Supervisory-learning microgrid control strategy in the presence of energy storage systems based on uncertainties and energy market interactions to improve stability

This investigation delineates an optimal power management platform that incorporates the involvement of plug-in hybrid electric vehicle units (PHEVs) alongside the demand response (DR) programs execution, which are founded on the adaptive neuro fuzzy inference system (ANFIS) algorithm, aimed at regulating power, voltage, and frequency. The proposed model is predicated on power injection and absorption control algorithms (day-ahead algorithm) that facilitate interactions between microgrids (MGs) and the main grid. In the proposed model, an optimal energy distribution strategy for PHEV units and renewable energy sources has been established, considering the state of charge of PHEV units and their contribution to the regulation of the MG frequency based on the central and local controller architectures. DR programs are delineated into two distinct categories, to alter the consumption subscriber behaviors and achieve a more uniform load profile. The day-ahead forecasting algorithm is implemented by the local controllers, which subsequently relay the outcomes to the central controller. In addition to determining the optimal distribution of power, the central controller also formulates the optimal power supply strategy predicated on day-ahead forecasting algorithms. The proposed method leads to a 16.61 % reduction in MG losses, an 18.49 % reduction in operational costs, a 23.38 % increase in profit and a 33.64 % reduction in uncertainty-related penalties, and flattening the load profile. The findings indicate a significantly enhanced efficiency of the proposed methodology in comparison to other methods.