class ProbabilisticObjectInput(object): def __init__(self, name): super(ProbabilisticObjectInput, self).__init__() self.frame = FrameInput('{}{}frame'.format( name, ControllerInputArray.separator)) self.dimensions = Vec3Input('{}{}dimensions'.format( name, ControllerInputArray.separator)) self.probability_class = ScalarInput('P_class') self.probability_pos = ScalarInput('P_trans') self.probability_rot = ScalarInput('P_rot') def get_update_dict(self, p_object): out_dict = self.frame.get_update_dict(*p_object.frame) out_dict.update(self.dimensions.get_update_dict(p_object.dimensions)) out_dict.update(self.probability_class.get_update_dict( p_object.pclass)) out_dict.update(self.probability_pos.get_update_dict(p_object.ppos)) out_dict.update(self.probability_rot.get_update_dict(p_object.prot)) return out_dict def get_frame(self): return self.frame.get_expression() def get_dimensions(self): return self.dimensions.get_expression() def get_class_probability(self): return self.probability_class.get_expression()
class EEFDiffController(QPController): def __init__(self, robot, builder_backend=None, weight=1): self.weight = weight super(EEFDiffController, self).__init__(robot, builder_backend) # @profile def add_inputs(self, robot): self.goal_diff = FrameInput(prefix='', suffix='goal') self.goal_weights = ScalarInput(prefix='', suffix='sc_w') # @profile def make_constraints(self, robot): t = time() eef1 = robot.end_effectors[0] eef1_frame = robot.frames[eef1] eef2 = robot.end_effectors[1] eef2_frame = robot.frames[eef2] eef_diff_pos = spw.pos_of(eef1_frame) - spw.pos_of(eef2_frame) goal_pos = self.goal_diff.get_position() dist = spw.norm((eef_diff_pos) - goal_pos) eef_diff_rot = spw.rot_of(eef1_frame)[:3, :3].T * spw.rot_of( eef2_frame)[:3, :3] goal_rot = self.goal_diff.get_rotation() goal_r = goal_rot[:3, :3].reshape(9, 1) dist_r = spw.norm(eef_diff_rot.reshape(9, 1) - goal_r) self._soft_constraints['align eefs position'] = SoftConstraint( lower=-dist, upper=-dist, weight=self.goal_weights.get_expression(), expression=dist) self._soft_constraints['align eefs rotation'] = SoftConstraint( lower=-dist_r, upper=-dist_r, weight=self.goal_weights.get_expression(), expression=dist_r) self._controllable_constraints = robot.joint_constraints self._hard_constraints = robot.hard_constraints self.update_observables( {self.goal_weights.get_symbol_str(): self.weight}) print('make constraints took {}'.format(time() - t)) def set_goal(self, goal): """ dict eef_name -> goal_position :param goal_pos: :return: """ self.update_observables(self.goal_diff.get_update_dict(*goal))
def __init__(self, name): super(ProbabilisticObjectInput, self).__init__() self.frame = FrameInput('{}{}frame'.format(name, ControllerInputArray.separator)) self.dimensions = Vec3Input('{}{}dimensions'.format(name, ControllerInputArray.separator)) self.probability_class = ScalarInput('P_class') self.probability_pos = ScalarInput('P_trans') self.probability_rot = ScalarInput('P_rot')
def add_inputs(self, robot): self.goal_eef = {} self.goal_weights = {} for eef in robot.end_effectors: self.goal_eef[eef] = FrameInput(prefix=eef, suffix='goal') self.goal_weights[eef] = ScalarInput(prefix=eef, suffix='sc_w')
def add_inputs(self, robot): self.goal_diff = FrameInput(prefix='', suffix='goal') self.goal_weights = ScalarInput(prefix='', suffix='sc_w')