A probabilistic long term prediction approach for highway scenarios

Risk estimation for the current traffic situation is crucial for safe autonomous driving systems. The computation of risk estimates however is always uncertain, especially if the behavior of traffic participants has to be taken into account. Besides risk estimation, knowledge about the future behavior of other traffic participants can be used for Adaptive Cruise Control Applications, helping to choose a driving strategy with more foresight, which is not only desirable under comfort aspects, but can also be used to reduce fuel consumption. In this publication we focus on highway scenarios, where possible behaviors consist of changes in acceleration and lane-change maneuvers. Based on this insight we present a novel approach for the prediction of future positions in highway scenarios.