Co-simulation techniques in assistance systems for process control

The penetration of production systems and common infrastructure with information and communication technologies (ICT) creates new systems, called Smart Grids in energy or Industry 4.0 in production. This offers unprecedented opportunities for flexible and customized production and distribution. However, these complex and interconnected systems need new approaches for operation. Assistance system similar to those known from the automotive sector will support operators in their task to provide a safe, efficient and undisturbed operation of production systems. The paper supports, that more complex functionalities of assistance systems are based on simulation. This paper discusses the question how well co-simulation approaches are suited for this task. Co-simulation is the joint simulation of independent simulators each representing a component or subsystem of the overall system. The modularity of co-simulation reflects the modularity of production systems. It is possible to compose a simulation by using existing models. It is easy to adapt co-simulation to different configurations of production systems or changes in topology. The inherent flexibility of co-simulation makes it possible to cover broad range of requirements regarding, among others, simulation speed and precision. Co-simulation further allows to include models as “black boxes” which can be used to protect intellectual property on the modeled components. However, co-simulation always comes with additional costs for communication. Data exchange between models sets upper bounds for simulation speed. This limits the application of co-simulation: If requirements on calculation speed are very high other approaches like hybrid simulations are more suitable. For very precise simulations parallel simulation may be appropriate.