placo::dynamics#

class placo.AvoidSelfCollisionsDynamicsConstraint#

configure(name: str, priority: any, weight: float = 1.0) → None#

Configures the object.

Parameters:

name – task name
priority – task priority (hard, soft or scaled)
weight – task weight

name: str#: Object name.

priority: any#: Object priority.

self_collisions_margin: float#: Margin for self collisions [m].

self_collisions_trigger: float#: Distance that triggers the constraint [m].

class placo.Contact#

active: bool#: true if the contact is active (ignored by the solver else, this allow to enable/disable a contact without removing it from the solver)

mu: float#: Coefficient of friction (if relevant)

weight_forces: float#: Weight of forces for the optimization (if relevant)

weight_moments: float#: Weight of moments for optimization (if relevant)

wrench: ndarray#: Wrench populated after the DynamicsSolver::solve call.

class placo.Contact6D(frame_task: DynamicsFrameTask, unilateral: bool)#

see DynamicsSolver::add_fixed_planar_contact and DynamicsSolver::add_unilateral_planar_contact

active: bool#: true if the contact is active (ignored by the solver else, this allow to enable/disable a contact without removing it from the solver)

length: float#: Rectangular contact length along local x-axis.

mu: float#: Coefficient of friction (if relevant)

orientation_task() → DynamicsOrientationTask#: Associated orientation task.

position_task() → DynamicsPositionTask#: Associated position task.

unilateral: bool#: true for unilateral contact with the ground

weight_forces: float#: Weight of forces for the optimization (if relevant)

weight_moments: float#: Weight of moments for optimization (if relevant)

width: float#: Rectangular contact width along local y-axis.

wrench: ndarray#: Wrench populated after the DynamicsSolver::solve call.

zmp() → ndarray#

Returns the contact ZMP in the local frame.

Returns:: zmp

class placo.DynamicsCoMTask(arg2: ndarray)#

A: ndarray#: A matrix in Ax = b, where x is the accelerations.

b: ndarray#: b vector in Ax = b, where x is the accelerations

configure(name: str, priority: any, weight: float = 1.0) → None#

Configures the object.

Parameters:

name – task name
priority – task priority (hard, soft or scaled)
weight – task weight

critically_damped: bool#: If this is true, kd will be computed from kp to have a critically damped system.

ddtarget_world: ndarray#: Target acceleration in the world.

derror: ndarray#: Current velocity error vector.

dtarget_world: ndarray#: Target velocity to reach in robot frame.

error: ndarray#: Current error vector.

kd: float#: D gain for position control.

kp: float#: K gain for position control.

mask: AxisesMask#: Axises mask.

name: str#: Object name.

priority: any#: Object priority.

target_world: ndarray#: Target to reach in world frame.

class placo.DynamicsConstraint#

configure(name: str, priority: any, weight: float = 1.0) → None#

Configures the object.

Parameters:

name – task name
priority – task priority (hard, soft or scaled)
weight – task weight

name: str#: Object name.

priority: any#: Object priority.

class placo.DynamicsFrameTask#

T_world_frame: any#

configure(name: str, priority: str = 'soft', position_weight: float = 1.0, orientation_weight: float = 1.0) → None#

Configures the frame task.

Parameters:

name – task name
priority – task priority
position_weight – weight for the position task
orientation_weight – weight for the orientation task

orientation() → DynamicsOrientationTask#

position() → DynamicsPositionTask#

class placo.DynamicsGearTask#

A: ndarray#: A matrix in Ax = b, where x is the accelerations.

add_gear(target: str, source: str, ratio: float) → None#

Adds a gear constraint, you can add multiple source for the same target, they will be summed.

Parameters:

target – target joint
source – source joint
ratio – ratio

b: ndarray#: b vector in Ax = b, where x is the accelerations

configure(name: str, priority: any, weight: float = 1.0) → None#

Configures the object.

Parameters:

name – task name
priority – task priority (hard, soft or scaled)
weight – task weight

critically_damped: bool#: If this is true, kd will be computed from kp to have a critically damped system.

derror: ndarray#: Current velocity error vector.

error: ndarray#: Current error vector.

kd: float#: D gain for position control.

kp: float#: K gain for position control.

name: str#: Object name.

priority: any#: Object priority.

set_gear(target: str, source: str, ratio: float) → None#

Sets a gear constraint.

