Ensemble-Based Solar Irradiance Forecasting for Decentralized Energy Systems in Ibadan, Nigeria

Surface solar irradiance volatility impairs battery scheduling and frequency stability in West African community minigrids. This study models hourly Global Horizontal Irradiance (GHI) using 8,015 ground-based observations from Ibadan, Nigeria. To overcome the limitations of standard empirical adaptations, we formulate a physics-informed machine learning framework benchmarking Random Forest and Gradient Boosting against Multiple Linear Regression, integrating trigonometric solar geometry and first-order autoregressive lags. The Gradient Boosting architecture achieved a superior Mean Absolute Error of 17.226 W/m² alongside a coefficient of determination of 0.9849. Split-criterion attribution demonstrates that solar hour cosine formulations govern 80 percent of feature importance, confirming the structural primacy of diurnal celestial mechanics over terrestrial noise. Operational dispatch simulations confirm that these tree ensembles anticipate convective cloud transients exceeding 340 W/m² per hour, providing the predictive horizon required to buffer lithium storage and prevent uncoordinated diesel generator activation.

Authors

Institutions

Publication Details

Journal
Nigerian Journal of Sustainability Research
Published
2026-11-16
DOI
https://doi.org/10.67163/njsr.4.2.2026.107
Primary Topic
Solar Radiation and Photovoltaics
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Ensemble-Based Solar Irradiance Forecasting for Decentralized Energy Systems in Ibadan, Nigeria

Adeleke ADENIYAN
2 links
Nigerian Journal of Sustainability Research
Solar Radiation and Photovoltaics
article

Ensemble-Based Solar Irradiance Forecasting for Decentralized Energy Systems in Ibadan, Nigeria

Adeleke ADENIYAN
article en

Abstract

Surface solar irradiance volatility impairs battery scheduling and frequency stability in West African community minigrids. This study models hourly Global Horizontal Irradiance (GHI) using 8,015 ground-based observations from Ibadan, Nigeria. To overcome the limitations of standard empirical adaptations, we formulate a physics-informed machine learning framework benchmarking Random Forest and Gradient Boosting against Multiple Linear Regression, integrating trigonometric solar geometry and first-order autoregressive lags. The Gradient Boosting architecture achieved a superior Mean Absolute Error of 17.226 W/m² alongside a coefficient of determination of 0.9849. Split-criterion attribution demonstrates that solar hour cosine formulations govern 80 percent of feature importance, confirming the structural primacy of diurnal celestial mechanics over terrestrial noise. Operational dispatch simulations confirm that these tree ensembles anticipate convective cloud transients exceeding 340 W/m² per hour, providing the predictive horizon required to buffer lithium storage and prevent uncoordinated diesel generator activation.

Nigerian Journal of Sustainability ResearchVol. 4(2)
Wesley University (NG)
Affordable and clean energy, Climate action, Industry, innovation and infrastructure
Solar Radiation and Photovoltaics
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon