← 返回
电动汽车驱动 ★ 4.0

事件触发的智能控制方案用于多区域电网中数据完整性攻击的缓解

Event-Triggered Intelligent Control Scheme for Data Integrity Attack Mitigation in Multi-Area Power Grids

作者 Vivek Kapil · Sheetla Prasad
期刊 IEEE Transactions on Power Systems
出版日期 2024年5月
技术分类 电动汽车驱动
相关度评分 ★★★★ 4.0 / 5.0
关键词 电网 物联网技术 网络攻击 事件触发智能控制 人工智能估计器
语言:

中文摘要

随着物联网技术的广泛应用,电力系统日益智能化。大量传感器通过通信信道实时传输状态估计数据,易导致通信过载并面临网络攻击风险。虚假数据注入攻击可能篡改系统频率,破坏负荷频率控制性能,甚至引发电网崩溃。本文提出一种结合人工智能估计器与非线性滑模控制器的事件触发式智能控制方案,有效降低通信负担并抑制网络入侵影响。AI观测器利用反向传播误差算法预测系统状态并检测攻击,事件触发SMC则抑制攻击对系统轨迹的不利影响。MATLAB仿真验证了该策略在应对网络攻击及分布式可再生能源接入时的有效性。

English Abstract

The power grids are becoming smarter with adoption of various IoT Technologies (Internet of Things). IoT calls for sensors, which are transacting real time data through communication channels, installed at various locations in power system. These communication channels not only become overloaded by high volume of state-estimation data from these sensors but are also prone to cyber-attacks. The cyber-attack may infirmly affect the load frequency control (LFC) schemes which are implemented across multi-area grids using state estimation techniques. The cyber-attack may modify system frequency via signal false data injections in system and therefore deteriorates system performance or may even lead to blackout. In this work, artificial intelligent (AI) estimator and nonlinear sliding mode controller (SMC) are combined together to form event-triggered intelligent control scheme which not only optimise the use of communication infrastructure but also attenuates cyber intrusions. The AI observer predicts grid states and also senses cyber intrusions via the back propagation error algorithm in presence of cyber-attacks. The proposed event-triggered SMC rejects the undesirable impact of cyber intrusions on the plant trajectories. The proposed event-triggered control strategy when subjected to cyber attack and power penetration from the distributed renewable resources is demonstrated through MATLAB simulations.
S

SunView 深度解读

该事件触发智能控制方案对阳光电源PowerTitan储能系统和iSolarCloud云平台具有重要应用价值。针对大规模储能电站面临的通信过载和网络攻击风险,文章提出的AI观测器结合滑模控制的架构可直接应用于ST系列储能变流器的网络安全防护。事件触发机制能有效降低储能集群通信负担,AI估计器可增强iSolarCloud平台的异常检测能力,识别虚假数据注入攻击对频率调节的干扰。该方案对阳光电源构网型GFM控制技术在多区域电网协同中的安全性提升具有启发意义,可融入智能运维系统实现预测性安全防护,保障分布式光储系统在复杂网络环境下的稳定运行。