A System-integrating and Energy Forecasting Tool for Renewable Energy Solutions
Abstract
Accurate forecasting of photovoltaic (PV) system performance is increasingly critical as energy markets evolve. Such predictions enable aggregators and energy communities to participate in short-term markets, enhancing revenue and supporting grid stability. This study introduces a performance model capable of predicting PV output within a cluster using only internal generation data. The model is designed to operate independently of external data sources, ensuring robustness and simplicity. The objective of this work is to support energy communities and grid operators in energy planning. Incorporating clear-sky baselines, real-time data, and machine learning, the developed model advances short-term spatial forecasting to support smart and sustainable energy systems. Furthermore, it aims to scale forecasting capabilities from local clusters to a regional or island-wide level, such as Cyprus. This effort aligns with broader goals of fostering smart, sustainable urban development.
Keywords
solar energy markets, grid stability, PV power output, spatiotemporal forecasting.