Skip to content

Latest commit

 

History

History
53 lines (45 loc) · 2.03 KB

README.md

File metadata and controls

53 lines (45 loc) · 2.03 KB
Raw

Cartesian Force Controller

This package provides a controller for end effector force control.

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_force_controller.

  2. Publish a geometry_msgs/WrenchStamped to /target_wrench with force x = 2 and watch the robot move.

  3. In rqt open the Dynamic Reconfigure plugin under Configuration. Play a little with the parameters of my_cartesian_force_controller (e.g. solver/error_scale) and observe the changes in RViz with publishing different values along various axes. Also try forcing the end effector through singularities and get a feeling how the controller behaves.

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_force_controller:
    type: "position_controllers/CartesianForceController"
    end_effector_link: "tool0"
    robot_base_link: "base_link"
    ft_sensor_ref_link: "sensor_link"
    joints:
    - joint1
    - joint2
    - joint3
    - joint4
    - joint5
    - joint6

    pd_gains:
        trans_x: {p: 0.05}
        trans_y: {p: 0.05}
        trans_z: {p: 0.05}
        rot_x: {p: 0.01}
        rot_y: {p: 0.01}
        rot_z: {p: 0.01}

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_force_controller.

Tips

Note that the controller does not strictly move only in the commanded direction. Sometimes there's a small drift in other axes. This is a feature of the forward dynamics solver. In fact, there is no error reduction on axes orthogonal to the target wrench. The benefit is that the controller finds approximate solutions near singular configurations.