A Python Test Environment for Multi-Agent Systems in a Large Electrical Distribution Grid Model

This paper provides a possible simulation approach for large multi-agent systems (MAS) with many electrical assets. Using public available data, an electrical 110 kV grid model of TEN Thüringer Energienetze GmbH & Co. KG (TEN) was imitated. Basic modelling assumptions are made for the available data. Machine controllers and parameters are derived from controllers of the European Network of Transmission System Operators for Electricity (ENTSO-E) transmission grid. Renewable energy sources (RES) are represented individually but also accumulated. The controls for RES are grid-following and are based on the controllers of the WECC Western Electricity Coordinating Council (WECC). Loads were distributed to the different substations via a range of key factors. RMS simulations show that the generated distribution grid can be operated stably even in fault cases. In the second part, Python test environment is introduced and the coupling with a commercial available power system software is shown. For this purpose, the co-simulation of a shared memory solution is used. The interaction between the MAS python test environment and PSS(r)SINCAL can be shown within a stable RMS simulation.