A comprehensive review of hybrid models for solar radiation forecasting

• Published in: Journal of cleaner production Vol. 258; p. 120357
• Main Authors: Guermoui, Mawloud, Melgani, Farid, Gairaa, Kacem, Mekhalfi, Mohamed Lamine
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 10.06.2020
• Subjects: Hybrid Models; Machine Learning; Solar Resource Estimation; Solar Resource Estimation; Machine Learning; Hybrid Models
• ISSN: 0959-6526; 1879-1786
• DOI: 10.1016/j.jclepro.2020.120357

Solar radiation components assessment is a highly required parameter for solar energy applications. Due to the non-stationary behavior of solar radiation parameters and variety of atmosphere conditions, stand-alone forecasting models are insufficient for providing accurate estimation in some cases. In this respect, different hybrid models have been proposed in recent years to overcome the limitations of single models and boost the forecasting precision. In this paper, acomprehensive literature review of the recent trends in hybrid model techniques for solar radiation components assessment is presented. The main objective behind this study is to present a comparative study between different hybrid models, explore their application, and identify promising and potential models for solar radiation application assessment. The performance ranking of each hybrid model is complicated due the diversity of the data length and scale, forecasting horizon, performance metrics, time step and climate condition. Overall, the presented study provides preliminary guidelines for a complete view of the hybrid models and tools that can be used in order to improve solar radiation assessment. •In-depth review of different hybrid methods for solar radiation forecasting is presented.•Forecasting horizon for each category is discussed.•Worldwide regions and data are investigated.