Lamb Wave Propagation-based Damage Identification for Quasi-isotropic CF/EP Composite Laminates Using Artificial Neural Algorithm: Part I - Methodology and Database Development
Lamb Wave Propagation-based Damage Identification for Quasi-isotropic CF/EP Composite Laminates Using Artificial Neural Algorithm: Part I - Methodology and Database Development
复制标题
DOI：
10.1177/1045389x05047599
复制
发表时间：
2005-02
影响因子：
2.7
通讯作者：
Z. Su;L. Ye
中科院分区：
文献类型：
--
作者：
Z. Su;L. Ye
提出了一种基于Lamb波的损伤识别方法和基于压电致动器-传感器网络的结构健康在线监测系统。将该方法应用于CF-EP准各向同性层压板的通孔型缺陷的定量诊断中，并应用了人工神经网络算法。为此，考虑了多种具有随机损伤的复合材料层合板，并进行了相应的三维动态有限元模拟。为了描述压电致动器激发的Lamb波，建立了压电致动器和传感器与复合材料层合板耦合的模型。设计了一种基于小波变换的信号处理包（SPP），对采集到的波信号进行净化，进一步在时间尺度域上提取Lamb波的能量谱特征。引入“数字损伤指纹”概念，构建损伤参数数据库（DPD），并利用该数据库离线训练多层前馈神经网络，并采用误差反向传播（BP）神经算法进行监督。这样的识别技术然后被验证，将在本研究的第二部分描述。
A guided Lamb wave-based damage identification scheme and an online structural health monitoring (online-SHM) system with an integrated piezoelectric actuator-sensor network are developed. The proposed methodology is applied to the quantitative diagnosis of through-hole-type defect in the CF-EP quasi-isotropic laminate with the aid of an artificial neural network algorithm. For this purpose, a variety of composite laminates with stochastic damages are considered, and the corresponding three-dimensional dynamic FEM simulations are conducted. To describe a Lamb wave excited by the PZT actuator, models for both the piezoelectric actuator and sensor coupled with the composite laminates are established. A wavelet transform-based signal processing package (SPP) is devised to purify the acquired wave signals, and further extract characteristics from the energy spectra of Lamb waves over the time-scale domain. A concept of ‘digital damage fingerprints’ is introduced, with which a damage parameters database (DPD) is constructed and used to offline train a multilayer feedforward neural network, supervised by an error-back propagation (BP) neural algorithm. Such an identification technique is then validated, to be described in the second part of this study.