Parameters:

target – target joint
source – source joint
ratio – ratio

class placo.DynamicsJointsTask#

A: ndarray#: A matrix in Ax = b, where x is the accelerations.

b: ndarray#: b vector in Ax = b, where x is the accelerations

configure(name: str, priority: any, weight: float = 1.0) → None#

Configures the object.

Parameters:

name – task name
priority – task priority (hard, soft or scaled)
weight – task weight

critically_damped: bool#: If this is true, kd will be computed from kp to have a critically damped system.

derror: ndarray#: Current velocity error vector.

error: ndarray#: Current error vector.

kd: float#: D gain for position control.

kp: float#: K gain for position control.

name: str#: Object name.

priority: any#: Object priority.

set_joint(joint: str, target: float, velocity: float = 0.0, acceleration: float = 0.0) → None#

Sets the target for a given joint.

Parameters:

joint – joint name
target – target position
velocity – target velocity
acceleration – target acceleration

set_joints(arg2: dict) → None#

set_joints_velocities(arg2: dict) → None#

class placo.DynamicsOrientationTask(arg2: int, arg3: ndarray)#

A: ndarray#: A matrix in Ax = b, where x is the accelerations.

R_world_frame: ndarray#: Target orientation.

b: ndarray#: b vector in Ax = b, where x is the accelerations

configure(name: str, priority: any, weight: float = 1.0) → None#

Configures the object.

Parameters:

name – task name
priority – task priority (hard, soft or scaled)
weight – task weight

critically_damped: bool#: If this is true, kd will be computed from kp to have a critically damped system.

derror: ndarray#: Current velocity error vector.

domega_world: ndarray#: Target angular acceleration.

error: ndarray#: Current error vector.

kd: float#: D gain for position control.

kp: float#: K gain for position control.

mask: AxisesMask#: Mask.

name: str#: Object name.

omega_world: ndarray#: Target angular velocity.

priority: any#: Object priority.

class placo.DynamicsPositionTask(arg2: int, arg3: ndarray)#

A: ndarray#: A matrix in Ax = b, where x is the accelerations.

b: ndarray#: b vector in Ax = b, where x is the accelerations

configure(name: str, priority: any, weight: float = 1.0) → None#

Configures the object.

Parameters:

name – task name
priority – task priority (hard, soft or scaled)
weight – task weight

critically_damped: bool#: If this is true, kd will be computed from kp to have a critically damped system.

derror: ndarray#: Current velocity error vector.

dtarget_world: ndarray#: Target velocity in the world.

error: ndarray#: Current error vector.

frame_index: any#: Frame.

kd: float#: D gain for position control.

kp: float#: K gain for position control.

mask: AxisesMask#: Mask.

name: str#: Object name.

priority: any#: Object priority.

target_world: ndarray#: Target position in the world.

class placo.DynamicsReactionRatioConstraint(arg2: Contact, arg3: float)#

configure(name: str, priority: any, weight: float = 1.0) → None#

Configures the object.

Parameters:

name – task name
priority – task priority (hard, soft or scaled)
weight – task weight

name: str#: Object name.

priority: any#: Object priority.

reaction_ratio: float#: Reaction ratio to be enforced.

class placo.DynamicsRelativeFrameTask#

T_a_b: any#

configure(name: str, priority: str = 'soft', position_weight: float = 1.0, orientation_weight: float = 1.0) → None#

Configures the relative frame task.

Parameters:

name – task name
priority – task priority
position_weight – weight for the position task
orientation_weight – weight for the orientation task

orientation() → DynamicsRelativeOrientationTask#: Orientation task.

position() → DynamicsRelativePositionTask#: Position task.

class placo.DynamicsRelativeOrientationTask(arg2: int, arg3: int, arg4: ndarray)#

A: ndarray#: A matrix in Ax = b, where x is the accelerations.

R_a_b: ndarray#: Target relative orientation.

b: ndarray#: b vector in Ax = b, where x is the accelerations

configure(name: str, priority: any, weight: float = 1.0) → None#

Configures the object.

