2015_i004_DI_MAGNA_BasicResearch_Magosi_VILAD: Contributions to the development of a full vehicle test bench for automated driving

Automated driving functions use digital information of the environment. This environment information can be provided by different types of sensors located either on the automated vehicle or on the road infrastructure. When integrating automated driving functions on full vehicle level using virtual validation methods, detailed sensor models are required to provide input data to the control algorithms of the automated driving functions. The present research deals with modelling of automotive radar sensors. Due to the complex physics of these sensors in the ranges of 24 and 77GHz, different approaches, ranging from phenomenological to a mixture between physical and phenomenological modelling is investigated. Driving tests with radar sensors are performed in order to acquire object data that are used for parametrization and verification of the sensor model. The model is implemented into a full vehicle virtual validation framework and is intended to be used on test benches for automated driving. For the latter, the model has to perform in real-time.