23 results on '"Veloso, Gustavo Vieira"'
Search Results
2. Integrating proximal geophysical sensing and machine learning for digital soil mapping: Spatial prediction and model evaluation using a small dataset
- Author
-
Mello, Danilo César de, Veloso, Gustavo Vieira, Mello, Murilo Ferre de, Lana, Marcos Guedes de, Oliveira, Isabelle de Angeli, Mello, Fellipe Alcantara de Oliveira, Siqueira, Rafael Gomes, Gomes, Lucas Carvalho, Fernandes-Filho, Elpídio Inácio, Schaefer, Carlos Ernesto Gonçalves Reynaud, Francelino, Márcio Rocha, Leite, Emilson Pereira, Ferreira, Tiago Osório, and Demattê, José Alexandre Melo
- Published
- 2025
- Full Text
- View/download PDF
3. The Brazilian semiarid region over the past 21,000 years: Vegetation dynamics in small pulses of higher humidity
- Author
-
Silva, Rafael César, Corrêa, Guilherme Resende, Arruda, Daniel Meira, Veloso, Gustavo Vieira, Fernandes-Filho, Elpídio Inácio, Cândido, Hugo Galvão, de Oliveira, Fábio Soares, Rosa, Gabriel Palucci, Behling, Hermann, and Schaefer, Carlos Ernesto Gonçalves Renaud
- Published
- 2023
- Full Text
- View/download PDF
4. Chemical weathering detection in the periglacial landscapes of Maritime Antarctica: New approach using geophysical sensors, topographic variables and machine learning algorithms
- Author
-
de Mello, Danilo César, Veloso, Gustavo Vieira, Moquedace, Cassio Marques, de Angeli Oliveira, Isabelle, Francelino, Márcio Rocha, de Oliveira, Fabio Soares, de Souza, José João Lelis Leal, Gomes, Lucas Carvalho, Schaefer, Carlos Ernesto Gonçalves Reynaud, Fernandes-Filho, Elpídio Inácio, de Medeiros Júnior, Edgar Batista, and Demattê, José Alexandre Melo
- Published
- 2023
- Full Text
- View/download PDF
5. Modeling and mapping of Inselberg habitats for environmental conservation in the Atlantic Forest and Caatinga domains, Brazil
- Author
-
da Silveira, Vitor Alves, Veloso, Gustavo Vieira, de Paula, Hígor Brandão, dos Santos, Alexandre Rosa, Schaefer, Carlos Ernesto Gonçalves Reynaud, Fernandes-Filho, Elpídio Inácio, and Francelino, Marcio Rocha
- Published
- 2022
- Full Text
- View/download PDF
6. Machine learning models applied to TSS estimation in a reservoir using multispectral sensor onboard to RPA
- Author
-
Dias, Rafael Luís Silva, da Silva, Demetrius David, Fernandes-Filho, Elpídio Inácio, do Amaral, Cibele Hummel, dos Santos, Erli Pinto, Marques, Juliana Fazolo, and Veloso, Gustavo Vieira
- Published
- 2021
- Full Text
- View/download PDF
7. Effect of environmental covariable selection in the hydrological modeling using machine learning models to predict daily streamflow
- Author
-
Reis, Guilherme Barbosa, da Silva, Demetrius David, Fernandes Filho, Elpídio Inácio, Moreira, Michel Castro, Veloso, Gustavo Vieira, Fraga, Micael de Souza, and Pinheiro, Sávio Augusto Rocha
- Published
- 2021
- Full Text
- View/download PDF
8. Machine learning models for streamflow regionalization in a tropical watershed
- Author
-
Ferreira, Renan Gon, Silva, Demetrius David da, Elesbon, Abrahão Alexandre Alden, Fernandes-Filho, Elpídio Inácio, Veloso, Gustavo Vieira, Fraga, Micael de Souza, and Ferreira, Lucas Borges
