发电技术

• •    下一篇

抵抗虚假数据注入攻击的综合能源系统弹性提升策略

吴丽珍1,张永朋1,魏建平1,2,陈伟1,3   

  1. 1.兰州理工大学电气工程与信息工程学院,甘肃省 兰州市,730050;2.北京交通大学国家能源主动配电网技术研发中心,北京 海淀区,100044;3.兰州理工大学国家级电气与控制工程实验教学中心,甘肃省 兰州市,730050
  • 基金资助:
    国家自然科学基金资助项目(62063016)

A Comprehensive Energy System Resilience Enhancement Strategy to Resist False Data Injection Attacks

WU Lizhen1,ZHANG Yongpeng1,WEI Jianping1,2,CHEN Wei1,3   

  1. 1. College of Electrical and Information Engineering, Lanzhou University of Technology, Lanzhou, 730050, Gansu Province, China;2. National Active Distribution Network Technology Research Center, Beijing Jiaotong University, Haidian District, Beijing, 100044, China;3. National Experimental Teaching Center of Electrical and Control Engineering, Lanzhou University of Technology, Lanzhou 730050, Gansu Province, China
  • Supported by:
    Project Supported by National Natural Science Foundation of China (62063016)

摘要: 【目的】为提升系统在虚假数据注入攻击下的弹性能力,研究综合能源系统(integrated energy system, IES)遭受虚假数据注入攻击(false data injection attack, FDIA)后的弹性提升问题。【方法】构造系统遭受FDIA攻击的弹性提升框架,建立考虑安全性及经济性的系统弹性评估方法。在分析能量流与信息流攻击机理的基础上,提出基于连续小波变换(continuous wavelet transform, CWT)+卷积神经网络(convolutional neural network, CNN)的能量流FDIA检测方法,以及基于CWT+广义回归神经网络模型(general regression neural network, GRNN)的信息流FDIA检测方法。进一步分析FDIA攻击对系统调度造成的影响,建立优化调度模型,提升系统受到网络攻击后的弹性。【结果】含CWY+GRNN检测模型的综合能源系统弹性提升策略比不含检测模型的提升策略更具优越性,安全可靠性高出22.79%,运行经济性高出12.89%,弹性提升水平高出19.82%。含检测模型的综合能源系统弹性提升策略接近于不受网络攻击运行时的水平。【结论】提出CWT+GRNN检测模型的综合能源系统弹性提升策略在系统受FDIA攻击后能够明显提升系统弹性,系统弹性提升后的效果接近于正常运行时的水平。

关键词: 电热气, 综合能源系统(IES), 虚假数据注入攻击(FDIA), 攻击检测, 优化调度, 弹性提升

Abstract:  [Objectives] To enhance the resilience of the integrated energy system (IES) under false data injection attacks, this study aims to investigate the issue of resilience improvement after the IES is subjected to false data injection attacks (FDIA). [Methods] Construct a resilience enhancement framework for systems subjected to FDIA attacks, and establish a system resilience assessment method that considers both security and economy. Based on analyzing the attack mechanisms of energy flow and information flow, this paper proposes an energy flow FDIA detection method based on continuous wavelet transform (CWT)+convolutional neural network (CNN), as well as an information flow FDIA detection method based on CWT+general regression neural network (GRNN) model. Further, analyze the impact of FDIA attacks on system scheduling, establish an optimized scheduling model, and enhance the resilience of the system after network attacks. [Results] The comprehensive energy system resilience enhancement strategy with CWY+GRNN detection model is superior to the enhancement strategy without detection model, with 22.79% higher safety and reliability, 12.89% higher operational economy, and 19.82% higher resilience enhancement level. The comprehensive energy system resilience enhancement strategy with detection models is close to the level of operation without network attacks. [Conclusions] The comprehensive energy system resilience enhancement strategy proposed by the CWT+GRNN detection model can significantly improve system resilience after an FDIA attack, and the effect of system resilience enhancement is close to the level of normal operation.

Key words: electricity-heat-gas, integrated energy system(IES), false data injection attack(FDIA), attack detection, optimal dispatch, resilience enhancement