Developing Effective Strategies for Enhancing Power System Resiliency in the Presence of Cyber-Physical Attacks

与传统电力系统相比，发展中的智能电网的主要特点之一是广泛部署更多的网络技术，以实现各种测量、通信和控制功能。然而，无数信息和通信技术的实施将不可避免地增加电网的网络脆弱性。除了网络入侵，电网还面临着来自物理攻击的新威胁。例如，攻击者可以对故意选择的设备进行物理攻击，以中断电力供应。以潜在恐怖活动为形式的人为破坏最近也受到更多关注。该项目将通过解决几个紧迫的智能电网网络物理安全问题带来变革性的结果。它将通过为网络物理风险评估和缓解计划的制定制定一个全面的框架，大大促进最先进的技术水平。开发的网络-物理安全分析模型将能够涵盖物理组件、网络脆弱性、物理脆弱性、人为因素等广泛的不确定性。这将通过综合考虑一系列不确定因素，有望在电力系统网络-物理安全、弹性和可信性领域取得重大进展。该项目中的定量研究将使电力系统运营商和规划者能够做出更明智的决策。该项目的成果将使公用事业公司、监管机构和政府机构能够评估和改善现代电力系统的网络物理安全和弹性。此外，研究成果将被纳入各种教育和外联活动，以使下一代工作人员认识到国家关键基础设施中新出现的网络物理脆弱性。在智能电网部署中实施各种先进的传感、通信和控制技术也增加了人们对网络物理脆弱性的担忧，从长远来看，这一领域的研究将对智能电网的成功至关重要。该项目的主要目标是提高电力系统在出现网络物理攻击时的弹性。为此，提出了三种新的策略，可在系统规划或运行阶段部署，包括有效识别与网络物理攻击有关的综合关键突发事件，面向弹性的安全约束最优潮流，以及考虑不确定性的基于稳健优化的电力系统保护。本项目的主要研究任务是：1)基于状态空间剪枝和智能搜索方法，开发一种有效的电力系统关键部件综合识别方法；2)开发一种新的面向弹性的最优潮流框架和相关的混合并行计算解决方案；3)基于多攻击场景的防御者-攻击者-防御者模型，针对进攻资源中的不确定性，提出一种新的防御资源分配方案，以最大限度地减少损害。这些拟议的防御战略可能会大大增强当代电网在出现不利的、重大的网络物理事件时的弹性。

As compared with the traditional power system, one of the major characteristics of the evolving smart grid is the widespread deployment of more cyber technologies for achieving various measurement, communication, and control functionalities. However, the implementation of myriad information and communication technologies will inevitably increase the cyber vulnerability of the power grid. Besides cyber intrusions, the power grid is also under emerging threats from physical attacks. For example, attackers could physically attack deliberately selected equipment to disrupt the power supply. Man-made sabotage in the form of potential terrorist activities are also receiving more attention recently. This project will lead to transformative results by addressing several pressing smart grid cyber-physical security problems. It will significantly advance the state of the art by developing a comprehensive framework for cyber-physical risk assessment and mitigation scheme development. The developed cyber-physical security analysis models will be able to cover a wide range of uncertainties in physical components, cyber vulnerability, physical vulnerability, human factors, etc. This is expected to lead to significant advancement in the field of power system cyber-physical security, resiliency, and trustworthiness by considering a comprehensive set of uncertain factors. The quantitative study in the project will enable more informed decision-making for power system operators and planners. The project outcomes will allow utilities, regulators, and government agencies to evaluate and improve cyber-physical security and resiliency of modern electric power systems. In addition, the research outcomes will be integrated into a wide variety of educational and outreach activities for equipping the next-generation workforce with the awareness on the emerging cyber-physical vulnerabilities in national critical infrastructures. Implementation of myriad advanced sensing, communication, and control technologies in smart grid deployment also increases concerns on cyber-physical vulnerability, and research in this field will be critical to the success of the smart grid in the long term. The major goal of this project is to enhance power system resiliency in the presence of emerging cyber-physical attacks. Three novel strategies are proposed for this purpose which can be deployed in either the system planning or operation stage, including the efficient identification of comprehensive critical contingencies relating to cyber-physical attacks, resiliency-oriented security-constrained optimal power flow, and robust optimization based power system protection accounting for uncertainties. The major research tasks of this project are: 1) developing an efficient procedure for comprehensively identifying the critical components in a power system based on an integrated state space pruning and intelligent search methodology; 2) developing a novel resiliency-oriented optimal power flow framework and an associated hybrid parallel computing solution; and 3) developing a novel scheme for allocating defensive resources to minimize damage considering the uncertainties in offensive resources based on a multiple-attack-scenario defender-attacker-defender model. These proposed defense strategies could result in major enhancement of the resiliency of contemporary power grids in the presence of adverse, significant cyber-physical events.