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.


Load Carrying using quadcopters and distributed overlapping controller and estimator



Yvonne Stürz, Paul Beuchat, Roy Smith, John Lygeros

The goal of this Master Thesis is to use a fleet of Crazyflie 2.0 quadcopters in the D-FaLL (the Distributed Flying and Localisation Lab at IfA) to lift cooperatively a load using a recently developed distributed control and estimation algorithm. This will be the extension of a previous Master project where a distributed control algorithm for cooperatively carrying a load, modelled as a point mass was implemented.

The first goal of this proposed Master project is the extension of the previous simplified model of a point mass towards a more realistic load of dimensions, inertia etc. and implement it in the simulation environment to model multiple quadcopters dynamically coupled via this hanging load. The model should be linearised to be used in the distributed controller and estimator design.

The second goal of the Master Thesis is the extension of the previous implementation of distributed control towards an underlying distributed estimation scheme. The distributed control and estimation scheme has been published in [1] and is able to make a trade off between communication, performance and computational effort. The estimation is done in an augmented state space, where overlapping parts of the entire state space are estimated locally. The local control actions of the single agents are based on these overlapping estimates. All the theory is provided in [1]. Experiments should be performed for comparing the influence of different design parameters.

As a third goal, this overlapping estimation scheme will be further extended by introducing additional communication between the crazyflies in order to obtain physically meaningful augmented estimates. The theory for this will be provided. Experiments will be performed for evaluating the control and estimation performance.

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

1) Extend the model and the simulation environment for multiple quadcopters carrying a realistic load.
2) Extend the implementation of the distributed controller for the task of multiple quadcopters carrying the hanging load towards the underlying overlapping estimation scheme as in [1].
3) Further extend the overlapping estimation scheme by additional communication to achieve physical estimates. Implement the extension in simulation and on the real-world system.
4) Perform experiments in order to compare the control performance, amount of communication, and other criteria for different design parameters of the implemented distributed control and estimation scheme.

[1] Yvonne R. Stürz, Annika Eichler, and Roy S. Smith. A framework for distributed control based on overlapping estimation for cooperative tasks. In IFAC Proc., pages 14861 – 14866, 2017.

Weitere Informationen

Roy Smith

Art der Arbeit:
Anzahl StudentInnen:
Status: open
Projektstart: February 2018