On the Value of Information forecasting with multi-information systems

Abstract

The deterioration of structural components together with the climate and mobility increasing demands undermine civil structuresﾒ safety and functionality performance. Planning of maintenance activities is necessary, along with budget optimization. In this context, inspections, destructive and non-destructive tests, and Structural Health Monitoring (SHM) can increase the knowledge about the structuresﾒ safety and functionality, improving prioritization and the selection of the optimal interventions.

This paper aims to present a method to investigate different information strategies for a generic multi-component redundant structure. A structural model capable of redistributing the loads and accounting for correlation, component behavior characteristics, and sensor configuration, is adopted in the study. The modeling of structural integrity management accounts for a set of actions, considered to be representative for the maintenance action domain. Information strategies are modelled by probabilistic structural system characteristics accounting for diverse measurement and monitoring strategies. Value of Information (VoI) from Bayesian decision theory is used to quantify the effects collected observations have in structural integrity management. The approach is exemplified with analyses of brittle and ductile Daniels systems revealing effects of structural system characteristics on the VoI.