机器学习算法在焊接领域中的应用  

Application of Machine Learning Algorithm in Welding Field

作　　者：李驰[1] 李广军[2] 

机构地区：[1]四川大学锦城学院计算机科学与软件工程系,成都611731 [2]江苏理工学院汽车与交通工程学院,江苏常州213001

出　　处：《焊管》2016年第5期25-30,共6页Welded Pipe and Tube

摘　　要：针对机器学习算法已经渗透到了焊接领域的情况,在对近年来焊接领域相关文献深入研究分析的基础上,重点阐述了在焊接领域运用较多的神经网络、支持向量机和遗传算法。分析总结了机器学习算法在焊接领域的应用方式,包括标准应用、改进应用和交叉混合应用。同时也介绍了机器学习算法在建模预测与参数优化、路径规划与焊接顺序、过程控制与质量监测和缺陷识别与分类判定方面的应用情况。对机器学习算法在焊接领域的未来发展进行了展望。The machine learning algorithm penetrated in welding field, based on deep analysis on the relevant literature in welding field in recent years, it emphatically expounded artificial neural network, support vector machine and genetic algorithm which are widely applied in the welding field. Analyzed and summarized the application mode of machine learning algorithm in the welding field, including standard application, improvement application and cross mixed application. And it also introduced the application of machine learning algorithm in the aspects of modeling prediction and parameter optimization, path planning and welding sequence, process control and quality monitoring, defect recognition and classification criteria. It also carried out future development outlook for machine learning algorithm in the welding field.

关 键 词：焊接 机器学习算法 人工神经网络 支持向量机 遗传算法 

分 类 号：TG409[金属学及工艺—焊接]