Controller¶

Every Robot is equipped with a controller, which determines both the action space as well as how its values are mapped into command torques. By default, all controllers have a pre-defined set of methods and properities, though specific controllers may extend and / or override the default functionality found in the base class.

Base Controller¶

class perls2.controllers.base_controller.Controller¶

General controller interface.

abstract run_controller(action)¶: Go from actions to torques

scale_action(action)¶: Scale the action based on max and min of action

Joint Impedance Controller¶

class perls2.controllers.joint_imp.JointImpController(robot_model, input_max=1, input_min=-1, output_max=1.0, output_min=-1.0, kp=50, kv=None, damping=1, control_freq=20, qpos_limits=None, interpolator_qpos=None)¶

Controller for joint impedance

set_goal(delta, set_qpos=None, **kwargs)¶

Set goal for controller. :param delta: (7f) list of deltas from current joint positions :type delta: list :param set_qpos: (7f) goal positions for the joints. :type set_qpos: list :param kwargs: additional keyword arguments. :type kwargs: dict

Returns: None

Examples::

$ self.controller.set_goal(delta=[0.1, 0.1, 0, 0, 0, 0], set_pos=None, set_ori=None)

$self.controller.set_goal(delta=None, set_pose=[0.4, 0.2, 0.4], set_ori=[0, 0, 0, 1])

run_controller()¶

Calculate torques to acheive goal state from current state.

Parameters: None –
Returns: 7f list of torques to command.
Return type: torques (list)

Joint Velocity Controller¶

class perls2.controllers.joint_vel.JointVelController(robot_model, input_max=1, input_min=-1, output_max=1, output_min=-1, kv=4.0, control_freq=20, interpolator=None)¶

Controller for joint velocity

set_goal(velocities, **kwargs)¶

Set goal joint velocities

Parameters

velocities (list) – 7f list of joint velocities to command to the robot.
kwargs (dict) – additional keyword arguments.

Returns

None

run_controller()¶

Run controller to calculate torques.

Returns: 7f list of torques to command.
Return type: torques (list)

Joint Torque Controller¶

class perls2.controllers.joint_torque.JointTorqueController(robot_model, input_max=1, input_min=-1, output_max=1.0, output_min=-1.0, control_freq=20, interpolator=None)¶

Controller for joint torque

robot_model¶

model of robot containing state and parameters.

Type: Model

input_max¶

Maximum above which an inputted action will be clipped.

Type: float or list of floats

input_min¶

Minimum below which an inputted action will be clipped.

Type: float or list of floats

output_max¶

Maximum which defines upper end of scaling range when scaling an input action.

Type: float or list of float

output_min¶

Minimum which defines lower end of scaling range when scaling an input action.

Type: float or list of float

interpolator¶

Interpolator object to be used for interpolating between consecutive goals..

Type: LinearInterpolator

goal_torque¶

goal torques for controller.

Type: list of float

current_torque¶

current torque being outputted.

Type: list of float

joint_dim¶

number of joints in the arm (usually 7.)

Type: int

torques¶

7f list of torques.

Type: list

set_goal(torques, **kwargs)¶

Set goal torques.

Parameters

torques (list) – 7f list of torques to command to the robot.
kwargs (dict) – additional keyword arguments.

Returns

None

run_controller()¶

Run controller to calculate torques.

Returns: 7f list of torques to command.
Return type: torques (list)

EEImpedance Controller¶

class perls2.controllers.ee_imp.EEImpController(robot_model, input_max=1, input_min=-1, output_max=1.0, output_min=-1.0, kp=50, kv=None, damping=1, control_freq=20, position_limits=None, orientation_limits=None, interpolator_pos=None, interpolator_ori=None, uncouple_pos_ori=False, **kwargs)¶

Class definition for End-effector Impedance Controller.

