Page - 40 - in Integration of Advanced Driver Assistance Systems on Full-Vehicle Level - Parametrization of an Adaptive Cruise Control System Based on Test Drives

3. Development Process requirements specification functional design module development component validation module validation system validation vehicle system level component level test cases test cases cases test full-vehicle level Figure 3.1.: V-Model, adapted from [Mau09] and [Rei12] software, theactuatorsandtheHuman-Machine Interface (HMI)arepartsof thesystem, see fig. 2.1. Each of them has its own sub-processes. 3.1. Full-Vehicle Level There are different ways to perform tests on the full-vehicle level. Since maintaining the safety of the people involved is the most important task, tests conducted in driving simulators are the safest choice. The advantage is that the repeatability is very good, and safety-critical situations can be simulated with no risk. The disadvantage is that the probands know that they are not really driving a car, so they may behave differently than they would in real traffic, [Bre09]. The second possibility is to perform test drives on closed test tracks. Here, the probands are driving real cars, but complex traffic situations cannot be simulated or require a very high effort, [Bre09]. Examples are given by Bock in [Boc09] and Schwab et al. in [SLZB14]. There, theprobandsdroveontest tracksandworeaspecialdevice, anOptical See Through Head-Mounted Display, which projects the traffic on the road in front of the vehicle. The position of the ego vehicle, which must be measured precisely, is sent to a simulation software. There, the relative movement between the virtual traffic and the ego vehicle is calculated by a multi-body simulation. Thereby, a sensor model measures the required quantities and sends them to the real ADAS controller. This method is suitable for evaluating safety-oriented systems, such as the Automatic Emergency Brake (AEB). One main disadvantage is that no real environmental-recognition sensor is used, which has a significant influence on the performance of the system. One method for performing tests on the full-vehicle level is described by Gietelink et al. in [GPSV06]. They built up a facility where the ego vehicle is driven on a dynamometer. The motion of the other vehicles is simulated by moving robot cars in front of the ego 40