Development of Gazebo-based high-fidelity simulation environment for morphing-wing UAVs

Flapping- and morphing-wing aerial robots are one of the emerging topics in the aerial robotics community. The problem with these types of aerial robots is the lack of control possibilities available for them. This severely limits the functionality of these robots and the applications they can be used for.

The Portwings project aims to understand and model the forces that work on birds and their artificial counterparts. Using this understanding they will make a birdlike robot with never before seen control functionality.

The making of a robot controller is frequently done with the use of high fidelity robotic simulators like Gazebo. The aerodynamic support of these simulators is limited to fixed-wing and multi-rotor aerial robots. The first goal of this research is to document the software architecture, functionalities and limitations of the current aerodynamics implementation used in Gazebo. The second goal is to use the gained understanding to find extension possibilities for flapping- and morphing-wing support.

During this research, file interaction and code analysis are used for the documentation of the current aerodynamics implementation. The problems found in this implementation are documented and discussed.

It is concluded that the current implementation can not be extended by simply changing the formulas used in the aerodynamics plug-in. It is recommended to copy the use of a model plug-in and method of force application upon the model links from the current implementation. It is also recommended that the forces that should be applied are calculated elsewhere and communicated through the API. Alternatively, it was recommended to convert a 20-sim model of the robot to C++ code, this code could then be edited to be a model plug-in.

To join the presentation via Microsoft Teams click here