Road regulation

A RegulatedRoad is a Road in which the behavior of vehicles take or give the right of way at an intersection based on the priority lane attribute.

On such a road, some rules are enforced:

• most of the time, vehicles behave as usual;

• however, they try to predict collisions with other vehicles through the is_conflict_possible() method;

• when it is the case, right of way is arbitrated through the respect_priorities() method, and the yielding vehicle target velocity is set to 0 until the conflict is resolved.

API

class highway_env.road.regulation.RegulatedRoad(network: highway_env.road.road.RoadNetwork = None, vehicles: List[highway_env.vehicle.kinematics.Vehicle] = None, obstacles: List[highway_env.road.objects.Obstacle] = None, np_random: numpy.random.mtrand.RandomState = None, record_history: bool = False)

YIELDING_COLOR: Tuple[float, float, float] = None

REGULATION_FREQUENCY: int = 2

YIELD_DURATION: float = 0.0

__init__(network: highway_env.road.road.RoadNetwork = None, vehicles: List[highway_env.vehicle.kinematics.Vehicle] = None, obstacles: List[highway_env.road.objects.Obstacle] = None, np_random: numpy.random.mtrand.RandomState = None, record_history: bool = False) → None

New road.

Parameters

• network – the road network describing the lanes

• vehicles – the vehicles driving on the road

• road_objects – the objects on the road including obstacles and landmarks

• np_random (np.random.RandomState) – a random number generator for vehicle behaviour

• record_history – whether the recent trajectories of vehicles should be recorded for display

step(dt: float) → None

Step the dynamics of each entity on the road.

Parameters

dt – timestep [s]

enforce_road_rules() → None

Find conflicts and resolve them by assigning yielding vehicles and stopping them.

static respect_priorities(v1: highway_env.vehicle.kinematics.Vehicle, v2: highway_env.vehicle.kinematics.Vehicle) → highway_env.vehicle.kinematics.Vehicle

Resolve a conflict between two vehicles by determining who should yield

Parameters

• v1 – first vehicle

• v2 – second vehicle

Returns

the yielding vehicle

static is_conflict_possible(v1: highway_env.vehicle.controller.ControlledVehicle, v2: highway_env.vehicle.controller.ControlledVehicle, horizon: int = 3, step: float = 0.25) → bool

__annotations__ = {'REGULATION_FREQUENCY': <class 'int'>, 'YIELDING_COLOR': typing.Tuple[float, float, float], 'YIELD_DURATION': <class 'float'>}

__module__ = 'highway_env.road.regulation'