← Back to all news
arXiv Machine Learning

Two-stage residual learning for calibrated uncertainty in solar irradiance forecasts

The paper presents a two‑stage residual learning approach for solar irradiance forecasting that separately predicts the expected irradiance with an LSTM and models the forecast error distribution with a second network. This method improves uncertainty estimation, reducing the deviation between predicted and historical standard deviations by 34.6 % for a standalone LSTM and 56.5 % when combined with Gaussian process regression. The approach captures total forecast error, providing more reliable uncertainty estimates for decision‑making.

Related stories

Hugging Face Trending Papers
Jul 14

Robustness of Deep Learning Models for PV Power Forecasting under NWP Forecast Errors: A Spatiotemporal and Physically Interpretable Analysis

Engineering use of AI forecasting models requires not only high nominal accuracy but also predictable behavior under uncertain inputs. In photovoltaic (PV) forecasting, this requirement is especially challenging because numerical weather prediction (NWP) errors are temporally correlated, state dependent, and physically coupled across variables.

More like this →
arXiv Machine Learning
Sep 23

PROSWIN: Probabilistic Solar Wind Speed Forecasting Using Deep Distributional Regression From Solar Images

PROSWIN is a probabilistic machine learning model that forecasts hourly solar wind speed at Earth up to four days ahead, using solar images and magnetograms processed by a deep neural network and distributional regression. It introduces a prediction score metric that rewards both timeline and high‑speed solar wind peak accuracy, achieving well‑calibrated uncertainties and superior performance on 14 years of data compared to existing models. The study highlights the importance of the 171 Å channel and demonstrates that probabilistic forecasts outperform single‑value models for both overall timelines and peak events.

By Daniel Collin, Yuri Shprits, Luca Chiarabini, Stefan J. Hofmeister, Nadja Klein, Guillermo Gallego
More like this →
arXiv AI
Sep 15

Horizon-specific Expert Fusion for Photovoltaic Power Forecasting

The paper presents a hierarchical ensemble for short‑term photovoltaic power forecasting that fuses temporal neural models, historical analogs, state climatology, and gradient‑boosted trees with horizon‑specific convex weights. Solar geometry and numerical weather forecasts provide expected generation conditions, while a calibration step corrects recent bias using historical forecast errors. Evaluated on PVDAQ and GEFCom2014 data, the ensemble reduces mean absolute error by up to 6% compared to strong individual models, though its advantage varies with dataset and horizon.

By Xu Yuqing, Zhou Liguo, Sun Ze, Yu Lei, Jiang Mingming
fine-tuningsafety
More like this →
arXiv AI
Jul 10

PARA-PV: Physics-Aware Retrieval-Augmented PV Prediction Based on Frozen Foundation Model and Distribution Shift Correction

arXiv:2607. 08079v1 Announce Type: new Abstract: Accurate photovoltaic (PV) power forecasting is essential for reliable grid dispatch and renewable energy integration, yet it remains challenging because PV generation is jointly shaped by weather variability, day-night transitions, regime-dependent dynamics, and strict physical constraints.

By Hang Fan, Weican Liu, Ying Lu, Dunnan Liu, Long Cheng, Wei Wei
rag
More like this →
arXiv AI
Oct 31

IMPACT: Inverter-Informed Multimodal Learning for Photovoltaic Generation Forecasting

The paper introduces IMPACT, a physics‑informed multimodal deep‑learning framework that forecasts photovoltaic power 10–110 minutes ahead with probabilistic outputs. It uses a four‑branch architecture to encode plant‑level temporal history, inverter‑level spatial cloud dynamics, known‑future solar geometry, and current meteorological data, fusing them via gated cross‑attention. The model outperforms 14 baselines, achieving a forecast skill score of 0.2950 and maintaining strong performance under overcast and variable sky conditions.

By Anushka Katiyar, Cong Feng, Jing Wang, Tristan Liu, Simon Julien, Zachary Jacobs
multimodalbenchmarks
More like this →
arXiv Machine Learning
Oct 22

Probabilistic Machine Learning for Short-Term Atmospheric Seeing Forecasting

The study evaluates machine learning models for forecasting atmospheric seeing up to two hours ahead, emphasizing probabilistic forecasting and uncertainty quantification. Gaussian Processes and a normalizing‑flow‑based model (FloTS) are compared against deterministic RNN and LSTM baselines using Maunakea summit data. Results show that deep learning models, particularly LSTM and FloTS, achieve the best predictive accuracy, with FloTS offering flexible and informative uncertainty estimates.

By Mary-Joe Medlej, Rahul Srinivasan, Simon Prunet, Aziz Ziad, Christophe Giordano
benchmarks
More like this →