Agent-based distributed hierarchical control of dc microgrid systems

In order to enable distributed control and management for microgrids, this paper explores the application of information consensus and local decisionmaking methods formulating an agent based distributed hierarchical control system. A droop controlled paralleled DC/DC converter system is taken as a case study. The objective is to enhance the system efficiency by finding the optimal sharing ratio of load current. Virtual resistances in local control systems are taken as decision variables. Consensus algorithms are applied for global information discovery and local control systems coordination. Standard genetic algorithm is applied in each local control system in order to search for a global optimum. Hardware-in-Loop simulation results are shown to demonstrate the effectiveness of the method.