- Published
- 2021
- Full Text
- View/download PDF
9. Relative Radiometric Normalization for the PlanetScope Nanosatellite Constellation Based on Sentinel-2 Images.
- Author
-
Dias, Rafael Luís Silva, Amorim, Ricardo Santos Silva, da Silva, Demetrius David, Fernandes-Filho, Elpídio Inácio, Veloso, Gustavo Vieira, and Macedo, Ronam Henrique Fonseca
- Subjects
MACHINE learning ,STANDARD deviations ,LATIN hypercube sampling ,RADIAL basis functions ,SURFACE of the earth ,DECISION trees - Abstract
Detecting and characterizing continuous changes on Earth's surface has become critical for planning and development. Since 2016, Planet Labs has launched hundreds of nanosatellites, known as Doves. Despite the advantages of their high spatial and temporal resolution, these nanosatellites' images still present inconsistencies in radiometric resolution, limiting their broader usability. To address this issue, a model for radiometric normalization of PlanetScope (PS) images was developed using Multispectral Instrument/Sentinel-2 (MSI/S2) sensor images as a reference. An extensive database was compiled, including images from all available versions of the PS sensor (e.g., PS2, PSB.SD, and PS2.SD) from 2017 to 2022, along with data from various weather stations. The sampling process was carried out for each band using two methods: Conditioned Latin Hypercube Sampling (cLHS) and statistical visualization. Five machine learning algorithms were then applied, incorporating both linear and nonlinear models based on rules and decision trees: Multiple Linear Regression (MLR), Model Averaged Neural Network (avNNet), Random Forest (RF), k-Nearest Neighbors (KKNN), and Support Vector Machine with Radial Basis Function (SVM-RBF). A rigorous covariate selection process was performed for model application, and the models' performance was evaluated using the following statistical indices: Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), Lin's Concordance Correlation Coefficient (CCC), and Coefficient of Determination (R
2 ). Additionally, Kruskal–Wallis and Dunn tests were applied during model selection to identify the best-performing model. The results indicated that the RF model provided the best fit across all PS sensor bands, with more accurate results in the longer wavelength bands (Band 3 and Band 4). The models achieved RMSE reflectance values of approximately 0.02 and 0.03 in these bands, with R2 and CCC ranging from 0.77 to 0.90 and 0.87 to 0.94, respectively. In summary, this study makes a significant contribution to optimizing the use of PS sensor images for various applications by offering a detailed and robust approach to radiometric normalization. These findings have important implications for the efficient monitoring of surface changes on Earth, potentially enhancing the practical and scientific use of these datasets. [ABSTRACT FROM AUTHOR]- Published
- 2024
- Full Text
- View/download PDF
10. Spatial Modeling of Groundwater Potential in the North of Minas Gerais, Brazil: An Integrated Approach Using Machine Learning and Environmental Data
- Author
-
Souza, Cristiano Marcelo Pereira, primary, Silva, Lucas Augusto Pereira, additional, Leite, Marcos Esdras, additional, Veloso, Gustavo Vieira, additional, and Fernandes-Filho, Elpídio Inácio, additional
- Published
- 2023
- Full Text
- View/download PDF
11. Application of machine learning to proximal gamma-ray and magnetic susceptibility surveys in the Maritime Antarctic: assessing the influence of periglacial processes and landforms.
- Author
-
Mello, Danilo César de, Baldi, Clara Glória Oliveira, Moquedace, Cássio Marques, Oliveira, Isabelle de Angeli, Veloso, Gustavo Vieira, Gomes, Lucas Carvalho, Francelino, Márcio Rocha, Schaefer, Carlos Ernesto Gonçalves Reynaud, Fernandes-Filho, Elpídio Inácio, Júnior, Edgar Batista de Medeiros, Oliveira, Fabio Soares de, Souza, José João Lelis Leal de Souza, Ferreira, Tiago, and Demattê, José A. M.
- Subjects
PERIGLACIAL processes ,HYDROGRAPHIC surveying ,MAGNETIC susceptibility ,SOLIFLUCTION ,RANDOM forest algorithms ,NATURAL radioactivity ,POTASSIUM channels ,MACHINE learning ,POTASSIUM - Abstract
Maritime Antarctica (M.A.) contains the most extensive and diverse lithological exposure compared to the entire continent. This lithological substrate reveals a rich history encompassing lithological, pedogeomorphological, and glaciological aspects of M.A., all influenced by periglacial processes. Although geophysical surveys can detect and provide valuable information to understand Antarctic lithologies and their history, such surveys are scarce on this continent and, in practice, almost non-existent. In this sense, we conducted a pioneering and comprehensive gamma-spectrometric (natural radioactivity) and magnetic susceptibility (κ) survey on various igneous rocks. The main objective was to create ternary gamma-ray and κ maps using machine learning algorithms, terrain attributes, and a nested-leave-one-out cross-validation method. Additionally, we investigated the relationship between the distribution of natural radioactivity and κ to gain insights into pedogeomorphological and periglacial processes and dynamics. For that, we used proximal gamma-spectrometric and κ data in different lithological substrates associated to