This package utilizes the pluginlib to make the
simulate
app from mujoco easily extendable. With this
plugin support, mujoco_ros2_control is integrated to enable a flexible
control interface of the simulated robot and sensors with the ros2_control framework.
You can use this package together with the mujoco_ros2_control package to get a similar setup as the gz_ros2_control package for Gazebo.
The following environment and software versions are supported and regularly tested:
- Ubuntu 24.04
- ROS2 Jazzy
- Mujoco 3.2.7
It's recommended to clone and build the mujoco_demo_robot
package to get a working example of the simulation setup.
The demo robot model test_cart.xml
is defined in mujoco_ros2_control
and the mujoco_demo_robot
package
demonstrates how to start the simulation as well as the ros2 controllers.
cd <colcon-ws>/src
git clone -b 3.2.7 https://github.com/google-deepmind/mujoco.git
git clone https://github.com/rxdu/mujoco_sim_ros2.git
git clone https://github.com/rxdu/mujoco_ros2_control.git
git clone https://github.com/rxdu/mujoco_demo_robot.git
cd <colcon-ws>
colcon build --symlink-install
Now you can start the simulation with the demo robot.
ros2 launch mujoco_demo_robot cart_effort.launch.py
You can check the ros2_control setup by listing available hardware and controllers:
You can test the controller by sending a command to the effort controller:
ros2 run mujoco_ros2_control_demos example_effort