DynSyst_Special_Topics: Multiscale Dynamics and Information in Data Collection and Assimilation for Complex Systems

本研究的重点是利用多学科方法从先进的测量能力、建模和模拟来理解有噪声的非线性动力系统。该问题的独特特征是信息收集、信息融合和信息同化之间的相互作用。这些效应之间微妙的相互作用将导致新的和新颖的分析技术。该方法是：(1)利用基于概率的决策方法开发新的数据驱动的低阶建模和多尺度非线性滤波框架，用于电力系统的状态估计；(2)开发分布式采集、共享、挖掘和利用数据的新算法和工具，用于电力系统协同健康监测和脆弱性实时评估；(3)通过新颖的数据融合技术和粒子方法来提高电力系统的整体动态性能和稳定性，这些技术和方法专门适应底层动力系统的复杂性。互联电力系统的日常运行必须防范自然灾害、人为攻击、老化和意外情况引起的故障。这项工作的核心是庞大的数据收集、分析和控制策略，时间跨度从微秒到分钟和小时。通过新技术的创造，该项目将促进我们对下一代电力系统关键操作问题的理解，并将导致更智能的电网，以提供更好的电力能源安全性、效率和可持续性。通过这项资助，我们将加强对工程本科学生的培训，培养他们在这个多学科领域攻读研究生的动力。PIs预期在培养精通工程、建模和应用数学的研究人员方面有直接的科学效益。共同负责人和他的团队开发了小程序，旨在教育K-12学生如何操作电力系统。建议工作的最终成果的一个组成部分将是为社会提供一种预测工具，以增强对噪音和不确定性对电力系统运行的影响的直观理解。

This research focuses on understanding noisy nonlinear dynamical systems using multidisciplinary approaches from advanced measurement capabilities, modeling, and simulations. The unique features of the problem are interactions between information collection, information fusion and information assimilation. The subtleties of the interaction between these effects will lead to new and novel analytical techniques. The approach is to (1) develop novel data-driven lower-order modeling and multi-scale nonlinear filtering frameworks using probability-based decision-making approaches for state estimation of the power system; (2) develop new algorithms and tools for the distributed collection, sharing, mining and harnessing of data for cooperative health monitoring and vulnerability assessment of power systems in real time; and (3) enhance overall power system dynamic performance and stability through novel data fusion techniques and particle methods which are specifically adapted to the complexities of the underlying dynamical system.The daily operation of interconnected electric power systems must guard against failures due to natural disasters, human attack, and failures due to aging and unexpected conditions. At the heart of this effort is huge data collection, analysis, and control strategies spanning time scales from microseconds to minutes and hours. Through the creation of new techniques this project will advance our understanding of the critical operational issues of the next-generation power system and will lead to smarter power grids to provide better electric energy security, efficiency, and sustainability.Through this grant we will foster intensive training of engineering undergraduate students, developing their motivation to pursue graduate studies in this multidisciplinary area. The PIs anticipate a direct scientific benefit in training researchers who are well-versed in engineering, modeling, and applied mathematics. The Co-PI and his group have developed Applets designed for education of K-12 students in how power systems operate. An integral part of the final outcome of the proposed work will be to provide the community a predictive tool that will enhance visual understanding of the effect of noise and uncertainty on power system operations.