terrain attributes. The geophysical variables were collected in the field from various lithological substrates, by use field portable equipment. The geophysical variables were collected in the field from various lithological substrates using portable equipment. These variables, combined with relief data and lithology, served as input data for modeling to predict and spatially map the content of radionuclides and κ by random forest algorithm (RF). In addition, we use nested-LOOCV as a form of external validation in a geophysical data with a small number of samples, and the error maps as evaluation of results. The RF algorithm successfully generated detailed maps of gamma-spectrometric and κ variables. The distribution of radionuclides and ferrimagnetic minerals was influenced by morphometric variables. Nested-LOOCV method evaluated algorithm performance accurately with limited samples, generating robust mean maps. The highest thorium levels were observed in elevated, flat, and west beach areas, where detrital materials from periglacial erosion came through fluvioglacial channels. Lithology and pedogeomorphological processes-controlled thorium contents. Steeper areas formed a ring with the highest uranium contents, influenced by lithology and geomorphological-periglacial processes (rock cryoclasty, periglacial erosion, and heterogeneous deposition). Felsic rocks and areas less affected by periglacial erosion had the highest potassium levels, while regions with sulfurization-affected pyritized-andesites near fluvioglacial channels showed the lowest potassium contents. Lithology and pedogeochemical processes governed potassium levels. The κ values showed no distinct distribution pattern. Pyritized-andesite areas had the highest levels due to sulfurization and associated pyrrhotite, promoting iron release. Conversely, Cryosol areas, experiencing freezing and thawing activity, had the lowest κ values due to limited ferrimagnetic mineral formation. Lithology and pedological-periglacial processes in Cryosols played a significant role in controlling κ values. In regions characterized by diverse terrain attributes and abundant active and intense periglacial processes, the spatial distribution of geophysical variables does not reliably reflect the actual lithological composition of the substrate. The complex interplay of various periglacial processes in the area, along with the morphometric features of the landscape, leads to the redistribution, mixing, and homogenization of surface materials, contributing to the inaccuracies in the predicted-spatialized geophysical variables. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
12. Soil Erosion Satellite-Based Estimation in Cropland for Soil Conservation
- Author
-
Gallo, Bruna Cristina, Magalhães, Paulo Sérgio Graziano, Demattê, José A.M., Cervi, Walter Rossi, Carvalho, João Luís Nunes, Barbosa, Leandro Carneiro, Bellinaso, Henrique, César de Mello, Danilo, Veloso, Gustavo Vieira, Alves, Marcelo Rodrigo, Fernandes-Filho, Elpídio Inácio, Francelino, Márcio Rocha, Schaefer, Carlos Ernesto Gonçalves Reynaud, Gallo, Bruna Cristina, Magalhães, Paulo Sérgio Graziano, Demattê, José A.M., Cervi, Walter Rossi, Carvalho, João Luís Nunes, Barbosa, Leandro Carneiro, Bellinaso, Henrique, César de Mello, Danilo, Veloso, Gustavo Vieira, Alves, Marcelo Rodrigo, Fernandes-Filho, Elpídio Inácio, Francelino, Márcio Rocha, and Schaefer, Carlos Ernesto Gonçalves Reynaud
- Abstract
Intensive cropland expansion for an increasing population has driven soil degradation worldwide. Modeling how agroecosystems respond to variations in soil attributes, relief and crop management dynamics can guide soil conservation. This research presents a new approach to evaluate soil loss by water erosion in cropland using the RUSLE model and Synthetic Soil Image (spectroscopy technique), which uses time series remotely sensed environmental, agricultural and anthropic variables, in the southeast region of São Paulo State, Brazil. The availability of the open-access satellite images of Tropical Rainfall Measuring Mission (TRMM) and Landsat satellite images provided ten years of rainfall data and 35 years of exposed soil surface. The bare soil surface and agricultural land use were extracted, and the multi-temporal rainfall erosivity was assessed. We predict soil maps’ attributes (texture and organic matter) through innovative soil spectroscopy techniques to assess the soil erodibility and soil loss tolerance. The erosivity, erodibility, and topography obtained by the Earth observations were adopted to estimate soil erosion in four scenarios of sugarcane (Saccharum spp.) residue coverage (0%, 50%, 75%, and 100%) in five years of the sugarcane cycle: the first year of sugarcane harvest and four subsequent harvesting years from 2013 to 2017. Soil loss tolerance means 4.3 Mg ha−1 exceeds the minimum rate in 40% of the region, resulting in a total soil loss of ~6 million Mg yr−1 under total coverage management (7 Mg ha−1). Our findings suggest that sugarcane straw production has not been sufficient to protect the soil loss against water erosion. Thus, straw removal is unfeasible unless alternative conservation practices are adopted, such as minimum soil tillage, contour lines, terracing and other techniques that favor increases in organic matter content and soil flocculating cations. This research also identifies a spatiotemporal erosion-prone area that requests an imme
