Multi-objective optimization for scheduling isolated microgrids

Abstract

Following up the recent innovations in smart microgrids as well as the continuous deployment of renewable energy resources (RES), the need for efficient operation of microgrids is increasing. Particularly, microgrid scheduling involves a challenging optimization problem, where continuous and discrete optimization tasks are to be solved. One of the challenges involved in this problem is to provide an optimal scheduling policy of the incorporated resources that can ensure maximum utilization of the RES and minimum operational cost of the conventional resources as well (which are not necessarily equivalent). From this perspective, this work proposes an optimal scheduling approach of a hybrid microgrid, incorporating battery storage system (BSS), conventional generation (e.g. diesel generator), and photovoltaic solar generators (PVs), by using dynamic programming (DP). The optimal switching strategy between conventional generation and BSS is explored firstly. Thereafter, the associated optimal charging and discharging policy of the BSS is investigated based on day-ahead power profiles of the load and renewable generation, which are all assumed predictable up to a realistic uncertainty. The proposed approach is compared with the conventional one which does not employ the feature of prediction.