EAGER: Uncertainty Quantification of Structural Systems: Generalized Information Theory

在评估风险、减轻灾害和脆弱性/复原力时，对结构和社区基础设施性能的评估和预测是固有的不确定性。结构工程需要反映多学科的社会问题，包括社会、政治和经济，这些问题引入了部分知识和不确定性的不同特征。虽然处理不确定性的传统概率方法可以处理部分和不相似的信息，但它们不是最适合全面结合反映诸如专家判断和不精确等来源的更广泛背景的方法。这个早期概念探索性研究（EAGER）项目对结构工程安全的不确定性方法的更广泛方面进行探索性研究，包括人类决策中的多属性判断和不可预测性。该研究将各种不确定性的来源引入到结构设计中，并将其与社区问题结合起来，评估结构对经济、环境和社会/政治可持续性的价值。数量和语言的不确定性被纳入一个综合的设计理念，能够处理模糊和歧义。这些结构工程和社区表现的扩展方法将通过改善建筑环境为社会带来巨大利益。该项目的主要目标是在广义信息论的框架下发展方法，将传统的概率方法扩展到结构和社区安全，通过结合无知、交互性和语言信息的概念来加强对不确定性的处理。该项目的范围包括概率论的概化与基于单调理论的不确定性度量，并结合语言模糊性，模糊性，证据概念（信念和合理性），相互冲突的专家判断，不精确，部分知识，信心和主观可能性。这一探索性项目的初步应用将是地震损害估计和易损性评估。结合社区实际数据，利用模糊模式识别技术对建筑损伤模式的存在性进行调查，并进行潜在地震易损性预测。

Uncertainty is inherent in the assessment and prediction of the performance of structures and community infrastructure when evaluating risk, hazard mitigation and vulnerability/resiliency. Structural engineering needs to reflect multi-disciplinary societal issues, including social, political and economic, which introduce partial knowledge and disparate characteristics of uncertainty. While traditional probabilistic methods to treat the uncertainty can address partial and dissimilar information, they are not the most appropriate for comprehensive incorporation of broader contexts reflecting sources such as expert judgement and imprecision. This EArly-concept Grant for Exploratory Research (EAGER) project undertakes exploratory research on the broader aspects of uncertainty approaches to structural engineering safety, incorporating multi-attribute aspects of judgement and unpredictability in human decision-making. The research introduces varied sources of uncertainty into structural design, melding this with community issues that assess the value of the structure to economic, environmental and social/political sustainability. Quantitative and linguistic uncertainties are incorporated under a comprehensive design philosophy capable of handling vagueness, and ambiguity. These broadened approaches to structural engineering and community performance will lead to great benefits to society through enhancements of the built environment. The primary objective of this project is the development of methods under the umbrella of Generalized Information Theory that expand traditional probabilistic approaches to structural and community safety, enhancing the treatment of uncertainty by incorporating notions of ignorance, interactivity, and linguistic information. The scope of the project includes generalization of probability theory with uncertainty measures based on monotone theory and incorporating linguistic vagueness, ambiguity, evidence concepts (belief and plausibility), conflicting expert judgement, imprecision, partial knowledge, confidence, and subjective possibility. An initial application in this exploratory project will be for earthquake damage estimation and vulnerability assessment. With actual community data, fuzzy pattern recognition will be used to investigate the existence of building damage patterns and potential seismic vulnerability prediction.