混合网络攻击下智能电网故障交互演化机理与攻击危害性能评估研究

Hybrid cyber attacks from information space have brought some new issues such as incomplete measurement information, multiple types of error information (e.g., false data, replay data) and system failures in the smart grid of dual network coupling (i.e., power grid and information network), which make analysis of interactive mechanism on the cascading failures and performance evaluation of attack damage more difficult than they should be. In this proposal, we intend to launch a study to performance evaluation of attack damage and interactive mechanism on failures in smart grid with hybrid cyber attacks. It aims at exploring the behavioral pattern and mechanism of cyber attacks subject to attack resource. After analyzing the relationship among multiple types of cyber attacks, the characteristic of cyber attacks behavior is firstly described. The dynamic interactive mechanism on multiple types of the cascading failures is then revealed, and the model of the cascading failures of dual network coupling is established. The novel state estimation methods in smart grid with hybrid cyber attacks are proposed and Craomér-Rao Bound is also analyzed. Furthermore, the model of probabilistic optimal power flow in smart grid with hybrid cyber attacks is developed, and the performance of grid security risk (e.g., overload risk and voltage exceeding risk, etc.) is evaluated. Finally, an experimental platform will be built for system validation. Eventually, the achievements of this project will enrich the theory of smart grid security and lay a solid foundation for the secure and economic operation of smart grid.

在双网（电力网与信息网）耦合的智能电网中，针对信息空间的混合网络攻击造成测量信息不完备、错误（如虚假、重复等）信息和系统故障类型更加多样等新问题，带来连锁故障演化机理与攻击危害性能评估更加困难的现实，本项目拟开展混合网络攻击下智能电网故障交互演化机理与攻击危害性能评估理论和实验研究：探索攻击资源约束下网络攻击的行为模式与攻击机理，分析多类型网络攻击之间关系，研究建立描述混合网络攻击行为特征的数学模型；针对多类型故障揭示连锁故障在双网中动态交互演化机理和传播规律，研究建立双网耦合连锁故障模型；考虑不完备与多样错误信息，提出适合混合网络攻击下智能电网状态估计方法并进行克拉美-罗界分析；研究适合描述混合网络攻击下智能电网的概率最优潮流模型并进行安全风险（过载风险、电压越限风险等）性能评估。最后，构建实验平台并进行实验验证。这将丰富智能电网安全理论体系，为智能电网的安全经济运行奠定理论与实验基础。

针对电力网与信息网耦合的智能电网中网络攻击造成测量信息不完备、错误（如虚假、重复等）信息和系统故障类型更加多样等导致连锁故障演化机理与攻击危害性能评估更加困难的新问题，本项目开展了网络攻击下智能电网故障交互演化机理与攻击危害性能评估理论和实验研究：建立了描述虚假数据注入攻击和连续拒绝服务攻击行为的数学模型，探索了相应的注入攻击策略和数据补偿策略，揭示了网络攻击的构造机理与特征；分析了电力网与信息网相互依存与交互作用机制，研究了网络攻击下的智能电网的连锁故障的动态交互演化机理、传播规律及特性，建立了网络攻击下双网耦合连锁故障模型；提出了混合网络攻击下的智能电网动态状态估计算法，理论证明了动态状态估计算法的收敛性，研究了状态估计算法的克拉美-罗界；建立了以运行风险值为目标的虚假数据注入攻击优化模型，获取了考虑运行风险的虚假数据注入攻击方案和网络-物理协同攻击方案，对网络攻击下电压越限风险、线路过载风险进行评估；构建了仿真和实际微电网系统安全实验平台并针对典型网络攻击进行了实验验证。这丰富了智能电网安全理论体系，为智能电网的安全经济运行提供支撑。