UAV swarm control strategies: A case study for leak detection

Rafael Braga da Silva (1) , Roberto C da Silva (1) , Alexandre Carlos Brandao-Ramos (2) , Felix Mora-Camino (3)

1 Federal University of Itajuba
2 UNIFEI
3 ENAC - Ecole Nationale de l'Aviation Civile

Disasters involving gas leakage can cause serious damage to properties, to the environment and even endanger human lives. In these situations it is very important to locate the source of gas leakage as fast as possible. A swarm of robots is a solution specially efficient in missions involving exploration and search in unknown environments. This work proposes a strategy using behavioral rules and an idea inspired by the Particle Swarm Optimization algorithm to control a group of Unmanned Aerial Vehicles (UAVs) while searching for the source of a gas leakage. The solution was implemented in C++ using the ROS platform and was tested in simulations using the Pixhawk SITL simulator. Results show that after some time the UAVs converge to the source of leaking gas.

Mots clés: Collision Avoidance; Robot Swarm; Unmanned Aerial Vehicles; ROS; Particle Swarm Optimization

Domaines: Automatique / Robotique

Rafael Braga da Silva, Roberto C da Silva, Alexandre Carlos Brandao-Ramos, Felix Mora-Camino. UAV swarm control strategies: A case study for leak detection. ICAR 2017, 18th International Conference on Advanced Robotics, Jul 2017, Hong Kong, China. pp 173-178 / ISBN: 978-1-5386-3158-4, ⟨10.1109/ICAR.2017.8023514⟩. ⟨hal-01626533⟩