Model-Free Data Interpretation for Continuous Monitoring of Complex Structures

• Unité de recherche: INNOSUISSE
• Numéro de projet: 8613.1;10 EPRP-IW
• Titre du projet anglais: Model-Free Data Interpretation for Continuous Monitoring of Complex Structures

Description succincte (Anglais)

Model-Free Data Interpretation for Continuous Monitoring of Complex Structures

Description succincte (Français)

Model-Free Data Interpretation for Continuous Monitoring of Complex Structures

Résumé des résultats (Abstract) (Anglais)

This is a proposal to develop and implement software able to analyze time series that are generated by data from monitoring full-scale civil-engineering structures. The main scientific objective of this study is to develop a methodology that helps establish whether or not a structure is behaving normally without having to employ expensive geometric and behavioral models. Although the methodology is general, for the purposes of this project, it is assumed that the structure is instrumented with long gauge fiber optic deformation sensors. Damage is thus any deterioration in structural integrity that influences these measurements. Early detection of such damage is not straightforward due to other phenomena that influence deformation measurements. For example, dynamic loading, such as wind and truck loading on bridges, as well as longer term cyclic deformations due to effects such as daily and seasonal temperature variations also influence measurements. Improving the efficiency of data interpretation and system identification will facilitate development of robust approaches for structural management. Currently few structures are monitored due, in large part, to the costs associated with data interpretation. An increase in efficiency is expected to encourage more engineers to carry out structural monitoring, thus lowering structural life-cycle costs and increasing revenues for SMARTEC¿s products and services.

Résumé des résultats (Abstract) (Français)

This is a proposal to develop and implement software able to analyze time series that are generated by data from monitoring full-scale civil-engineering structures. The main scientific objective of this study is to develop a methodology that helps establish whether or not a structure is behaving normally without having to employ expensive geometric and behavioral models. Although the methodology is general, for the purposes of this project, it is assumed that the structure is instrumented with long gauge fiber optic deformation sensors. Damage is thus any deterioration in structural integrity that influences these measurements. Early detection of such damage is not straightforward due to other phenomena that influence deformation measurements. For example, dynamic loading, such as wind and truck loading on bridges, as well as longer term cyclic deformations due to effects such as daily and seasonal temperature variations also influence measurements. Improving the efficiency of data interpretation and system identification will facilitate development of robust approaches for structural management. Currently few structures are monitored due, in large part, to the costs associated with data interpretation. An increase in efficiency is expected to encourage more engineers to carry out structural monitoring, thus lowering structural life-cycle costs and increasing revenues for SMARTEC¿s products and services.