基于压电阻抗与改进马氏距离的木试件损伤检测试验研究

发布时间：2018-04-29 18:26

本文选题：压电阻抗 + 木试件　；参考：《华中科技大学》2015年硕士论文

【摘要】：压电阻抗技术近年来被广泛应用于土木工程结构损伤识别与健康监测领域,包括钢结构、混凝土结构以及铝合金结构的损伤检测。但是,关于将其应用于木结构的损伤检测方面的研究比较少。本文在压电阻抗技术已有的研究成果上,通过木试件损伤检测试验研究,探索了将压电阻抗技术应用于木结构损伤识别的可行性。本文的试验材料包括针叶类代表木材芬兰松和阔叶类代表木材巴劳木,损伤类型包括横纹损伤、顺纹损伤以及孔洞损伤。通过分别测量健康与损伤状态下试件的导纳曲线,研究压电阻抗技术对于木结构损伤识别的可行性。为定性判断损伤程度,除使用常规的相关系数(CC)、均方根偏差(RMSD)指标以外,还引入了马氏距离(Mahalanobis Distance)。由于马氏距离指标在本试验数据处理中存在计算结果不可靠等问题,本文对马氏距离进行了基于主成分分析的合理改进,建立了基于改进马氏距离(Modified Mahalanobis Distance)的损伤程度评价指标。根据本文研究内容,可得出以下结论:1、压电阻抗技术对芬兰松的损伤比巴劳木更为敏感。通过观察发现,芬兰松损伤检测试验中所有频率段的CC和RMSD均可有效反映损伤变化,而巴劳木损伤检测试验中存在少量工况的部分频率段不能反映损伤变化。2、不同损伤类型的损伤敏感频率段不一样。分析CC、RMSD指标可知,横纹、顺纹损伤在30kHz-50k Hz的损伤指标变化更显著;而孔洞损伤在30kHz-50kHz和150kHz-500kHz的指标变化更为显著。3、改进马氏距离融合了多频率段损伤信息,比CC和RMSD指标包含了更多的统计信息,可以从整体上衡量各工况之间的差异,为基于压电阻抗技术的木结构损伤检测提供更可靠的评价。综上所述,将压电阻抗技术应用于木试件损伤检测是可行的。
[Abstract]:Piezoelectric impedance technology has been widely used in the field of damage identification and health monitoring of civil engineering structures in recent years, including steel structure, concrete structure and aluminum alloy structure damage detection. However, there is little research on the application of it to the damage detection of wood structures. In this paper, based on the research results of piezoelectric impedance technology, the feasibility of applying piezoelectric impedance technology to the damage identification of wood structures is explored by means of the damage detection test of wood specimens. The experimental materials in this paper include coniferous pine and broad-leaved wood, and the damage types include transverse damage, collateral damage and hole damage. The feasibility of piezoelectric impedance technique for wood structure damage identification was studied by measuring the admittance curves of the specimens under the condition of health and injury respectively. In order to judge the degree of damage qualitatively, Mahalanobis distance was introduced in addition to the conventional correlation coefficient (CCN) and root mean square deviation (RMSD). Because the Mahalanobis distance index is not reliable in the data processing of this experiment, this paper improves the Markov distance reasonably based on principal component analysis, and establishes the damage degree evaluation index based on the improved Markov distance modified Mahalanobis distance. According to the present study, the following conclusions can be drawn: 1, piezoelectric impedance technique is more sensitive to the damage of Pinus Finland than Baraum. It was found that CC and RMSD of all frequency bands in the damage detection test of Finnish pine could effectively reflect the damage changes. However, in the damage detection test of Baroum, a small number of operating conditions can not reflect the damage change. 2. The damage sensitive frequency of different damage types is different. According to the analysis of RMSD, the damage index of cross stripe and parallel grain is more significant in 30kHz-50k Hz, while the index of hole damage in 30kHz-50kHz and 150kHz-500kHz is more significant. 3. The improved Markov distance fuses multi-frequency damage information. The index of CC and RMSD contains more statistical information, which can measure the difference between different working conditions on the whole, and provide a more reliable evaluation for damage detection of wood structure based on piezoelectric impedance technology. In conclusion, it is feasible to apply piezoelectric impedance technique to the damage detection of wood specimens.
【学位授予单位】：华中科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TU317

【引证文献】