Uma solução IoT voltada à identificação de equipamentos elétricos com o uso da Computação de Borda

• Autor(a) principal: Amurim, Antonio David Tavares
• Data de Publicação: 2022
• Tipo de documento: Trabalho de conclusão de curso
• Idioma: por
• Título da fonte: Repositório Institucional da Universidade Federal do Ceará (UFC)
• Texto Completo: http://www.repositorio.ufc.br/handle/riufc/65340
• topic: Internet das Coisas; Aprendizado do computador; Fiware; Computação em Nuvem

Poor management of energy resources has become a major problem in urban centers, as it harms not only the environment, but also financial resources, worsening in large constructions. More specifically, the lack of management of which electrical appliances are in operation can lead to unnecessary electrical energy expenditures. In addition, the uncontrolled use of some devices can reduce their useful life by causing even more financial losses. However, the Internet of Things (IoT) combined with Artificial Intelligence brings the idea of intelligent environments, which insert computing into the physical world in order to monitor physical variables, automate everyday activities and make predictions. In this work, an IoT solution is proposed focusing on the identification of electrical appliances by monitoring the electrical current consumption signal. In this way, the construction procedures of the different parts of the project are treated, such as the development of the IoT device responsible for collecting the electrical signal, the implementation of Cloud services using Edge Computing using the FIWARE platform, the construction of the set of data with the training and evaluation of classification models based on Machine Learning, as well as the development of the dashboard to visualize the monitoring data. In addition, tests were carried out and promising results were presented, indicating that the solution has a quality monitoring with a hit rate ranging between 90% and 100% in the identification of most of the monitored electrical equipment.