Physics-Informed State Space Models for Reliable Solar Irradiance Forecasting in Off-Grid Systems
2026-04-13 • Machine Learning
Machine LearningArtificial Intelligence
AI summaryⓘ
The authors developed a new solar forecasting model to improve how off-grid solar power systems predict sunlight and power generation. Their method combines weather data and precise celestial mechanics to avoid errors like predicting power at night or delays when clouds move. They tested their model over five years in a dry climate and found it accurately follows sunlight changes without impossible predictions. This lightweight model can run efficiently on small solar controllers to help maintain reliable solar power.
solar forecastingphotovoltaic systemscelestial mechanicsdeep learningKoopman operatorRiemannian manifoldatmospheric opacityclear-sky modelsphase lagmicrogrid control
Authors
Mohammed Ezzaldin Babiker Abdullah
Abstract
The stable operation of autonomous off-grid photovoltaic systems dictates reliance on solar forecasting algorithms that respect atmospheric thermodynamics. Contemporary deep learning models consistently exhibit critical anomalies, primarily severe temporal phase lags during cloud transients and physically impossible nocturnal power generation. To resolve this divergence between data-driven modeling and deterministic celestial mechanics, this research introduces the Thermodynamic Liquid Manifold Network. The proposed methodology projects 15 meteorological and geometric variables into a Koopman-linearized Riemannian manifold to systematically map complex climatic dynamics. The architecture integrates a Spectral Calibration unit and a multiplicative Thermodynamic Alpha-Gate. This system synthesizes real-time atmospheric opacity with theoretical clear-sky boundary models, structurally enforcing strict celestial geometry compliance. This completely neutralizes phantom nocturnal generation while maintaining zero-lag synchronization during rapid weather shifts. Validated against a rigorous five-year testing horizon in a severe semi-arid climate, the framework achieves an RMSE of 18.31 Wh/m2 and a Pearson correlation of 0.988. The model strictly maintains a zero-magnitude nocturnal error across all 1826 testing days and exhibits a sub-30-minute phase response during high-frequency transients. Comprising exactly 63,458 trainable parameters, this ultra-lightweight design establishes a robust, thermodynamically consistent standard for edge-deployable microgrid controllers.