Partager

Document IIF

Causality-preserving short-term cooling-load forecasting with online correction for building cooling systems.

Auteurs : LI D., WANG Y., SHEN Y., WAN C., LIN S., ZHOU B.

Type d'article : Article de la RIF

Résumé

Accurate short-term cooling-load forecasting is essential for cooling-capacity reservation, peak-demand management, and reliable operation of building cooling systems. However, many decomposition-based forecasting methods rely on offline full-series decomposition, which may introduce look-ahead information leakage, and their open-loop forecasts are sensitive to changes in operating conditions. This study develops a causality-preserving forecasting framework with online residual correction for building cooling load. At each prediction step, only available historical data are decomposed to separate slow cooling-demand trends from fast load disturbances. Frequency-aware predictors generate a prior forecast, and selected operational measurements are assimilated to correct residual errors online. Tests on office-building cooling-load data achieve MAE/RMSE/𝑅2 values of 15.73 RT/23.94 RT/0.887 and 17.17 RT/22.54 RT/0.952 under hourly and 30-min scenarios, respectively. The results demonstrate improved tracking of cooling-load variations and reduced load-estimation errors under the tested scenarios, providing more accurate load information for subsequent operational decision-making.

Documents disponibles

Format PDF

Pages : 12 p.

Disponible

  • Prix public

    20 €

  • Prix membre*

    Gratuit

* meilleur tarif applicable selon le type d'adhésion (voir le détail des avantages des adhésions individuelles et collectives)

Détails

  • Titre original : Causality-preserving short-term cooling-load forecasting with online correction for building cooling systems.
  • Identifiant de la fiche : 30035299
  • Langues : Anglais
  • Sujet : Technologie
  • Source : International Journal of Refrigeration - Revue Internationale du Froid - vol. 190
  • Date d'édition : 10/2026
  • DOI : http://dx.doi.org/https://doi.org/10.1016/j.ijrefrig.2026.107061

Liens


Voir d'autres articles du même numéro (29)
Voir la source

Indexation