A Sparse Regularization Based Method Of Bridge Influence Line Identification And Its Engineering Application

The bridge structure has high construction cost and long service life.During its service period of decades or even hundreds of years,due to the coupling effects of various unfavorable factors,the bridge structure generates damage accumulation and resistance attenuation inevitably,which even results in catastrophic accidents in extreme cases.The bridge influence line(IL)shows its promising development prospects in the fields of bridge weigh-in-motion(BWIM),model updating,damage detection and conditon assessment of bridge.However,the premise for realizing these applications is that the bridge ILs can be obtained quickly and accurately.The IL identification from the measured bridge responses is an essentially inverse problem of parameter identification.In order to avoid the IL identification being too sensitive to disturbance factors such as measurement noise,it is crucial to handle the ill-posedness of the inverse problem properly.This paper focuses on the topic of "A Sparse Regularization Based Method of Bridge Influence Line Identification and Its Engineering Application",and mainly discusses the following three aspects:(1)The bridge IL identification based on the basis function representation and sparse regularization;(2)The adaptive knot placement of IL identification;(3)The development and engineering application of IL identification system for middle-small-scale bridges.Based on the basis function representation and sparse regularization,a new method of accurate bridge IL identification is proposed in this paper.In this method,the new expression form of ILs is established based on the basis function expansion method,then the optimal function is established based on the sparse regularization method and the weight coefficient of the basis function is solved.Then,through the numerical example of Hong Kong Stonecutters bridge and the scale model test of reinforced concrete(RC)continuous beam bridge in laboratory,it is proved that the proposed method has good feasibility and high accuracy.In order to optimize the bridge IL identification,the method of adaptive knot placement is proposed to realize the self-adaptive optimal placement of the knots between different ILs and within the IL.Through the finite element numerical simulation,by taking the stress ILs of the key bars on the main truss of Hong KongTsing Ma Bridge(TMB)as an example,the good effect of adaptive knot placement on improving the IL identification is verified.Then,the stress response and train load information measured by the structural health monitoring system(SHMS)of TMB are used to identify the stress ILs of different measuring points,which further validates the feasibility of adaptive knot placement for identifying the stress ILs of local components on large-span bridges.Finally,this paper introduces the development and engineering application of IL identification system for middle-small-scale bridges.Through the system hardware,the dynamic acquisition,wireless transmission and signal synchronization of various signals are realized,including the dynamic strain response,dynamic deflection response of bridge and the real-time position of mobile vehicle.The system software including multi-source signal display,IL identification and result visualization is formed by integrating the bridge IL identification method.Then,the feasibility of the practical application of the system is proved through field verifications on multiple middle-small-scale bridges.