Fully Automatic Modal Identification And Tracking Of Bridgs And Its Application In Bridge Engineering

Modal parameters are the inherent indices that can reflect the state of bridge structures.Continuously obtaining the long-term modal parameters is one of the the core requirements in bridge health monitoring.This paper focuses on the modal parameters based engineering application of bridge health monitoring.This study focuses on modal parameters based application and tries to improve the degree of automation,feasibility and robustness of modal identification and tracking.The main works in this study are summarizing the core problems of aitomatic modal identification and tracking and proposing new index and method to solve the problems in traditional methods.This study tries to remove the obstacles between theoretical study and engineering application and vlidate the feasibility through real engineering application.The main research contents and conclusions are summarized as follows:(1)An automatic modal identification method based on automatic interpretation of stabilization diagram is proposed.The whole process includes automatic clearing of stabilization diagram,automatic interpretation of stabilization diagram and automatic elimination of outliers in the identification results.Firstly,the theoretical proportional relationship between the number of real and spurious modes in the stabilization diagram is analyzed,and a new method of automatic clearing of the spurious modes is proposed.Then,an improved FCM method for automatic interpretation of stabilization diagram is proposed to automatically extract the physical modes in the stabilization diagram.Finally,the adjusted box plot is introduced to remove the outliers in the identification results.(2)The modal similarity criterion(MSC)for measuring modal similarity is proposed to solve the problem that the existing automatic modal identification method based on modal assurance criterion(MAC)might fail when few sensors are installed.The automatic clearing and interpretation of stabilization diagram are also improved to make the proposed method suitable for bridges with any stiffness and with any sensors deployed in monitoring.(3)A unified method of automatic modal tracking for brifges with ant sensors deployed based on MSC index is proposed.Three steps are including in the proposed automatic modal tracking method,which are updating the reference mode,evaluating the modal similarity based on MSC index and searching the maximum value of MSC for modal connection.The feasibility of the proposed modal tracking method is proved by theoretical analysis,and the feasibility of the proposed method is validated by long-term identification results of suspension bridge and continuously supported steel truss footbridge under different sensor deployments.(4)The causes of modal uncertainties are summarized.The modal uncertainty is divided into internal uncertainty from identification algorithmand external uncertainty from environmental variables.Based on long-term modal parameters of a suspension bridge,the statistical characteristics of operational modal uncertainties are analyzed,and adaptive bandwith kernel density estimation is introduced to analyse the probability distribution characteristics with consideration of internal and external uncertainties.(5)A method for automatic identification of vortex induced vibration(VIV)is proposed.A software module for automatic modal identification and tracking and automatic VIV identification is developed and applied to practical bridge projects.The feasibility of the proposed method is verified through the application to Hong Kong Zhuhai Macao Bridge,continuous automatic modal identification and tracking before,during and after the VIV of a suspension bridge and the continuous automatic modal identification and tracking of a suspension bridge.