Hochgenaue und robuste odometriebasierte Lokalisierung in einem Parkvorgang

Abstract

As part of this work, a special vehicle movement model is being developed for future parking applications. It uses standard sensor technology and requires hardly any additional computing power. An adaptive localization filter with a robust information filter structure processes redundant motion data and reliably estimates wheel slip and sensor errors. The method significantly increases the accuracy and robustness of localization.