Polynomial Chaos Expansion Models for the Monitoring of Structures under Operational Variability

Abstract

AbstractThis study overviews the development and implementation of a stochastic framework efficiently fusing operational response data with external influencing agents, namely environmental conditions, for representing structural behavior in its complete operational spectrum. The delivered methodology results in the formulation of a structural performance indicator serving as a precursor of potential damage or deterioration. The proposed approach relies on the combined use of (1) a polynomial chaos expansion method, able to account for structural variations tied to the variation of the acting agents; and (2) an independent component analysis algorithm, for extraction of salient behavioral features. The efficacy of the introduced method is demonstrated on field data from actual large-scale systems in both operational and damaged states.