Reliability-based metrics for structural health monitoring information quality assessment

Abstract

One of the primary challenges in civil engineering today is the safety assessment and management of existing infrastructures. In this context, structural health monitoring (SHM) plays a fundamental role in data acquisition and assessment of the safety of a given infrastructure. SHM data are instrumental for quantifying uncertainties, informing structural models, and updating structural reliability over time. In principle, an SHM system should provide the best information possible for assessing and updating structural reliability. Since infrastructure management resources are inevitably limited, SHM information quality assessment becomes a key issue. Several metrics are available in the scientific literature to measure the SHM information quality. Most of these are based on the accuracy of SHM information or the impact of monitoring on decision-making. This article introduces novel information quality metrics that quantify how SHM affects the structural reliability assessment. First, we state a comprehensive framework based on Bayesian networks for structural reliability updating. Based on this, we define a metric based on the measurement resolution concept. Next, according to structural reliability theory, we present an importance measure of the SHM observations. Finally, we propose two explanatory case studies.