Sparse Regularization Identification Of Structural Damages Based On Frequency Response Function

Structural damage identification(SDI)is one of important contents in field of structural health monitoring(SHM).At present,there are many SDI methods.But,the accuracy of damage identification results is still needed to be improved under environmental noise.In this study,the sparse regularization method is adopted to identify structural damage based on the theory of frequency response function(FRF).The main research contents are as follows:(1)SDI methods in recent years are reviewed.The general ideas and development statuses of common SDI methods are introduced and summarized,especially,FRF-based SDI methods are reviewed in detail.(2)A sparse regularization method is proposed for SDI based on FRF.Acceleration FRF is deduced in detail based on the basic dynamic equation,the FRF matrix is composed of FRF amplitude values at frequency points selected.Further,a relationship between global stiffness variation before and after structural damages and FRF is established.A new objective function is defined by combining the relationship with the sparse regularization method,and it is used to determine the locations and degrees of structural damages by estimating the element damage coefficients in the objective function via the particle swarm optimization(PSO)algorithm.The structural response data in the vertical direction of beam-like structure are usually measured.Combining the acceleration FRF with the dynamic reduction theory is proposed here.The dimension of both system stiffness matrix and mass matrix is reduced and equal to the total element number of structural finite element model.This is not only beneficial to collecting the measured response data,but also to improving the calculation efficiency.(3)Numerical simulations are carried out to assess the effectiveness of the proposed SDI method.Some finite element models about beam-like structures,such as cantilever beam,simply-supported beam and truss structure,are established respectively.A few structural damage cases are simulated and effectively assessed bythe proposed SDI method through numerical simulations under different parameters,such as measurement noise,sparse regularization method and PSO parameter selection.(4)Some experiment studies are carried out for further assessment of the proposed SDI method.A cantilever beam model is fabricated in laboratory,and different damage cases are created on the cantilever beam structure.The structural response data are measured under each damage case by LMS system and used to update the intact finite element models.The effectiveness of the proposed SDI method is verified based on experimental data in laboratory.(5)The advantages and disadvantages of the proposed SDI method are summarized and the prospect are given for future work.