1.云南电网有限责任公司电力科学研究院,云南省 昆明市 650217
云南省绿色能源与数字电力量测及控保重点实验室,云南省 昆明市 650217
3.云南电网有限责任公司怒江供电局,云南省 怒江市 673100
4.输变电装备技术全国重点实验室(重庆大学),重庆市,400044
刘斯扬(1992—),男,博士研究生,工程师,研究方向:物联网与智能量测技术,E-mail:lsyfd27@sina.com ;
李博,高级工程师,研究方向:电能计量,E-mail:49923387@qq.com。
收稿:2025-11-26,
修回:2026-03-08,
录用:2026-05-18,
移动端阅览
刘斯扬, 李博, 朱萌瑶, 等. 基于GAN数字孪生框架的电力物联网边缘智能对抗性测试方法[J/OL]. 电信科学, 2026.
Liu Siyang, Li Bo, Zhu Mengyao, et al. A GAN-based Adversarial Testing Method for Edge Intelligence in Power Internet of Things via Digital Twin Framework[J/OL]. Telecommunications Science, 2026.
刘斯扬, 李博, 朱萌瑶, 等. 基于GAN数字孪生框架的电力物联网边缘智能对抗性测试方法[J/OL]. 电信科学, 2026. DOI: 10.11959/j.issn.1000-0801.DXKX250682.
Liu Siyang, Li Bo, Zhu Mengyao, et al. A GAN-based Adversarial Testing Method for Edge Intelligence in Power Internet of Things via Digital Twin Framework[J/OL]. Telecommunications Science, 2026. DOI: 10.11959/j.issn.1000-0801.DXKX250682.
针对电力物联网边缘节点面临的复杂对抗攻击识别难、实时性要求高及资源约束严等问题,提出一种基于GAN数字孪生框架的边缘智能对抗性测试方法。首先构建融合物理约束与时序一致性的GAN模型,生成符合边缘设备运行规律的对抗样本,结合数字孪生技术搭建边缘节点虚拟镜像,实现攻击场景的精准复现;其次设计常规工况、单一对抗、混合对抗三类测试场景,建立包含检测准确率、响应时延、虚假阳性率及算力占用率的四维评价体系;最后,案例分析结果表明,该方法在混合攻击场景下检测准确率达89.5%,响应时延控制在3.3-5.2ms,虚假阳性率低至0.5%,算力占用维持在22%-23%区间,较传统规则匹配法与单一机器学习模型,在复杂攻击识别能力、实时性及资源适配性上均有显著提升,为电力物联网边缘智能安全测试提供了高效可行的解决方案。
To address the challenges faced by edge nodes in the Power Internet of Things (PIoT)
such as difficulty in identifying complex adversarial attacks
high real-time requirements
and strict resource constraints
an adversarial testing method for edge intelligence based on a GAN-driven digital twin framework is proposed. Firstly
a GAN model integrating physical constraints and temporal consistency is constructed to generate adversarial samples that conform to the operating rules of edge devices. Combined with digital twin technology
a virtual mirror of edge nodes is built to achieve accurate reproduction of attack scenarios. Secondly
three types of test scenarios—normal operating conditions
single adversarial attacks
and mixed adversarial attacks—are designed
and a four-dimensional evaluation system including detection accuracy
response delay
false positive rate
and computing power occupancy rate is established. Finally
case analysis results show that the proposed method achieves a detection accuracy of 89.5% in mixed attack scenarios
controls the response delay within the range of 3.3–5.2 ms
reduces the false positive rate to as low as 0.5%
and maintains the computing power occupancy between 22% and 23%. Compared with traditional rule-based matching methods and single machine learning models
it exhibits significant improvements in complex attack identification capability
real-time performance
and resource adaptability
providing an efficient and feasible solution for the security testing of edge intelligence in the PIoT.
丰雷 , 谢坤宜 , 朱亮 , 等 . 面向电网业务质量保障的 5G 高可靠低时延通信资源调度方法 [J ] . 电子与信息学报 , 2021 , 43 ( 12 ): 3418 - 3426 .
FENG Lei , XIE Kunyi , ZHU Liang , et al . 5G Ultra - reliable and low latency communication resource scheduling for power business quality assurance [J ] . Journal of Electronics & Information Technology , 2021 , 43 ( 12 ): 3418 - 3426 .
缪巍巍 , 曾铮 , 张明轩 , 等 . 基于多智能体强化学习的边缘物联代理资源分配算法 [J ] . 电力信息与通信技术 , 2021 , 19 ( 12 ): 9 - 15 .
MIAO Weiwei , ZENG Zheng , ZHANG Mingxuan , et al . Edge IoT agent resource allocation algorithm based on multi-agent reinforcement learning [J ] . Electric Power Information and Communication Technology , 2021 , 19 ( 12 ): 9 - 15 (in Chinese) .
韩富佳 , 王晓辉 , 乔骥 , 等 . 基于人工智能技术的新型电力系统负荷预测研究综述 [J ] . 中国电机工程学报 , 2023 , 43 ( 22 ): 8569 - 8591 .
