Parameterization of automated driving functions in virtual environments based on characteristic test scenarios

In order to cope with the increasing amount of testing with regard to autonomous driving, the number of test kilometers required with the real vehicle must be drastically reduced. A crucial tool for achieving this challenging goal is simulation. Realistic scenarios are necessary for effective testing of a driving function in virtual environments. This paper presents a method how relevant test cases can be generated. Furthermore, an approach is proposed how the parameterization of a driving function in virtual driving tests can be performed. First, scenarios are selected from a traffic flow simulation in which a collision occurs or a vehicle brakes strongly. These traffic situations are used to test the driving function with an empirically chosen parameterization. Based on the performance shown by the driving function, characteristic scenarios are identified. This reduced set of test scenarios is used to determine the optimal parameterization. The procedure is exemplified for an emergency brake function. Finally, the result of the optimization process is evaluated based on the frequency distribution of the performance values.