Simulation study of freeway work zone scenes for autonomous driving using Simulink and PreScan.

Abstract

To solve the problem of high traffic accident rates in freeway work zones, an in-depth analysis of freeway work zone scenes under an autonomous driving environment was conducted, and traffic conflicts in work zones were analyzed, as were warning areas and upstream transition areas, which are the key road sections where the most potential conflicts occur. Using the National Vehicle Accident In-depth Investigation System (NAIS) database, representative accident cases were selected and studied for various work zones, and specific scenes of the work zone were summarized accordingly. A joint simulation platform was constructed using PreScan and Simulink software, and the behaviors of the main vehicle with lane changes and without lane changes were simulated. The simulation results reveal the main factors influencing of the potential conflict between the warning area and upstream transition area under the condition of no lane change behavior, which is comprehensively analyzed via correlation analysis and the gray correlation analysis method. Specific requirements for the length of the zone and the length of the upstream transition zone are proposed accordingly. In the case of lane change behavior, the analysis identified the key influencing factors for potential conflicts when the target vehicle changes lanes to the front of the main vehicle and when the main vehicle changes lanes to the front of the target vehicle. Based on this foundation, appropriate recommendations are proposed for the speed limit of borrowing from the opposite lane work zone.