Implementation of a Multi-Agent control scheme for autonomous migrogrids

This thesis describes the implementation of a Multi-Agent control scheme for autonomous microgrids. This is realized by a Multi-Agent system, which executes a three-stage algorithm to control the voltage at the nodes and the power balance of the microgrid. The power system is simulated by using a Real Time Digital Simulator (RTDS). The Gauss method is used for distributed calculation of power flows. Additionally, using a Multi-Agent system in a low-voltage distribution network to fulfill voltage constraint, power balance and frequency droop for low voltage are also discussed. In the next step a script feature of RSCAD and file sharing protocol over the local area network (LAN) is adopted for establishing of necessary communication between the agents and the RTDS. Numerous hardware configurations for the interface were examined during the research work. This interfacing shows a flexible communication structure to exchange the data. The link between the agents and the RTDS has been examined and the results are given. The final stages of the project an interface using the DAQ toolbox was investigated. The communication speed among the agents, the RTDS, and the hardware available were a constraint that needed to be considered for constructing the framework. However, the communication between the agents and the system has been established with regard to the communication speed.