Fotos das Defesas
DEFESAS 2023
Título: Análise de Arquiteturas de Redes Neurais Para Segmentação Semântica de Dados Sísmicos
Autor: Gabriel Danilo Figueiredo da Silva Data da defesa: 31/05/2023
Qualificação
Título: Desenvolvimento de um sensor ótico de umidade do solo para aplicaçoes agrícolas
Autor: Gorsi Mawuli Atiglo Data da qualificação: 18/12/2023
DEFESAS 2022
Autor: Higor Oliveira da Cunha Data da Defesa: 24/03/2022
Autor: Arcano Matheus Bragança Leite Data da Defesa: 23/03/2022
DEFESAS 2021
Aluno: Gabriel Gelard Reis de Castro
Título: Dynamic Path Planning based on Neural Networks for Aerial Inspection
Resumo:
Unmanned Aerial Vehicles (UAVs) are a suitable solution to automate inspections on large structures that require periodic inspections due to their flexibility of flight and preventing risks to humans. However, the inspected environment could be complex and highly dynamic. Therefore, tridimensional path planners are crucial in aerial tasks. Different kinds of 3D path planning were developed in the last years, and most of them are classified as heuristic methods. A drawback of these traditional methods is that computational time depends on the environment’s complexity and scale. Path planning based on Neural Networks (NNs) arises from overcoming these algorithms’ bottlenecks. In this sense, this work proposes a novel deep learning structure to perform real-time path planning in an unknown environment. The strategy was developed to generate an initial database used for training and validate the results in a simulated environment using Gazebo/ROS software. The results showed that the network topology and the training method obtained satisfactory performance to be applied in a real-world scenario.
Banca: Presidente, Milena Faria Pinto, D.Sc. (orientadora) (CEFET/RJ), Diego Barreto Haddad, D.Sc. (CEFET/RJ), Ana Lucia Ferreira de Barros, PhD. (CEFET/RJ), André Luis Marques Marcato, D.Sc. (UFJF), e Raul Queiroz Feitosa, D.Sc. (PUC-RJ)
Aluno: Gabryel Silva Ramos
Título: A Framework for Autonomous UAV in Offshore Mooring Tasks
Resumo:
As oil and gas exploration goes toward deeper fields in the Brazilian industry scenario, the offloading operations consolidate themselves as the most viable option to drain production. However, this operation demands expensive resources, such as shuttle tankers and support boats, and presents operational risks, which despite managed, limited and mitigated to be as low as reasonably possible, are still present. Therefore, this research proposes to use Unmanned Aerial Vehicles (UAVs) in an autonomous mode to carry out the messenger line from the shuttle tanker to the oil rig instead of using a Line Handling (LH) boat (for conventional operations that use those resources) or the messenger cable launching guns (for dynamic positioning operations). This represents a viable alternative solution to reduce costs and risks in these tasks and a possibility to eliminate some meteorologic and oceanographic limiting conditions to operations, since the UAV will be susceptible only to wind conditions, and not to sea and visibility conditions like LHs are. It is presented in this work a hybrid navigation methodology based on computer vision and sensor fusion technique with Extended Kalman Filter for autonomous Unmanned Aerial Vehicles (UAV). The proposed framework was developed in Robotics Operating System (ROS), tested in a realistic simulated environment considering several practical operational constraints. The same controller was tested on a physical UAV using scale ship models and also in a field test aboard a production facility to validate simulation results. The developed controller was noted to behave very well and simulations and physical tests showed robust results. Based on these tests, an economical evaluation was conducted by simulating the application of this technology in two distinct Brazilian deep water oilfield scenarios.
Banca: Presidente, Milena Faria Pinto, D.Sc. (orientadora) (CEFET/RJ), Diego Barreto Haddad, D.Sc. (CEFET/RJ), Ana Lucia Ferreira de Barros, PhD. (CEFET/RJ), Leonardo de Melo Honório (UFJF), e Cristina Urdiales (Universidad de Málaga)
DEFESAS 2020
Redes Sociais