Search Results

Typical co-simulation platforms cannot handle co-simulations under hard real-time conditions. The most important issue is the coupling of simulation tools together with real-time systems. To solve this problem a real-time coupling mechanism has to be developed which ensures the required hard real-time conditions. This coupling technology has to solve additional problems compared to non-real-time coupling mechanisms. First the most important aspect is a time-correct coupling with the involved real-time systems. Another important task is to keep the round trip time between the coupled systems as small as possible. This is relevant for stability aspects of the coupled simulation. Furthermore sensor signals of such real-time systems are typically corrupted by noise so the real-time coupling technology must be able to handle these noisy coupling signals.

In the automotive industry well-established different simulation tools targeting different needs are used to mirror the physical behavior of domain specific components. To estimate the overall system behavior coupling of these components is necessary. As systems become more complex, simulation time increases rapidly by using traditional coupling approaches. Reducing simulation time by still maintaining accuracy is a challenging task. Thus, a coupling methodology for co-simulation using adaptive macro step size control is proposed. Convergence considerations of the used algorithms and scheduling of domain specific components are also addressed. Finally, the proposed adaptive coupling methodology is examined by means of a cross-domain co-simulation example describing a hybrid electric vehicle. Considerable advantages in terms of simulation time reduction are presented and the trade-off between simulation time and accuracy is depicted.