DC micro — Grid pricing and market model

Abstract

The fundamental roots of micro-grids are different types of renewable energy sources. There are two broad and distinctive control set ups for power systems. They are centralized and decentralized (hierarchical) controls. In market models of micro-grids there are normally groups of electricity sources and loads that operate in synch with a centralized grid or macro-grid. This paper studies the functionality and ideas of micro-grids. Then implementing Artificial Neural Network (ANN) model for the proposed micro-grid in very precise manner is established. It proposes general simulation modeling for micro-grid using MATLAB, Simulink and (ANN). Its goal is to connect between the most important parameters in DC-Microgrid and price. This modeling approach proposes general Modeling and simulation at more probable situations for variable values at each bus. The ANN model for the proposed range of Different parametric characteristics is presented for Extended Analysis on IEEE 14-Bus Test System. Finally, algebraic equations for the ANN model are deduced in order to optimize them in the future for optimal micro-grid's performance. The training, testing and validating data for this ANN model is extracted from a real micro-grid to connect between numbers of units at each DG source (Distributed Generation), Loads, Minimum/ Maximum Power, Marginal Loss Factor and Time (Hour) over 24 hours as inputs, with Cost ($), Saving ($), Revenue ($), Profit ($) as outputs. So, it helps the humanity to understand more about renewable energy sources and techniques. Moreover, it presents an excellent model to predict the price and saving with this trend in power systems especially from the side of humans or customers. The work is useful for creating sustainable business model for energy access to energy deprived population. The paper's presentation includes examples and comparisons for approach's validity. Now, there is a running real-time validation for the work via OPAL real-digital-simulator.