基于小波包能量曲率差法的桥梁损伤识别试验研究  

Experimental study on bridge damage identification based on wavelet packet energy curvature difference method

作　　者：余竹[1,2] 夏禾[1] Goicolea J M 战家旺[1] 

机构地区：[1]北京交通大学土木建筑工程学院,北京100044 [2]Technical University of Madrid, Madrid 28040

出　　处：《振动与冲击》2013年第5期20-25,共6页Journal of Vibration and Shock

基　　金：国家自然科学基金(51078029);国家重点基础研究计划973项目(2013CB036203);中央高校基本科研业务费专项资金资助项目(2012JBM080)

摘　　要：针对各种损伤识别指标中,小波包能量曲率差法仅有数值仿真、未经试验验证,尤其缺少实际工程验证问题,用数值模拟验证该方法识别结构损伤的有效性;利用沧州子牙河新桥替换下的梁体,进行两种工况损伤模拟。通过测试完好与损伤状态各点加速度响应,用小波包能量曲率差法识别损伤,考察小波函数和分解层数对识别效果的影响。结果表明,该方法有效,并可应用于实际工程。Among various damage identification methods, the Wavelet Packet Energy Curvature Difference （WPECD） method is an effective one. As a regret, most of the existing methods rely only on numerical simulation and are unverified via experiment, and moreover very few of them has been applied to practice. In the paper, the validity of WPECD in structural damage identification was verified by a numerical example. A damage simulation experiment was taken on a real replaced girder at the Ziya River New Bridge in Cangzhou. Two damage cases were simulated and the acceleration responses at the measuring points were obtained, based on which the damages were identified with the WPECD method, and the influences of wavelet function and decomposition level were studied. The results show that the WPECD method can identify structure damage efficiently and can be put into practice.

关 键 词：小波包分析 小波包能量曲率差 损伤模拟 损伤识别 试验 

分 类 号：O327[理学—一般力学与力学基础]