基于EMD-ARMA模型的飞行器健康诊断  

Health diagnosis for aircraft based on EMD-ARMA

作　　者：刘延军[1] 刘铁良[1] 郑新起[2] 张庆[3] 

机构地区：[1]大庆石油学院计算机与信息技术学院,黑龙江大庆163318 [2]沈阳航空工业学院自动化学院,辽宁沈阳110136 [3]大庆油田技术培训中心学生工作部,黑龙江大庆163255

出　　处：《大庆石油学院学报》2010年第2期108-112,共5页Journal of Daqing Petroleum Institute

基　　金：航空科学基金资助项目(2007ZD54006)

摘　　要：为了有效地对飞行器的健康状况进行诊断,将EMD-ARMA模型引入到飞行器健康诊断中,提出基于AR-MA(n,n-1)模型参数的健康诊断方法.采用EMD模型将飞行器关键部件的声发射信号进行分解,得到多个内禀模态分量IMF,选取包含主要信息的IMF分量建立ARMA(n,n-1)模型,采用长自回归模型法进行参数估计,得到模型主要的自回归参数,绘出模型参数的细化图用以健康诊断.对某型号真实飞行器关键结构部件的健康监测实验表明,该方法可以有效地诊断出飞行器关键结构部件的疲劳裂纹.To effectively diagnose the aircraft structure components health status,a new kind of health diagnosis approach for the aircraft,based on EMD and ARMA model,is proposed in this paper.The advanced acoustic emission technique is used to monitor the aircraft stabilizer health state and get the AE information.And the AE signal is decomposed into the limited IMF by the EMD.Then the first two IMF components were used to set up ARMA (n,n-1) model with the method of long autoregressive model.Then we can get the auto-regressive parameters of the ARMA model,further we can draw the thinned figures of the auto-regressive parameters to diagnose the health status of aircraft.Experiments show that this method can effectively monitor the fatigue crack of the aircraft structure components.

关 键 词：ARMA模型 EMD 长自回归算法 健康诊断 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]