空军工程大学学报2024，Vol.25Issue(2)：85-94,10.DOI:10.3969/j.issn.2097-1915.2024.02.010
基于DBSCAN算法与GMM理论的铝合金板孔边疲劳裂纹萌生监测
A Methodof Fatigue Crack Initiation-Monitoring at Hole Edge of Aluminum Alloy Plate Based on DBSCAN Algorithm and GMM Theory
摘要
Abstract
In view of hole edge crack initiation-monitoring of porous aluminum alloy plates in aviation structures under fatigue load,a method of hole edge crack initiation-monitoring is proposed based on the fiber optic sensing system in combination with wavelet decomposition,density based spatial clustering of applications with noise and Gaussian mixture model.Firstly,a fiber optic grating sensing system is used to collect the micro strain at the edge of a porous aluminum alloy plate during the entire process from crack initiation to structural fracture under cyclic loading conditions,and a micro strain curve at the edge of the hole is constructed.Wavelet decomposition on the micro strain curve at the hole edge is performed to ob-tain the low-frequency and high-frequency components of the micro strain curve,and the minimum value of the low-frequency component and the sudden change of the high-frequency component are taken as the characteristics of crack initiation at the hole edge.The crack initiation being analyzed,the number of fa-tigue loading cyclic number is calculated by DBSCAN algorithm and GMM theory when the crack at the hole edge is initiated,and the initiation position of the main crack at the hole edge of the porous aluminum alloy plate and the number of fatigue loading cyclic number are compared and analyzed when the main crack at the hole edge is initiated.The test results show that this monitoring method is capable of accu-rately locating the initiation position of the main crack at the hole edge,and calculating the fatigue loading cyclic number when the main crack at the hole edge initiates,and the calculation error of the fatigue load-ing cyclic number is within the range of 5％.In the future,the method can be applied to the ground fa-tigue test of a whole-aircraft,structural health monitoring and other scenarios.关键词
多孔铝合金板/光纤传感/孔边裂纹/裂纹萌生/DBSCAN算法/GMM理论Key words
porous aluminum alloy plate/fiber optic sensing/cracks at the edge of the hole/crack initia-tion/DBSCAN algorithm/GMM theory分类
航空航天引用本文复制引用
于翀,宋昊,刘春红,赵启迪,付佳豪..基于DBSCAN算法与GMM理论的铝合金板孔边疲劳裂纹萌生监测[J].空军工程大学学报,2024,25(2):85-94,10.基金项目
航空科学基金(20185644003) （20185644003）
