针对智能电网状态信息流的数据完整性攻击检测和安全控制策略研究

In the smart grid with highly integration of cyber and physical systems, the data integrity attacks against state information flow causes the deviation of the state estimation results by tampering with the measurements, thereby affecting the security and stable operation of smart grid. The increasing popularity of Wide Area Measurement System (WAMS) presents a multi-spatiotemporal coupling of state information flows to the smart grid, posing a more serious data integrity attack threat. To this end, based on multi-space-time coupling measurement information, this project first formalize data integrity attacks model and analyzes the evolutionary principle of the state estimation bias. Secondly, based on the spatial-temporal coupling of measurement information characteristics, this project designs an online data integrity attack detection mechanism based on deep feature of the state information. Thirdly, from the aspect of cyber physical system, this project designs adaptive secure control strategy which aims to ensure the security elasticity of smart grid when suffered data integrity attack. Finally, experimental verification is conducted based on the simulation platform. The research of this project can further enrich the information security and comprehensive security technology system of smart grid from the perspective of cyber physical systems, and provide support for the application of data integrity attack detection and secure control technology in smart grid.

在信息物理高度融合的智能电网中，针对状态信息流的数据完整性攻击通过篡改量测信息对状态估计结果造成偏差，进而影响电网安全、稳定运行。广域测量系统（Wide Area Measurement System, WAMS）的不断普及，使智能电网状态信息流呈现多时空尺度耦合特性，从而面临更为严峻的数据完整性攻击威胁。为此，本项目首先在研究基于多时空尺度耦合量测信息的智能电网状态信息流机理基础上，构建数据完整性攻击模型并分析电网性能损失演化机理；其次，借鉴量测信息时间和空间耦合特性，设计基于深度模型特征的数据完整性攻击在线检测机制；再次，从信息物理融合系统层面，设计自适应安全控制策略保证电网受到数据完整性攻击时的安全弹性；最后，构建半实物仿真平台并进行实验验证。本项目的研究能够进一步丰富信息物理融合视角下智能电网信息安全和综合安全技术体系，为智能电网数据完整性攻击检测和安全控制技术的应用提供支撑。

在信息物理高度融合的智能电网中，针对状态信息流的数据完整性攻击通过篡改量测信息对状态估计结果造成偏差，进而影响电网安全、稳定运行。广域测量系统（Wide Area Measurement System, WAMS）的不断普及，使智能电网状态信息流呈现多时空尺度耦合特性，从而面临更为严峻的数据完整性攻击威胁。为此，本项目首先在研究基于多时空尺度耦合量测信息的智能电网状态信息流机理基础上，构建了数据完整性攻击模型并分析电网性能损失演化机理；其次，借鉴量测信息时间和空间耦合特性，设计基于深度模型特征和深度强化学习的数据完整性攻击在线检测机制；再次，从信息物理融合系统层面，设计自适应安全控制策略保证电网受到数据完整性攻击时的安全弹性；最后，构建半实物仿真平台并进行实验验证。本项目的研究能够进一步丰富信息物理融合视角下智能电网信息安全和综合安全技术体系，为智能电网数据完整性攻击检.测和安全控制技术的应用提供支撑。