CPS:Medium:Collaborative Research: High-Fidelity High-Resolution and Secure Monitoring and Control of Future Grids: a synergy of AI, data science, and hardware security

CPS：中：协作研究：未来电网的高保真高分辨率和安全监控：人工智能、数据科学和硬件安全的协同作用

• 批准号: 1932196
• 负责人: Meng Wang
• 金额: $ 45万
• 依托单位: Rensselaer Polytechnic Institute
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-09-15 至 2023-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1932196&HistoricalAwards=false
• 关键词: CPS; Medium; Collaborative; Research; Fidelity

The increasing presence of renewable generations and distributed energy resources in transmission systems heightens the need for fast-timescale situational awareness for system reliability, resiliency, and both the operational and cyber security. Despite the invention of phasor measurement units that promised close-to-real-time monitoring of the system states, the limited deployment of phasor measurement units had hampered the ability of the system operator to uncover trends of instability, react to system contingencies, and detect malicious attacks on the power grid. This research develops new hardware and software solutions for high-fidelity, high-resolution, and secure monitoring and control of the future grid. By harnessing and exploiting the increasingly abundant and diverse data sources and through novel applications of machine learning and artificial intelligence, this research advances the state-of-the-art monitoring of cyber-physical systems in three fronts. First, this research develops machine learning approaches to high-resolution state estimation for power systems that are unobservable by existing phasor measurement units. Second, this research offers new solutions to detecting and mitigating data anomaly caused by malfunctions of sensors, communications systems, and cyber attacks by adversarial agents. Third, this research develops a new hardware architecture and prototypes for future digital substations that provide hardware-based security. This research has broader impacts on enhancing national security in critical infrastructures, promoting economic competitiveness through accelerated adoption of phasor measurement technology, and broadening participation of women and under-represented minority groups in science, technology, engineering, and mathematics.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.

传输系统中可再生世代和分布的能源的存在日益增加，增强了对系统可靠性，弹性以及运营和网络安全性的快速限制情况意识的需求。 尽管有保证对系统状态进行近实时监测的相量测量单元发明，但相量测量单元的部署有限，这阻碍了系统操作员发现不稳定趋势的能力，对系统突发事件做出了反应，并检测到对电动机网格的恶意攻击。 这项研究开发了新的硬件和软件解决方案，以实现高保真，高分辨率和对未来网格的安全监控和控制。 通过利用和利用越来越丰富和多样化的数据源，并通过机器学习和人工智能的新应用，这项研究可以在三个方面对网络物理系统进行最新监测。 首先，这项研究开发了机器学习方法，以实现现有相量测量单元无法观察到的电力系统的高分辨率状态估计。 其次，这项研究提供了新的解决方案，可检测和减轻由传感器，通信系统和对抗剂的网络攻击引起的数据异常。 第三，这项研究开发了一种新的硬件体系结构和原型，用于将来提供基于硬件安全性的数字变电站。 这项研究对增强国家安全基础设施的安全性具有更广泛的影响，通过加速采用拟法测量技术来促进经济竞争力，扩大妇女的参与以及代表性不足的科学，技术，工程和数学中代表性不足的少数群体。这一奖项反映了NSF的法定任务和审查的范围，这是通过评估的范围来进行的。

项目成果

Learning and generalization of one-hidden-layer neural networks, going beyond standard Gaussian data

单隐层神经网络的学习和泛化，超越标准高斯数据

• DOI: 10.1109/ciss53076.2022.9751184
• 发表时间: 2022
• 期刊: Proc. 2022 56th Annual Conference on Information Sciences and Systems (CISS
• 作者: Li, Hongkang;Zhang, Shuai;Wang, Meng Wang
• 通讯作者: Wang, Meng Wang

Patch-level Routing in Mixture-of-Experts is Provably Sample-efficient for Convolutional Neural Networks

• DOI: 10.48550/arxiv.2306.04073
• 发表时间: 2023-06
• 期刊: ArXiv
• 作者: Mohammed Nowaz Rabbani Chowdhury;Shuai Zhang;M. Wang;Sijia Liu;Pin-Yu Chen

Bayesian High-Rank Hankel Matrix Completion for Nonlinear Synchrophasor Data Recovery

用于非线性同步相量数据恢复的贝叶斯高阶 Hankel 矩阵补全

• DOI: 10.1109/tpwrs.2023.3254909
• 发表时间: 2023
• 期刊: IEEE Transactions on Power Systems
• 影响因子: 6.6
• 作者: Yi, Ming;Wang, Meng;Hong, Tianqi;Zhao, Dongbo
• 通讯作者: Zhao, Dongbo

Real-Time Energy Disaggregation at Substations With Behind-the-Meter Solar Generation

• DOI: 10.1109/tpwrs.2020.3035639
• 发表时间: 2020-11
• 期刊: IEEE Transactions on Power Systems
• 影响因子: 6.6
• 作者: Wenting Li;Ming Yi;Meng Wang;Yishen Wang;Di Shi;Zhiwei Wang

Joint Edge-Model Sparse Learning is Provably Efficient for Graph Neural Networks

• DOI: 10.48550/arxiv.2302.02922
• 发表时间: 2023-02
• 期刊: ArXiv
• 作者: Shuai Zhang;M. Wang;Pin-Yu Chen;Sijia Liu;Songtao Lu;Miaoyuan Liu