Reachable Set Analysis for the Situation Evaluation in Driver Assistance Systems of active Safety in the Commercial Vehicle

The protection of vulnerable road users gets increasingly in focus of current developments of advanced driver assistance systems. The complex situation evaluation requires an approach that decides about the application relevance of a detected pedestrian. The future positions of the road users are predicted probabilistically using a combination of Monte Carlo simulation, stochastic reachable sets and Markov chain abstraction, which accounts for the uncertainty in the behavior and the measurement. Previous measurement cycles serve the assignment of probable prototype paths to the ego vehicle and the pedestrian movement. The conditioned collision probability of the ego vehicle with a pedestrian is computed efficiently by intersecting their stochastic reachable sets and considering the path probabilities.