ITR/NGS: Collaborative Research: DDDAS: Data Dynamic Simulation for Disaster Management

该项目将开发先进的信息技术工具，数学模型和原型基础设施，用于灾害建模和管理。该项目将把全面的信息和数值预报带到需要的地方，在灾害指挥中心，在真实的时间。该系统将包含来自数据流的大量信息，例如，地图，传感器，测量和天气数据，以及数学模型将在远程超级计算机上运行。该模型将由动态可用的数据控制，结果将在分布式设备中可视化，如通过无线以太网和宽带卫星链路连接到互联网的笔记本电脑和掌上电脑。其他传感器和机载成像仪也将无线联网。 设想的建模系统的核心将是该地区周围荒地和天气的现有计算机模型，这些模型将使用现代软件工程方法作为数据驱动的应用程序从头开始重写，并通过新的数学建模技术以及先进的统计技术来增强，以管理不确定性。高性能的中间件将被用来连接计算节点，传感器节点，监测节点和可视化节点到一个分布式信息和建模系统。新的网络软件技术将使节点之间的通信安全，并提供服务质量保证。该系统将被设计为容忍通信中断，增加延迟和节点消失。

This project will develop advanced InformationTechnology tools, mathematical models, and prototype infrastructure for disaster modeling and management. The project will bring comprehensive information and numerical prediction where it is needed, at the disaster command center, in real time. The system will incorporate large volume of information from data streams, e.g., as maps, sensor, surveyance, and weather data, and the mathematical model will run on remote supercomputers. The model will be controlled by dynamically available data, and the results visualized in distributed devices, like laptops and palmtops connected to the Internet by wireless ethernet and a broadband satellite link. Other sensors and airborne imagers will also be networked wirelessly. The core of the envisioned modeling system will be an existing computer model of wildland and weather around the area, adn these models will be rewritten from scratch using modern software engineering methodology as a data driven application, and enhanced by new mathematical modeling techniques together with advanced statistical techniques will be used to manage uncertainty. High-performance middleware will be used to connect omputational nodes, sensor nodes, surveyance nodes, and visualization nodes into a distributed information and modeling system. New network software technologies will make the communication between the nodes secure and provide quality of service guarantees. The system will be designed to tolerate interruptions of communication, increased latencies, and node disappearances.