Skip to content

Latest commit

 

History

History
67 lines (58 loc) · 2.82 KB

README.md

File metadata and controls

67 lines (58 loc) · 2.82 KB
Raw

Cartesian Motion Controller

This package provides a controller for following Cartesian target poses.

This type of controller automatically interpolates between two target poses. The amount of how much it interpolates can be set with

  • the PD gains individually for each Cartesian axis
  • the error_scale for all axes alike
  • the number of iterations per control cycle.

As an example, the CartesianMotionController can be given target poses, that pop up somewhere with a relatively low frequency. Depending on how responsive the system is, users can obtain anything in between fast jumps or smooth, directed motion, depending on their use case.

Getting Started

  1. In a sourced terminal, run
roslaunch cartesian_controller_examples examples.launch

  1. In another sourced terminal, open rqt and navigate to the Controller Manager plugin under Robot Tools. Select /controller_manager as namespace and activate my_cartesian_motion_controller. Also activate my_motion_control_handle.

  2. Switch to the RViz window and tick the box Interactive Markers. Use the interactive marker to guide the end effector of the robot. The robot should follow slowely.

  3. In rqt open the Dynamic Reconfigure plugin under Configuration. Play a little with the parameters of my_cartesian_motion_controller (e.g. solver/error_scale) and observe the changes in RViz using the interactive marker as handle.

Example Configuration

Below is an example entry for a controller specific configuration. Also see cartesian_controller_examples/config/example_controllers.yaml for further tips.

my_cartesian_motion_controller:
    type: "position_controllers/CartesianMotionController"
    end_effector_link: "tool0"
    robot_base_link: "base_link"
    target_frame_topic: "target_frame"
    joints:
    - joint1
    - joint2
    - joint3
    - joint4
    - joint5
    - joint6

    pd_gains:
        trans_x: {p: 10.0}
        trans_y: {p: 10.0}
        trans_z: {p: 10.0}
        rot_x: {p: 1.0}
        rot_y: {p: 1.0}
        rot_z: {p: 1.0}

The controller configuration must be loaded to the ros parameter server and is accessed by the controller manager when looking for configuration for the loaded controller my_cartesian_motion_controller.

Tips

Note, that the maximal joint velocities of the robot usually represent the limiting factor for speed. If you notice that increasing error_scale does not make the robot move as fast as you would like, then probably the joint velocity limits in your ROS controlled hardware interface are already reached for this motion. Reaching these limits on individual joints also leads to a distortion of the Cartesian path. Also note that when using this controller in inverse kinematics mode (very high gains), then you must also publish high-frequently sampled targets in order to avoid jumps on joint control level.