- Published
- 2023
13. Soil Erosion Satellite-Based Estimation in Cropland for Soil Conservation
- Author
-
Gallo, Bruna Cristina, primary, Magalhães, Paulo Sérgio Graziano, additional, Demattê, José A. M., additional, Cervi, Walter Rossi, additional, Carvalho, João Luís Nunes, additional, Barbosa, Leandro Carneiro, additional, Bellinaso, Henrique, additional, Mello, Danilo César de, additional, Veloso, Gustavo Vieira, additional, Alves, Marcelo Rodrigo, additional, Fernandes-Filho, Elpídio Inácio, additional, Francelino, Márcio Rocha, additional, and Schaefer, Carlos Ernesto Gonçalves Reynaud, additional
- Published
- 2022
- Full Text
- View/download PDF
14. Characterizing and Modeling Tropical Sandy Soils through VisNIR-SWIR, MIR Spectroscopy, and X-ray Fluorescence
- Author
-
Di Raimo, Luis Augusto Di Loreto, primary, Couto, Eduardo Guimarães, additional, de Mello, Danilo Cesar, additional, Demattê, José Alexandre Mello, additional, Amorim, Ricardo Santos Silva, additional, Torres, Gilmar Nunes, additional, Bocuti, Edwaldo Dias, additional, Veloso, Gustavo Vieira, additional, Poppiel, Raul Roberto, additional, Francelino, Márcio Rocha, additional, and Fernandes-Filho, Elpídio Inácio, additional
- Published
- 2022
- Full Text
- View/download PDF
15. Maps of soil organic carbon stocks in Brazil
- Author
-
Gomes, Lucas Carvalho, Faria, Raiza Moniz, de Souza, Eliana, Veloso, Gustavo Vieira, Schaefer, Carlos Ernesto G.R., Filho, Elpídio Inácio Fernandes, Gomes, Lucas Carvalho, Faria, Raiza Moniz, de Souza, Eliana, Veloso, Gustavo Vieira, Schaefer, Carlos Ernesto G.R., and Filho, Elpídio Inácio Fernandes
- Abstract
This database was created by Gustavo Vieira Veloso and Lucas Carvalho Gomes 04/06/2022. Contact: gustavo.v.veloso@gmail.com and lucascarvalhogomes15@hotmail.com Maps of soil organic carbon (SOC) stocks in Brazil of the article: "Modeling and mapping soil organic carbon stocks in Brazil" (doi: 10.1016/j.geoderma.2019.01.007) The dataset is composed of five folders of SOC stocks maps at the standard depths (0–5, 5–15, 15–30, 30–60, and 60–100 cm). The maps are in Geotif format (EPSG 102015) with a spatial resolution of approximately 1 km and include the mean SOC stocks, standard deviation (SD), coefficient of variation (CV), 0.05 and 0.95 quantiles. The maps are free to use and please cite also the article: Gomes, L.C., Faria, R.M., de Souza, E., Veloso, G.V., Schaefer, C.E.G., & Fernandes Filho, E.I. (2019). Modeling and mapping soil organic carbon stocks in Brazil. Geoderma, 340, 337-350., This database was created by Gustavo Vieira Veloso and Lucas Carvalho Gomes 04/06/2022. Contact: gustavo.v.veloso@gmail.com and lucascarvalhogomes15@hotmail.com ------------------------------------------------------------------------------------ Maps of soil organic carbon (SOC) stocks in Brazil of the article: "Modeling and mapping soil organic carbon stocks in Brazil" (doi: 10.1016/j.geoderma.2019.01.007) The dataset is composed of five folders of SOC stocks maps at the standard depths (0–5, 5–15, 15–30, 30–60, and 60–100 cm). The maps are in Geotif format (EPSG 102015) with a spatial resolution of approximately 1 km and include the mean SOC stocks, standard deviation (SD), coefficient of variation (CV), 0.05 and 0.95 quantiles. The maps are free to use and please cite also the article: Gomes, L.C., Faria, R.M., de Souza, E., Veloso, G.V., Schaefer, C.E.G., & Fernandes Filho, E.I. (2019). Modeling and mapping soil organic carbon stocks in Brazil. Geoderma, 340, 337-350.
- Published
- 2022
16. A new methodological framework for geophysical sensor combinations associated with machine learning algorithms to understand soil attributes
- Author
-
Mello, Danilo César de, primary, Veloso, Gustavo Vieira, additional, Lana, Marcos Guedes de, additional, Mello, Fellipe Alcantara de Oliveira, additional, Poppiel, Raul Roberto, additional, Cabrero, Diego Ribeiro Oquendo, additional, Di Raimo, Luis Augusto Di Loreto, additional, Schaefer, Carlos Ernesto Gonçalves Reynaud, additional, Filho, Elpídio Inácio Fernandes, additional, Leite, Emilson Pereira, additional, and Demattê, José Alexandre Melo, additional
- Published
- 2022
- Full Text
- View/download PDF
17. MULTICRITERIA ANALYSIS AND MACHINE LEARNING ALGORITHM FOR DEFINITION OF AREAS FOR MICRO-DAM, SOUTHEASTERN BRAZIL
- Author
-
Souza, Cristiano Marcelo Pereira, primary, Veloso, Gustavo Vieira, additional, Faria, André Luiz Lopes, additional, Leite, Marcos Esdras, additional, and Fernandes Filho, Elpídio Inácio, additional
- Published
- 2021
- Full Text
- View/download PDF
18. MODELAGEM ESPAÇO-TEMPORAL DE ILHAS DE CALOR URBANA E CORRELAÇÃO COM VARIÁVEIS- MONTES CLAROS, MG, BRASIL
- Author
-
Souza, Cristiano Marcelo Pereira, primary, Filho, Raul Magalhães, additional, Passos, Renata Silva, additional, Leite, Marcos Esdras, additional, Veloso, Gustavo Vieira, additional, and França, Iara Soares, additional
- Published
- 2020
- Full Text
- View/download PDF
19. A new methodological framework by geophysical sensors combinations associated with machine learning algorithms to understand soil attributes.