End effector impedance uses PD control to reach desired end-effector position and orientation.

input_max¶

Maximum above which an inputted action will be clipped. Can be either be a scalar (same value for all action dimensions), or a list (specific values for each dimension). If the latter, dimension should be the same as the control dimension for this controller

Type: float or list of float

input_min¶

Minimum below which an inputted action will be clipped. Can be either be a scalar (same value for all action dimensions), or a list (specific values for each dimension). If the latter, dimension should be the same as the control dimension for this controller

Type: float or list of float

output_max¶

Maximum which defines upper end of scaling range when scaling an input action. Can be either be a scalar (same value for all action dimensions), or a list (specific values for each dimension). If the latter, dimension should be the same as the control dimension for this controller

Type: float or list of float

output_min¶

Minimum which defines upper end of scaling range when scaling an input action. Can be either be a scalar (same value for all action dimensions), or a list (specific values for each dimension). If the latter, dimension should be the same as the control dimension for this controller

Type: float or list of float

kp¶

positional gain for determining desired torques based upon the pos / ori errors. Can be either be a scalar (same value for all action dims), or a list (specific values for each dim)

Type: float or list of float

kv¶

velocity gain for determining desired torques based on vel / ang vel errors. Can be either a scalar (same value for all action dims) or list (specific values for each dim). If kv is defined, damping is ignored.

Type: float or list of float

damping¶

used in conjunction with kp to determine the velocity gain for determining desired torques based upon the pos / ori errors. If kv is defined, damping is ignored.

Type: float or list of float

policy_freq¶

Frequency at which actions from the robot policy are fed into this controller

Type: int

position_limits¶

Limits (m) below and above which the magnitude of a calculated goal eef position will be clipped. Can be either be a 2-list (same min/max value for all cartesian dims), or a 2-list of list (specific min/max values for each dim)

Type: 2-list of float or 2-list of list of floats

orientation_limits¶

[Lower_bounds, upper_bounds]. Limits (rad) below and above which the magnitude of a calculated goal eef orientation will be clipped. Can be either be a 2-list (same min/max value for all joint dims), or a 2-list of list (specific min/mx values for each dim)

Type: 2-list of float or 2-list of list of floats

interpolator_pos¶

Interpolator object to be used for interpolating from the current position to the goal position during each timestep between inputted actions

Type: Interpolator

interpolator_ori¶

Interpolator object to be used for interpolating from the current orientation to the goal orientation during each timestep between inputted actions

Type: Interpolator

control_ori¶

Whether inputted actions will control both pos and ori or exclusively pos

Type: bool

uncouple_pos_ori¶

Whether to decouple torques meant to control pos and torques meant to control ori

Type: bool

**kwargs: Does nothing; placeholder to “sink” any additional arguments so that instantiating this controller via an argument dict that has additional extraneous arguments won’t raise an error

set_goal(delta, set_pos=None, set_ori=None, **kwargs)¶

Set goal for controller.

Parameters

delta (list) – (6f) list of deltas from current position [dx, dy, dz, ax, ay, az]. Deltas expressed as position and axis-angle in orientation.
set_pos (list) – (3f) fixed position goal. [x, y, z] in world frame. Only used if delta is None.
set_ori (list) – (4f) fixed orientation goal as quaternion in world frame. Only used if delta is None.
kwargs (dict) – additional keyword arguments.

Returns

None

Examples::

$ self.controller.set_goal(delta=[0.1, 0.1, 0, 0, 0, 0], set_pos=None, set_ori=None)

$ self.controller.set_goal(delta=None, set_pose=[0.4, 0.2, 0.4], set_ori=[0, 0, 0, 1])

run_controller()¶

Calculate torques to acheive goal.

See controllers documentation for more information on EEImpedance: control. Set goal must be called before running controller.

Parameters: None –
Returns: 7f list of torques to command.
Return type: torques (list)

Run impedance controllers.

EEPosture Controller¶

class perls2.controllers.ee_posture.EEPostureController(robot_model, input_max=1, input_min=-1, output_max=1.0, output_min=-1.0, kp=50, kv=None, damping=1, posture_gain=0, posture=[0, -1.18, 0.0, 2.18, 0.0, 0.57, 3.3161], policy_freq=20, control_freq=500, position_limits=None, orientation_limits=None, interpolator_pos=None, interpolator_ori=None, uncouple_pos_ori=False, **kwargs)¶

Class definition for End-effector Impedance Controller.

