Identificação de zonas de manejo com a experiência do produtor

• Ano de defesa: 2017
• Autor(a) principal: Schenatto, Kelyn
• Orientador(a): Souza, Eduardo Godoy de
• Banca de defesa: Pinheiro Neto, Raimundo , Gonçalves, Antonio Carlos Andrade , Maggi, Marcio Furlan , Mercante, Erivelto
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual do Oeste do Parana
• Programa de Pós-Graduação: Programa de Pós-Graduação "Stricto Sensu" em Engenharia Agrícola
• Departamento: Engenharia
• País: BR
• Palavras-chave em Português: Unidades de manejo; Agricultura de precisão; Agrupamento; Percepção do produtor
• Palavras-chave em Inglês: Management units; Precision agriculture; Clustering; Farmer feeling
• Área do conhecimento CNPq: CNPQ::CIENCIAS AGRARIAS::ENGENHARIA AGRICOLA
• Link de acesso: http://tede.unioeste.br:8080/tede/handle/tede/277

The process of generating management zones (MZs) comprises several steps, among them the choice of variables to be used in the grouping process, the choice of the best interpolator, the use of a data normalization method and the choice of an algorithm suitable for the grouping process, as well as adequate evaluation indices. In this work, it was attempted to improve techniques used in the MZs generation process. First, the influence of data normalization techniques in the clustering process was studied using the Fuzzy C-Means algorithm data proposed in the literature. It was also evaluated the use of the producer s experience as a new variable for the generation of the MZs and a piece of software was developed for the purpose of capturing this variable and managing agricultural data. The experiments were carried out with data obtained between the years of 2011 and 2016 from four commercial agricultural areas, located in the State of Paraná, where corn and soybean had been grown. With the experiments concluded, it was possible to assert that the technique of normalization of data influences the process of grouping when more than one variable is used with different scales of values and that the technique that presented better results was the normalization by amplitude. The proposed variable "farmer s experience" showed good results in the generation of the MZs and, with such, it was possible to generate management areas with different productive potential between the classes and thus, it can be used as a new variable to be used combined with the variables of stable soil conditions. Based on the positive results generated by the variable experience of the producer in three agricultural areas, a piece of software was developed that allows the insertion of this variable by the producer. The software called AgData-Mobile also allows the management of field data, allowing the producer to insert data of all the operations performed in each field and to keep a record of their areas. In addition this tool stands out for being free and for using clod storage