- Author
-
de Mello, Danilo César, Veloso, Gustavo Vieira, de Lana, Marcos Guedes, de Oliveira Mello, Fellipe Alcantara, Poppiel, Raul Roberto, Cabrero, Diego Ribeiro Oquendo, Di Raimo, Luis Augusto Di Loreto, Schaefer, Carlos Ernesto Gonçalves Reynaud, Filho, Elpídio Inácio Fernandes, Leite, Emilson Pereira, and Demattê, José Alexandre Melo
- Subjects
- *
MACHINE learning , *STANDARD deviations , *ELECTRIC conductivity , *SOILS , *SOIL testing - Abstract
Geophysical sensors combined with machine learning algorithms have been used to understand the pedosphere system, landscape processes and to model soil attributes. In this research, we used parent material, terrain attributes and data from geophysical sensors in different combinations, to test and compare different and novel machine learning algorithms to model soil attributes. Also, we analyzed the importance of pedoenvironmental variables in predictive models. For that, we collected soil physico-chemical and geophysical data (gamma-ray emission from uranium, thorium and potassium, magnetic susceptibility and apparent electric conductivity) by three sensors, gamma-ray spectrometer - RS 230, susceptibilimeter KT10 - Terraplus and Conductivimeter - EM38 Geonics) at 75 points and, we performed soil analysis afterwards. The results showed varying models with the best performance (R2 > 0.2) for clay, sand, Fe2O3, TiO2, SiO2 and Cation Exchange Capacity prediction. Modeling with selection of covariates at three phases (variance close to zero, removal by correction and removal by importance), demonstrated to be adequate to increase the parsimony. The prediction of soil attributes by machine learning algorithms demonstrated adequate values for field collected data, without any sample preparation, for most of the tested predictors (R2 ranging from 0.20 to 0.50). Also, the use of four regression algorithms proved important, since at least one of the predictors used one of the tested algorithms. The performances of the best algorithms for each predictor were higher than the use of a mean value for the entire area comparing the values of Root Mean Square Error (RMSE) and Mean Absolute Error (MAE). The best combination of sensors that reached the best model performance to predict soil attributes were gamma-ray spectrometer and susceptibilimeter. The most important variables were parent material, digital elevation model, standardized height and magnetic susceptibility for most predictions. We concluded that soil attributes can be efficiently modelled by geophysical data using machine learning techniques and geophysical sensors combinations. The technique can bring light for future soil mapping with gain of time and environment friendly. [ABSTRACT FROM AUTHOR]
- Published
- 2021
- Full Text
- View/download PDF
20. Space distribution of metals and metaloids in the soils of the state of minas gerais using geostatistics and machine learning methods
- Author
-
Veloso, Gustavo Vieira, Santos, Nerilson Terra, Abrahão, Walter Antônio Pereira, and Mello, Jaime Wilson Vargas de
- Subjects
Solos - Análise ,Ciência do Solo ,Química do solo ,Solos - Composição ,Solos - Metais pesados - Minas Gerais - Abstract
O Estado de Minas Gerais tem uma área de 588.384 km² que corresponde a um percentual de 6,9% do território brasileiro e de 63,5% da região sudeste do Brasil. Minas Gerais apresenta uma grande diversidade geológica. Essa variedade geológica sugere que os solos derivados desses materiais possuem elementos químicos nos mais diferentes teores no Estado. Nesse sentido, o objetivo desta pesquisa foi gerar mapas dos teores de metais e metaloides na camada superficial dos solos do Estado de Minas Gerais, utilizando-se de técnicas de aprendizado de máquina e krigagem Ordinária e comparar os resultados dos métodos. Os mapas de teores de 13 elementos traço (Al, As, B, Ba, Cd, Co, Cr, Cu, Fe, Hg, Mn, Mo, Ni, Pb, Sr, V e Zn) foram gerados a partir da Krigagem Ordinária desse elementos e de 5 outros constituintes do solo. Foi utilizado neste estudo um total de 648 amostras de solos georreferenciadas obtidas do „Banco de Solos‟ do Estado de Minas Gerais. A seleção das variáveis para a predição do teor de metais e metaloides seguiu procedimentos que visaram identificar as covariáveis com maior relevância, sendo usados métodos correlação linear e não linear. Para identificar a dependência espacial dos dados, foram realizados os seguintes testes de modelos teóricos de semivariograma: esférico, exponencial, estável e gaussiano. Seis algoritmos de predição dos teores foram usados neste estudo para o aprendizado de máquina, sendo estes: Cubist, SVMRadialSigma, Random Forest, ExtraTree, Ranger e Rborist. A verificação do desempenho desses algoritmos foi executada pela validação cruzada. Os resultados da seleção de covariáveis para o aprendizado de máquinas mostram coeficiente