Since 1988 the offshore platform VEGA-A is operating in the Sicily channel. The platform, one of the largest built in Italy, is equipped with a monitoring system that collects both environmental (a current meter, a depth gauge and a system for detecting speed and direction of wind) and structural (by means of 9 linear accelerometers) data. This paper, after a description of the recording system, provides a statistical interpretation of the structural response of the platform over 30 years of activity. The structural monitoring system is aimed to perform dynamic identification (i.e. to evaluate main frequencies and mode shapes), and part of the acquired data is directly (and automatically, by spectral analysis) processed on the platform, in order to constantly check their representativeness in relation to the structural control of the platform. The variation of main frequencies and mode shapes over time may be related to both changes of environmental loads/conditions or to a structural damage; in this respect, the Principal Components Analysis (PCA) and the Novelty Index Detection were employed to clean the variation of the dynamic properties from the environmental loads. The final goal of the structural monitoring system is, in fact, to provide an early warning system. Variations of the main frequencies over these 30 years of activity of the platform are discussed (and correlated with the storms that have affected the structural system) and, eventually, a statistical analysis of the accelerometer signals performed with the aim to identify the warning thresholds and alarm is presented.
[1]
Andrea Vignoli,et al.
Monitoring and Dynamic Identification of VEGA-A Platform
,
2015
.
[2]
Sabine Van Huffel,et al.
Application of Kernel Principal Component Analysis for Single-Lead-ECG-Derived Respiration
,
2012,
IEEE Transactions on Biomedical Engineering.
[3]
Edwin Reynders,et al.
Output-only structural health monitoring in changing environmental conditions by means of nonlinear system identification
,
2014
.
[4]
Guido De Roeck,et al.
REFERENCE-BASED STOCHASTIC SUBSPACE IDENTIFICATION FOR OUTPUT-ONLY MODAL ANALYSIS
,
1999
.
[5]
Sang Joon Kim,et al.
A Mathematical Theory of Communication
,
2006
.
[6]
Michele Betti,et al.
Damage detection on a three-storey steel frame using artificial neural networks and genetic algorithms
,
2015
.
[7]
John H. Holland,et al.
Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence
,
1992
.