CommonRoad Helper Functions

Installation as package

Install this package with the following command:

pip install commonroad-helper-functions

CommonRoad Scenarios

In order to test the provided functions, download the scenarios by cloning the commonroad-scenarios repository:

git clone https://gitlab.lrz.de/tum-cps/commonroad-scenarios.git

Make sure to place the folder on the same directory level as this repository:

|
|-- commonroad_helper_functions (This repository)
|-- commonroad_scenarios
|

Functionalities

Sensor Occlusion Model

As part of this repository a sensor model is implemented. Based on the position of an ego vehicle, the visible area is determined and all dynamic objects within this area are returned. Occlusions by dynamic objects can be taken into account, as well as by the road geometry. The two examples show the visible area in green. In contrast to the left, the right video uses a wall buffer of 5m. This means that we assume a buffer of 5m around the lane geometry that can be seen through.

A dynamic objects is considered as visible as soon as any part of it can be physically seen from the ego position.

Deployment

1. Import sensor model function

from commonroad_helper_functions.sensor_model import get_visible_objects

2. Call function

visible_object_ids, visible_area = get_visible_objects(scenario, time_step, ego_pos)

Optional Arguments

• sensor_radius: Float number that defines the sensor radius in meter
• occlusion: True if occlusions by dynamic objects should be considered
• wall_buffer: Float number that defines the wall buffer in meter (see example above)