de determinação do algoritmo de treinamento (R 2 ) utilizado na seleção de covariáveis variaram de 0,03 a 0,43. As covariáveis categóricas litologia, geomorfologia e classe de solo apresentaram a maior importância na predição de 17 elementos. Covariáveis bioclimáticas foram importantes na predição dos teores de 16 elementos. Enquanto que, As covariáveis relacionadas à gamaespectrometria auxiliaram na predição dos teores de dez dos elementos analisados. Por sua vez, a magnetometria apresentou baixa importância para a predição do teor da maioria dos elementos nos solos, sendo importante somente para predição do Sr. Em geral, a transformação dos dados não promoveu um aumento na capacidade de predição dos elementos pelos modelos de aprendizado de máquinas, exceto para três elementos (Al, Sr e Zn). Na Krigagem Ordinária houver aumento do número de elementos com boa capacidade preditiva em alguns dos métodos de krigagem (Al, As, Co, Hg, Mn, Sb, Se, Sr e Zn. OS resultados de nRMSE variam entre 10% e 20%, obtendo resultados considerados bons, exceto por B e Mo. Os melhores resultados de nRMSE foram obtidos na predição com os dados sem transformações. A transformação dos dados em log1p gerou melhores resultados de treinamento apenas para Al e Zn. Para os elementos com os melhores resultados no método de aprendizado de máquinas foram: Ranger, SVMRadialSigma, RandomFlorest, Extratree. Os algoritmos Cubist e Rborist não alcançaram desempenho satisfatório na predição para nenhum elemento analisado. Os mapas espacializados pelos métodos de krigagem e aprendizado de máquinas apresentaram características similaridades para os elementos avaliados, com estimativas dos teores próximos. Os mapas especializados mostram maiores teores de As, Cd, Cr e Ni estão localizados na região do Quadrilátero Ferrífero, enquanto os elementos Cu, Fe, Mn, V e Zn apresentaram os maiores teores na região do triângulo mineiro. O elemento Co apresentou teores altos nas duas regiões, Quadrilátero Ferrífero e Triângulo mineiro. Os resultados mostraram que 11 elementos demostraram melhores pelos algoritmos de aprendizado de máquinas (Al, As, B, Co, Fe, Hg, Mn, Ni, Pb, Se e Zn). A espacialização dos teores dos metais e metaloides pelos algoritmos de aprendizado de máquinas, gerou mapas de 11 elementos com desempenhos superiores à krigagem. Os mapas gerados pelo método de espacialização por krigagem apresentam teores máximos superiores aos apresentados todos os elementos tirando o Pb. Os métodos de aprendizado de máquinas geram mapas com maior nível de detalhamento quando comparados com mapas gerados pela krigagem. The State of Minas Gerais has an area of 588,384 km² which corresponds to a percentage of 6.9% of Brazilian territory and 63.5% of the southeastern region of Brazil. Minas Gerais presents a wide lithological diversity, which suggests that the soils derived from these materials have large variation of chemical elements content in this State. In this sense, the goal of this research was to generate prediction maps of metals and metalloids contents in the surface layer of the soils of the State of Minas Gerais, using techniques of machine learning and Ordinary kriging and compare the results of these methods. The maps of 13 trace elements (Al, As, B, Ba, Cd, Co, Cr, Cu, Fe, Hg, Mn, Mo, Ni, Pb, Sr, V, Zn) was generated from Ordinary kriging of these elements and other soil constituents. A total of 648 geo-referenced soil samples obtained from the „Soil data bank‟ of the State of Minas Gerais, were used in this study. The selection of variables for the prediction of metal and metalloid contents followed procedures that aimed to identify covariates with great relevance, using linear and nonlinear correlation methods for removal of collinearity. After confirmation of data spatial dependence, the following tests of theoretical semivariogram models, such as: spherical, exponential, stable and Gaussian were performed. Six predicting algorithms of machine learningwere used in this study: Cubist, SVMRadialSigma, Random Forest, ExtraTree, Ranger and Rborist. The algorithms performance was verified by cross validation. The results of the covariates selection by machine learning showed coefficient of the training algorithm (R2) used in the selection of covariates ranged from 0.03 to 0.43. The categorical covariates lithology, geomorphology and soil class presented the highest importance in the prediction of 17 elements. Bioclimatic covariates were important in predicting the contents of 16 elements. Meanwhile, covariates related to gamma spectrometry helped to predict the contents of ten elements analyzed. In addition, magnetometry was low importance for the content prediction of most elements in soils, being important only for the prediction of Sr. In general, the transformation of the data did not promote an increase in the