Skip to main content

Integrating Custom Controllers

Blue integrates auv_controllers for vehicle control. This offers the benefit of supporting the development of custom control algorithms and enabling benchmarking against existing controllers. The following tutorial describes how to use custom controllers with Blue. The files used for this tutorial are available in the blue_demos/control_integration directory. Prior to starting this tutorial, you should have completed the Running Blue in Simulation tutorial.

Dependencies​

The following ROS 2 dependencies are required for this tutorial:

These dependencies will be met by default if you have installed Blue with Docker as described in the installation instructions.

Tutorial steps​

  1. Launch the demo BlueROV2 in simulation by running the following command:

    ros2 launch blue_demos bluerov2_demo.launch.yaml use_sim:=true

  2. Wait for Gazebo, ArduSub, and MAVROS to fully load. Typically, the param MAVROS plugin is the last to fully load, which is indicated by the following message:

    [mavros.param]: PR: parameters list received

  3. Open a new terminal and launch the demo control framework:

    ros2 launch blue_demos bluerov2_controllers.launch.py use_sim:=true

    This should launch a set of cascading controllers, concluding with the integral_sliding_mode_controller (ISMC) controller. The control framework uses the thruster_hardware/ThrusterHardware system interface to enable thruster-level control with ArduSub.

    tip

    On occasion, the ros2_control controller manager will fail to load all of the controllers. If this happens, stop the launch and try again.

  4. Verify that the controllers are running with the following command:

    ros2 control list_controllers

    The output should resemble:

    integral_sliding_mode_controller[velocity_controllers/IntegralSlidingModeController] active
    thruster_allocation_matrix_controller[thruster_allocation_matrix_controller/ThrusterAllocationMatrixController] active
    thruster_1_controller[thruster_controllers/PolynomialThrustCurveController] active
    thruster_2_controller[thruster_controllers/PolynomialThrustCurveController] active
    thruster_3_controller[thruster_controllers/PolynomialThrustCurveController] active
    thruster_4_controller[thruster_controllers/PolynomialThrustCurveController] active
    thruster_5_controller[thruster_controllers/PolynomialThrustCurveController] active
    thruster_6_controller[thruster_controllers/PolynomialThrustCurveController] active
    tip

    Many of the controllers implemented in auv_controllers support dynamically modifying controller parameters and gains. We encourage you to take advantage of this feature to tune the controllers to your specific vehicle. This helps avoid the need to recompile and relauch the system for each change that you make.