Probabilistic energy management of a renewable microgrid with hydrogen storage using self-adaptive charge search algorithm

Micro Grids (MGs) are clusters of the DER (Distributed Energy Resource) units and loads which can operate in both grid-connected and island modes. This paper addresses a probabilistic cost optimization scheme under uncertain environment for the MGs with several multiple Distributed Generation (DG) units. The purpose of the proposed approach is to make decisions regarding to optimizing the production of the DG units and power exchange with the upstream network for a Combined Heat and Power (CHP) system. A PEMFCPP (Proton Exchange Membrane Fuel cell power plant) is considered as a prime mover of the CHP system. An electrochemical model for representation and performance of the PEMFC is applied. In order to best use of the FCPP, hydrogen production and storage management are carried out. An economic model is organized to calculate the operation cost of the MG based on the electrochemical model of the PEMFC and hydrogen storage. The proposed optimization scheme comprises a self-adaptive Charged System Search (CSS) linked to the 2m + 1 point estimate method. The 2m + 1 point estimate method is employed to cover the uncertainty in the following data: the hourly market tariffs, electrical and thermal load demands, available output power of the PhotoVoltaic (PV) and Wind Turbines (WT) units, fuel prices, hydrogen selling price, operation temperature of the FC and pressure of the reactant gases of FC. The Self-adaptive CSS (SCSS) is organized based on the CSS algorithm and is upgraded by some modification approaches, mainly a self-adaptive reformation approach. In the proposed reformation method, two updating approaches are considered. Each particle based on the ability of those approaches to find optimal solutions in the past iterations, chooses one of them to improve its solution. The effectiveness of the proposed approach is verified on a multiple-DG MG in the grid-connected mode.