prediction capacity of the elements at using machines learning , except for three elements (Al, Sr, Zn). With Ordinary kriging there was an increase in the number of elements with “good” predictive capacity in some of the kriging methods (Al, As, Co, Hg, Mn, Sb, Se, Sr and Zn). In general, the best results (nRMSE) range from 10% to 20%. These results are considered good, except for B and Mo. The best results of nRMSE were obtained for data, without transformations. The best results using machine learning method were obtained with the algorithms: Ranger, SVMRadialSigma, Random Forest and Extratree. The algorithms Cubist and Rborist did not reach satisfactory performance in the prediction for any element analyzed. The spatialized maps by kriging and machine learning methods presented similar characteristics for the evaluated elements, with estimates of the next levels. The Cu, Fe, Mn, V and Zn elements presented the highest levels in the “Triângulo Mineiro” region. The element Co, presented high levels in the two regions, “Quadrilátero Ferrífero” and “Triângulo Mineiro”. The spatialization of metal and metalloid contents by machine learning algorithms generated 11 elements with higher performance than kriging. The maps generated by the kriging spatialisation method have higher maximum levels than all the elements showing Pb. Machine learning methods generate maps with a greater level of detail when compared to maps generated by kriging.
- Published
- 2017
21. Multivariate Analysis and Machine Learning in Properties of Ultisols (Argissolos) of Brazilian Amazon
- Author
-
Souza, Cristiano Marcelo Pereira de, primary, Thomazini, André, additional, Schaefer, Carlos Ernesto Gonçalves Reynaud, additional, Veloso, Gustavo Vieira, additional, Moreira, Guilherme Musse, additional, and Fernandes Filho, Elpídio Inácio, additional
- Published
- 2018
- Full Text
- View/download PDF
22. Reference Values of Soil Quality for the Rio Doce Basin
- Author
-
Guevara, Yang Zumbo Coronel, primary, Souza, José João Lelis Leal de, additional, Veloso, Gustavo Vieira, additional, Veloso, Renato Welmer, additional, Rocha, Pablo Azevedo, additional, Abrahão, Walter Antônio Pereira, additional, and Fernandes Filho, Elpídio Inácio, additional
- Published
- 2018
- Full Text
- View/download PDF
23. Automation system direction of a coffee harvester
- Author
-
Veloso, Gustavo Vieira, Fernandes, Haroldo Carlos, Valente, Domingos Sárvio Magalhães, Teixeira, Mauri Martins, and Monteiro, Paulo Marcos de Barros
- Subjects
Harvester ,Colhedora ,Café ,CIENCIAS AGRARIAS::ENGENHARIA AGRICOLA [CNPQ] ,Sistema de direção ,System direction ,Coffee - Abstract
Conselho Nacional de Desenvolvimento Científico e Tecnológico Coffee is one of the most important products exported by Brazil, it is responsible for many employments, and keeps a positive balance of trade, as well. The harvest is the most complex and expensive operation on grain production process, principally in relation to the labor. As a result, it is increasing the necessity of mechanizing.There are three types of harvest: manual, semi-manual and mechanical. The mechanical coffee harvesting decreases the labor necessary on the process, which causes reduction of production costs. However, there is not fully executable solution for mechanical harvesting in mountainous region. For this reason, the costs related to the manual harvesting are transferred to the consumers in these regions. In order to reduce the harvesting costs, many researches have been developed to adjust coffee harvesters to this type of relief. The biggest obstacles are the risk of coffee harvester tipping and the difficult to maneuver the machine.Therefore, it is necessary to build a machinery system adapted to the terrain slope and it will demand an innovative steering wheel system. The concept of automatic vehicle guidance has already happened in other types of monocrops, using electronic components associated to the agricultural machinery. All machine control systems are connected to the central command, which applies algorithms stored in the memory, to perform pre-established tasks by the operator. For example, functions of the machine such as guidance and control on the harvesting area. For this reason, there are PLC systems (Programmable Logic Controller) that use microcontrollers to control the guidance systems of agricultural machinery. Among the microcontrollers systems, the Arduino platform is highlighted due to its low cost of components and simplicity of programming language. The main objective of this work was to develop two control systems to the coffee harvester guidance, based on PCL system and Arduino platform. Both systems were evaluated and compared in laboratory. The systems were assembled in the coffee harvester and potentiometric sensors were used to control the steering wheel. The steering