Scholarly record

ANALYSIS OF THE FANTANELE-COGEALAC WIND FARM PERFORMANCE BASED ON A NEURAL-NETWORK APPROACH

Roberto-Adrian Dobri, Florin Onea

Abstract

The present work investigates the performance of the Fantanele-Cogealac wind farm through the application of a neural-network-based modeling approach, aiming to capture the correlations between various meteorological parameters and the performances of a wind turbine. As a first step, the meteorological conditions from the vicinity of this site were evaluated from a statitiscal and seasonal point of view, by considering MERRA-2 dataset covering the time interval 2018-2025. The results are indicated in terms of surface pressure, wind speed measured at 10 m height and air temperature. As a next step, the performance of the Cogealac wind farm was assessed from a theorethical point of view, the obtained results being adjusted with some real data indicated in some previous reports. According to these results the simulated annual energy production is significantly much lower than the ones provided in the literature review. In the final part of this work, the EassyNN software was considered to simulate the correlation between the meteorological data and the performance of the Cogealac wind farm. The contribution of this research lies in offering a methodological framework that can be further applied to similar wind energy systems, supporting decision-making processes related to energy management, forecasting, and operational optimization.

Publication details

Back to publication list