Development of a Multi-Agent System for Optimal Sizing of a Commercial Complex Microgrid

In this paper, a novel intelligent method based on a multi-agent system (MAS) is applied to the problem of optimal sizing in a stand-alone office complex microgrid such that the electricity demand of the office building and the charging demand of the plug-in hybrid electric vehicle (PHEV) charging station are met. The proposed MAS-based architecture consists of five different agents, namely generation agent (GA), electrical load agent (LA), PHEV charging station agent (SA), control agent (CA), and design agent (DA) which are organized in three levels. In the proposed MAS, control agent coordinates the interactions between the generation agent and electrical load and charging station agents and following a request from design agent, sends the information on the operation of the microgrid with determined sizes to the design agent. According to the received information, design agent finds the optimal sizes of the system's components such that the electricity and charging demands are met considering two reliability indices to investment decisions. This study is performed for Kish Island in south Iran.