Data Acquisition In Bridge Health Monitoring System: Study Of Theory And Methods

Bridges are key elements in the transportation system, but aging bridges which are normally designed to have long life span are facing a severe crisis. It is an important issue how to monitor these widely used bridges in order to prevent potential catastrophic events.Continuous health monitoring or regular condition assessment of important bridges is necessary so that early identification and localization of any potential damage can be made. Before this work can be done, however, the bridge data must be acquired reliably and precisely. This thesis is focused mainly on the precise acquisition of bridge deflections, followed by a brief discussion of the crack detection and identification of a bridge.In this thesis, we first introduce the theory of the generalized structural element as well as that of the extended morphological filtering. On this basis, four deflection-measuring systems are designed, which are outlined below.(1) A non-contact weighted-stretched-wire system is developed. The principle, organization and structure of the system are discussed. Using a stretched wire as the stationary reference line, this system employs the video capturing, digital image processing and recognition techniques, simultaneously. Some possible sources of the measurement error produced by the system are analyzed, and the corresponding strategies are proposed. Through a series of laboratory experiments concerning the precision achievable by the system as well as the coherence between the measured values obtained by our system and by a deflection gauge, it is shown that the system can acquire the deflections in a bridge reliably and with a high precision. The proposed system has been applied to a real-life box-girder bridge.(2) An adaptive, all-weather self-calibration system is advised, which can monitor and detect both the displacements and the deflections in a bridge, simultaneously. One remarkable advantage of this system is that the calibration process is automated.(3) A system is designed for bridge alignment. The system configuration and hardware are described, followed by a description of the alignment algorithm as well as the relevant filtering algorithms used. In this system, the deflection values are acquired rapidly with the aid of tiltmeter and displacement sensing module.(4) A bridge (deflection, displacement)-measuring system is realized based on the laser alignment radiation and video equipment.The crack detection and damage identification of a bridge are also studied in this thesis, which is summarized as follows.(1) An algorithm is proposed for identifying and reconstructing the cracks in a bridge from a set of images of the bridge surface, which is accompanied with a detailed description of the methods for crack image preprocessing, crack extraction, storage and crack image reconstruction.(2) According to the principle of PCA (Principle Component Analysis), a PCA based bridge health model identification method is presented. Some issues related to the collection of the measured data and the establishment of model are also discussed.In summary, this thesis addresses the design of four novel systems for the precise acquisition of bridge deflections and/or displacements. New methods for the crack detection and damage identification of a bridge are also proposed.