Collaborative Research: Learning and Optimizing Power Systems: A Geometric Approach

电网的变革给系统运营商和公用事业公司带来了过多的挑战。他们必须适应管理一系列高度不确定和分布式的资源，如电动汽车和太阳能PV，同时运营几十年前设计的电网基础设施。这些挑战在分配系统中尤其严重，传统上网络不受密切监控，运营商缺乏必要的信息来获得系统的准确实时运行状态。与此同时，随着系统老化，配电系统中的停电次数开始增加，负荷变得更加动态。这项提议的目标是通过开发新的算法和新的见解来克服这些挑战，以提高分配系统的效率和弹性。将围绕这些研究项目开展教育活动，以确保不同的学生参与和接触更广泛的社区。该项目的重点是三个方面：i)利用智能电表和其他传感器提供的大量数据进行系统拓扑估计，其中网络可能包含回路，数据可能高度异质；ii)使用对潮流的新的几何理解来表征运行点的可行性，从而产生经证明有效和优化的算法；iii)利用前两个方面的结果，在线路切换后立即通过线路切换恢复服务。这些研究带来了电力系统分析、优化和统计学习的工具，使配电系统运行取得了根本性的进步。特别是，这些推力使我们能够利用最近在技术和理论上的进步来开发及时和严格的算法，为电网解决一些紧迫的工程问题。我们提议的项目的成功应用将使配电系统运营商能够回答那些具有大量分布式资源的高度不稳定的电网所产生的各种“现在”和“如果”问题。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

The transformations of the electrical grid present a plethora of challenges to system operators and utilities. They must adapt to manage a set of highly uncertain and distributed resources such as electric vehicles and solar PVs, while at the same time operating a grid infrastructure that was designed decades ago. These challenges are particularly acute in the distribution system, where the networks are traditionally not monitored closely, and operators lack the essential information to obtain an accurate real-time operational state of the system. At the same time, the number of outages in distribution systems has started to increase as the system ages, and the loads become more dynamic. The goal of this proposal is to overcome these challenges by developing novel algorithms and new insights that increase the efficiency and resilience of the distribution systems. Educational activities would be developed around these research thrusts to ensure diverse student participation and outreach to the broader community. The project focuses on three thrusts: i) system topology estimation using the wealth of data made available by smart meters and other sensors, where the network may contain loops and the data may be highly heterogeneous; ii) characterization of the feasibility of operating points using a new geometric understanding of power flow that leads to provably efficient and optimal algorithms; and iii) restoration of service right after outages through line switching by using the results from the first two thrusts. These investigations bring in tools from power system analysis, optimization, and statistical learning to enable fundamental advances in the distribution system operations. In particular, these thrusts allow us to leverage recent advances in both technology and theory to develop timely and rigorous algorithms that solve some pressing engineering problems for the power grids. Successful application of our proposed project will allow distribution system operators to answer various "what now" and "what if" questions deriving from those highly volatile grids with large amounts of distributed resources.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.