The Research On The Damage Identification Of The Girder Bridge Under Vehicle Loads Based On Lifting Wavelet

The traditional damage identification methods, which need artificial incentive,blocking traffic, a large number of sensors, cannot be widely applied in the actualproject. To overcome these disadvantages, the methods based on the lifting wavelet,using the acceleration response of the bridge under vehicle loads, are proposed in thispaper, which are convenient and suitable. The main contents include the followingaspects:1) According to the latest achievements in research at home and abroad, the bridgedamage detection method based on coupled vibration of vehicle-bridge can bedivided into three categories: the method using bridge dynamic response, the methodusing vehicle dynamic response, the method using both bridge dynamic response andvehicle dynamic response. The basic principles and engineering applicability ofvarious methods are introduced and evaluated in detail. Finally, the key problems anddevelopment direction of the future study of damage detection method based onvehicle-bridge are pointed out.2) The basic principle of wavelet transform and lifting wavelet are discussed indetail, and both advantages and disadvantages of them are introduced, as well as theapplication of wavelet analysis in signal processing and the selection of wavelet basisfunction.3) According to the dynamic response characteristics of damage bridges undervehicles load function, and magnifying function of lifting wavelet transform to signalbreak point, the structural damage identification method has been proposed by thedistribution characteristics of lifting wavelet transform coefficient of accelerationresponse of bridge under a moving load. Firstly, the acceleration response signal ofthe bridge subjected to a moving load is captured. Then the signal is transformed bylifting wavelet. The damage location is determined by spatial variation of liftingwavelet transform coefficients of acceleration response and acceleration responsedifference respectively. At the same time, the damage identification effect have been analyzed and discussed under the different speed of vehicle, the different damagelocation, the different damage severity and noise level. The following studied resulthas been shown. When the speed of vehicle is less than8m/s and the noise level isless than5%, the multiple site damage may been detected and determined by thelifting wavelet coefficients of an dynamic response signal of bridge under vehiclesload.4) Lifting wavelet energy spectrum entropy and lifting wavelet coefficient ofvariation are proposed combined with mutual entropy and correlation function theory,and the numerical example is done to validate the damage index, which suggests thatthe occurrence of damage and the damage location of bridge structure can be found.What’s more, lifting wavelet coefficient of variation, to some extent, can be used toindicate the damage degree, and finally the influence of measurement noise ondamage index is analyzed.