multi-rate methods for the real-time co-simulation in the virtual commissioning of industrial robots

• Applicant: Professor Dr.-Ing. Alexander Verl
• Subject Area: Production Automation and Assembly Technology
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 573320699

Real-time co-simulation can be used to increase the model depth in virtual commissioning, which increases the informative value of the simulation and thus enhances the added value of virtual commissioning. If sub-simulations with different step sizes are used, this is referred to as multi-rate simulation. This leads to discontinuous signals in the faster sub-simulation and thus to errors in the simulation result. The signals can be processed using multi-rate methods. Signal-based or model-based methods can be used for this purpose. Preliminary work has shown that multi-rate methods can improve the simulation results in real-time co-simulation during virtual commissioning. However, simplified application examples from material flow simulation and industrial robotics are used in this preliminary work. There is currently no suitable solution for the practice-relevant real-time dynamic simulation of industrial robots in virtual commissioning. The goal of this project is to be able to calculate precise and therefore computationally intensive dynamic models of industrial robots within a real-time co-simulation with high accuracy. To this end, signal-based multi-rate methods will be investigated and compared in terms of robustness and accuracy. Model-based multi-rate methods are then designed and adjusted using various model reduction methods. Additionally, the combination of signal-based and model-based methods will be made possible for the first time, thereby combining the advantages of both methods. To ensure the applicability of the results, a recommendation for action is to be derived depending on the modelling effort and achievable accuracy.

• DFG Programme: Research Grants
• Co-Investigator: Dr.-Ing. Armin Lechler