http://repositorio.ufc.br/handle/riufc/24023 Thu, 02 Apr 2026 14:48:23 GMT 2026-04-02T14:48:23Z http://repositorio.ufc.br/handle/riufc/84976 Título: Comparação de algoritmos de machine learning para predição de vibração em moinhos de rolos verticais aplicados à manutenção preditiva industrial Autor(es): Silva Júnior, Paulo César da Abstract: Predictive maintenance based on vibration monitoring has become a fundamental strategy for optimizing operational costs and increasing the availability of critical equipment in the cement industry. Vertical roller mills exhibit characteristic vibrational patterns that can indicate the development of operational anomalies before critical failures occur. This work presents a comparative analysis of six machine learning algorithms for vibration prediction in vertical roller mills, using real industrial data collected over 82 days of continuous operation, totaling 23,538 samples after preprocessing. The evaluated models were: K-Neighbors, Linear Regression, XGBoost, CatBoost, Huber Regressor, and Random Forest. The methodology included careful selection of 15 predictor variables based on operational knowledge and statistical analysis, feature engineering with categorical variable encoding, temporal dataset splitting for realistic validation, and comparative evaluation using quantitative metrics (R², RMSE, MAE, MAPE) complemented by visual analysis of time series and residuals. The results revealed that KNeighbors showed superior performance with R² of 0.7679, RMSE of 0.8792 mm/s, and MAE of 0.66 mm/s, surpassing traditionally more robust ensemble models. Visual analysis identified a critical regression-to-the-mean phenomenon in five of the six models, where high vibration values are systematically underestimated, a limitation that compromises applicability in alarm systems for predictive maintenance. K-Neighbors was the only model capable of preserving the natural variability of vibration, an essential characteristic for reliable detection of operational anomalies. This work demonstrates that model selection for critical industrial applications cannot be based exclusively on aggregate metrics, requiring qualitative evaluation of predictive behavior across the entire operational range. The developed methodology represents a replicable process for predictive systems in critical rotating equipment, with potential for significant reduction of unplanned shutdowns and optimization of operational costs. Tipo: TCC Thu, 01 Jan 2026 00:00:00 GMT http://repositorio.ufc.br/handle/riufc/84976 2026-01-01T00:00:00Z http://repositorio.ufc.br/handle/riufc/84946 Título: A influência do arqueamento do perfil de aerofólio nos coeficientes aerodinâmicos para aplicação em aerodesign: uma análise numérica Autor(es): Silva, José Joabe de Andrade Abstract: This study aims to investigate the influence of airfoil profile geometry variation on key aerodynamic coefficients, particularly how the variation of an airfoil’s camber affects the lift, drag, and moment coefficients, with a focus on applications in the aerodesign niche. Through computational numerical analysis, using the panel method due to its lower computer processing requirements via the XFLR-5 software. Derived profiles based on the NACA 0012 airfoil were simulated, with the new profiles having cambers of 2%, 4%, 6%, and 8%, respectively, conducting quantitative and qualitative analysis of the obtained values for a fixed angle of attack at 10°. The results show that the lift coefficient increases proportionally with camber and can be modeled by a recurrence formula based on the statistical tool control chart. On the other hand, the drag and moment coefficients also increase, but in a non-linear manner and without a pattern predictable by the same methodology. The study concludes that the lift coefficient can be mathematically modeled as a function of the airfoil’s camber; however, further complementary studies are recommended for the drag and moment coefficients. Tipo: TCC Thu, 01 Jan 2026 00:00:00 GMT http://repositorio.ufc.br/handle/riufc/84946 2026-01-01T00:00:00Z http://repositorio.ufc.br/handle/riufc/84777 Título: Implementação de um sistema de análise de vibrações mecânicas: estudo de caso Autor(es): Pereira, Thomaz Éderson dos Santos Abstract: This work presents the implementation of an intelligent mechanical vibration analysis system, utilizing IoT (Internet of Things) sensors and artificial intelligence to enhance predictive maintenance in critical industrial equipment, with a focus on cranes. Such knowledge is essential to anticipate failures before they occur, thereby reducing operational costs and increasing safety. Accelerometer sensors were installed and connected to a digital platform, enabling continuous monitoring and automatic alert generation based on vibration patterns. The results demonstrated the system's effectiveness in detecting real-time anomalies that are more difficult to identify using traditional methods, validating its technical efficiency. In the case study, the preventive replacement of a bearing avoided estimated losses of approximately R$ 12.9 million, reinforcing the positive economic impact of the solution. It is concluded that the integration of precise data and human technical expertise is essential to transform maintenance into a strategic process aligned with Industry 4.0 demands, contributing to the field of mechanical engineering by validating the application of IoT and AI technologies in industry, and providing information for replicability of the application. Tipo: TCC Thu, 01 Jan 2026 00:00:00 GMT http://repositorio.ufc.br/handle/riufc/84777 2026-01-01T00:00:00Z http://repositorio.ufc.br/handle/riufc/84737 Título: Dimensionamento de um mezanino metálico em perfil W laminado Autor(es): Silva, Felipe dos Santos Abstract: O avanço nos processos de produção do aço possibilitou sua aplicação em larga escala como elemento principal na construção de estruturas nos mais diversos setores industriais, em razão de características que o tornam uma opção mais vantajosa quando comparado ao concreto armado, tais como velocidade de execução, obra limpa e elevada resistência mecânica. Atualmente, observa-se que o uso de softwares para o dimensionamento de estruturas está cada vez mais comum, devido à sua rapidez e precisão nos cálculos. Contudo, é fundamental que o projetista seja capaz de realizar os cálculos de forma analítica, a fim de interpretar adequadamente os resultados obtidos por meio dessas ferramentas computacionais. Dessa forma, este trabalho tem como objetivo explanar o processo de dimensionamento de um mezanino metálico em perfis laminados, tendo como sistema de piso painéis Wall. O mezanino é uma estrutura que subdivide um pavimento, bastante utilizado para aumentar a área útil de um ambiente, principalmente no ambiente fabril. O desenvolvimento do trabalho baseou-se na aplicação do método dos estados limites, adotando-se um procedimento de cálculo misto, no qual os esforços solicitantes de cálculo foram obtidos por meio de uma análise estrutural via software, enquanto as verificações dos esforços resistentes foram realizadas a partir de rotinas desenvolvidas em planilhas, em conformidade com as normas técnicas vigentes pertinentes ao escopo do projeto. Tipo: TCC Thu, 01 Jan 2026 00:00:00 GMT http://repositorio.ufc.br/handle/riufc/84737 2026-01-01T00:00:00Z