class FF:
def __init__(self):
robot = self.robot = loadTalosArm(freeFloating=True)
rmodel = self.rmodel = robot.model
qmin = rmodel.lowerPositionLimit
qmin[:7] = -1
rmodel.lowerPositionLimit = qmin
qmax = rmodel.upperPositionLimit
qmax[:7] = 1
rmodel.upperPositionLimit = qmax
State = self.State = StatePinocchio(rmodel)
actModel = self.actModel = ActuationModelFreeFloating(rmodel)
contactModel = self.contactModel = ContactModelMultiple(rmodel)
contact6 = ContactModel6D(rmodel,
rmodel.getFrameId(contactName),
ref=pinocchio.SE3.Identity(),
gains=[0., 0.])
contactModel.addContact(name='contact', contact=contact6)
costModel = self.costModel = CostModelSum(rmodel, nu=actModel.nu)
self.cost1 = CostModelFrameTranslation(
rmodel,
nu=actModel.nu,
frame=rmodel.getFrameId(opPointName),
ref=np.array([.5, .4, .3]))
stateWeights = np.array([0] * 6 + [0.01] * (rmodel.nv - 6) +
[10] * rmodel.nv)
self.cost2 = CostModelState(rmodel,
State,
ref=State.zero(),
nu=actModel.nu,
activation=ActivationModelWeightedQuad(
stateWeights**2))
self.cost3 = CostModelControl(rmodel, nu=actModel.nu)
costModel.addCost(name="pos", weight=10, cost=self.cost1)
costModel.addCost(name="regx", weight=0.1, cost=self.cost2)
costModel.addCost(name="regu", weight=0.01, cost=self.cost3)
self.dmodel = DifferentialActionModelFloatingInContact(
rmodel, actModel, contactModel, costModel)
self.model = IntegratedActionModelEuler(self.dmodel)
self.data = self.model.createData()
self.cd1 = self.data.differential.costs.costs['pos']
self.cd2 = self.data.differential.costs.costs['regx']
self.cd3 = self.data.differential.costs.costs['regu']
self.ddata = self.data.differential
self.rdata = self.data.differential.pinocchio
self.x = self.State.rand()
self.q = a2m(self.x[:rmodel.nq])
self.v = a2m(self.x[rmodel.nq:])
self.u = np.random.rand(self.model.nu)
def calc(self, x=None, u=None):
return self.model.calc(self.data, x if x is not None else self.x,
u if u is not None else self.u)
def runningModel(contactIds, effectors, com=None, integrationStep=1e-2):
'''
Creating the action model for floating-base systems. A walker system
is by default a floating-base system.
contactIds is a list of frame Ids of points that should be in contact.
effectors is a dict of key frame ids and SE3 values of effector references.
'''
actModel = ActuationModelFreeFloating(rmodel)
State = StatePinocchio(rmodel)
# Creating a 6D multi-contact model, and then including the supporting foot
contactModel = ContactModelMultiple(rmodel)
for cid in contactIds:
contactModel.addContact('contact%d' % cid,
ContactModel6D(rmodel, cid, ref=None))
# Creating the cost model for a contact phase
costModel = CostModelSum(rmodel, actModel.nu)
wx = np.array([0] * 6 + [.1] * (rmodel.nv - 6) + [10] * rmodel.nv)
costModel.addCost('xreg',
weight=1e-1,
cost=CostModelState(
rmodel,
State,
ref=rmodel.defaultState,
nu=actModel.nu,
activation=ActivationModelWeightedQuad(wx)))
costModel.addCost('ureg',
weight=1e-4,
cost=CostModelControl(rmodel, nu=actModel.nu))
for fid, ref in effectors.items():
if not isinstance(ref, SE3):
ref = SE3(eye(3), a2m(ref))
costModel.addCost("track%d" % fid,
weight=100.,
cost=CostModelFramePlacement(rmodel, fid, ref,
actModel.nu))
if com is not None:
costModel.addCost("com",
weight=100.,
cost=CostModelCoM(rmodel, ref=com, nu=actModel.nu))
# Creating the action model for the KKT dynamics with simpletic Euler
# integration scheme
dmodel = DifferentialActionModelFloatingInContact(rmodel, actModel,
contactModel, costModel)
