摘要
目的为了充分利用来自多传感器的冗余、带噪声数据,提高结构损伤识别的精度.方法利用小波包良好的时-频特性,首先用小波包分解对结构响应进行处理,提取信号的不同特征参数,然后利用不同的特征向量对结构分别进行损伤识别,最后应用融合技术对不同的识别结果进行融合处理.并用一个七层钢结构框架的多损伤识别验证了该方法的有效性.结果结果表明,该方法能够极大地提高了结构损伤识别精度.结论运用小波包分析提取信号的特征参数与数据融合技术进行损伤识别,并使二者有机的结合是结构健康监测与检测的有效途径与发展趋势.
In order to make full use of the redundant and noisy data from the multi-resources, and enhance the damage identification accuracy, a new damage identification method is proposed which takes advantage of the good time-frequency characteristic of the wavelet packet. In this method, the response signal is firstly decomposed so as to extract feature parameters through the use of wavelet packet. Then different feature vectors are adopted to identify the structural damage patterns. Finally, data fusion technology is employed to fuze the different identification results. A 7-floor steel structure is adopted to illustrate the efficiency of this method. The result shows that this method can enhance the identification accuracy greatly. Extracting the feature parameters by wavelet packet and identifying structural damages by data fusion as well as combining both of them are the effective way and prospective trend for structural health monitoring.
出处
《沈阳建筑大学学报（自然科学版）》
EI
CAS
2006年第6期881-884,共4页
Journal of Shenyang Jianzhu University：Natural Science
基金
国家自然科学基金(50408033)
辽宁省高等学校优秀人才支持计划(RC-05-16)
沈阳建筑大学省级重点实验室基金
关键词
小波包分析
特征向量
损伤识别
欧氏距离
数据融合
wavelet packet analysis
feature vector
damage identification
Euclid-distance
data fusion
