Principles of co-simulation

As we said above, co-simulation means that two models (at least) are simulated within their own simulation environments (including solvers), and are communicating at fixed time steps (see figure Figure 4). Communication between models takes place only at fixed time steps. This is why each model is seen as discrete by the other one, the two models being coupled through invisible zero order samples and holds.

Figure 4: Co-simulation principle

When performing co-simulation, one model is identified as master and imposes the communication time step on the slave model.

Note

The fact that the simulations communicate at fixed time steps does not imply that fixed-step solvers have to be used by the two simulations tools. Depending on available software capabilities, there are four supported possibilities for the master / slave solver association, namely:

• Fixed-Step / Fixed-Step

• Fixed-Step / Variable-Step

• Variable-Step / Fixed-Step

• Variable-Step / Variable-Step

The Generic Co-simulation is primarily intended for establishing co-simulation with third-party or proprietary / in-house code, but is also usable for a Simcenter Amesim - Simcenter Amesim co-simulation.

For further information concerning the technical terms used in this manual, please refer to the Glossary at the end of the document.