Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.


Extension and implementation of a distributed formation controller


Daniele Speziali

Paul Beuchat, Yvonne Stürz

The goal of D-FaLL (the Distributed Flying and Localisation Lab at IfA) is the development of a distributed autonomous testbed of a fleet of nano-quadcopters. The Crazyflie 2.0 quadcopters will be controlled in a distributed way to fly in formation or to perform tasks together, such as lifting a load cooperatively.

Previous student projects have developed a simulation environment for proving the potential of distributed control and estimation algorithms on the real system. The simulation environment is based on the open-source simulation-software Gazebo and utilises an interface with the Robot Operating System (ROS) to mirror the distributed architecture.

The goal of the first phase of this semester project is to extend the formation controller described in [1] to allow for trajectories to be described as piecewise splines, and to simulate a trajectory-tracking controller that is favourable for implementation on the real-system. The previous semester project that investigated this formation control algorithm concluded that describing the trajectory with a single spline greatly restricted the possible trajectories. Additionally, the previous semester project implemented an acceleration based controller that requires measurements that are not readily available for implementation. It is expected that a Real-Time-Iteration Model-Predictive-Controller (RTI-MPC) is a favourable choice for trajectories-tracking on the real system.

The second phase of the semester project is to implement the developed distributed formation controller on the real system. This will draw from the work of a previous semester project that setup up a framework for stable flight of a single quadcopter. To implement a distributed formation controller, the student will need to extend this framework to multi-agent support, and implement the developed control algorithms.

Coding will be in C for the on-board implementation, Python and C++ inside ROS for the off-board implementation, and Gazebo for simulation.

1) Extend the formation control algorithm from [1] to allow for trajectories to be described by piecewise splines
2) Derived and simulate a RTI-MPC controller for trajectory tracking
3) Implement the developed formation controller on the real multi-quadcopter system

Weitere Informationen

John Lygeros

Art der Arbeit:
Anzahl StudentInnen:
Status: taken
Projektstart: Feb2017
Semester: FS 2017