stable_learning_control.disturbers

Contains the disturbers that are available in the SLC package.

Note

These disturbers are implemented as gymnasium:gymnasium wrappers <api/wrappers/>. Because of this, they can be used with any gymnasium environment. If you want to add a new disturber, you only have to ensure that it is a Python class that inherits from the gym.Wrapper class.

Submodules

• stable_learning_control.disturbers.action_impulse_disturber
• stable_learning_control.disturbers.action_random_noise_disturber
• stable_learning_control.disturbers.env_attributes_disturber
• stable_learning_control.disturbers.observation_random_noise_disturber

Classes

ActionImpulseDisturber	A gymnasium wrapper that can be used to disturb the action of a gymnasium
ActionRandomNoiseDisturber	A gymnasium wrapper that can be used to disturb the action of a gymnasium
EnvAttributesDisturber	A gymnasium wrapper that can be used to disturb a physics parameter of a
ObservationRandomNoiseDisturber	A gymnasium wrapper that can be used to disturb the observation of a gymnasium

Package Contents

class stable_learning_control.disturbers.ActionImpulseDisturber(env, magnitude, time)

Bases: gymnasium.ActionWrapper

A gymnasium wrapper that can be used to disturb the action of a gymnasium environment with a impulse applied at a certain time step.

impulse_magnitude

The magnitude of the impulse.

Type:

float

impulse_time

The time (s) at which to apply the impulse.

Type:

float

Initialise the ActionImpulseDisturber object.

Parameters:

• env (gym.Env) – The gymnasium environment.

• magnitude (float) – The impulse magnitude to apply.

• time (float) – The time (s) at which to apply the impulse.

time_warning = False

time_step_warning = False

_track_time = False

_time_step_attr = None

_time_attr = None

magnitude

impulse_time

max_episode_steps

track_time()

Track the time of the environment.

action(action)

Add a impulse to the action.

Parameters:

action (np.ndarray) – The action.

Returns:

The action with added impulse.

Return type:

np.ndarray

class stable_learning_control.disturbers.ActionRandomNoiseDisturber(env, mean, std)

Bases: gymnasium.ActionWrapper

A gymnasium wrapper that can be used to disturb the action of a gymnasium environment with normally distributed random noise.

mean

The mean of the noise normal distribution.

Type:

float

std

The standard deviation of the noise normal distribution.

Type:

float

Initialise the ActionRandomNoiseDisturber object.

Parameters:

• env (gym.Env) – The gymnasium environment.

• mean (float) – The mean of the noise normal distribution.

• std (float) – The standard deviation of the noise normal distribution.

mean

std

action(action)

Add normally distributed random noise to the action.

Parameters:

action (np.ndarray) – The action.

Returns:

The action with added noise.

Return type:

np.ndarray

class stable_learning_control.disturbers.EnvAttributesDisturber(env, attributes, values)

Bases: gymnasium.Wrapper

A gymnasium wrapper that can be used to disturb a physics parameter of a gymnasium environment.

disturbance_label

A label for the disturbance that can be used for logging or plotting. Used in the eval_robustness utility.

Type:

str

Initialise the EnvAttributesDisturber object.

Parameters:

• env (gym.Env) – The gymnasium environment.

• attributes (list) – A list of attributes to disturb.

• values (list) – A list of values to set the parameters to.

Raises:

• ValueError – The number of parameters and values must be the same.

• AttributeError – The parameter does not exist in the environment.

attributes

values

class stable_learning_control.disturbers.ObservationRandomNoiseDisturber(env, mean, std)

Bases: gymnasium.ObservationWrapper

A gymnasium wrapper that can be used to disturb the observation of a gymnasium environment with normally distributed random noise.

mean

The mean of the noise normal distribution.

Type:

float

std

The standard deviation of the noise normal distribution.

Type:

float

Initialise the ObservationRandomNoiseDisturber object.

Parameters:

• env (gym.Env) – The gymnasium environment.

• mean (float) – The mean of the noise normal distribution.

• std (float) – The standard deviation of the noise normal distribution.

mean

std

observation(observation)

Add normally distributed random noise to the observation.

Parameters:

observation (np.ndarray) – The observation.

Returns:

The observation with added noise.

Return type:

np.ndarray