Multi-Vehicle Simulation Framework for Heterogeneous Unconventional MAVs

Overview & Synopsis

The growing interest in using aerial vehicles for diverse missions creates the need for higher autonomy and dexterity. Unconventional configurations can potentially satisfy such requirements, at a cost of new challenges in design and control. Simulation environments are often employed to tackle such challenges while minimizing resources expenditure. In this paper, an existing simulation framework has been extended to allow for the modeling of new UAV configurations. Aerodynamic forces and moments generated by propellers and airframes are defined in separate functions to ease the understanding and allow future addition of new vehicle configurations. Incremental nonlinear dynamic inversion and vector field guidance methods are implemented as vehicle agnostic control and guidance solutions. Vehicle models are extracted and implemented into Paparazzi open-source autopilot system, which makes a seamless switch towards real flights and hardware in the loop tests, facilitating control law development. The simulator can also be used within vehicle and mission design environments. With that, the entire closed loop behavior of the system can be addressed from the design stage, allowing for the vehicle and fleet optimization while considering control and operation constraints. Finally, the whole codebase is made available for the community and can be reached at: github.com/enac-drones/dronesim