01. Universidade Federal Rural de Pernambuco - UFRPE (Sede)
URI permanente desta comunidadehttps://arandu.ufrpe.br/handle/123456789/1
Navegar
6 resultados
Resultados da Pesquisa
Item Automatização dos serviços de coleta domiciliar com o QGIS e Python(2024-10-04) Prado, Artillis Henrique Mendes do; Medeiros, Victor Wanderley Costa de; http://lattes.cnpq.br/7159595141911505; http://lattes.cnpq.br/7294722691197017This work describes the development of an innovative tool for monitoring household waste collection services in the city of Recife, Brazil, using QGIS and Python. The main objective is to optimize the management of the Operational Control Center (CCO) of TPF, which operates within Emlurb, allowing for better monitoring of collection operations and analysis of vehicle routes. The developed platform collects and processes geospatial data, providing accurate and timely information about the trucks’ paths and generating reports that assist in service management. By integrating these technologies, the tool offers features such as identifying unserviced areas, analyzing response times by sector, and evaluating vehicle productivity. The generated reports provide a comprehensive view of operational efficiency, facilitating strategic decision-making, including route adjustments and the reallocation of more productive vehicles. Additionally, the tool allows filtering data by speed and distance tolerance, helping to verify whether the trucks adhered to their planned routes. Implementing this solution yields significant benefits for public management, including enhanced operational control, data accuracy, and transparency in waste collection activities. By generating real-time information and rapid results, the tool contributes to the efficiency of operations and the continuous improvement of services, fostering a more effective and data-driven administration. Thus, the developed system helps streamline the waste collection process, making it more agile and sustainable.Item Da propagação do calor à construção de desenhos: uma aplicação das séries de Fourier com Python(2024) Domingos, Cleianderson Paz; Freitas, Lorena Brizza Soares; http://lattes.cnpq.br/2302580820419163; http://lattes.cnpq.br/8909785797719318This work aims to present an application of Fourier series in generating figures. To do so, we first study the problem of heat conduction in a finite rod, as well as the equation that models it and its solution, both proposed by Joseph Fourier in the early 19th century. Initially, a historical note is presented, exhibiting some facts that lead to the motivation for studying heat propagation. Then, we derive the Heat Equation from two physical laws and study how Fourier series emerge in an attempt to solve this equation. Subsequently, through convergence theorems, we study necessary conditions for a function to be represented by its Fourier series. Finally, we explore an application of Fourier series in figure generation using epicycles and develop a Python algorithm to visualize this application.Item Uso de Machine Learn para classificação de lançamentos financeiro: estudo comparativo entre modelo AutoML e Redes MLP(2022-10-10) Silva, Vinicius Mateus Mendonça da; Monteiro, Cleviton Vinicius Fonsêca; http://lattes.cnpq.br/9362573782715504; http://lattes.cnpq.br/6180002649065928The study of this work aims to help companies in their financial management by generating models based on Machine Learning to classify financial releases. With the help of libraries developed in the Python language, it was possible to train AutoML models and Multilayer Perceptron Neural Networks responsible for data classification. With results above 85% in the metrics of Accuracy, Recall, F-measure and Precision for both models, using them brings the possibility of better management of financial releases with less effort.Item Processo de ETL voltado para jurimetria(2021-12-20) Cristovam, Edson Fagner da Silva; Bocanegra, Silvana; http://lattes.cnpq.br/4596111202208863; http://lattes.cnpq.br/1519472768897299Large companies may have difficulty managing numerous court cases they are defendants in, as a team of the size needed to manage these cases can be very costly. With the digitization of judicial processes, a strategy that automates the adjustment of process information and is easy to consult can be very useful. Therefore, the data warehouse seems to be a very suitable technique for this, as it normalizes and standardizes information from different sources that are contained in it. But to load the data into a data warehouse, it’s necessary to prepare it with a method commonly known as ETL. With this method we can create data extraction, transformation and load flows that will be written in the data warehouse. With a data warehouse filled with legal data, we can use it to help decision-making in the legal sector of companies through jurimetry, which consists of applying statistical tools to Law, providing new points of view based on data.Item Uso da ciência de dados para estudo de falhas e fraudes dos abastecimentos de postos de gasolina(2019-12-19) Arruda, Luiz Felipe Ribeiro de; Albuquerque Júnior, Gabriel Alves de; Roullier, Ana; http://lattes.cnpq.br/1399502815770584; http://lattes.cnpq.br/1825682578554550Item Técnicas de Modelagem Matemática e os Métodos de Runge-Kutta(2021-07-23) Silva, Angelo Antunes da Rocha; Didier, Maria Ângela Caldas; Gondim, João Antônio Miranda; http://lattes.cnpq.br/2674397127545655; http://lattes.cnpq.br/9721552594807972; http://lattes.cnpq.br/9069459979748516This work consists in the study of Mathematical Modeling with numerical analysis of the models. In it we present the steps of a modeling process, define and evaluate a mathematical model and also discuss the technique of modeling by fitting curves through the Minimal Squares Method, as well as by differential equations where we approach some models, among them those which describe a populational growth dynamic and epidemiological models. We also present the methods from Taylor Series and Runge-Kutta for the construction of numeric solutions for a initial value problem. As the main contribution we simulated analytical and numerical solutions for four problems of initial value, analysing the error linked to numerical solutions, aiming to answer questions related to the general formula of the Runge-Kutta method of order 2. To calculate error for a certain range we used L2 norm and a closed formula from Newton-Cotes. The purpose here is to offer material in the subject of Mathematical Modeling which can be used by Mathematics Graduation students as another area that utilizes Differential Calculus as a tool. The simulations were coded using Python and the code may be accessed through the link in the beggining of this essay.