Untersuchung geeigneter MPC-Ansätze, um eine reaktive Umgebung für AVs zu emulieren und verschiedene Szenarien zu testen.

As autonomous vehicles (AV) are coming closer road admission, suitable methods to test AVs are in rising demand. Specifically, the behavior of an AV in specific traffic situations needs to be evaluated to address safety concerns. Model predictive control (MPC) is of particular interest for this purpose as as MPC is capable to represent realistic traffic scenarios via MPC’s easy incorporation of constraints and a certain prediction of future behaviors. This project aims to investigate and extend suitable MPC appraoches to emulate a reactive environment for testing AVs in different scenarios.