model = IntegratedActionModelEuler(dmodel)
model.timeStep = integrationStep
return model
class Fix:
def __init__(self):
robot = self.robot = loadTalosArm()
rmodel = self.rmodel = robot.model
rmodel.armature = np.matrix([0] * rmodel.nv).T
for j in rmodel.joints[1:]:
if j.shortname() != 'JointModelFreeFlyer':
rmodel.armature[j.idx_v:j.idx_v + j.nv] = 1
State = self.State = StatePinocchio(rmodel)
self.cost1 = CostModelFrameTranslation(
rmodel,
nu=rmodel.nv,
frame=rmodel.getFrameId(opPointName),
ref=np.array([.5, .4, .3]))
self.cost2 = CostModelState(rmodel,
State,
ref=State.zero(),
nu=rmodel.nv)
self.cost3 = CostModelControl(rmodel, nu=rmodel.nv)
costModel = CostModelSum(rmodel)
costModel.addCost(name="pos", weight=10, cost=self.cost1)
costModel.addCost(name="regx", weight=0.1, cost=self.cost2)
costModel.addCost(name="regu", weight=0.01, cost=self.cost3)
self.dmodel = DifferentialActionModelFullyActuated(rmodel, costModel)
self.model = IntegratedActionModelEuler(self.dmodel)
self.data = self.model.createData()
self.cd1 = self.data.differential.costs.costs['pos']
self.cd2 = self.data.differential.costs.costs['regx']
self.cd3 = self.data.differential.costs.costs['regu']
self.ddata = self.data.differential
self.rdata = self.data.differential.pinocchio
self.x = self.State.rand()
self.q = a2m(self.x[:rmodel.nq])
self.v = a2m(self.x[rmodel.nq:])
self.u = np.random.rand(self.model.nu)
def calc(self, x=None, u=None):
return self.model.calc(self.data, x if x is not None else self.x,
u if u is not None else self.u)
def createModels(self, timeStep, supportFootIds, comTask=None):
# Creating the action model for floating-base systems
actModel = ActuationModelFreeFloating(self.rmodel)
# Creating a 3D multi-contact model, and then including the supporting
# feet
contactModel = ContactModelMultiple(self.rmodel)
for i in supportFootIds:
supportContactModel = ContactModel3D(self.rmodel,
i,
ref=[0., 0., 0.],
gains=[0., 0.])
contactModel.addContact('contact_' + str(i), supportContactModel)
# Creating the cost model for a contact phase
costModel = CostModelSum(self.rmodel, actModel.nu)
# CoM tracking cost
if isinstance(comTask, np.ndarray):
comTrack = CostModelCoM(self.rmodel, comTask, actModel.nu)
costModel.addCost("comTrack", comTrack, 1e2)
# State and control regularization
stateWeights = np.array([0] * 6 + [0.01] * (self.rmodel.nv - 6) +
[10] * self.rmodel.nv)
stateReg = CostModelState(self.rmodel, self.state,
self.rmodel.defaultState, actModel.nu,
ActivationModelWeightedQuad(stateWeights**2))
ctrlReg = CostModelControl(self.rmodel, actModel.nu)
costModel.addCost("stateReg", stateReg, 1e-1)
costModel.addCost("ctrlReg", ctrlReg, 1e-4)
# Creating the action model for the KKT dynamics with simpletic Euler
# integration scheme
dmodel = DifferentialActionModelFloatingInContact(
self.rmodel, actModel, contactModel, costModel)
model = IntegratedActionModelEuler(dmodel)
model.timeStep = timeStep
return model
def createMultiphaseShootingProblem(rmodel, rdata, csw, timeStep):
"""
Create a Multiphase Shooting problem from the output of the centroidal solver.
:params rmodel: robot model of type pinocchio::model
:params rdata: robot data of type pinocchio::data
:params csw: contact sequence wrapper of type ContactSequenceWrapper
:params timeStep: Scalar timestep between nodes.
:returns list of IntegratedActionModels
"""
# -----------------------
# Define Cost weights
class Weights:
com = 1e1
regx = 1e-3
regu = 0.
swing_patch = 1e6
forces = 0.
