IoT-Assisted Power Monitoring in Smart Grids using Progressive Graph Convolutional Network

Smart grids (SGs) benefit from power monitoring through Internet of Things (IoT) technology which establishes continuous information collection systems that advance energy management (EM) features alongside power system (PS) reliability. To defeat security threats against data privacy and cyber risk...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:2025 7th International Conference on Inventive Material Science and Applications (ICIMA) s. 406 - 411
Hlavní autori: Askar, Sami, Mahaboob Basha, A. M., Sethi, Gaurav, Siva Ramkumar, M., Dhivya, S., Kameswara Rao, Nynalasetti Kondala
Médium: Konferenčný príspevok..
Jazyk:English
Vydavateľské údaje: IEEE 28.05.2025
Predmet:
Graph convolutional networks
Internet of Things
Monitoring
Photovoltaic
Power system stability
Progressive Graph Convolutional Network
Reliability
Scalability
Security
Smart grids
Spatial resolution
Stability analysis
Wind
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Smart grids (SGs) benefit from power monitoring through Internet of Things (IoT) technology which establishes continuous information collection systems that advance energy management (EM) features alongside power system (PS) reliability. To defeat security threats against data privacy and cyber risks from IoT technology integration requires substantial amount of work. The paper introduces Progressive Graph Convolutional Network (PGCN) as a solution for power monitoring systems operated by IoT technologies in SGs. The proposed method builds an adaptive learning system to enhance stability and predictive forecasting alongside operational performance in power grids. PGCN is used to predict load variations and detect anomalies in SGs by effectively capturing spatio-temporal dependencies in IoT-sensed power data. An implementation of PGCN on the MATLAB platform to evaluate it against including Multi-Layer Deep Recurrent Neural Network (MLDRNN), Chebyshev Nonspiritual Network Model (CN2M), and Binary Orientation Search Algorithm (BOSA) as existing techniques. The evaluation results show PGCN delivers superior accuracy than MLDRNN and BOSA and CN2M because it achieves 98% success rate while they reach 83%, 79% and 87% respectively. The research outcomes established PGCN's capability to utilize spatial-temporal relationship patterns for developing enhanced power monitoring systems in SGs.
DOI:10.1109/ICIMA64861.2025.11074202