Parameters:

name – task name
priority – task priority (hard, soft or scaled)
weight – task weight

critically_damped: bool#: If this is true, kd will be computed from kp to have a critically damped system.

derror: ndarray#: Current velocity error vector.

domega_a_b: ndarray#: Target relative angular velocity.

error: ndarray#: Current error vector.

kd: float#: D gain for position control.

kp: float#: K gain for position control.

mask: AxisesMask#: Mask.

name: str#: Object name.

omega_a_b: ndarray#: Target relative angular velocity.

priority: any#: Object priority.

class placo.DynamicsRelativePositionTask(arg2: int, arg3: int, arg4: ndarray)#

A: ndarray#: A matrix in Ax = b, where x is the accelerations.

b: ndarray#: b vector in Ax = b, where x is the accelerations

configure(name: str, priority: any, weight: float = 1.0) → None#

Configures the object.

Parameters:

name – task name
priority – task priority (hard, soft or scaled)
weight – task weight

critically_damped: bool#: If this is true, kd will be computed from kp to have a critically damped system.

ddtarget: ndarray#: Target relative velocity.

derror: ndarray#: Current velocity error vector.

dtarget: ndarray#: Target relative velocity.

error: ndarray#: Current error vector.

kd: float#: D gain for position control.

kp: float#: K gain for position control.

mask: AxisesMask#: Mask.

name: str#: Object name.

priority: any#: Object priority.

target: ndarray#: Target relative position.

class placo.DynamicsSolver(robot: RobotWrapper)#

add_avoid_self_collisions_constraint() → AvoidSelfCollisionsDynamicsConstraint#

Adds a constraint to the solver.

Returns:: constraint

add_com_task(target_world: ndarray) → DynamicsCoMTask#

Adds a center of mass (in the world) task.

Parameters:: target_world – target (in the world)
Returns:: center of mass task

add_constraint(constraint: any) → any#

Adds a constraint to the solver.

Parameters:: constraint – constraint
Returns:: reference to internal constraint

add_external_wrench_contact(frame_name: str) → ExternalWrenchContact#

Adds an external wrench.

Parameters:: frame_name – frame
Returns:: external wrench contact

add_fixed_contact(frame_task: DynamicsFrameTask) → Contact6D#

Adds a fixed contact.

Parameters:: frame_task – the associated frame task
Returns:: fixed contact

add_frame_task(frame_name: str, T_world_frame: ndarray) → DynamicsFrameTask#

Adds a frame task, which is a pseudo-task packaging position and orientation, resulting in a decoupled task.

Parameters:

frame_index – target frame
T_world_frame – target transformation in the world

Returns:

frame task

add_gear_task() → DynamicsGearTask#

Adds a gear task, allowing replication of a joint. This can be used to implement timing belt, if coupled with an internal force.

Returns:: gear task

add_joints_task() → DynamicsJointsTask#

Adds a joints task.

Returns:: joints task

add_orientation_task(frame_name: str, R_world_frame: ndarray) → DynamicsOrientationTask#

Adds an orientation (in the world) task.

Parameters:

frame_index – target frame
R_world_frame – target world orientation

Returns:

orientation task

add_planar_contact(frame_task: DynamicsFrameTask) → Contact6D#

Adds a planar contact, which is unilateral in the sense of the local body z-axis.

Parameters:: frame_task – associated frame task
Returns:: planar contact

add_point_contact(position_task: DynamicsPositionTask) → PointContact#

Adds a point contact.

Parameters:: position_task – the associated position task
Returns:: point contact

add_position_task(frame_name: str, target_world: ndarray) → DynamicsPositionTask#

Adds a position (in the world) task.

Parameters:

frame_index – target frame
target_world – target position in the world

Returns:

position task

add_puppet_contact() → PuppetContact#

Adds a puppet contact, this will add some free contact forces for the whole system, allowing it to be controlled freely.

Returns:: puppet contact

add_reaction_ratio_constraint(contact: Contact, reaction_ratio: float) → DynamicsReactionRatioConstraint#

Adds a constraint enforce reaction ratio.

Parameters:

contact – contact
reaction_ratio – reaction ratio to enforce

Returns:

reaction ratio constraint

add_relative_fixed_contact(frame_task: DynamicsRelativeFrameTask) → Relative6DContact#

Adds a relative fixed contact, can be used for fixed closed loops.

