This paper presents a methodology for sizing and operating microgrids using an energy management system (EMS) strategy embedded in the software HOMER Pro. The proposed EMS-based strategy, represented by a mixed-integer linear model, determines the optimal day-ahead operation of a grid-connected microgrid, which considers photovoltaic generation, battery storage system, and energy purchases and sales from/to the grid while minimizing the total operational costs. The integration of the optimal EMS-based dispatch strategy to HOMER Pro is done through the AMPL API for Matlab and the Matlab Link module. The proposed methodology was tested using real data from the CampusGrid microgrid that will be deployed at the UNICAMP campus. Results show that the proposed strategy presents economical advantages over the default dispatch strategies provided by HOMER.