新一代能源电力系统信息安全风险传播机理与动态评估方法研究

It is the current trend in energy sectors to link up several separate energy networks by the aid of advanced information and communication technology to construct a new generation of electricity-centric, power-grid-backboned energy and power systems (EPS). In EPSs, information and energy flows are tightly coupled, which engenders cyber security risks not only in cyber networks but also across cyber networks and energy networks, even cause large-scale cascading faults in EPSs. The clarification of cyber security risk propagation mechanisms will form a basis for risk prevention and control in EPSs. This research starts with risk factor evaluations in both cyber attack and cyber system failure scenarios, based on the fusion of multi-source heterogeneous monitoring data and non-credibility assessment of applications, respectively. Then the multi-agent based cellular automata theory is introduced into the field of EPS cyber security to build a risk propagation model. This model makes use of the synergistic ability of the agent clusters to clarify the propagation mechanisms of cyber security risks through accurate simulation and dynamic risk evaluation. Finally, this research plans to design and develop a prototype of a cyber security virtual sandbox system, which can give out decision-making references for the planning and design, security analysis, scheduling management, and emergency control in EPSs. On light of this, the research is of significant theoretical and application values in related fields.

借助先进信息技术将多种能源系统联结起来，构建以电力为中心、电网为主干的新一代能源电力系统是当下能源领域的发展趋势。能源电力系统中信息流与能量流紧密耦合，信息安全风险除了在信息系统中蔓延外，也能在信息系统与能源系统间交互传播，甚至引发能源电力系统大规模连锁故障。厘清能源电力系统中信息安全风险的传播机理，是有针对性地开展信息安全风险防控的基础。本项目基于多信源异构监控数据融合和应用软件不可信性度量，感知信息攻击与信息系统功能失效风险因素，并将多智能体元胞自动机集成理论引入能源电力系统信息安全领域，构建风险传播模型，利用智能体集群的协同能力实现风险传播的准确推演和动态评估，以期明确能源电力系统信息安全风险的传播机理，并以此为指导设计开发信息安全虚拟沙盘原型系统，为新一代能源电力系统的规划设计、安全分析、调度管理、紧急控制提供决策参考，具有重要的理论和应用价值。

信息流与能量流的紧密耦合与相互作用是新一代能源电力系统的基本特征。随着先进信息通信技术在能源电力系统中逐步推广，信息安全风险事件的发生率也在无形之中提高。本项目重点关注新一代能源电力系统信息安全风险的感知、建模、推演与防御，形成了一套数据驱动的能源电力系统量测数据篡改攻击检测与恢复方法体系，涵盖风险数据感知与提取、数据预处理与特征增强、受污染数据检测定位与恢复等多方面功能。考虑到云计算环境中不同参与主体的可信性难以充分保证，本项目还提出了一套适用于不可信环境的隐私保护式分层状态估计架构，允许参与方在不知道其他参与方敏感数据的前提下协同完成状态估计所需计算。此外，本项目还研发了一套能源电力系统信息物理联合仿真平台，支持信息层与物理层的协同建模，可对能源电力系统中的信息安全风险进行虚拟沙盘推演。项目研究成果具有较高的应用潜力，可为新一代能源电力系统的规划设计、安全分析、调度管理、紧急控制提供决策参考。