Parameters:: frame_task – the associated relative frame task
Returns:: relative fixed contact

add_relative_frame_task(frame_a_name: str, frame_b_name: str, T_a_b: ndarray) → DynamicsRelativeFrameTask#

Adds a relative frame task, which is a pseudo-task packaging relative position and orientation tasks.

Parameters:

frame_a_index – frame a
frame_b_index – frame b
T_a_b – target transformation value for b frame in a

Returns:

relative frame task

add_relative_orientation_task(frame_a_name: str, frame_b_name: str, R_a_b: ndarray) → DynamicsRelativeOrientationTask#

Adds a relative orientation task.

Parameters:

frame_a_index – frame a
frame_b_index – frame b
R_a_b – target value for the orientation of b frame in a

Returns:

relative orientation task

add_relative_point_contact(position_task: DynamicsRelativePositionTask) → RelativePointContact#

Adds a relative point contact, which can be typically used for internal forces like loop-closing.

Parameters:: position_task – associated relative position task
Returns:: relative point contact

add_relative_position_task(frame_a_name: str, frame_b_name: str, target_world: ndarray) → DynamicsRelativePositionTask#

Adds a relative position task.

Parameters:

frame_a_index – frame a
frame_b_index – frame b
target – target value for AB vector, expressed in A

Returns:

relative position task

add_task(task: any) → any#

Adds a task to the solver.

Parameters:: task – task
Returns:: reference to internal task

add_task_contact(task: DynamicsTask) → TaskContact#

Adds contact forces associated with any given task.

Parameters:: task – task
Returns:: task contact

add_torque_task() → DynamicsTorqueTask#

Adds a torque task.

Returns:: torque task

add_unilateral_point_contact(position_task: DynamicsPositionTask) → PointContact#

Adds an unilateral point contact, in the sense of the world z-axis.

Parameters:: position_task – the associated position task
Returns:: unilateral point contact

clear() → None#: Clears the internal tasks.

count_contacts() → int#

dt: float#: Solver dt (seconds)

dump_status() → None#: Shows the tasks status.

enable_joint_limits(enable: bool) → None#: Enables/disables joint limits inequalities.

enable_torque_limits(enable: bool) → None#: Enables/disables torque limits inequalities.

enable_velocity_limits(enable: bool) → None#: Enables/disables joint velocity inequalities.

enable_velocity_vs_torque_limits(enable: bool) → None#: Enables the velocity vs torque inequalities.

friction: float#: Global friction that is added to all the joints.

get_contact(arg2: int) → Contact#

gravity_only: bool#: Use gravity only (no coriolis, no centrifugal)

mask_fbase(masked: bool) → None#: Decides if the floating base should be masked.

problem: Problem#: Instance of the problem.

qdd_safe: float#: The value of qdd safe.

remove_constraint(constraint: DynamicsConstraint) → None#

Removes a constraint from the solver.

Parameters:: constraint – constraint

remove_contact(contact: Contact) → None#

Removes a contact from the solver.

Parameters:: contact

remove_task(task: DynamicsFrameTask) → None#

Removes a frame task from the solver.

Parameters:: task – frame task

reset_joint(joint_name: str) → None#

Resets a given joint so that its torque is no longer overriden.

Parameters:: joint_name – the joint

robot: RobotWrapper#

set_passive(joint_name: str, kp: float = 0.0, kd: float = 0.0) → None#

Sets a DoF as passive, the corresponding tau will be fixed in the equation of motion it can be purely passive joint or a spring-like behaviour.

Parameters:

joint_name – the joint
is_passive – true if the should the joint be passive
kp – kp gain if the joint is a spring (0 by default)
kd – kd gain if the joint is a spring (0 by default)

set_tau(joint_name: str, tau: float) → None#

Sets a custom torque to be applied by a given joint.

Parameters:

joint_name – the joint
tau – torque

solve(integrate: bool = False) → DynamicsSolverResult#

class placo.DynamicsSolverResult#

qdd: ndarray#

success: bool#

tau: ndarray#

tau_dict(arg2: RobotWrapper) → dict#

class placo.DynamicsTask#

A: ndarray#: A matrix in Ax = b, where x is the accelerations.

b: ndarray#: b vector in Ax = b, where x is the accelerations

configure(name: str, priority: any, weight: float = 1.0) → None#

Configures the object.

