ADAPTIVE REACHABILITY ANALYSIS FOR FORMAL VERIFICATION OF AUTONOMOUS DRIVING SYSTEMS
Keywords:
Reachability, autonomous driving, decision-making process, Monte-Carlo simulations, collision avoidance.Abstract
This paper investigates the application of adaptive reachability analysis in the navigation and decision-making processes of autonomous driving systems, emphasizing its role in ensuring safe and efficient vehicle operation in complex environments. Reachability, defined as the set of all possible states a vehicle can attain within given physical and operational constraints, serves as a fundamental tool for evaluating safe trajectories in dynamic and uncertain scenarios. The proposed approach is systematically compared with prior methodologies [1,2,3,4,5] in reachability analysis, demonstrating notable improvements in computational efficiency and accuracy.
