Ir directamente a la navegación principal Ir directamente a la búsqueda Ir directamente al contenido principal

Hydropower generation forecasting with Artificial Intelligence and Machine Learning: A systematic review of modeling approaches, input variables, and validation practices

  • Universidad de Cuenca

Producción científica: Contribución a una revista › Artículo de revisión › revisión exhaustiva

Resumen

Hydropower plays a central role in renewable energy generation, yet its forecasting remains challenging due to the nonlinear, temporally dependent, and highly variable nature of hydrological and operational systems. Although recent advances in Artificial Intelligence (AI) and Machine Learning (ML) have shown promising results, the literature remains heterogeneous in terms of modeling approaches, input variables, and validation practices. This study presents a systematic review of studies published between 2014 and 2024 on AI/ML applications to hydropower generation and operational forecasting. Following a domain-specific reassessment of 51 studies, 13 studies — nine Core and four Transferable — met the criteria for inclusion in the main analytical synthesis, while 18 additional studies were retained exclusively as contextual methodological evidence. Within the analytical corpus, Artificial Neural Networks were the most frequently identified algorithmic family (53.8%), hybrid modeling was reported in 61.5% of studies, and historical records were the dominant data source (84.6%). Monthly (53.8%) and daily (46.2%) temporal resolutions were the most common. Holdout validation predominated (69.2%), whereas rolling-origin and walk-forward validation were each identified in only 7.7% of studies. Explicit climate adaptability was incorporated in 23.1% of the analytical corpus, while 53.8% did not explicitly address this dimension. The findings reveal persistent limitations in time-aware validation, operational data integration, climate adaptability, and cross-study comparability. Although the restricted English-language search and domain-specific selection criteria limit the comprehensiveness of the evidence base, the review provides a more focused synthesis of hydropower-specific forecasting research and identifies priorities for developing more robust, reproducible, climate-aware, and operationally relevant AI/ML forecasting frameworks.

Idioma originalInglés
Número de artículo109789
PublicaciónEnergy Reports
Volumen16
DOI
EstadoPublicada - dic 2026

ODS de las Naciones Unidas

Este resultado contribuye a los siguientes Objetivos de Desarrollo Sostenible

  1. ODS 7: Energía asequible y no contaminante
    ODS 7: Energía asequible y no contaminante

Huella

Profundice en los temas de investigación de 'Hydropower generation forecasting with Artificial Intelligence and Machine Learning: A systematic review of modeling approaches, input variables, and validation practices'. En conjunto forman una huella única.

Citar esto