Performance Assessment Based on Evidence Theory and Fuzzy Logic: Application to Building and Dam Performance

Abstract

In the current economic, sociological, and environmental context, the durable performance of civil engineering works is a keystone of risk management. The data available for assessing performance are often numerous, heterogeneous, and imperfect. To cope with these difficulties, a four-phase method is proposed, comprised of (1) performance assessment modeling, (2) the formalization of heterogeneous information from different sources, (3) the propagation of imperfections in performance assessment models, and (4) the expression of outputs of these models in suitable formats to facilitate decision making and improve communication. Two applications of this methodology have been developed for buildings and dams and are used as illustrations. This methodology is supported by various methods that must be selected as a function of the case studied. Systemic approaches, knowledge-based methods, evidence theory, and fuzzy logic were used in the applications.