Open-Access Platform for the Simulation of Aerial Robotic Manipulators

Abstract

Recent technological advances have brought increased attention to aerial robotic manipulators (ARMs), particularly in applications that involve physical interactions using tools such as welding and drilling, as well as in the autonomous pickup and transport of objects. However, translating control algorithms into real-world applications for aerial robotic manipulators may prove challenging, given the potential for accidents and the time-consuming nature of experiments; furthermore, the acquisition of aerial robotic manipulators could impose a substantial financial burden on universities, research centers, and companies. Therefore, this work addresses these issues by developing an open access platform to simulate aerial robotic manipulators and test control strategies. The presented simulator is based on the kinematics and dynamics of the Matrice-100 aerial platform equipped with a 3 DOF robotic arm, where the mathematical formulation was developed using the Euler-Lagrange formalism. In addition, optimization techniques were used to perform the parameter identification procedure, ensuring the development of an accurate model for the open-access platform. The simulator platform is built upon the integration of Python, the Robot Operating System (ROS), and Unity 3D. These components collaborate to describe and demonstrate the behavior of the aerial robotic manipulator during the test process of control system algorithms. Simple tests were conducted to validate the open-access simulator platform. The proposed approach ensures the evaluation, testing of control strategies, and the ability to conduct experiments before hardware implementations. Finally, the proposal was published as an open source platform available in the following Code.

Keywords

• Robotic simulation
• aerial robotic manipulator dynamics
• open source
• dynamic parameters identification
• control algorithms