F Lanata

Model-free data interpretation for continuous monitoring of complex structures

Civil engineering structures are difficult to model accurately and this challenge is compounded when structures are built in uncertain environments. As consequence, their real behavior is hard to predict; such difficulties have important effects on the reliability of damage detection. Such situations encourage the enhancement of traditional approximate structural assessments through in-service measurements and interpretation of monitoring data. While some proposals have recently been made, in general, no current methodology for detection of anomalous behavior from measurement data can be reliably applied to complex structures in practical situations. This paper presents two new methodologies for model-free data interpretation to identify and localize anomalous behavior in civil engineering structures. Two statistical methods (i) moving principal component analysis and (ii) moving correlation analysis have been demonstrated to be useful for damage detection during continuous static monitoring of civil structures. The algorithms are designed to learn characteristics of time series generated by sensor data during a period called the initialization phase where the structure is assumed to behave normally. This phase subsequently helps identify those behaviors which can be classified as anomalous. In this way the new methodologies can effectively identify anomalous behaviors without explicit (and costly) knowledge of structural characteristics such as geometry and models of behavior. The methodologies have been tested on numerically simulated elements with sensors at a range of damage severities. A comparative study with wavelet and other statistical analyses demonstrates superior performance for identifying the presence of damage.

Damage detection and localization for continuous static monitoring of structures using a proper orthogonal decomposition of signals

In recent years, several structures have been equipped with permanent monitoring systems, able to record the response in terms of displacements and strains over very long periods of time and, theoretically, for the entire life of the structure. Despite the number of applications, very few studies have been presented aimed at interpreting the data and at providing a means for detecting and localizing the insurgence of damage or material degradation from the measurements. The paper provides an approach to an algorithmic treatment of the data flows produced by continuous static monitoring systems. The proposed algorithm is the proper orthogonal decomposition (POD), a statistical method particularly suitable and versatile in multivariate analysis. However, its application in the field of damage detection for static monitoring is a new approach. The discussion will be based on data obtained by means of a computer simulation of a typical bridge structure subjected to environmental loadings. The POD will be used for studying a temporal and spatial correlation analysis between the structural responses measured at different sensor locations. The comparison in terms of damage indices of the responses from different sensors in time has been shown to be able to detect damaged sections.

Model free interpretation of monitoring data

No current methodology for detection of anomalous behavior from continuous measurement data can be reliably applied to complex structures in practical situations. This paper summarizes two methodologies for model-free data interpretation to identify and localize anomalous behavior in civil engineering structures. Two statistical methods i) moving principal component analysis and ii) moving correlation analysis have been demonstrated to be useful for damage detection during continuous static monitoring of civil structures. The algorithms memorize characteristics of time series generated by sensor data during a period called the initialisation phase where the structure is assumed to behave normally. This phase subsequently helps identify anomalous behavior. No explicit (and costly) knowledge of structural characteristics such as geometry and models of behaviour is necessary. The methodologies have been tested on numerically simulated elements with sensors at a range of damage severities. A comparative study with wavelets and other statistical analyses demonstrates superior performance for identifying the presence of damage.

A Multi-Algorithm Procedure for Damage Location and Quantification in the Field of Continuous Static Monitoring of Structures

In recent years, several structures have been equipped with permanent monitoring systems, able to record the response both in terms of displacements and strains over very long periods of time and, theoretically, for the entire life of the structure. Despite of the number of applications, very few studies have been presented focused on the interpretation of the data without the study of a numerical model of the structure. Since an optimal and unique algorithm cannot be proposed depending on the variety of applications, the aim of the work is to propose a multi-algorithm methodology as a tool for detecting and localizing the insurgence of damage or material degradation from the measurements taken during a continuous static monitoring of civil structures. A method based on Principal Component Analysis will be proposed in order to compare the responses and detect the insurgence of anomalous behaviors. The algorithm will be first tested on simulated data deriving from a numerical benchmark with sensors and different damage scenarios, then the proposed methodology will be validated on a real structure. In this second application, due to the great number of installed sensors, the algorithm will be integrated with a preliminary analysis in order to cluster and gather together the sensors with a comparable behavior and a similar sensitivity to damage.

Analysis of long-term deformation data from the San Giorgio harbor pier in Genoa

In the framework of a large-scale monitoring program conducted by the Port Authority of Genoa, the east quay wall of the San Giorgio pier has been equipped with an array of more than 60 SOFO fiber optic sensors for continuous monitoring. These sensors allow the measurement of the pier displacements during the dredging works, ship docking and in the long term. The sensors measure the curvature changes in the horizontal and vertical planes and allow a localization of settlements with a spatial resolution of 10 m over a total length of 400 m. The system is in operation since fall 1999, and data has been collected automatically and continuously since then. This paper is intended to present the first analyses and interpretations performed on the monitoring data. Correlation of raw data and curvature analysis to environmental conditions is also presented.