Parameters:

name – task name
priority – task priority (hard, soft or scaled)
weight – task weight

critically_damped: bool#: If this is true, kd will be computed from kp to have a critically damped system.

derror: ndarray#: Current velocity error vector.

error: ndarray#: Current error vector.

kd: float#: D gain for position control.

kp: float#: K gain for position control.

name: str#: Object name.

priority: any#: Object priority.

class placo.DynamicsTorqueTask#

A: ndarray#: A matrix in Ax = b, where x is the accelerations.

b: ndarray#: b vector in Ax = b, where x is the accelerations

configure(name: str, priority: any, weight: float = 1.0) → None#

Configures the object.

Parameters:

name – task name
priority – task priority (hard, soft or scaled)
weight – task weight

critically_damped: bool#: If this is true, kd will be computed from kp to have a critically damped system.

derror: ndarray#: Current velocity error vector.

error: ndarray#: Current error vector.

kd: float#: D gain for position control.

kp: float#: K gain for position control.

name: str#: Object name.

priority: any#: Object priority.

set_torque(joint: str, torque: float) → None#

Sets the target for a given joint.

Parameters:

joint – joint name
torque – target torque

class placo.ExternalWrenchContact(frame_index: any)#

see DynamicsSolver::add_external_wrench_contact

active: bool#: true if the contact is active (ignored by the solver else, this allow to enable/disable a contact without removing it from the solver)

frame_index: any#

mu: float#: Coefficient of friction (if relevant)

w_ext: ndarray#

weight_forces: float#: Weight of forces for the optimization (if relevant)

weight_moments: float#: Weight of moments for optimization (if relevant)

wrench: ndarray#: Wrench populated after the DynamicsSolver::solve call.

class placo.PointContact(position_task: DynamicsPositionTask, unilateral: bool)#

see DynamicsSolver::add_point_contact and DynamicsSolver::add_unilateral_point_contact

active: bool#: true if the contact is active (ignored by the solver else, this allow to enable/disable a contact without removing it from the solver)

mu: float#: Coefficient of friction (if relevant)

position_task() → DynamicsPositionTask#: associated position task

unilateral: bool#: true for unilateral contact with the ground

weight_forces: float#: Weight of forces for the optimization (if relevant)

weight_moments: float#: Weight of moments for optimization (if relevant)

wrench: ndarray#: Wrench populated after the DynamicsSolver::solve call.

class placo.PuppetContact#

see DynamicsSolver::add_puppet_contact

active: bool#: true if the contact is active (ignored by the solver else, this allow to enable/disable a contact without removing it from the solver)

mu: float#: Coefficient of friction (if relevant)

weight_forces: float#: Weight of forces for the optimization (if relevant)

weight_moments: float#: Weight of moments for optimization (if relevant)

wrench: ndarray#: Wrench populated after the DynamicsSolver::solve call.

class placo.Relative6DContact(frame_task: DynamicsRelativeFrameTask)#

see DynamicsSolver::add_relative_fixed_contact

active: bool#: true if the contact is active (ignored by the solver else, this allow to enable/disable a contact without removing it from the solver)

mu: float#: Coefficient of friction (if relevant)

weight_forces: float#: Weight of forces for the optimization (if relevant)

weight_moments: float#: Weight of moments for optimization (if relevant)

wrench: ndarray#: Wrench populated after the DynamicsSolver::solve call.

class placo.RelativePointContact(position_task: DynamicsRelativePositionTask)#

see DynamicsSolver::add_relative_point_contact

active: bool#: true if the contact is active (ignored by the solver else, this allow to enable/disable a contact without removing it from the solver)

mu: float#: Coefficient of friction (if relevant)

weight_forces: float#: Weight of forces for the optimization (if relevant)

weight_moments: float#: Weight of moments for optimization (if relevant)

wrench: ndarray#: Wrench populated after the DynamicsSolver::solve call.

class placo.TaskContact(task: DynamicsTask)#

see DynamicsSolver::add_task_contact

active: bool#: true if the contact is active (ignored by the solver else, this allow to enable/disable a contact without removing it from the solver)

mu: float#: Coefficient of friction (if relevant)

weight_forces: float#: Weight of forces for the optimization (if relevant)

weight_moments: float#: Weight of moments for optimization (if relevant)

wrench: ndarray#: Wrench populated after the DynamicsSolver::solve call.