Test Case Selection Method for the Verification of Automated Driving Systems

摘要

Assuring the safety of Automated Driving Systems (ADSs) can be achieved by a scenario-based approach as suggested by [1], but needs a solid rationale on the selection process for the test cases to be conducted. To build such a solid and traceable test case selection process a method consisting of four sequential steps is proposed in the following, FIGURE 1. In the first step, the ADS use case is used to define a set of the related so-called logical scenarios [2]. Existing datasets are then used to populate the parameter space of these logical scenarios. Parameter spaces are built as distributions for the considered parameters that give information about which parameter combinations could occur under real-world operational conditions for the logical scenario (and how often).