Robust energy management of micro‐grid system using an efficient RNN‐MDA approach

Abstract

This paper introduces a hybrid methodology of robust energy management for system modeling and configuration of Microgrid linked system. The proposed system is combined performance of both the Recurrent Neural Network (RNN) and Modified Dragonfly algorithm (MDA), hence it is known as RNN‐MDA. Microgrid connected system like photovoltaic, wind turbine, fuel cell, micro turbine, and battery. RNN predicts the optimal load demand of proposed system. MDA optimally configures the microgrid combination. Cost function of micro turbine, fuel cell, operation, and maintenance cost and the constraints involving power balance and generation capacity of each MG sources are involved. At last, the proposed model is instigated on MATLAB/Simulink working platform and performance equates through existing performances. Cost savings of proposed and existing for 24 h and 1 year is also analyzed in paper. The cost savings of the proposed technique for 24 h and 1 year are 54.8387% and 82.77%. Thus the proposed technique shows the better performance over other procedures based on cost savings, power generation, fitness function, error reduction, computation time and statistical parameters, cost accuracy percentage.