Publicação
Planning and control for cooperative launch and capture of drones
| Resumo: | With the popularity in today’s world of using Unmanned Aerial Vehicles (UAVs) for different applications and the presence of these machines in various industries, the focus on building vehicles adapted for specific tasks has increased. The work presented was developed within the scope of Project CAPTURE which aims to tackle the challenges of using rotary-wing shuttle drones to aid in the deployment and landing maneuvers of fixed-wing target drones. This objective emerges as an alternative to using higher complexity vehicles for specific applications, making use of the benefits of each type of drone for their usual missions. This dissertation proposes a solution for the planning and control of a shuttle drone under a cooperative capture scenario and explores the challenges of implementing such a system, both in a simulation environment and in real-world hardware. The developed system was built around the PX4 ecosystem and is based on the Robot Operating System (ROS) framework to implement a state machine-based controller for the shuttle drone, which includes a Model Predictive Control (MPC) strategy for the autonomous execution of the capture maneuver portion of the mission. To test the developed solution on a real system, this work also describes the instrumentation process of a real-world shuttle drone. Simulation results show the success of the cooperative capture maneuver under the considered scenario, whereas preliminary experimental trials indicate that further re- search is necessary to actually achieve a capture maneuver. Nonetheless, these experi- mental trials validate the hardware and software integration of the developed system and the obtained results showed promising outcomes. |
|---|---|
| Autores principais: | Oliveira, Diogo Nunes de |
| Assunto: | Cooperative Planning UAV MPC State Machine Drone Instrumentation PX4 |
| Ano: | 2024 |
| País: | Portugal |
| Tipo de documento: | dissertação de mestrado |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade Nova de Lisboa |
| Idioma: | inglês |
| Origem: | Repositório Institucional da UNL |
| Resumo: | With the popularity in today’s world of using Unmanned Aerial Vehicles (UAVs) for different applications and the presence of these machines in various industries, the focus on building vehicles adapted for specific tasks has increased. The work presented was developed within the scope of Project CAPTURE which aims to tackle the challenges of using rotary-wing shuttle drones to aid in the deployment and landing maneuvers of fixed-wing target drones. This objective emerges as an alternative to using higher complexity vehicles for specific applications, making use of the benefits of each type of drone for their usual missions. This dissertation proposes a solution for the planning and control of a shuttle drone under a cooperative capture scenario and explores the challenges of implementing such a system, both in a simulation environment and in real-world hardware. The developed system was built around the PX4 ecosystem and is based on the Robot Operating System (ROS) framework to implement a state machine-based controller for the shuttle drone, which includes a Model Predictive Control (MPC) strategy for the autonomous execution of the capture maneuver portion of the mission. To test the developed solution on a real system, this work also describes the instrumentation process of a real-world shuttle drone. Simulation results show the success of the cooperative capture maneuver under the considered scenario, whereas preliminary experimental trials indicate that further re- search is necessary to actually achieve a capture maneuver. Nonetheless, these experi- mental trials validate the hardware and software integration of the developed system and the obtained results showed promising outcomes. |
|---|