Damage Identification of Large Generator Stator Insulation Based on PZT Sensor Systems and Hybrid Features of Lamb Waves.

Damage Identification of Large Generator Stator Insulation Based on PZT Sensor Systems and Hybrid Features of Lamb Waves.
复制标题

基于PZT传感器系统和兰姆波混合特征的大型发电机定子绝缘损伤识别

DOI:
10.3390/s18092745
复制
发表时间:
2018-08-21
期刊:
Sensors (Basel, Switzerland)
影响因子:
--
通讯作者:
Hu B
Hu B
中科院分区:
其他
文献类型:
--
作者:
Li R;Li H;Hu B

文献摘要

相似文献

大型发电机是电力系统中的主要设备,其运行可靠性在很大程度上取决于定子绝缘。定子绝缘的损坏将逐渐导致发电机的故障和击穿。鉴于Lamb波在结构健康监测中的优势,本研究采用分布式压电(PZT)传感器系统和Lamb波的混合特征来识别大型发电机定子绝缘损伤。针对定子绝缘材料的非均匀性和各向异性,提出了一种分层概率损伤成像算法。该方法包括三个步骤:利用相关系数进行全局检测,利用飞行时间(ToF)随损伤散射兰姆波振幅进行局部检测,最后进行图像融合。利用小波变换在时频域上提取Lamb波的ToF。采用有限元模拟和实验方法,对定子绝缘内部空洞、内部分层、击穿和裂纹等4种典型损伤进行了验证。结果表明,该方法能准确识别定子绝缘损伤的位置,并能利用重建图像识别定子绝缘损伤的大小。
Large generators are the principal pieces of equipment in power systems, and their operation reliability critically depends on the stator insulation. Damages in stator insulation will gradually lead to the failure and breakdown of generator. Due to the advantages of Lamb waves in Structural health monitoring (SHM), in this study, a distributed piezoelectric (PZT) sensor system and hybrid features of the Lamb waves are introduced to identify stator insulation damage of large generator. A hierarchical probability damage-imaging (PDI) algorithm is proposed to tackle the material inhomogeneity and anisotropy of the stator insulation. The proposed method includes three steps: global detection using correlation coefficients, local detection using Time of flight (ToF) along with the amplitude of damage-scattered Lamb wave, and final images fusion. Wavelet Transform was used to extract the ToF of Lamb wave in terms of the time-frequency domain. Finite Element Modeling (FEM) simulation and experimental work were carried out to identify four typical stator insulation damages for validation, including inner void, inner delamination, puncture, and crack. Results show that the proposed method can precisely identify the location of stator insulation damage, and the reconstruction image can be used to identify the size of stator insulation damage.