Collision Avoidance Driving Control for Highway Traffic Simulators

Open Access
Wu, Haochen
Area of Honors:
Mechanical Engineering
Bachelor of Science
Document Type:
Thesis Supervisors:
  • Sean N Brennan, Thesis Supervisor
  • Daniel Humberto Cortes Correales, Honors Advisor
  • Highway Traffic Simulator
  • Human Driving Behavior
  • Decision Making
  • Autonomous Vehicles
  • Potential Fields
  • Preview Control
  • Collision Avoidance
Compared to field experimentation, simulation testing for autonomous vehicles provides controllable and reproducible testing scenarios and ensures human safety in emergency situations. This thesis introduces a collision avoidance driving control algorithm for automated and connected vehicles that emulates human driving behavior under severe events. The goal of the algorithm is to exhibit behavior that prevents potential collisions and yet satisfies human comfort. The algorithm is developed by considering the highway environment, static obstacles, and moving objects as hazards interpreted by potential fields, where steering control is determined by the gradient of potential fields and vehicle dynamics is incorporated. A preview control is used to assess the hazards in the future, where autonomous vehicles react to the prediction. In cases where intended vehicle trajectories produce a conflict between vehicles, a decision management process is used to avoid potential collisions and minimize the deviation in overall traffic acceleration. The effectiveness of the algorithm is demonstrated by simulations under different emergency and traffic situations.