Adaptive Matrix Decomposition for False Data Injection Attack Detection in Consensus-Based Home Energy Management Systems Integrating Electric Vehicles
针对集成电动汽车的家庭能量管理系统，提出一种增强型快速Go分解技术，通过自适应参数选择动态检测虚假数据注入攻击，在FREEDM微电网仿真中达到99.96%检测准确率，仅0.02%误报率。
Modern home energy management systems (HEMSs) increasingly employ consensus-based algorithms for distributed energy resource (DER) planning. However, these cyber–physical systems face emerging security vulnerabilities that traditional detection methods cannot adequately address. Despite extensive research on consensus-based energy management and cyberattack detection independently, significant limitations persist in their integration, particularly in residential networks with bidirectional electric vehicles (EVs) and vehicle-to-grid (V2G) capabilities. This article addresses this research gap by developing a comprehensive framework that integrates consensus-based home energy management optimization with an advanced false data injection attack (FDIA) detection methodology. Our primary innovation is the enhanced fast go decomposition (EFGD) technique. EFGD employs adaptive parameter selection mechanisms that dynamically determine matrix rank and detection thresholds based on statistical properties of power exchange data. This enables more accurate separation of legitimate system behavior from malicious data injections compared to fixed-parameter approaches. Simulation results on the future renewable electric energy delivery and management (FREEDM) microgrid system with three distributed energy storage devices (DESDs), a wind turbine, a photovoltaic panel, a bidirectional EV, and residential loads over 24-h operation cycles demonstrate superior performance. The consensus-based optimization effectively coordinates multiple energy resources, while the integrated security framework successfully detects attacks that cause 0.52% daily operational cost increases without compromising distributed management efficiency. The EFGD method achieves 99.96% detection accuracy with only 0.02% false positive rate and 98.7% precision. This approach addresses critical vulnerabilities in residential energy systems with V2G integration while maintaining practical implementation feasibility.