upkie.envs.wheeled_inverted_pendulum.WheeledInvertedPendulum Class Reference

Wheeled inverted pendulum model. More...

Classes
class	Uncertainty
	Biases and noise levels on modeled system uncertainties. More...

Public Member Functions
def	__init__ (self, float fall_pitch=1.0, float frequency=200.0, bool frequency_checks=True, float length=0.6, float max_ground_accel=10.0, float max_ground_velocity=1.0, bool regulate_frequency=True, Optional[str] render_mode=None, Optional["WheeledInvertedPendulum.Uncertainty"] uncertainty=None)
	Initialize a new environment. More...
Tuple[np.ndarray, dict]	reset (self, *Optional[int] seed=None, Optional[dict] options=None)
	Resets the spine and get an initial observation. More...
Tuple[np.ndarray, float, bool, bool, dict]	step (self, np.ndarray action)
	Run one timestep of the environment's dynamics. More...
bool	detect_fall (self, float pitch)
	Detect a fall based on the base-to-world pitch angle. More...
def	render (self)
	Render state to a live plot.
def	log (self, str name, Any entry)
	Log a new entry to the "log" key of the action dictionary. More...

Public Attributes
	observation_space
	Observation space.
	action_space
	Action space.
	dt
	Period of the control loop in seconds.
	fall_pitch
	Fall pitch angle, in radians.
	length
	Length of the inverted pendulum.
	model
	Robot model read from its URDF description.
	plot
	Optional plot used for rendering.
	render_mode
	Inherited from gymnasium.Env.
	uncertainty
	Biases and noise levels on modeled system uncertainties.

Static Public Attributes
dictionary	metadata = {"render_modes": ["plot"]}
	Metadata of the environment containing rendering modes.
int	version = 1
	Environment version number.

Detailed Description

Wheeled inverted pendulum model.

This environment helps test and debug behaviors in a perfect or noise-controlled setting. It has the same observation and action spaces as upkie.envs.upkie_ground_velocity.UpkieGroundVelocity. Model assumptions and discretization are summed up in this note.

Action space

The action corresponds to the ground velocity resulting from wheel velocities. The action vector is simply:

Index	Description
0	Ground velocity in [m] / [s].

Observation space

Vectorized observations have the following structure:

Index	Description
0	Pitch angle of the base with respect to the world vertical, in radians. This angle is positive when the robot leans forward.
1	Position of the average wheel contact point, in meters.
2	Body angular velocity of the base frame along its lateral axis, in radians per seconds.
3	Velocity of the average wheel contact point, in meters per seconds.

Constructor & Destructor Documentation

__init__()

def upkie.envs.wheeled_inverted_pendulum.WheeledInvertedPendulum.__init__	(		self,
		float	fall_pitch = 1.0,
		float	frequency = 200.0,
		bool	frequency_checks = True,
		float	length = 0.6,
		float	max_ground_accel = 10.0,
		float	max_ground_velocity = 1.0,
		bool	regulate_frequency = True,
		Optional[str]	render_mode = None,
		Optional["WheeledInvertedPendulum.Uncertainty"]	uncertainty = None
	)

Initialize a new environment.

Parameters

fall_pitch	Fall detection pitch angle, in [rad].
frequency	Regulated frequency of the control loop, in Hz.
frequency_checks	If regulate_frequency is set and this parameter is true (default), a warning is issued every time the control loop runs slower than the desired frequency. Set this parameter to false to disable these warnings.
length	Length of the pole.
max_ground_velocity	Maximum commanded ground velocity in [m] / [s].
max_ground_accel	Maximum acceleration of the ground point, in [m] / [s]².
regulate_frequency	Enables loop frequency regulation.
render_mode	Rendering mode, set to "plot" for live plotting.
uncertainty	Uncertainty biases and noise magnitudes.

Member Function Documentation

detect_fall()

bool upkie.envs.wheeled_inverted_pendulum.WheeledInvertedPendulum.detect_fall	(		self,
		float	pitch
	)

Detect a fall based on the base-to-world pitch angle.

Parameters

pitch

Pitch angle of the rotation from base to world frame.

Returns

True if and only if a fall is detected.

log()

def upkie.envs.wheeled_inverted_pendulum.WheeledInvertedPendulum.log	(		self,
		str	name,
		Any	entry
	)

Log a new entry to the "log" key of the action dictionary.

Parameters

name	Name of the entry.
entry	Dictionary to log along with the actual action.

reset()

Tuple[np.ndarray, dict] upkie.envs.wheeled_inverted_pendulum.WheeledInvertedPendulum.reset	(		self,
		*Optional[int]	seed = None,
		Optional[dict]	options = None
	)

Resets the spine and get an initial observation.

Parameters

seed	Number used to initialize the environment’s internal random number generator.
options	Currently unused.

Returns

• observation: Initial vectorized observation, i.e. an element of the environment's observation_space.
• info: Dictionary with auxiliary diagnostic information. For Upkie this is the full observation dictionary sent by the spine.

step()

Tuple[np.ndarray, float, bool, bool, dict] upkie.envs.wheeled_inverted_pendulum.WheeledInvertedPendulum.step	(		self,
		np.ndarray	action
	)

Run one timestep of the environment's dynamics.

When the end of the episode is reached, you are responsible for calling reset() to reset the environment's state.

Parameters

action

Action from the agent.

Returns

• observation: Observation of the environment, i.e. an element of its observation_space.
• reward: Reward returned after taking the action.
• terminated: Whether the agent reached a terminal state, which can be a good or a bad thing. When true, the user needs to call reset().
• truncated: Whether the episode is reaching max number of steps. This boolean can signal a premature end of the episode, i.e. before a terminal state is reached. When true, the user needs to call reset().
• info: Dictionary with auxiliary diagnostic information. For us this is the full observation dictionary coming from the spine.

---

The documentation for this class was generated from the following file:

• upkie/envs/wheeled_inverted_pendulum.py