Dynamic system simulation has gained an important role in automotive engineering since powertrains and power management systems are getting more and more complex. The goal is to establish dynamic system simulation as an integral part of the vehicle development process for the purpose of saving development time, improving subsystem and component specifications, and reducing of development cost. A crucial prerequisite for this approach is that the model quality has been identified for the utilized vehicle simulation models. It determines the applicability of the simulation results since the reliability of simulation results cannot be judged without an identified model quality. The identification of model qualities requires an established and standardized verification and validation process adapted to dynamic vehicle simulation, which is introduced in this paper.
The here introduced process for the verification and validation of dynamic vehicle simulation models is structured by certain elements, starting with the definition of unambiguous terms for model validation. It is followed by the description of validation load cases, since the reliability of simulation data strongly depends on the quantity and quality of the load cases and the available reference data that is used for the validation. The introduction of appropriate metrics for the assessment of correlation between simulated data and reference data finally leads to the determination of the model quality for a dynamic vehicle simulation model.
This paper describes a standard validation and verification process and its application to simulation and test data of a fuel cell vehicle system, but the process is defined in a generic way and, therefore, directly applicable to any dynamic vehicle or powertrain simulation model.