Autor: Johanna Rieke

Abstract Positioning is a central component of many driver assistance systems and automated driving. Requirements for the position solution vary depending on the application in the vehicle. In the theoretical part of this thesis first deals with requirements possibly arising in road vehicles and which realistic application scenarios exist for a GNSS-based location module. The current developments of further satellite navigation systems result in new options for these GNSS-based solutions. Urban area in particular is still a challenge due to shadowing and multi-path effects that influence the reception of satellite signals. In this work it is examined what influence a higher number of satellites and the combination of systems have on localization. The challenge of using multi-constellation GNSS consists in the synchronisation of the data. Each system refers to a different time base. In addition, the signal frequencies and signal codings differ. This is accompanied by different signal processing by the hardware of the receiver. As already known from the literature the intersystem bias sent along compensates the offset between system clocks whereas receiver-specific delays between various systems are not compensated. In this thesis different possibilities are beeing analyzed to integrate several satellite navigation systems into a tightly coupled positioning system with inertial sensors and odometers. The position solution with the intersystem bias provided is benchmarked against a separate clock error estimation and a separate intersystem error estimation. Thereby the receiver’s clock error estimation is extended by corresponding variables. Furthermore, a modularisation of state estimation in the Kalman-filter fusion using multi-constellation is investigated combined with the intersystem bias provided as well as the separate estimated clock error and separate estimated intersystem bias. The results show the influence of the intersystem error on position solutions and the benefits of the intersystem error estimation. Due to the modularization in combination with several systems only marginal changes in the position solution can be generated. Finally, the results of the position calculation are evaluated considering the requirements analysis.