Use of Stimulation-Independent Evaluation in the Context of Vehicle Body Domain

Abstract:

Currently established automotive test approaches in the field of vehicle hardware-in-the-loop （HiL） testing use predefined test structures. The test case structure couples a specific test stimulation with an individual test step to a fixed， predefined system validation with pass / fail result. Each test case has its own specific functional focus. If more than one functional aspect is to be tested， the current methods will not work anymore. The aim of future test methods is to evaluate the system as comprehensively and resource-efficiently as possible. If possible， several sub-functionalities should be evaluated at the same time in a randomly generated， realistic environment. A different validation approach is required for the evaluation of randomly generated stimulations sequences. Previously they are unknown to the evaluation environment. It is no longer possible to validate each test step individually. The exact sequence of the stimulation is not known at the beginning of the test run. This should simulate a behavior that is as realistic as possible. For this purpose， a methodology for the generation of system evaluation for randomly generated stimulations was introduced. Combinatorial and model-based approaches were combined to support the creation of system evaluation for the vehicle body domain （describes the passenger functions that can be experienced within a vehicle with its control units and functionalities， no direct influence on driving dynamics） and link them to the corresponding system requirements. The approach supports the existing methods and achieve a wider and deeper coverage of system assessments than a normal test catalogue implementation would give. This will be shown in a proof of concept with a vehicle comfort function on a HiL system at a German car manufacturer.