Pinheiro, Ricardo FerreiraGonçalves, Flávio Henriques2024-11-072024-11-072024-08-23GONÇALVES, Flávio Henriques. Desenvolvimento da plataforma hórus para previsão de falha estudo de caso: conjunto de motores de "pitch" em aerogerador. Orientador: Dr. Ricardo Ferreira Pinheiro. 2024. 105f. Dissertação (Mestrado Profissional em Energia Elétrica) - Centro de Tecnologia, Universidade Federal do Rio Grande do Norte, Natal, 2024.https://repositorio.ufrn.br/handle/123456789/60573Efficient management of wind farms is essential to ensure sustainability and maximize the production of renewable energy. One of the main challenges is preventive maintenance and predicting failures in critical components, such as the bearing gearbox motor assembly of the blade angle in wind turbines, commonly referred to as "pitch motors". Predictive maintenance is an emerging field promising greater operational efficiency and cost reduction. In this context, the Hórus platform, which is the product of this work, emerges as a viable alternative aiming to contribute to the management of wind turbine health. The Hórus platform was developed to predict failures in wind turbine pitch motors, with a flexible architecture that allows the inclusion of other equipment and predictive maintenance techniques. The choice of the name Hórus, inspired by the Egyptian god of vision, reflects the essence of the platform: to offer a clear and early view of maintenance needs, avoiding unexpected interruptions and maximizing operational efficiency. The technological structure of Hórus combines data analysis and visualization tools, such as Power BI for processing large volumes of data and Python for complex algorithms. The visual interface is developed with HTML and CSS, providing a friendly and interactive experience. Data collection by Hórus involves multiple points, such as machine failure data, alarms, operating guidelines, wind history, among others. The integration of these data allows for comprehensive and detailed analysis, essential for accurate predictions tailored to the specificities of each wind farm. The user-friendly interface facilitates data interpretation by managers, while the integration capability with other management systems ensures a holistic view of operations. The failure prediction methodology includes the Jackknife Diagram, Weibull Distribution, and an empirical method of weighting and penalization based on the author's experience in the wind sector, allowing for effective classification and prioritization of failures. The Hórus platform combines advanced and low-cost analytical tools; customization of analysis and reporting screens; integration of multiple data sources for analysis; continuous evolution in beta mode; the initial results of the Hórus platform implementation, started in 2022, show a satisfactory accuracy rate in the predictions made. The purpose of this work is to offer a "predictive vision" that allows managers and technical staff of wind farms to identify and address potential problems before they affect the production of electrical energy, thus promoting operational efficiency and sustainability of wind farms.Acesso AbertoAerogeradoresÂngulo de páMotores de pitchManutenção preditivaPrevisãoConfiabilidademasterThesisCNPQ::ENGENHARIAS::ENGENHARIA ELETRICA