超声衍射时差法近表面盲区减小算法研究  

Research on algorithm for reducing subsurface blind zone in TOFD technique

作　　者：程江 张旭 涂君[1] 廖春晖 CHENG Jiang;ZHANG Xu;TU Jun;LIAO Chunhui(Hubei Key Laboratory of Modern Manufacturing Quantity Engineering,School of Mechanical Engineering,Hubei University of Technology,Wuhan 430068,China)

机构地区：[1]湖北省现代制造质量工程重点实验室湖北工业大学机械工程学院,湖北武汉430068

出　　处：《传感器与微系统》2022年第6期48-51,60,共5页Transducer and Microsystem Technologies

基　　金：国家自然科学基金青年科学基金资助项目(51807052,51707058);湖北省自然科学基金创新群体资助项目(2019CFA021)。

摘　　要：针对由于近表面盲区的存在,导致超声衍射时差(TOFD)技术无法检出近表面缺陷的问题,提出利用经验模态分解(EMD)和希尔伯特(Hilbert)变换对近表面缺陷信号进行提取,并针对传统经验模态分解存在的端点发散问题提出了基于自回归(AR)模型的端点延拓解决算法。利用本征模态函数(IMF)的能量和频率分布特点选择重构模态函数进行缺陷信号重构。利用改进后的经验模态分解和缺陷重构算法进行仿真设计和实验验证,结果表明:该算法可辨识的最小近表面缺陷深度可达4 mm,测量相对误差小于3.7%。Aiming at the problem that due to the existence of subsurface blind zone,traditional time of flight diffraction(TOFD)technique can not detect defects near the surface.A method based on empirical mode decomposition(EMD)and the Hilbert transform is proposed to extract defect diffraction signal from the received signal.To solve the endpoint divergent problem of traditional empirical mode decomposition,an advanced endpoint extension approach based on auto regressive(AR)forecasting model is put forward.The choosing of reconstruction signal is based on energy distribution and frequency characteristics of intrinsic mode function(IMF).With the improved empirical mode decomposition and defect reconstruction algorithm,simulation work and experimental verification are done.Results show the algorithm can identify the subsurface defect depth of up to 4 mm,while the relative measurement error is less than 3.7%.

关 键 词：超声衍射时差 自回归模型 重构算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TH878[自动化与计算机技术—计算机科学与技术]