基于小波包和频响函数随机—模糊统计的斜拉桥损伤识别

Damage Identification of Cable-Stayed Bridge Based on Wavelet Packet and Random-Fuzzy Statistical of Frequency Response Function

针对单一方法损伤识别的缺点,提出了基于频响函数和小波包能量谱的斜拉桥损伤识别指标,结构损伤前后的频响函数作为信号,进行小波包能量谱分析。由于试验中的频响函数存在一定的随机模糊性,对完好状态下的10次频响函数值进行了随机模糊均值处理,得到的频响函数作为基准。独塔斜拉桥模型实验表明,斜拉桥的一处和两处的损伤都可以识别,损伤位置也能确定,锤击实验效果好于敲击桥墩激励的识别效果,通过仿真证明了连续梁桥跨中和支座附近的损伤也可以识别,并且外界激励对损伤指标的识别没有影响,为大型桥梁结构在线监测提供了一种方法。

Single method for structural damage identification is not sensitive to some damages,a comprehensive damage identification method in combination with wavelet packet and frequency response function（FSF）is proposed to cable-stayed bridge.FSF which comes from damaged and undamaged structure as signal to be carried out by wavelet packet.As there is random-fuzzy in FSF,10 times FSF from undamaged bridge are put into random-fuzzy mean processing,through experiment of one cable-stayed bridge,one and two damage can be identified,position can also be located,experiment effect of hammer is better than bridge pier excitation.The simulation proved that the damage either in support or in middle of beam can all be identified,and a way for the large bridge online monitoring is provided.