基于经验模态分解和相关系数对玻璃纤维增强聚合物复合材料板的损伤识别及扫查成像

Damage identification and scanning imaging of glass fiber reinforced polymer composite plates based on empirical mode decomposition and correlation coefficient

  • 摘要: 针对外界环境噪声等因素造成损伤因子不敏感,导致复合材料损伤识别困难和成像误差大等问题,提出了一种基于经验模态分解(Empirical mode decomposition, EMD)和相关系数的损伤因子。用空气耦合探头采集损伤前后的Lamb波信号进行EMD分解获取多个本征模态分量(Intrinsic mode function, IMF)。根据相关系数获取与信号相关性最大的IMF分量,并定义其能量值的相对变化为损伤因子。在模拟噪声环境前后,分别对玻璃纤维增强聚合物复合材料(GFRP)板中的分层缺陷进行识别和扫查成像,验证了该损伤因子的有效性。结果表明:信号经过EMD分解后,与其相关性最大的IMF分量对损伤最敏感,能够定义为识别损伤的损伤因子。将该损伤因子结合概率成像方法进行空耦Lamb波扫查,不仅能够有效对复合材料中的缺陷进行成像,而且在模拟强噪声环境中具有良好的抗噪性。

     

    Abstract: Aiming at the problem that the influence of environmental noise and insensitive damage factor make damage identification difficult and imaging error large for composite plate, a damage factor based on empirical mode decomposition(EMD) and correlation coefficient was proposed. The air-couple probe was used to obtain Lamb wave signal before and after the damage, and a group of intrinsic mode function (IMF) components of signal was obtained by EMD. According to the correlation coefficient, the IMF component which has the greatest correlation with the signal was obtained, and the relative change of its energy value was defined as the damage factor. The validity of the proposed algorithm was assessed by identifying damage and scaning imaging at composite plate. The results show that after EMD, the IMF component with the greatest correlation with the original signal is most sensitive to damage, and can be used as a damage factor to identify damage. Combining this damage factor with probability imaging algorithm for the air-coupled Lamb wave scanning can not only effectively image the defect in composite plate, but also has good noise resistance in simulated strong noise environment.

     

/

返回文章
返回