameters on porosity, 18 sets of orthogonal experiments are designed followed by the analysis of variance. In addition, the welding parameters are optimized by the analysis of variance. The multi-objective optimization of weld performance and welding efficiency are carried out by using genetic algorithm combined with BP neural network. The tensile strength of the weld optimized by genetic algorithm is increased by 3.02% and the welding efficiency is increased by18.3% compared with the orthogonal optimization. Therefore, the combined model of BP neural network and genetic algorithm can improve welding efficiency while ensuring welding performance.

关 键 词：激光摆动焊 正交试验 BP神经网络 遗传算法 参数优化 

分 类 号：TG457.14[金属学及工艺—焊接]