基于机器学习的复材装配制孔微损伤监测方法研究  

Research on Monitoring Method of Micro-damage in Composite Assembly Drilling Holes based on Machine Learning

作　　者：乐洪博 王宇宁 谢大叶 LE Hongbo;WANG Yuning;XIE Daye(The First Military Representative Office of Ministry of Air Force Equipment in Shenyang,Shenyang 110850,China;Engineering Technology Center,AVIC Shenyang Aircraft Industry(Group)Co.,Ltd.,Shenyang 110850,China)

机构地区：[1]空装驻沈阳局驻沈阳地区第一军事代表室,辽宁沈阳110850 [2]航空工业沈阳飞机工业(集团)有限公司工程技术中心,辽宁沈阳110850

出　　处：《新技术新工艺》2024年第8期49-53,共5页New Technology & New Process

摘　　要：研究旨在开发一种基于机器学习的复合材料装配制孔微损伤监测方法,针对碳纤维增强复合材料(CFRP)在制孔加工过程中容易产生的分层损伤等问题,通过采集力学信号、声发射信号和温度信号方式,利用机器学习模型实现对碳纤维增强复合材料(CFRP)制孔微损伤的实时监控。研究内容包括理论研究、实验研究、信号处理、模型构建及监测系统设计等,从而形成一套基于工艺参数及刀具角度耦合约束下的复材制孔微损伤控制方法,并达到一定的技术指标,为后续研究指明了方向。Aimed to develop a composite material assembly hole micro-damage monitoring method based on machine learning.In view of the problems of delamination damages which were easily caused during the process of drilling holes in carbon fiber reinforced composite materials(CFRP),it was collected that mechanical signals,acoustic emission signals,and temperature signals,and machine learning models were used to achieve real-time monitoring of micro-damage in the drilling process of carbon fiber reinforced composite materials(CFRP).The research contents included theoretical research,experimental research,signal processing,model construction,monitoring system design and so on,so as to form a set of composite material micro-damage control methods based on process parameters and tool angle coupling constraints.It achieved certain technical indicators,and pointed out the direction for future research.

关 键 词：碳纤维增强复合材料(CFRP) 制孔微损伤监测 机器学习 损伤抑制方法 神经网络 遗传算法 

分 类 号：TG659[金属学及工艺—金属切削加工及机床] TB33[一般工业技术—材料科学与工程] TH162[机械工程—机械制造及自动化] TP206[自动化与计算机技术—检测技术与自动化装置]