Submission 121
A Virtual Power Plant Novel Approach: Spatiotemporal Graph Neural Network Smart-Contract Energy Management
WISO25-121
Presented by: Ignacio Smith Salazar
The world for the last 15 years has seen the increase in the deployment of Distributed Energy Resources (DERs); this rapid growth in areas where there is an existing grid has caused significant grid instability. Consequently, optimizing energy dispatch requires the implementation of innovative solutions. To achieve a smooth energy delivery curve, we need a coordinated effort among stakeholders and resources.
Our paper introduces an innovative way to solve the problem; it entails using a virtual power plant system that employs spatio-temporal graph neural networks (GNNs) to accurately predict how distributed energy resources (DERs) behave, along with smart contracts that use energy tokens for better management.
The automated dispatch process and smart contracts work like a flexible priority system that considers the available extra energy from the sources and how close the independent power producers are. The main advantage is a decentralized, secure, and intelligent virtual power plant (VPP) delivery system that aims to balance local energy and use resources wisely.
Our concept proposes better utilization and energy use while also offering a secure and clear way to handle transactions, all while recognizing the challenges of scaling and performing in real-time.