The adoption of simulation is critical to reducing development time and enhancing system robustness for advanced driver assistance systems (ADAS). Automotive companies typically have an abundance of real world data recorded from a vehicle that is suitable for open-loop simulations. However, recorded data is often not suitable for testing closed-loop control systems since the recorded data cannot react to changes in vehicle movement.
This paper introduces a methodology to create virtual driving scenarios from recorded vehicle data to enable closed-loop simulation. This methodology is applied to test a lane centering application. A lane centering application helps a driver control steering to stay in the current lane and control acceleration and braking to maintain a set speed or to follow a preceding vehicle. The driver’s vehicle is referred to as the ego vehicle. Other vehicles on the road are referred to as target vehicles. To test the lane centering system in simulation, engineers must model the ego vehicle (sensors and dynamics) as well as the scenario (roads and target vehicles). A virtual driving scenario is created by reconstructing roads and target vehicles using GPS, camera-based lane detections, radar-based vehicle detections, and map data. The virtual driving scenario is integrated into a closed-loop simulation to assess the behavior of a lane centering system.
Copyright © 2020 by The MathWorks, Inc. Published by SAE International, with permission.
This paper was presented at the WCX Digital Summit 2020.