ABSTRACT
Over the last two decades, structural health monitoring (SHM) technologies have emerged, creating an attractive field within bridge engineering. Application of bridge health monitoring (BHM) has been recognized as an attractive tool for early warnings on structural damage or deterioration before costly maintenance and repair, or even unexpected bridge collapse (Brownjohn 2007). To successfully accomplish the purposes of BHM, several challenging tasks from the selection of proper sensors to the development of effective algorithms for monitored data analysis and interpretation were investigated. The need for a bridge health paradigm that helps to effectively manage bridge performance was highlighted (Ko and Ni 2005; Catbas et al. 2008, 2013a, 2013b; Frangopol 2011). With a sufficient amount of monitored data, bridge performance under uncertainty can be predicted reliably and, furthermore, bridge maintenance management can be optimized to allocate limited financial resources and extend the service life of bridges (Strauss et al. 2008; Messervey et al. 2011; Orcesi and Frangopol 2013). Therefore, probabilistic and statistical concepts and methods, time-dependent bridge performance under uncertainty, and bridge life-cycle cost analysis and performance prediction should be well understood to achieve the recent need for a bridge health paradigm.
