An AI-Based Decision-Support Pipeline for Day-Ahead Photovoltaic Forecasting
2026-08-03 • Machine Learning
Machine LearningArtificial Intelligence
AI summaryⓘ
The authors focus on improving day-ahead solar power forecasts at a UK charging station, where short and imperfect data records make this challenging. They build a forecasting system that fixes timing errors, uses solar and weather information safely, and combines different prediction models to improve accuracy. Their approach outperforms simple reference methods and individual models, reducing forecast errors notably. They show that mixing physics-based features with machine learning can help even when site data is limited, but the benefits depend on how models are tested and used.
photovoltaic forecastingday-ahead forecastsolar irradiancemachine learning ensemblesolar geometryclearness indexforecast validationrolling-origin evaluationenvironmental AIpersistence forecasting
Authors
Fariba Dehghan, Sebastian Stein, Vahid Yazdanpanah, Stephanie Gauthier, Masood Nazari
Abstract
Reliable photovoltaic (PV) forecasts are needed for low-carbon energy systems, but newly deployed sites often have short, imperfect records. This makes standard day-ahead forecasting difficult: persistence and physical baselines can be sensitive to calibration and timestamp alignment, while single machine-learning models may capture only one structure in the data and overstate skill under non-temporal validation. We study this problem at a United Kingdom charging-station site, where PV forecast errors affect charging availability, storage scheduling, and downstream control. Using measured inverter output and publicly available meteorological inputs, we develop a deployment-oriented environmental-AI pipeline for day-ahead hourly PV forecasting. The pipeline corrects timestamp conventions, constructs leakage-safe solar-geometry and clearness-index features, adds short-term atmospheric context, and combines complementary predictors through validation-learned stacking. Against smart persistence, a clear-sky baseline that adjusts recent PV output using expected clear-sky irradiance, the best ensemble reduces daylight normalised RMSE by about 32% under random day-blocked evaluation and 9% under the stricter rolling-origin protocol. It also reduces daylight RMSE relative to the strongest individual machine-learning baseline by 6.6% and 6.4%, respectively. The results show that physics-aware stacking can support PV forecasts from limited site data, but its value depends on model class, evaluation protocol, and deployment context.