A Probabilistic Approach for Optimal Sensor Placement of Bridge Structures for the Purpose of Structural Model Updating

A probabilistic methodology is developed in this paper for identifying the most effective way to install a limited number of sensors on the continuous-coordinated system to extract as much information as possible. This is performed by minimizing the information entropy associated with the identified bridge modeling parameters. Although the commonly used discrete-coordinated systems (such as the finite element method) provide a convenient and practical approach to the dynamic response analysis of arbitrary structures, the solutions obtained can only approximate their actual dynamic behavior, because the motions are represented by a limited number of displacement coordinates. This would undoubtedly influence the precision of dynamical analysis results, which, in turn, influence the sensor configuration results. For bridge structures having typical distributed properties, it is more reasonable and accurate to be modeled as distributed-parameter system with analytical formulation. This is employed in this paper for investigating the sensor placement problem with the information entropy-based method through the numerical simulation of a three-span continuous bridge model.