double_pendulum.controller.inverse_dynamics
Submodules
double_pendulum.controller.inverse_dynamics.computed_torque_controller
- class double_pendulum.controller.inverse_dynamics.computed_torque_controller.ComputedTorqueController(mass=[0.5, 0.6], length=[0.3, 0.2], com=[0.3, 0.2], damping=[0.1, 0.1], coulomb_fric=[0.0, 0.0], gravity=9.81, inertia=[None, None], torque_limit=[0.0, 1.0], model_pars=None, dt=0.01, csv_path=None)
Bases:
AbstractController
Computed torque controller. Controller which computes torque from the inverse dynamics and PID. Only useful for the fully actuated double pendulum.
- Parameters:
- massarray_like, optional
shape=(2,), dtype=float, default=[0.5, 0.6] masses of the double pendulum, [m1, m2], units=[kg]
- lengtharray_like, optional
shape=(2,), dtype=float, default=[0.3, 0.2] link lengths of the double pendulum, [l1, l2], units=[m]
- comarray_like, optional
shape=(2,), dtype=float, default=[0.3, 0.3] center of mass lengths of the double pendulum links [r1, r2], units=[m]
- dampingarray_like, optional
shape=(2,), dtype=float, default=[0.1, 0.1] damping coefficients of the double pendulum actuators [b1, b2], units=[kg*m/s]
- gravityfloat, optional
default=9.81 gravity acceleration (pointing downwards), units=[m/s²]
- coulomb_fricarray_like, optional
shape=(2,), dtype=float, default=[0.0, 0.0] coulomb friction coefficients for the double pendulum actuators [cf1, cf2], units=[Nm]
- inertiaarray_like, optional
shape=(2,), dtype=float, default=[None, None] inertia of the double pendulum links [I1, I2], units=[kg*m²] if entry is None defaults to point mass m*l² inertia for the entry
- torque_limitarray_like, optional
shape=(2,), dtype=float, default=[0.0, 1.0] torque limit of the motors [tl1, tl2], units=[Nm, Nm]
- model_parsmodel_parameters object, optional
object of the model_parameters class, default=None Can be used to set all model parameters above If provided, the model_pars parameters overwrite the other provided parameters
- dtfloat
timestep, unit=[s]
- csv_pathstring or path obj
path to csv file containing a reference trajectory (Default value=None)
- get_control_output_(x, t=None)
The function to compute the control input for the double pendulum’s actuator(s).
- Parameters:
- xarray_like, shape=(4,), dtype=float,
state of the double pendulum, order=[angle1, angle2, velocity1, velocity2], units=[rad, rad, rad/s, rad/s]
- tfloat, optional
time, unit=[s] (Default value=None)
- Returns:
- array_like
shape=(2,), dtype=float actuation input/motor torque, order=[u1, u2], units=[Nm]
- get_init_trajectory()
Get the initial (reference) trajectory used by the controller.
- Returns:
- numpy_array
time points, unit=[s] shape=(N,)
- numpy_array
shape=(N, 4) states, units=[rad, rad, rad/s, rad/s] order=[angle1, angle2, velocity1, velocity2]
- None
Does not return reference torques
- init_()
Initialize controller.
- save_(save_dir)
Save the energy trajectory to file.
- Parameters:
- pathstring or path object
directory where the parameters will be saved
- set_parameters(Kp, Ki, Kd)
Set PID parameters.
- Parameters:
- Kpfloat
Gain proportional to position error
- Kifloat
Gain proportional to integrated error
- Kdfloat
Gain proportional to error derivative
double_pendulum.controller.inverse_dynamics.inverse_dynamics_ff_controller
- class double_pendulum.controller.inverse_dynamics.inverse_dynamics_ff_controller.InverseDynamicsController(mass=[0.5, 0.6], length=[0.3, 0.2], com=[0.3, 0.2], damping=[0.1, 0.1], coulomb_fric=[0.0, 0.0], gravity=9.81, inertia=[None, None], torque_limit=[0.0, 1.0], model_pars=None, csv_path=None)
Bases:
AbstractController
InverseDynamicsController. Controller which computes torque from the inverse dynamics. Only useful for the fully actuated double pendulum.
- Parameters:
- massarray_like, optional
shape=(2,), dtype=float, default=[0.5, 0.6] masses of the double pendulum, [m1, m2], units=[kg]
- lengtharray_like, optional
shape=(2,), dtype=float, default=[0.3, 0.2] link lengths of the double pendulum, [l1, l2], units=[m]
- comarray_like, optional
shape=(2,), dtype=float, default=[0.3, 0.3] center of mass lengths of the double pendulum links [r1, r2], units=[m]
- dampingarray_like, optional
shape=(2,), dtype=float, default=[0.1, 0.1] damping coefficients of the double pendulum actuators [b1, b2], units=[kg*m/s]
- gravityfloat, optional
default=9.81 gravity acceleration (pointing downwards), units=[m/s²]
- coulomb_fricarray_like, optional
shape=(2,), dtype=float, default=[0.0, 0.0] coulomb friction coefficients for the double pendulum actuators [cf1, cf2], units=[Nm]
- inertiaarray_like, optional
shape=(2,), dtype=float, default=[None, None] inertia of the double pendulum links [I1, I2], units=[kg*m²] if entry is None defaults to point mass m*l² inertia for the entry
- torque_limitarray_like, optional
shape=(2,), dtype=float, default=[0.0, 1.0] torque limit of the motors [tl1, tl2], units=[Nm, Nm]
- model_parsmodel_parameters object, optional
object of the model_parameters class, default=None Can be used to set all model parameters above If provided, the model_pars parameters overwrite the other provided parameters
- csv_pathstring or path obj
path to csv file containing a reference trajectory (Default value=None)
- get_control_output_(x, t=None)
The function to compute the control input for the double pendulum’s actuator(s).
- Parameters:
- xarray_like, shape=(4,), dtype=float,
state of the double pendulum, order=[angle1, angle2, velocity1, velocity2], units=[rad, rad, rad/s, rad/s]
- tfloat, optional
time, unit=[s] (Default value=None)
- Returns:
- array_like
shape=(2,), dtype=float actuation input/motor torque, order=[u1, u2], units=[Nm]
- get_init_trajectory()
Get the initial (reference) trajectory used by the controller.
- Returns:
- numpy_array
time points, unit=[s] shape=(N,)
- numpy_array
shape=(N, 4) states, units=[rad, rad, rad/s, rad/s] order=[angle1, angle2, velocity1, velocity2]
- None
Does not return reference torques
- init_()
Initialize controller.
- save_(save_dir)
Save the energy trajectory to file.
- Parameters:
- pathstring or path object
directory where the parameters will be saved