Assets Forecasting and Power Management of DC-Based MG under Dynamic Pricing for Smart Cities

Abstract

Due to the penetration of different renewable energy sources and DC storage systems, DC microgrids (MGs) have been studied extensively. The optimal planning of DC microgrids directly impacts the operation and control algorithms; thus, coordination among them is required. The safe operation of DC microgrids has been ensured by different control techniques, such as centralized, decentralized, distributed and hierarchical control. This paper studies a DC-MG planning model. A time series and feedforward nonlinear autoregressive neural network (NARX) have been used for forecasting the load demand and level of solar Irradiance. Furthermore, the total operating cost has been minimized via optimization of the unit commitment problem while including the dynamic pricing of grid power. The system performance have been evaluated using MATLAB.