uav_simulator

This project facilitates simulating arduCopter enabled UAVs for swarms using SITL and Gazebo. This simulator extends the steps provided by the Intelligent Quads (https://github.com/Intelligent-Quads) developers by providing a containerized solution to their work.

Getting started

It is recommended that this simulator be run on Ubuntu 18.04+

Install Docker Engine

Docker Engine needs to be installed on the host machine, here is a link to the instructions for installing Docker Engine on Ubuntu:

https://docs.docker.com/engine/install/ubuntu/

Clone the repository and navigate to the swarm_simulator directory

cd uav_simulator/swarm_simulator

Generate the docker-compose file

docker-compose is a powerful tool that is utilized here to launch the simulation all at once, however, there is very little control over the "flow" of a docker-compose file. As such, a python script generate_compose.py is used to generate the docker-compose file based on the type of simulation that is desired. At this point the user may specify the number of UAV's to simulate, if gazebo should be used, and if any companion process should be used. Below is an example to generate a compose file for a single arduCopter simulation.

python3 generate_compose.py 1

Number of Vehicles

The first option specifies the number of UAVs to simulate (1-12). For instance, to generate a compose file for 5 SITL instances would be:

python3 generate_compose.py 5

then to start the simulation:

sudo docker-compose up

Include a Companion process -c

To include a ROS-based companion process use the -c flag as follows:

python3 generate_compose.py 4 -c

then to start the simulation:

sudo docker-compose up

There is a default companion process specified in generate_compose.py that starts the NIMBUS clustering control. A companion process can be used without the -c flag. By default, the SITL ports exposed for companion processes are 14XX1 starting at 14561 for agent 1, 14571 for agent 2 and so on.

MAVROS

In order to connect a MAVROS node to a STIL instance use:

agent_1: fcu_url:=udp://127.0.0.1:14561@14565

agent_2: fcu_url:=udp://127.0.0.1:14571@14575

and increment the port by 10 for each subsequent agent

Gazebo

Gazebo provides a more realistic simulation environment as well as the ability to include additional sensors, obstacles, and provides a nice visual component to your simulation.

Installing Gazebo Natively (recommended)

Gazebo and ROS can be installed on the host machine (recommended) by following the steps outlined in https://github.com/Intelligent-Quads/iq_tutorials/blob/master/docs/installing_gazebo_arduplugin.md and https://github.com/Intelligent-Quads/iq_tutorials/blob/master/docs/installing_gazebo_arduplugin.md (you can skip step 3 in the second README).

Containerized Gazebo (not recommended)

Alternativly a seperate docker compose file can be used to start a container with Gazebo already installed, this can be launched by navigating to the gazebo_docker directory and running the command:

sudo docker-compose up

The container can be acessed by launching a VNC client (Remmina works fine for Ubuntu) and opening up localhost:5901 and the password is password. Again, it is recommneded that Gazebo be run natively as the containerized version performs significantly worse.

Proceeding with Gazebo

In order to generate the gazebo portion of the compose file, specify the -g flag:

python3 generate_compose.py 5 -g

DO NOT run the compose file yet. Go to the /home/catkin_ws folder either on the host machine or in the VNCd container and run catkin build. Once the catkin build finishes, run . devel/setup.bash. NOTE if you are not using the container but you are using Ubuntu on a virtual machine, run export SVGA_VGPU10=0 to prevent erros in the future. Finally launch Gazebo with:

roslaunch iq_sim multi_drone.launch

After a Gazebo window appears displaying the drones, proceed as normal:

cd ~/uav_simulator/swarm_simulator

and

sudo docker-compose up