Ensemble Score Filtering for Real-Data Energy Consumption Forecast Correction

Ruoyu Hu, Dahai Yu, Feng Bao, Guang Wang, Guannan Zhang · 2026

Read Ensemble Score Filtering for Real-Data Energy Consumption Forecast Correction by Ruoyu Hu, Dahai Yu, Feng Bao, Guang Wang, Guannan Zhang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-29072.txt.

Description

Ruoyu Hu and team study data assimilation for real-world energy forecasting with partial, noisy, or delayed readings. A ready-made black-box space-time forecast model supplies forward predictions as the state driver. The Ensemble Score Filter folds in partial and noisy observations to correct the forecast path over time, using score-based diffusion models with a closed-form score and Monte Carlo sampling. Open-loop spread grows shaky over long horizons, while filter correction greatly lifts state estimation and beats the Ensemble Kalman Filter under nonlinear readings.

Catalog record

CREATORRuoyu Hu, Dahai Yu, Feng Bao, Guang Wang, Guannan Zhang
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE53205
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.29072

Browse the full library · About the Tower of Babel