Photovoltaic Knowledge-Informed Neural Network (PKINN)
Description
This dataset contains synchronized photovoltaic (PV) power measurements and environmental variables collected from multiple silicon-based PV modules under real outdoor operating conditions. The data include irradiance, temperature, electrical characteristics, and module-level power outputs recorded at a 5-minute resolution. These measurements capture a wide range of natural fluctuation behaviors, including both gradual variations and abrupt power ramps caused by dynamic weather changes. The dataset is intended to support research on PV power forecasting, data-driven modeling, Ensemble methods, and fluctuation analysis under realistic environmental variability. It enables benchmarking of forecasting models, evaluation of robustness under fluctuating conditions, and development of interpretable prediction frameworks for PV systems.
Files
Institutions
- Xi'an Jiaotong-Liverpool University