Predictive Trajectory Planning in Situations with Hidden Road Users Using Partially Observable Markov Decision Processes

Philip Schörner, Lars Töttel, Jens Doll and J. Marius Zöllner
IEEE Intelligent Vehicles Symposium (IV), Paris
30th IEEE Intelligent Vehicles Symposium (IV), Paris
State of the art emergency brake assistant systems solely based on sensor measurements reduced the number of traffic accidents and casualties drastically in recent years. In order to be able to react on road users who elude a vehicle’s field of view because of sensor limits or occlusions, this paper presents an approach to anticipate potential hidden traffic participants in occluded areas in the decision making process of an autonomous vehicle. A Partially Observable Markov Decision Process is used to determine the vehicle’s longitudinal motion. Observations are made using the vehicle’s field of view. Therefore the field of view is calculated with a generic model of a sensor setup in dependence of the current or predicted the environment. In this way, the vehicle can either observe that it detects a previously hidden road user or receives information that the road is clear. In total, that allows the vehicle to better anticipate future developments. Therefore, assumptions about vehicles that may be located in hidden areas need to be made. We demonstrate the approach in two scenarios. Firstly in a scenario, where the vehicle has to move cautiously into the intersection with a minimum number of actions and secondly in a typical scenario for urban traffic. Evaluation shows, that the approach is able to anticipate hidden road users correctly and act accordingly.
Download .bib
Download .bib
Eingetragen von
Lars Töttel