Solar Energy Harvesting Dataset (2023–2025)

This dataset contains over 2.5 million real-world measurements collected from a photovoltaic (PV) inverter system between 2023 and 2025. The data were recorded as part of research on solar energy forecasting, energy-aware intermittent computing, and predictive checkpointing for IoT nodes.

Measurements were acquired from a SH10RT inverter system equipped with dual MPPT channels and three-phase output monitoring. Each record includes synchronized current and voltage readings captured at high temporal resolution, supporting studies in energy forecasting, digital twins, and low-power system optimization.

The dataset is divided into two major components merged by timestamp:Current dataset — MPPT and phase current readingsVoltage dataset — MPPT, phase, and battery voltage readings

File Structure- `solar_data_2023_2025.csv.gz` — full merged dataset (compressed)- `sample_10k.csv` — 10,000-row subset for testing and analysis- All timestamps are stored as UTC datetime converted from nanosecond epoch.

Applications- Solar energy forecasting (short-term and day-ahead)- Energy-aware and intermittent system modeling- Battery and inverter performance analytics- Edge intelligence and digital twin simulation

CitationIf you use this dataset, please cite:> Abubakar, J. A. (2025) et al. Solar Energy Harvesting Dataset (2023–2025). Zenodo. https://doi.org/10.5281/zenodo.17400949

LicenseReleased under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. This allows for free use, redistribution, and adaptation with appropriate credit.

AcknowledgmentData collected within the research framework on predictive energy-aware computing and solar-powered IoT systems, University of Bremen.