Energy Applications

Featured Articles - Volume 1, Number 1

• Graph Neural Network-Based Battery State-of-Health Estimation with Multi-Scale Temporal Feature Extraction for Grid-Scale Energy Storage

  Yifei Yin

  Grid-scale energy storage systems face significant challenges in maintaining operational reliability due to the complex degradation patterns of individual battery cells within large-scale configurations. Existing state-of-health estimation methods struggle with the interconnected nature of battery deterioration processes...

  Keywords: State-Of-Health, Graph Neural Networks, Multi-Scale Feature Extraction, Grid-Scale Energy Storage

• Ultra-Short-Term Solar Power Forecasting Under Extreme Weather Using Multi-Scale Temporal Fusion Transformers

  Yanling Liu

  Solar power forecasting accuracy drops significantly during extreme weather events, creating substantial challenges for grid operators and energy traders. Current forecasting models struggle with rapid weather transitions and complex atmospheric dynamics...

  Keywords: Solar Power Forecasting, Transformer Neural Networks, Extreme Weather, Multi-Scale Temporal Analysis

• Multi-Agent Deep Reinforcement Learning for Decentralized Energy Trading in Peer-to-Peer Networks with Blockchain Integration

  Ling Liu

  The energy sector faces unprecedented challenges due to increasing renewable energy integration and the need for decentralized market structures. Traditional centralized energy trading systems struggle with scalability issues and single points of failure...

  Keywords: Multi-agent reinforcement learning, Blockchain, Peer-to-peer energy trading, Decentralized markets

• Edge-Cloud Collaborative Framework for Ultra-Low Latency Energy Demand Forecasting in Smart Buildings Using Lightweight Neural Networks

  Yanlun Chen

  Modern smart buildings require precise energy demand forecasting to optimize their operations and reduce costs. Traditional cloud-based prediction systems face latency issues that limit their effectiveness in real-time applications...

  Keywords: Edge Computing, Energy Forecasting, Lightweight Neural Networks, Smart Buildings

• Generative Adversarial Networks for Synthetic Load Profile Generation in Privacy-Preserving Smart Meters

  Xingwang Pi

  Smart meter data contains sensitive information about household energy consumption patterns that can reveal private details about residents' daily activities and lifestyle habits. Current privacy protection methods for smart meter data sharing often compromise data utility...

  Keywords: Smart Meters, Load Profile Generation, Generative Adversarial Networks, Privacy Preservation