drone-oculus-won / adscenariosimrefex

Agent based simulation framework for validating planners in automated driving synthetic scenarios. This examples demonstrates overtake maneuver.

Automated Driving Scenario Simulation Reference Example (Overtake Maneuver)

Overtake maneuver strategy allows a smart vehicle (an independent agent), to safely overtake the vehicles ahead of it by taking decisions based on the belief/knowledge it has about its environment.

This demo showcases a Simulink model architecture for creating and simulating synthetic scenarios. It reads as input the scenario file saved using the Driving Scenario Designer (DSD) application. This architecture allows creation of synthetic scenarios, by:

• Marking an actor in the scenario as an autonomous smart actor.
• Installing car-following (driver) model on some of the actors.
• Introducing rogue actors (actors devoid of any intelligence) in the scenario.

Driving scenario designer (DSD) application is part of Automated Driving System Toolbox (ADST). Refer to the documentation here for more information.

Configuration parameters can be set for individual actors to observe the variations in the behavior. The plan algorithm in the smart actor supports overtake maneuver on straight road.

MATLAB Toolbox dependencies

This model has been tested with MATLAB R2019b. The version tested with MATLAB R2018b and R2019a is available with release tag v1.0-MATLAB19a.

To run this model, you need: MATLAB, Automated Driving System Toolbox (ADST), Model Predictive Control Toolbox, Simulink, Simulink Coder, and Stateflow.

Running the model

The model needs a scenario file saved using the DSD application. A sample scenario file scenarioInput.mat is already included in the repository.

Run the Simulink model mOvertakeManeuver.slx and see the visualization.

Further documentation

See the file Overtake-Maneuver.pdf for detailed documentation covering - Scenario description, new scenario generation, Simulink model description, configuration and scenario visualization.