Skip to content

This s a Team project about simple swarm behavior simulations with drones

Notifications You must be signed in to change notification settings

Nils-Treuheit/swarmBehaviours

Repository files navigation

Swarm Behaviours

This is a repository where we are going to share our swarm behaviour implementations.

In the file helpful_links.md we are going to post links to websites, videos or documents, that can help us figuring out how to implement the behaviours. This behaviours will all be implemented assuming that we have a working interface and enough knowledge about the API to receive and send the needed data. Most C and Python Code will be specific to the Software Project that we are working on. This will help migrating the code later on to the actual Software Project.

In this Repository you will also find a simplified 2D Java Demo of our Swarm Behaviors.
To view the demo coded in Java you need to download an IDE called Processing and load the project.

The files testParameters_attRep.py and testParameters_conSteer.py are tools, that help to determine theoretically good hyperparameters for the different swarm behaviors - but of course you should test them thoroughly in your specific simulation setup and see how they perform.

We will provide you with a table full of tested parameter values for our paparazzi simulation setup, that we found useful for certain scenarios for attraction and repulsion here and for context steering here. The documentation of our project can be found here. The documented code of our project can be found here. The code documentation generated by doxygen in a website format can be found here.


Swarm Tasks/Arena Scenarios

-> First Scenario: simple_flight_plan.csv

This file resembles a simple flight plan scenario in which the swarm has to fly to different locations one after another by reaching goal/attraction points and avoiding repell points. The repell point corrdinates, attraction point coordinates and spawn point coordinates can be found each in one line. The first row gives each collumn a name. Coordinates can also be found here.

-> Second Scenario: circle.csv

This file resembles a simple flight plan scenario in which the swarm is placed in a cirlce 5 meters away from a repell point, which is in the center of the scene. Also 10 attraction points set in a circle with a radius of 50 meters around the repell point. The first row gives each collumn a name. Coordinates can also be found here.

-> Third Scenario: tube.csv

This file resembles a simple flight plan scenario in which the swarm has to fly through a tube of repell points to a horde of attraction points at the other side of the tube. The 10 repell point corrdinates, attraction point coordinates and spawn point coordinates can be found each in one line. The first row gives each collumn a name. Coordinates can also be found here.

In addition we created three random scenarios with the createRandom.py script, that can be found here.

Simulation Runs & Results

We provide some exemplary log files from our paparazzi simulation together with an executeable evaluation script version in the Log-Files folder. In the Results folder we published our offical final simulation evaluation results, that can also be found in the Documentation. These results were achieved using the code files in the Project-Code folder and the final parameters from the linked tables for the paparazzi simulation. All of our simulations were run on an Ultrabook from Razer within a Virtual Machine.
The machine has the following technical specifications:
-> CPU: 11th Gen Intel(R) Core(TM) i7-1165G7 @ 2.80GHz
-> GPU: Intel(R) Iris(R) Xe Graphics (integrated)
-> RAM: 16,0 GB 4267MHz
-> SSD: SSSTC CA5 M.2 2280 256GB
The VM was configured to use all 4 CPU cores and 6 GB of RAM while reserving about 40 GB of the SSD's capacity.

About

This s a Team project about simple swarm behavior simulations with drones

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published