A5002514177- Wind and Air Flow Studies
- Air Quality and Health Impacts
- Atmospheric chemistry and aerosols
- Meteorological Phenomena and Simulations
- Air Quality Monitoring and Forecasting
- Fractional Differential Equations Solutions
- Solar Radiation and Photovoltaics
- Differential Equations and Numerical Methods
- Geography and Environmental Studies
- Atmospheric aerosols and clouds
- Fluid Dynamics and Turbulent Flows
- Vehicle emissions and performance
- Energy Load and Power Forecasting
- Climate variability and models
- Complex Systems and Time Series Analysis
- Iterative Methods for Nonlinear Equations
- Environmental and biological studies
- Urban Heat Island Mitigation
- Hydrological Forecasting Using AI
- Wind Energy Research and Development
- Environmental Sustainability and Education
- Particle Dynamics in Fluid Flows
- Precipitation Measurement and Analysis
- Solar Thermal and Photovoltaic Systems
- Probabilistic and Robust Engineering Design
Serviço Nacional de Aprendizagem Industrial
2015-2024
Universidade Federal do Espírito Santo
2013-2024
Pontifícia Universidade Católica do Paraná
2024
University of Lisbon
2023
Mathematics Research Center
2015-2022
National Institute of Industrial Technology
2022
NSF National Center for Atmospheric Research
2021
Serviço Nacional de Aprendizagem Comercial
2016-2020
Unichristus
2020
Universidade do Estado da Bahia
2016
This work presents a novel transformer-based deep neural network architecture integrated with wavelet transform for forecasting wind speed and energy (power) generation the next 6 h ahead, using multiple meteorological variables as input multivariate time series forecasting. To evaluate performance of proposed model, different case studies were investigated, data collected from anemometers installed in three regions Bahia, Brazil. The model was compared an LSTM (Long Short Term Memory)...
In this paper, we use the conformable fractional derivative to discuss some linear differential equations with constant coefficients. By applying similar arguments theory of ordinary equations, establish a sufficient condition guarantee reliability solving coefficient by Laplace transform method. Finally, analytical solution for class models associated logistic model, von Foerster model and Bertalanffy is presented graphically various orders. The corresponding classical recovered as particular case.
In the present work, we investigate potential of fractional derivatives to model atmospheric dispersion pollutants. We propose simple differential equation models for steady state spatial distribution concentration a non-reactive pollutant in Planetary Boundary Layer. solve these and compare solutions with real experiment. found that derivative perform far better than traditional Gaussian even literature where it is considered diffusion coefficient function position order deal anomalous diffusion.
Short-term wind speed forecasting for Colonia Eulacio, Soriano Department, Uruguay, is performed by applying an artificial neural network (ANN) technique to the hourly time series representative of site. To train ANN and validate technique, data one year are collected tower, with anemometers installed at heights 101.8, 81.8, 25.7, 10.0 m. Different configurations applied each site height; then, a quantitative analysis conducted, statistical results evaluated select configuration that best...
Accounting for the current knowledge of convective boundary layer structure and characteristics, a new formulation eddy diffusivities to be used in atmospheric dispersion models has been derived. That is, expressions diffusivities, depending on source distance, inhomogeneous turbulence are proposed. The classical statistical diffusion theory, observed spectral properties, characteristics energy-containing eddies estimate these parameters. In addition, vertical diffusivity was introduced into...