contactv = 1e3
# Define state cost vector for WeightedActivation
ff_orientation = 1e1
xweight = np.array([0] * 3 + [ff_orientation] * 3 + [1.] * (rmodel.nv - 6) + [1.] * rmodel.nv)
xweight[range(18, 20)] = ff_orientation
# for patch in swing_patch: w.swing_patch.append(100.);
# Define weights for the impact costs.
imp_state = 1e2
imp_com = 1e2
imp_contact_patch = 1e6
imp_act_com = m2a([0.1, 0.1, 3.0])
# Define weights for the terminal costs
term_com = 1e8
term_regx = 1e4
w = Weights()
# ------------------------
problem_models = []
actuationff = ActuationModelFreeFloating(rmodel)
State = StatePinocchio(rmodel)
active_contact_patch = set()
active_contact_patch_prev = set()
for nphase, phase in enumerate(csw.cs.contact_phases):
t0 = phase.time_trajectory[0]
t1 = phase.time_trajectory[-1]
N = int(round((t1 - t0) / timeStep)) + 1
contact_model = ContactModelMultiple(rmodel)
# Add contact constraints for the active contact patches.
# Add SE3 cost for the non-active contact patches.
swing_patch = []
active_contact_patch_prev = active_contact_patch.copy()
active_contact_patch.clear()
for patch in csw.ee_map.keys():
if getattr(phase, patch).active:
active_contact_patch.add(patch)
active_contact = ContactModel6D(rmodel,
frame=rmodel.getFrameId(csw.ee_map[patch]),
ref=getattr(phase, patch).placement)
contact_model.addContact(patch, active_contact)
# print nphase, "Contact ",patch," added at ", getattr(phase,patch).placement.translation.T
else:
swing_patch.append(patch)
# Check if contact has been added in this phase. If this phase is not zero,
# add an impulse model to deal with this contact.
new_contacts = active_contact_patch.difference(active_contact_patch_prev)
if nphase != 0 and len(new_contacts) != 0:
# print nphase, "Impact ",[p for p in new_contacts]," added"
imp_model = ImpulseModelMultiple(
rmodel, {
"Impulse_" + patch: ImpulseModel6D(rmodel, frame=rmodel.getFrameId(csw.ee_map[patch]))
for patch in new_contacts
})
# Costs for the impulse of a new contact
cost_model = CostModelSum(rmodel, nu=0)
# State
cost_regx = CostModelState(rmodel,
State,
ref=rmodel.defaultState,
nu=actuationff.nu,
activation=ActivationModelWeightedQuad(w.xweight))
cost_model.addCost("imp_regx", cost_regx, w.imp_state)
# CoM
cost_com = CostModelImpactCoM(rmodel, activation=ActivationModelWeightedQuad(w.imp_act_com))
cost_model.addCost("imp_CoM", cost_com, w.imp_com)
# Contact Frameplacement
for patch in new_contacts:
cost_contact = CostModelFramePlacement(rmodel,
frame=rmodel.getFrameId(csw.ee_map[patch]),
ref=SE3(np.identity(3), csw.ee_splines[patch].eval(t0)[0]),
nu=actuationff.nu)
cost_model.addCost("imp_contact_" + patch, cost_contact, w.imp_contact_patch)
imp_action_model = ActionModelImpact(rmodel, imp_model, cost_model)
problem_models.append(imp_action_model)
# Define the cost and action models for each timestep in the contact phase.
# untill [:-1] because in contact sequence timetrajectory, the end-time is
# also included. e.g., instead of being [0.,0.5], time trajectory is [0,0.5,1.]
for t in np.linspace(t0, t1, N)[:-1]:
cost_model = CostModelSum(rmodel, actuationff.nu)
# For the first node of the phase, add cost v=0 for the contacting foot.
if t == 0:
for patch in active_contact_patch:
cost_vcontact = CostModelFrameVelocity(rmodel,
frame=rmodel.getFrameId(csw.ee_map[patch]),
ref=m2a(Motion.Zero().vector),
nu=actuationff.nu)
cost_model.addCost("contactv_" + patch, cost_vcontact, w.contactv)
# CoM Cost
cost_com = CostModelCoM(rmodel, ref=m2a(csw.phi_c.com_vcom.eval(t)[0][:3, :]), nu=actuationff.nu)
cost_model.addCost("CoM", cost_com, w.com)
# Forces Cost
for patch in contact_model.contacts.keys():
cost_force = CostModelForce(rmodel,
contactModel=contact_model.contacts[patch],
ref=m2a(csw.phi_c.forces[patch].eval(t)[0]),
nu=actuationff.nu)
cost_model.addCost("forces_" + patch, cost_force, w.forces)
# Swing patch cost
for patch in swing_patch:
cost_swing = CostModelFramePlacement(rmodel,
frame=rmodel.getFrameId(csw.ee_map[patch]),
ref=SE3(np.identity(3), csw.ee_splines[patch].eval(t)[0]),
nu=actuationff.nu)
cost_model.addCost("swing_" + patch, cost_swing, w.swing_patch)
# print t, "Swing cost ",patch," added at ", csw.ee_splines[patch].eval(t)[0][:3].T
# State Regularization
cost_regx = CostModelState(rmodel,
State,
ref=rmodel.defaultState,
nu=actuationff.nu,
activation=ActivationModelWeightedQuad(w.xweight))
cost_model.addCost("regx", cost_regx, w.regx)
# Control Regularization
cost_regu = CostModelControl(rmodel, nu=actuationff.nu)
cost_model.addCost("regu", cost_regu, w.regu)
dmodel = DifferentialActionModelFloatingInContact(rmodel, actuationff, contact_model, cost_model)
imodel = IntegratedActionModelEuler(dmodel, timeStep=timeStep)
problem_models.append(imodel)