wheel angle run from algorithm was the criterion used to evaluate the developed system. This study considered two movements of the machine. The first was realized with tires in parallel while the second was with tires in curvilinear motion. It was considered twelve angles of execution to the parallel system motion and to the curvilinear motion was considered eight angles of execution. The entire random delineation was applied with four replications to the parallel motion and five to the curvilinear motion. Both steering wheel systems were compared through their results in each tire and motion (parallel and curvilinear). The steering wheel systems of coffee harvesters were developed and the results showed great accuracy for both systems, as much to the parallel motion as to all tires. The systems obtained great precision and were practically equals statistically on the curvilinear motion. However, the PLC system was more accurate than the Arduino system. O café é um dos principais itens de exportação do Brasil, como um dos líderes na balança comercial, e responsável por gerar um grande número de empregos. A colheita é a operação mais complexa e dispendiosa dentro do processo produtivo deste grão, principalmente em relação à mão de obra. Assim, cada vez mais, se faz necessário mecanizar esse processo. A colheita pode ser realizada de três formas: a manual, a semi- mecanizada e a mecanizada. A mecanização da coleta do café promove a diminuição da mão-de-obra necessária, que incorre na redução dos custos de produção. Entretanto, não há uma solução plenamente executável para a mecanização da colheita do café em regiões montanhosas. Assim sendo, os altos custos associados com a colheita manual, nestas regiões, são repassados ao consumidor. Para reduzir os custos da colheita, pesquisas estão sendo realizadas para o desenvolvimento de colhedoras adaptadas para este tipo de relevo. O grande empecilho está relacionado à manobrabilidade e ao risco de tombamento da colhedora. Assim, há a necessidade da construção de um sistema de máquina adaptado à declividade do terreno, e isso irá demandar o desenvolvimento de um sistema de direção das rodas desta máquina em questão que seja inovador. Em outras monoculturas, o conceito de orientação automática já vem sendo realizado, com o uso de componentes eletrônicos associados com o maquinário agrícola. Todos os sistemas de controle da máquina são ligados uma central de comando. Por sua vez, a central utiliza algoritmos armazenados em sua memória, para executar funções preestabelecidas pelo operador. Como, por exemplo, a execução as funções de direção e controle da máquina na área de colheita. Para isso existem sistemas CLP (controlador lógico programável) que empregam microcontroladores responsáveis pelos controles dos sistemas de direção de máquinas agrícolas. Entre os sistemas microcontrolados, tem-se a plataforma Arduino, que vem ganhando força devido ao seu baixo custo de aquisição dos componentes e a simplicidade de programação. Desse modo, o objetivo desse trabalho foi desenvolver dois sistemas de controle da direção de uma colhedora de café para regiões montanhosas, com base em CLP e Arduino. Os dois sistemas de controle da colhedora foram avaliados e comparados em laboratório. Os sistemas foram montados na colhedora de café, e foram utilizados sensores potenciométricos para controlar a direção das rodas. O ângulo de esterçamento das rodas executado a partir de do algoritmo foi o critério utilizado para avaliação dos sistemas desenvolvidos. Durante o estudo foram considerados dois tipos de movimentos da máquina. O primeiro foi realizado com as rodas em paralelo, enquanto o segundo com as rodas em movimento curvilíneo. Para o sistema de movimentação em paralelo, foram considerados doze (12) ângulos de execução e para o sistema em movimento curvilíneo, foram considerados oito (8) ângulos de execução. O delineamento inteiramente casualizado foi aplicado na realização desse estudo, com quatro repetições para a movimentação em paralelo e cinco repetições para o curvilíneo. Para os dois sistemas foram feitas validações, sendo comparados os resultados em cada roda e em cada movimento (paralelo e curvilíneo). Os resultados mostraram uma grande exatidão para os dois sistemas desenvolvidos, quanto ao movimento em paralelo, para todas as rodas. Sendo que, o sistema CLP foi mais preciso que o sistema Arduino. Para o movimento curvilíneo, os dois sistemas apresentaram grande exatidão, praticamente iguais estatisticamente. Os resultados nos movimento em paralelo mostram que o sistema CLP e mais preciso que o sistema Arduino, sendo mais recomendado para aplicação na colhedora. Para o movimento em curvilíneo não ouve diferença estatística, assim poderá ser usado qual um dos dois sistemas.
- Published
- 2013
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.