A multiagent system for residential DC microgrids

This paper proposes a multiagent system (MAS) modeling and control architecture for a residential DC microgrid (RDCMG). The RDCMG mainly consists of a solid state transformer (SST) and DC smart homes equipped with DC loads, a home photovoltaic system, an energy storage system, electric vehicle (EV)/plug-in hybrid EV and other controllable loads, and advanced sensing and communication devices. The proposed MAS consisting of individual smart home agents and a control agent aims to minimize the electricity costs for the smart homes and alleviate the peak load of the SST during operation. These are achieved by agent utility functions and the best operating time algorithm in the MAS. The proposed MAS is validated by simulation studies for a RDCMG with five smart homes.