부산대학교 도서관

Self-sustainability of micro grids is an important challenge to be addressed, particularly when they operate disconnected from the utility grid. The multi-agent systems paradigm allows implementing intelligent and predictive strategies towards an efficient management of the power flow. In this paper, an agent-based model is introduced, optimizing the load demand, through load prioritization, by shedding and rescheduling, and using forecasting mechanisms to improve the micro grids self-sustainability. The designed agent-based model and strategies were implemented and tested in the Repast agent-based modelling and simulation framework. Experimental results show improvements in the self-sustainability of the micro grid under analysis when operating in critical mode.