paralleldomain.data_lab.behaviors.vehicle

class DrivewayCreepBehavior(reference_line, behavior_duration, agent_length=5.5)

Controls the behavior of vehicles that perform the driveway creeping behavior

Parameters:

  • reference_line (ndarray) – The driveway reference line along which the vehicle should perform the creeping behavior. Numpy array should be n x 3 in size containing n points along the reference line where each point is the [x, y, z] location of that point

  • behavior_duration (float) – Length of time in seconds in which the vehicle should travel from the start to end of the driveway

  • agent_length (float) – Length of the agent (in meters) to which we are assigning the Driveway Creeping Behavior

class RenderEgoBehavior

When used in conjunction with RenderEgoGenerator, allows the Ego Vehicle to be rendered

class VehicleBehavior(*, proto=None, start_speed=None, target_speed=None, ignore_speed_limit=None, lane_offset=None, lane_drift_scale=None, lane_drift_amplitude=None, lane_change_probability=None, lane_change_cooldown=None, enable_dynamic_lane_selection=None, start_gear=None, start_separation_time=None, target_separation_time=None, vehicle_aggression=None, ignore_obstacle_types=None, parking_scenario_goal=None, parking_scenario_time=None)

Parameters which direct a vehicle’s movement behavior (eg. lane change, speed, etc.).

Note that in some cases, the Vehicle Behavior exhibited in the simulation may differ from that which the user has directed. Vehicle Behavior is affected by obstacles near the vehicle, curvature of the road, end-of-lanes the vehicle is approaching and the slope of the road.

Parameters:
start_speed

Specifies the start speed of a vehicle at the beginning of a scenario. Value is specified as a ContinousUniformDistribution which is randomly sampled from during scenario generation.

target_speed

Specifies the target speed of the vehicle. Value is specified as a ContinousUniformDistribution which is randomly sampled from during scenario generation.

Will not cause vehicle to exceed the speed limit of a road unless ignore_speed_limit is set to True.

ignore_speed_limit

Flag to control whether vehicle will ignore speed limits in attempting to achieve the target_speed. Default: False

lane_offset

Specifies the lateral offset of the vehicle in the vehicle’s motion plan relative to lane center line. Possible values are floats from -1.0 to 1.0, and represent the proportion of the lane’s half width to offset laterally. Positive values offset to the right and negative values offset to the left. Offset distance is capped so vehicle does not protrude laterally out of the lane edges.

If not provided, will default to 0.0. Value is specified as a ContinousUniformDistribution which is randomly sampled from during scenario generation.

lane_drift_scale

Controls the frequency of weaving within the vehicle’s lanes during travel. Weaving is modelled as a perlin noise driven wave. If not provided, will default to 0.0. Value is specified as a ContinousUniformDistribution which is randomly sampled from during scenario generation.

lane_drift_amplitude

Controls the amplitude of weaving within the vehicle’s lanes during travel. Weaving is modelled as a perlin noise driven wave. If not provided, will default to 0.0. Value is specified as a ContinousUniformDistribution which is randomly sampled from during scenario generation.

lane_change_probability

Specifies the probability that a vehicle will make a lane change. Evaluated once every second. If evaluated to true, vehicle will lane change if conditions allow (having lanes to change to, no obstables in those lanes or in lane change path) Values can range from 0.0 to 1.0. If not provided, defaults to zero. Value is specified as a ContinousUniformDistribution which is randomly sampled from during scenario generation.

lane_change_cooldown

Specifies the time in seconds during which another lane change cannot occur after the vehicle undergoes a lane change. If not provided, a default value of 10 is used. Value is specified as a ContinousUniformDistribution` which is randomly sampled from during scenario generation.

enable_dynamic_lane_selection

Boolean flag to specify whether lane change decisions are made dynamically while the scenario is being simulated. Default: False

start_gear

Specifies the gear that the vehicle should start in at the beginning of scenario generation. Default: Gear.DRIVE

start_separation_time

Specifies the spacing, in seconds, that should exist between vehicles in the same lane at the start of the scenario. Value is specified as a ContinousUniformDistribution which is randomly sampled from during scenario generation.

target_separation_time

Specifies the spacing, in seconds, that targeted between vehicles in the same lane over the course of the scenario. Value is specified as a ContinousUniformDistribution which is randomly sampled from during scenario generation.

vehicle_aggression

Specifies a float measurement of the level of aggression with which the vehicle should drive. Impacts values such as lane change speed, abruptness of turns. Also multiplies the set target_speed to define the speed that will be targeted by the vehicle during the scenario.

Must be a float greater than 0.0 if specified. If not specified, a default value is used.

ignore_obstacle_types

Specifies a list of ObstacleType that the vehicle should ignore in driving.

parking_scenario_goal

Specifies the goal position for the vehicle to target in a parking/unparking scenario. Only AbsolutePositionRequest and LaneSpawnPolicy position request types are supported.

Must be provided if a parking/unparking scenario is desired. Will result in a scenario with standard vehicle driving behavior if not provided.

To specify a parking scenario, set a lane_spawn_policy that specifies a ParkingSpace as the goal. To control whether the parking occurs in a parking lot or on the street, specify a road_type within the lane_spawn_policy as appropriate.

To specify an unparking scenario, set a position_request in the generator to a LaneSpawnPolicy that specifies a parking space, and use this parking_scenario_goal to specify only a road_type that corresponds to the type of road the vehicle should exit onto. Unparking is not supported on streets.

Examples:

>>> # Parking in Parking Lot
>>> {
>>>    lane_spawn_policy: {
>>>        lane_type: {
>>>            probabilities: {
>>>                "ParkingSpace": 1.0
>>>            }
>>>        },
>>>        road_type: {
>>>            probabilities: {
>>>                "Parking_Aisle": 1.0
>>>            }
>>>        }
>>>    }
>>> }
>>>
>>> # Parking on street
>>> {
>>>     lane_spawn_policy: {
>>>         lane_type: {
>>>             probabilities: {
>>>                 "ParkingSpace": 1.0
>>>             }
>>>         },
>>>         road_type: {
>>>             probabilities: {
>>>                 "Primary": 1.0
>>>             }
>>>         }
>>>     }
>>> }
>>>
>>> # Unparking in Parking Lot
>>> {
>>>     lane_spawn_policy: {
>>>         lane_type: {
>>>             probabilities: {
>>>                 "ParkingSpace": 1.0
>>>             }
>>>         },
>>>         road_type: {
>>>             probabilities: {
>>>                 "ParkingAisle": 1.0
>>>             }
>>>         }
>>>     }
>>> }
parking_scenario_time

Specifies the time (in seconds) it takes to reach the end of the parking maneuver assuming constant acceleration from the initial speed to the targe speed. Does not account for vehicle slow downs due to obstacles etc. If not provided, will use a value 5s.

class Gear(value)

Enum containing the different types of vehicle transmission gears.

DRIVE = 4
NEUTRAL = 3
PARKED = 1
REVERSE = 2