Uncertainty-based condition assessment of existing concrete structures

Assessment of a structure condition is a complex subject imbued with vagueness and ambiguity. This complexity arises from the use of subjective opinion and imprecise numerical data since routine assessments have been so far carried out on the basis of intuition and engineering judgement of experienced engineers. The current practice in assessing the condition of existing structures consists of examining available documentation, visual inspection, field testing, and laboratory testing. Data collected as results of these investigations were called evidential elements, and they appear in two forms either numerical measurements or linguistic observations. An assessment of the condition of an existing concrete structure is accomplished using combined observational and measuremental information. Expert opinions on structural condition were used to develop the nonnumeric portion of the evidence. Fuzzy sets were used to quantify linguistic variables since this type of information is inherently vague. A knowledge base contains rules and graphs that relate the evidence space, e.g., observed distress or laboratory test results, to the hypothesis space, i.e., the quality conditions of a concrete structure. The processing of this information was controlled through a quantitative reasoning treatment, which retrieves necessary facts from the knowledge base using an appropriate reasoning treatment, i.e., numerical and/or linguistic. The beliefs in the collected hypotheses were discounted as a result of unreliability that might be associated with the source of evidence. Since the information being conveyed by a piece of evidence is often not only imprecise, but also incomplete, and vague. A hypotheses belief structure model that can cope with all these different kinds of uncertainties in a sound manner was developed. It is served as a theoretical base needed for developing the methodology for concrete structure overall quality condition assessment, diagnostic and decision analyses. This study also examined a diagnostic model based on the concepts of evidence theory, fuzzy set theory, Bayesian theorem, and finally open world assumption, which contributes a new methodology for diagnosing deterioration problems. The Dempster-Shafer evidential theory lacks a formal methodology for making decisions. Herein, frameworks have been proposed to model a decision-maker's degree of belief about which state of affairs will prevail.