EAGER: SSDIM: Superimposed Simulations: Fast Generation of Synthetic Data of Interdependent Critical Infrastructures

为了指导现代相互依赖的关键基础设施（ICI）系统的设计、建设、运营和更好的理解，决策者需要能够捕捉他们与他们的参与者（即，公民、公司、开发商和政府）。详细的计算机模拟对于生成此类数据至关重要，特别是当ICI设计的物理实现尚不存在时。然而，由于必须建模的系统的复杂性和所需的大量计算实验，从模拟中导出的数据的生成是耗时的。为了应对这一挑战，将开发利用现代高性能计算技术的新计算机算法。在EARLY概念探索性研究资助（EAGER）项目中，算法将用于创建测试用例的数据集，包括亚特兰大市水，能源，交通基础设施和人的相互作用。这些新算法将提供有关如何评估ICI内部和跨ICI的威胁和机会、分析优势和劣势以及设计满足社会需求和目标的解决方案、产品和服务的见解。目前用于评估ICI内的选项、各种技术组合和人类交互的计算方法非常耗时且计算密集。将开发一种称为叠加模拟的新方法，以快速创建合成数据，捕捉关键基础设施和人员的相互依赖性和动态。通过利用与并行计算技术相结合的许多模拟运行之间的相似性，新算法旨在将涉及多个相互作用和相互依赖的关键基础设施系统和人类参与者的模拟加速一到两个数量级。这些新开发的算法将允许快速生成大型数据集，这些数据集可用于确定ICI组合、人类行为和政策工具对能源、水、碳排放和人类健康的影响。

To guide the design, construction, operation and improved understanding of modern interdependent critical infrastructure (ICI) systems, decision-makers require data that capture the interactions among them their actors (i.e., citizens, firms, developers, and governments). Detailed computer simulations are essential to generating such data, particularly when physical implementations of ICI designs do not yet exist. However, generation of data derived from simulations is time consuming due to the complexity of the systems that must be modeled and the large number of computational experiments that are required. To address this challenge new computer algorithms exploiting modern high performance computing technologies will be developed. In this EArly-concept Grant for Exploratory Research (EAGER) project, algorithms will be applied to create data sets for test cases including the interaction of water, energy, transportation infrastructures and people in the city of Atlanta. The new algorithms will provide insights into how to evaluate threats and opportunities within and across ICIs, to analyze strengths and weaknesses, and to design solutions, products, and services that meet societal needs and goals.Current computational methods to assess options, various technology combinations, and human interactions within ICIs are time-consuming and computationally intensive. A new approach, termed superimposed simulations, will be developed to rapidly create synthetic data that captures the interdependencies and dynamics of critical infrastructures and people. By exploiting similarities among the many simulation runs in conjunction with parallel computing techniques, the new algorithms aim to accelerate simulations involving multiple interacting and interdependent critical infrastructure systems and human actors by one or two orders of magnitude. These newly developed algorithms will allow the quick generation of large data sets that can be used to determine the effects of ICI combinations, human behaviors, and policy tools on energy, water, carbon emissions, and human health.