No abstract

No abstract

No abstract

No abstract

In order to cope with the increasing number of test kilometers required to validate novel vehicle safety functions in the future, the majority of test scenarios must no longer be performed with the real vehicle but in virtual environments. To ensure that the results obtained there can also be transferred to the real vehicle, test cases are necessary that represent the traffic as realistically as possible. For this reason, a traffic flow simulation is used in which critical scenarios arise due to the driving errors of other road users. The quality with which these traffic scenarios are handled by the vehicle safety function is assessed by an objective evaluation standard. This takes into account the minimum distance between the vehicles as well as the maximum acceleration occurring during the braking or evasive maneuver. Based on this evaluation standard, the influence of different parameter settings on the behavior of the safety function can be quantified. The selection of the best configuration is based on statistical methods. This can significantly reduce the application effort in the real vehicle. The results of the performance evaluation with the best parameter setting are used afterwards to select characteristic test cases for the real vehicle test from the multitude of traffic situations analyzed in the simulation. The approach with which this is realized in this paper is based on an analogy to the experienced driver, who always manages simple scenarios safely and only reaches its limits in very complex traffic situations. Applied to the algorithms of the vehicle safety function to be tested, this means that if the scenarios with a low performance can be handles successfully, then those with a higher performance should also be covered. This allows a lower limit to be specified for the functionalities of the vehicle safety function that are validated in the simulation.