Sistema de monitoramento para industria agro utilizando IoT
Carregando...
Citações na Scopus
Tipo de produção
Trabalho de Conclusão de Curso
Data
2021-06-27
Autores
Pires, Carlos Rodrigues
Sena, Guilherme de Carvalho Souto
Faria, Leonardo Lopes
Souza, Paulo Henrique de Melo
Moreira, Vitor Heleno
Sena, Guilherme de Carvalho Souto
Faria, Leonardo Lopes
Souza, Paulo Henrique de Melo
Moreira, Vitor Heleno
Orientador
Melo, Marco Antonio Assis de
Periódico
Título da Revista
ISSN da Revista
Título de Volume
Citação
Texto completo (DOI)
Palavras-chave
agronegócio,hangraves-Samani,machine learning,agribusiness
Resumo
Devido ao seu grande poderio no agronegócio, o Brasil é reconhecido mundialmente não apenas por tal feito, mas também por ter o grande consumo de água como um de seus reveses. Setenta porcentode toda a água utilizada no país tem como destino a irrigação de lavouras, além da pecuária e agricultura (De acordo com o Fundo das Nações Unidas para Agricultura e Alimentação e a Agência Nacional de Águas).Este grande consumo de água do setor, deve-se em maior parte a desperdícios e problemas técnicos nas lavouras. Após a realização de um estudo, o grupo responsável por este trabalho observou que em modelos matemáticos utilizados, a evapotranspiração (evaporação de água através do solo e transpiração das plantas) é citada com frequência e influencia diretamente o consumo de água. Entretanto, informações relacionadas a evapotranspiração real são raras e difíceis de obter, pois demanda um alto custo e grande tempo de observação. Com o propósito de contornar o problema, modelos baseados em empirismos e leis físicas foram desenvolvidos. Neste documento, será utilizado o modelo de Hargreaves-Samani, relacionando a temperatura máxima, mínima e ângulo da radiação solar incidente. Foiutilizadoum banco de dados baseando-se na região de Mirante São Paulo, com o foco de estipular o comportamento gráfico utilizando machine learning e “ARIMA” (Método estatístico). O local escolhido é resultado de ser o local mais antigo do estado de São Paulo, tornando seu histórico de dados mais robusto. Por outro lado, os pontos utilizados pelo INMET (Instituto Nacional de Meteorologia) localizam-se em pontos distantes e alguns inclusive, por serem mais recentes, não possuem dados mais antigos.Este estudo propõe uma medição local no cultivo para obtenção de dados precisos da região e variáveis meteorológicas. Com estas informações em mãos, o grupo pode iniciar o desenvolvimento de um banco de dados próprio, permitindo uma previsão da evapotranspiração mais precisa e confiável. Para fim de protótipo,será utilizado o micro controlador ESP8266 com capacidade de conexão Wi-Fi, sensores térmicos precisos e uma plataforma analítica IoT, Thingspeak da Mathworks, permitindo agregar, visualizar e analisar fluxos de dados postados pelo dispositivo cadastrado.Desta forma, o agricultor conseguirá acompanhar seus dados de forma online em qualquer momento e lugar, podendo prever o consumo de água mês a mês.
Due to its great power in agribusiness, Brazil is recognized worldwide not only for this feat, but also for having the large consumption of water as one of its setbacks. 70% of all water used in the country is destined to irrigation of crops, in addition to livestock and agriculture (According to the United Nations Fund for Agriculture and Food and the National Water Agency).This large consumption of water in the sector is due in large part to waste and technical problems in crops. After conducting a study, the group responsible for this work observed that in the mathematical models used, evapotranspiration (evaporation of water through the soil and transpiration of plants) is frequently cited and directly influences water consumption. However, information related to actual evapotranspiration is rare and difficult to obtain, as it possesses a high cost and long observation time. In order to circumvent the problem, models based on empiricism and physical laws were developed. In this document, the Hargreaves-Samani model will be used, relating the maximum, minimum temperature and angle of the incident solar radiation. A database will be developed based on the region of Mirante São Paulo, with the focus on stipulating graphic behavior using machine learning and “ARIMA” (Statistical method).The location chosen is the result of being the oldest location in the state of São Paulo, making its data history more robust. On the other hand, the points used by INMET (National Institute of Meteorology) are located in distant points and some, being more modern, do not have older data.This study proposes a local measurement in the crop to obtain accurate measurements of the region and time. With this information in hand, the group can begin the development of its own database, allowing a more accurate and reliable forecast of evapotranspiration. The ESP8266 microcontroller with Wi-Fi connection capability, precise thermal sensors and an IoT analytical platform, Thingspeak by Mathworks, will be used, allowing the aggregation, visualization and analysis of data streams posted by the registered device. In this way, the farmer will be able to track his data online at any time and place, being able to forecast water consumption month by month.
Due to its great power in agribusiness, Brazil is recognized worldwide not only for this feat, but also for having the large consumption of water as one of its setbacks. 70% of all water used in the country is destined to irrigation of crops, in addition to livestock and agriculture (According to the United Nations Fund for Agriculture and Food and the National Water Agency).This large consumption of water in the sector is due in large part to waste and technical problems in crops. After conducting a study, the group responsible for this work observed that in the mathematical models used, evapotranspiration (evaporation of water through the soil and transpiration of plants) is frequently cited and directly influences water consumption. However, information related to actual evapotranspiration is rare and difficult to obtain, as it possesses a high cost and long observation time. In order to circumvent the problem, models based on empiricism and physical laws were developed. In this document, the Hargreaves-Samani model will be used, relating the maximum, minimum temperature and angle of the incident solar radiation. A database will be developed based on the region of Mirante São Paulo, with the focus on stipulating graphic behavior using machine learning and “ARIMA” (Statistical method).The location chosen is the result of being the oldest location in the state of São Paulo, making its data history more robust. On the other hand, the points used by INMET (National Institute of Meteorology) are located in distant points and some, being more modern, do not have older data.This study proposes a local measurement in the crop to obtain accurate measurements of the region and time. With this information in hand, the group can begin the development of its own database, allowing a more accurate and reliable forecast of evapotranspiration. The ESP8266 microcontroller with Wi-Fi connection capability, precise thermal sensors and an IoT analytical platform, Thingspeak by Mathworks, will be used, allowing the aggregation, visualization and analysis of data streams posted by the registered device. In this way, the farmer will be able to track his data online at any time and place, being able to forecast water consumption month by month.