Improve irrigation timing decision for agriculture using real time data and machine learning

• Autor(a) principal: Cardoso, J.
• Data de Publicação: 2020
• Outros Autores: Glória, A., Sebastião, P.
• Idioma: eng
• Título da fonte: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Texto Completo: http://hdl.handle.net/10071/23417
• topic: Machine learning; Neural network; Decision tree; Support vector machine; XGBoost; Random forest; Sustainability; Smart irrigation

With the constant evolution of technology and the constant appearance of new solutions that, when combined, manage to achieve sustainability, the exploration of these systems is increasingly a path to take. This paper presents a study of machine learning algorithms with the objective of predicting the most suitable time of day for water administration to an agricultural field. With the use of a high amount of data previously collected through a Wireless Sensors Network (WSN) spread in an agricultural field it becomes possible to explore technologies that allow to predict the best time for water management in order to eliminate the scheduled irrigation that often leads to the waste of water being the main objective of the system to save this same natural resource.