Half-hour global solar radiation forecasting based on static and dynamic multivariate neural networks

• DOI: 10.36909/jer.v9i2.8455
• Publication Date: 2021-05-29T10:09:35Z
• Authors: Mohammed Ali Jallal; Samira Chabaa; ; Abdelouhab Zeroual
• Keywords: Robustness; Empirical modelling

Precise global solar radiation (GSR) measurements in a given location are very essential for designing and supervising energy systems. In the case of rarity or absence these measurements, it is important to have theoretical empirical model compute GSR values. Therefore, main goal this work offer, designers engineers systems, an appropriate accurate way predict half-hour (HHGSR) time series from some available meteorological parameters (relative humidity, air temperature, wind speed, precipitation, acquisition vector scale). For that purpose, two intelligent models developed: first one multivariate dynamic neural network with feedback connection, second static network. The database used build was recorded Agdal’s station Marrakesh, Morocco, during years 2013 2014, divided into subsets. subset training validating models, testing efficiency robustness developed models. obtained results, terms statistical performance indicators, demonstrate forecasting accurately HHGSR parameter city Morocco.