首页|时变影响下导波健康HMM的损伤定量诊断方法

时变影响下导波健康HMM的损伤定量诊断方法

扫码查看
原文链接
  • 万方数据
  • 维普
疲劳裂纹是导致航空飞行器结构失效最主要的损伤形式之一,及时、有效地对疲劳裂纹扩展进行在线监测和诊断对保障结构的完整性及安全性具有重要意义.由于实际服役的航空结构往往经历复杂的服役环境和操作条件,这些时变因素相互耦合,容易引起监测信号特征发生明显不确定性变化,严重影响结构损伤诊断的可靠性.针对时变服役条件影响下航空结构损伤的可靠评估,提出一种基于导波健康隐马尔科夫模型(Hidden Markov model,HMM)的结构损伤定量诊断方法.该方法通过HMM模型表征航空结构健康状态下的导波信号特征,采用实时监测导波信号特征与HMM模型的概率迁移关系表示结构损伤状态,实现结构损伤预警及定量诊断.在批量耳片的疲劳裂纹扩展试验中进行方法验证,结果显示损伤预警诊断误差平均值为0.4 mm,损伤定量诊断评估的最大绝对误差平均值为0.9 mm,表明该方法能够及时预警监测结构中疲劳裂纹的产生,同时对裂纹扩展进行可靠的连续定量诊断.
Quantitative Diagnostic Method of Damage to the Health Baseline GW-HMM under Time-varying Influence
Fatigue crack is one of the most important forms of damage that cause the failure of aircraft structures,and timely and effective online monitoring and diagnosis of fatigue crack propagation is of great significance to ensure the integrity and safety of the structure.Due to the complex service environment and operating conditions,these time-varying factors are coupled with each other,which are easy to cause obvious uncertain changes in the characteristics of monitoring signals,and seriously affect the reliability of structural damage diagnosis.Aiming at the reliable evaluation of aircraft structural damage under the influence of time-varying service conditions,a method of structural damage quantitative diagnosis based on guided wave health Hidden Markov model(HMM)was proposed.In this method,the guided wave signal characteristics under the health state of aircraft structures are characterized by HMM model,and the structural damage state is represented by the probabilistic migration relationship between the real-time monitoring of guided wave signal characteristics and HMM model,so as to realize structural damage warning and quantitative diagnosis.In the batch bolts of fatigue crack propagation experiments verified the method,the results show that the damage to the early warning and diagnosis of average error is 0.4 mm,damage quantitative diagnostic assessment of maximum average absolute error is 0.9 mm,shows that the method can timely early warning and monitoring of fatigue cracks in structure,reliable continuous quantitative diagnosis for the crack extension.

structural health monitoringguided wavestime-varying conditionsquantitative diagnosis of injurieshidden Markov modellikelihood probability

任元强、严嘉慧、袁慎芳、张巾巾

展开 >

南京航空航天大学机械结构力学及控制国家重点实验室 南京 210016

结构健康监测 导波 时变条件 损伤定量诊断 隐马尔科夫模型 似然概率

国家自然科学基金创新群体国家自然科学基金青年国家自然科学基金面上江苏省自然科学基金青年江苏高校优势学科建设工程资助项目

519210035190526652275153BK20190418

2024

10.3901/JME.2024.16.054

机械工程学报
中国机械工程学会

机械工程学报

CSTPCD北大核心
影响因子：1.362
ISSN：0577-6686
年,卷(期)：2024.60(16)