Methodik zur effizienten Applikation automatisierter Fahrfunktionen

Abstract

The work introduces a method for an efficient calibration of automated driving functions. The method is based on an initial limitation of the relevant parameter space for the optimization of function behaviour. Furthermore, a sensitivity based optimization is presented, which provides optimal solution with a high maturity level. Finally, a hybrid vehicle model (based on a single track model and a neuronal network) is described, which enables a high accuracy and short simulation times.