Banca de QUALIFICAÇÃO: AMANDA FERREIRA SAMPAIO

Uma banca de QUALIFICAÇÃO de DOUTORADO foi cadastrada pelo programa.
STUDENT : AMANDA FERREIRA SAMPAIO
DATE: 21/02/2022
TIME: 14:00
LOCAL: CCET - Video Conferência
TITLE:

HYBRID PREDICTION OF ATMOSPHERIC VARIABLES TO CALCULATE THE WIND POTENTIAL FOR GREEN HYDROGEN PRODUCTION.


KEY WORDS:

Time Series Modeling, Forecasting; Holt-Winters, Box-Jenkins, Box-Jenkins-Tiao, Probabilistic Simulation, Northeast Brazil


PAGES: 150
BIG AREA: Ciências Exatas e da Terra
AREA: Geociências
SUBÁREA: Meteorologia
SPECIALTY: Climatologia
SUMMARY:

The study of the wind potential presents a significant importance in relation to the estimation of electricity production, helping companies that intend to invest in wind energy to choose the most appropriate location for the best use of this type of energy resource. In addition, the electrical energy from these large wind turbines can be used for the production of hydrogen. The study aims to use stochastic modeling to forecast wind speed, temperature and relative humidity with short and long term time series, in order to analyze the wind potential for five cities in the coastal region of the Northeast and estimate the production of hydrogen for the ammonia-based nitrogen fertilizer market. The wind speed data comes from five meteorological stations of the National Institute of Meteorology (INMET), which uses an anemometer with a height of 10 m. These stations are located in Northeastern Brazil: Fortaleza, Natal, São Luís, Recife, and Aracaju. These weather stations have periods of approximately 30 years, initially using monthly data. The analysis, modeling, simulation and forecasting with atmospheric variables are part of the study of energy potential, as it allows a better understanding of the nature of data variability, including seasonality, trend and randomness, and can detect small variations. For the analysis of climatological data, the complete time series without a time interval is often used. This implies that regardless of the size of the series, it is not appropriate to have missing data and it is necessary to use the data imputation technique. In this study, missing value imputation was evaluated by including a structural time series model. The wind speed, temperature, and relative humidity data from these coastal cities went through data analysis methods. These methods are composed of the data that will be modeled and the observed data. These will be compared with each other. In addition, the climatological data used in this study have undergone residual tests, accuracy tests, and forecasts and estimates for hydrogen production will be made from the hybrid forecast results. The project aims to contribute to the private wind energy sector, with the availability of a tool that helps investors in this area to choose the most appropriate location for the best use of this type of energy resource.


BANKING MEMBERS:
Presidente - 320597 - PAULO SERGIO LUCIO
Interna - 049.212.764-63 - DANIELE TÔRRES RODRIGUES - UFPI
Interno - 1164414 - WEBER ANDRADE GONCALVES
Externo à Instituição - JOÃO BOSCO VERÇOSA LEAL JUNIOR - UECE
Externo à Instituição - MARIA LUCIENE DIAS DE MELO - UFAL
Notícia cadastrada em: 14/02/2022 15:35
SIGAA | Superintendência de Tecnologia da Informação - (84) 3342 2210 | Copyright © 2006-2024 - UFRN - sigaa07-producao.info.ufrn.br.sigaa07-producao