HAN Fujia , WANG Xiaohui , QIAO Ji , et al . Review on artificial intelligence based load forecasting research for the new - type power system [J ] . Proceedings of the CSEE , 2023 , 43 ( 22 ): 8569 - 8591 .
黄兴 , 李立 , 王维 , 等 . 面向电力物联网的认知反向散射通信鲁棒资源分配算法 [J ] . 电力信息与通信技术 , 2023 , 21 ( 10 ): 35 - 40 .
HUANG Xing , LI Li , WANG Wei , et al . Robust resource allocation algorithm in cognitive backscatter communication networks for power internet of things [J ] . Electric Power Information and Communication Technology , 2023 , 21 ( 10 ): 35 - 40 (in Chinese) .
张琦 , 李志浩 , 范叶平 , 等 . 新型电力系统中一种基于 LSTM 和 CNN 的倾斜样本预测算法 [J ] . 电信科学 , 2025 , 41 ( 10 ): 211 - 221 .
ZHANG Qi , LI Zhihao , FAN Yeping , et al . An unbalanced sample prediction algorithm based on LSTM and CNN in the new power systems [J ] . Telecommunications Science , 2025 , 41 ( 10 ): 211 - 221 .
蒋守花 , 冯军 , 舒晖 , 等 . 基于深度强化学习的数据传输策略优化研究 [J ] . 电信科学 , 2025 , 41 ( 8 ): 149 - 162 .
JIANG Shouhua , FENG Jun , SHU Hui , et al . Research on optimization of data transmission strategies based on deep reinforcement learning [J ] . Telecommunications Science , 2025 , 41 ( 8 ): 149 - 162 .
周江 , 陈扬 , 凌凌云 . 基于效用函数的无源光网络动态带宽分配算法 [J ] . 计算机与现代化 , 2023 ( 4 ): 106 - 110 .
ZHOU Jiang , CHEN Yang , LING Lingyun . Dynamic bandwidth allocation algorithm for passive optical networks based on utility function [J ] . Computer and Modernization , 2023 ( 4 ): 106 - 110 (in Chinese) .
严兴煜 , 高赐威 , 陈涛 , 等 . 数字孪生虚拟电厂系统框架设计及其实践展望 [J ] . 中国电机工程学报 , 2023 , 43 ( 2 ): 604 - 618 .
YAN Xingyu , GAO Ciwei , CHEN Tao , et al . Framework design and application prospect for digital twin virtual power plant system [J ] . Proceedings of the CSEE , 2023 , 43 ( 2 ): 604 - 618 .
王日宁 , 武一 , 魏浩铭 , 等 . 基于智能终端特征信号的配电网台区拓扑识别方法 [J ] . 电力系统保护与控制 , 2021 , 49 ( 6 ): 83 - 89 .
WANG Rining , WU Yi , WEI Haoming , et al . Topology identification method for a distribution network area based on the characteristic signal of a smart terminal unit [J ] . Power System Protection and Control , 2021 , 49 ( 6 ): 83 - 89 .
卢宇亭 , 张乃平 , 高险峰 , 等 . 考虑频谱利用率的无源光网络多业务带宽动态分配方法 [J ] . 自动化与仪器仪表 , 2023 ( 7 ): 55 - 59 .
LU Yuting , ZHANG Naiping , GAO Xianfeng , et al . Dynamic bandwidth allocation method for passive optical networks considering spectrum utilization [J ] . Automation & Instrumentation , 2023 ( 7 ): 55 - 59 (in Chinese) .
赵鹏 , 蒲天骄 , 王新迎 , 等 . 面向能源互联网数字孪生的电力物联网关键技术及展望 [J ] . 中国电机工程学报 , 2022 , 42 ( 2 ): 447 - 458 .
ZHAO Peng , PU Tianjiao , WANG Xinyin , et al . Technologies and perspectives of power Internet of Things facing with digital twins of the energy Internet [J ] . Proceedings of the CSEE , 2022 , 42 ( 2 ): 447 - 458 .
刘林 , 祁兵 , 李彬 , 等 . 面向电力物联网新业务的电力通信网需求及发展趋势 [J ] . 电网技术 , 2020 , 44 ( 8 ): 3114 - 3128 .
LIU Lin , QI Bing , LI Bin , et al . Requirements and developing trends of electric power communication network for new services in electric internet of things [J ] . Power System Technology , 2020 , 44 ( 8 ): 3114 - 3128 .
张宁 , 杨经纬 , 王毅 , 等 . 面向泛在电力物联网的 5G 通信:技术原理与典型应用 [J ] . 中国电机工程学报 , 2019 , 39 ( 14 ): 4015 - 4024 .
ZHANG Ning , YANG Jingwei , WANG Yi , et al . 5G Communication for the ubiquitous internet of things in electricity:technical principles and typical applications [J ] . Proceedings of the CSEE , 2019 , 39 ( 14 ): 4015 - 4024 .
0
浏览量
0
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621