# Create Terminal Model.
contact_model = ContactModelMultiple(rmodel)
# Add contact constraints for the active contact patches.
swing_patch = []
t = t1
for patch in csw.ee_map.keys():
if getattr(phase, patch).active:
active_contact = ContactModel6D(rmodel,
frame=rmodel.getFrameId(csw.ee_map[patch]),
ref=getattr(phase, patch).placement)
contact_model.addContact(patch, active_contact)
cost_model = CostModelSum(rmodel, actuationff.nu)
# CoM Cost
cost_com = CostModelCoM(rmodel, ref=m2a(csw.phi_c.com_vcom.eval(t)[0][:3, :]), nu=actuationff.nu)
cost_model.addCost("CoM", cost_com, w.term_com)
# State Regularization
cost_regx = CostModelState(rmodel,
State,
ref=rmodel.defaultState,
nu=actuationff.nu,
activation=ActivationModelWeightedQuad(w.xweight))
cost_model.addCost("regx", cost_regx, w.term_regx)
dmodel = DifferentialActionModelFloatingInContact(rmodel, actuationff, contact_model, cost_model)
imodel = IntegratedActionModelEuler(dmodel)
problem_models.append(imodel)
problem_models.append
return problem_models
dnum = mnum.createData()
model.calcDiff(data, x, u)
mnum.calcDiff(dnum, x, u)
assertNumDiff(data.Fx, dnum.Fx,
NUMDIFF_MODIFIER * mnum.disturbance) # threshold was 1e-4, is now 2.11e-4 (see assertNumDiff.__doc__)
assertNumDiff(data.Fu, dnum.Fu,
NUMDIFF_MODIFIER * mnum.disturbance) # threshold was 1e-4, is now 2.11e-4 (see assertNumDiff.__doc__)
assertNumDiff(data.Rx, dnum.Rx,
NUMDIFF_MODIFIER * mnum.disturbance) # threshold was 1e-4, is now 2.11e-4 (see assertNumDiff.__doc__)
assertNumDiff(data.Ru, dnum.Ru,
NUMDIFF_MODIFIER * mnum.disturbance) # threshold was 1e-4, is now 2.11e-4 (see assertNumDiff.__doc__)
dmodel = DifferentialActionModelPositioning(rmodel)
ddata = dmodel.createData()
model = IntegratedActionModelEuler(dmodel)
data = model.createData()
x = model.State.zero()
u = np.zeros(model.nu)
xn, c = model.calc(data, x, u)
model.timeStep = 1
model.differential.costs
for k in model.differential.costs.costs.keys():
model.differential.costs[k].weight = 1
model.calcDiff(data, x, u)
mnum = ActionModelNumDiff(model, withGaussApprox=True)
dnum = mnum.createData()
SolverBoxKKT, SolverDDP, SolverKKT) from crocoddyl.qpsolvers import quadprogWrapper # --- TEST DDP vs KKT LQR --- # --------------------------------------------------- # --------------------------------------------------- # --------------------------------------------------- np.set_printoptions(linewidth=400, suppress=True) np.random.seed(220) nq = 4 nu = 2 nv = nq dmodel = DifferentialActionModelLQR(nq, nu) model = IntegratedActionModelEuler(dmodel, withCostResiduals=False) nx = model.nx nu = model.nu T = 1 problem = ShootingProblem(model.State.zero() + 2, [model] * T, model) xs = [m.State.zero() for m in problem.runningModels + [problem.terminalModel]] us = [np.zeros(m.nu) for m in problem.runningModels] x0ref = problem.initialState xs[0] = np.random.rand(nx) xs[1] = np.random.rand(nx) us[0] = np.random.rand(nu) # xs[1] = model.calc(data,xs[0],us[0])[0].copy() ddpbox = SolverBoxDDP(problem)