End effector impedance uses PD control to reach desired end-effector position and orientation, torques are projected to nullspace to maintain posture.

input_max¶

Maximum above which an inputted action will be clipped. Can be either be a scalar (same value for all action dimensions), or a list (specific values for each dimension). If the latter, dimension should be the same as the control dimension for this controller

Type: float or list of float

input_min¶

Minimum below which an inputted action will be clipped. Can be either be a scalar (same value for all action dimensions), or a list (specific values for each dimension). If the latter, dimension should be the same as the control dimension for this controller

Type: float or list of float

output_max¶

Maximum which defines upper end of scaling range when scaling an input action. Can be either be a scalar (same value for all action dimensions), or a list (specific values for each dimension). If the latter, dimension should be the same as the control dimension for this controller

Type: float or list of float

output_min¶

Minimum which defines upper end of scaling range when scaling an input action. Can be either be a scalar (same value for all action dimensions), or a list (specific values for each dimension). If the latter, dimension should be the same as the control dimension for this controller

Type: float or list of float

kp¶

positional gain for determining desired torques based upon the pos / ori errors. Can be either be a scalar (same value for all action dims), or a list (specific values for each dim)

Type: float or list of float

kv¶

velocity gain for determining desired torques based on vel / ang vel errors. Can be either a scalar (same value for all action dims) or list (specific values for each dim). If kv is defined, damping is ignored.

Type: float or list of float

damping¶

used in conjunction with kp to determine the velocity gain for determining desired torques based upon the pos / ori errors. If kv is defined, damping is ignored.

Type: float or list of float

posture_gain¶

gains for determining torques to maintiain desired joint posture.

Type: list

posture¶

7f joint positions for desired nullspace posture to maintain.

Type: list

control_freq¶

frequency at which controller sends to robot.

Type: float

position_limits¶

Limits (m) below and above which the magnitude of a calculated goal eef position will be clipped. Can be either be a 2-list (same min/max value for all cartesian dims), or a 2-list of list (specific min/max values for each dim)

Type: 2-list of float or 2-list of list of floats

orientation_limits¶

[Lower_bounds, upper_bounds]. Limits (rad) below and above which the magnitude of a calculated goal eef orientation will be clipped. Can be either be a 2-list (same min/max value for all joint dims), or a 2-list of list (specific min/mx values for each dim)

Type: 2-list of float or 2-list of list of floats

interpolator_pos¶

Interpolator object to be used for interpolating from the current position to the goal position during each timestep between inputted actions

Type: Interpolator

interpolator_ori¶

Interpolator object to be used for interpolating from the current orientation to the goal orientation during each timestep between inputted actions

Type: Interpolator

control_ori¶

Whether inputted actions will control both pos and ori or exclusively pos

Type: bool

uncouple_pos_ori¶

Whether to decouple torques meant to control pos and torques meant to control ori

Type: bool

**kwargs: Does nothing; placeholder to “sink” any additional arguments so that instantiating this controller via an argument dict that has additional extraneous arguments won’t raise an error

set_goal(delta, set_pos=None, set_ori=None, **kwargs)¶

Set goal for controller.

Parameters

delta (list) – (6f) list of deltas from current position [dx, dy, dz, ax, ay, az]. Deltas expressed as position and axis-angle in orientation.
set_pos (list) – (3f) fixed position goal. [x, y, z] in world frame. Only used if delta is None.
set_ori (list) – (4f) fixed orientation goal as quaternion in world frame. Only used if delta is None.
kwargs (dict) – additional keyword arguments.

Returns

None

Examples::

$ self.controller.set_goal(delta=[0.1, 0.1, 0, 0, 0, 0], set_pos=None, set_ori=None)

$ self.controller.set_goal(delta=None, set_pose=[0.4, 0.2, 0.4], set_ori=[0, 0, 0, 1])

run_controller()¶

Run controller to calculate torques to acheive desired pose.

Args: None :returns: 7f list of torques to command. :rtype: torques (list)

Runs impedance controllers and adds nullspace constrained posture compensation.