Estudio, control e implementación de sistemas robóticos avanzados : Aplicación de algoritmos iterativos de planificación de rutas (path planning) en un sistema multi-agente dentro del ambiente de ROS Gazebo.

Abstract

In this project, a Multi-agent System (SMA) is designed and path-planning algorithms are tested. This application is critical, for example, within the exploration and recognition of areas through mobiles robots. As a first approach, a bibliography study of the robot operating system (ROS) is carried out. It helps currently many software developers to create several robotic applications by employing open-source platforms. Furthermore, it has many complements and libraries, for example, Gazebo and Rviz that are tools for 3D simulation and visualization, respectively. In this project, the creation of a virtual world is developed in Gazebo and then this is mapped from Rviz with the SLAM (Simultaneous Localization and Mapping) technique. Finishing the designing of the virtual world, the rapidly exploring random trees (RRT) algorithm and the probabilistic roadmap (PRM) algorithm are implemented. The generation of paths by both algorithms are different; the first one traces the paths in form of branches while the second takes samples from the virtual environment. The tests of these algorithms with the SMA are carried out by modifying their parameters and including several obstacles to the virtual environment. As results, different paths are found and compared, for example, through the computation of the navigation distance and their simulation time.