Multi-Objective Non-Dominated Sorting Genetic Algorithm Optimization for Optimal Hybrid (Wind and Grid)-Hydrogen Energy System Modelling

Abstract

In this paper, an optimal hybrid (wind and grid)-hydrogen energy system (H-HES) is proposed using multi-objective non-dominated sorting algorithm (NSGA-II) optimization. The H-HES consists of the main energy system; wind-energy system (W-ES), which supplies a proton exchange membrane (PEM) electrolyzer via an energy management system (EMS) and a rectifier. In addition, the grid-energy system (G-ES) is available to support the W-ES to meet the PEM electrolyzer’s energy demand, and the EMS facilitates control between the W-ES and G-ES. The W-ES is modelled using wind data from Wind Atlas South Africa (WASA) for six Renewable Energy Development Zones (REDZs) in South Africa and their appropriate wind turbine models. The selection of appropriate wind turbine models is guided by the optimal wind turbine variables obtained from NSGA-II corresponding to the optimal H-HES model. The optimal H-HES model is developed using two objective functions: cost of electricity and efficiency, which are minimized and maximized respectively and evaluated using NSGA-II available in Pymoo framework. NSGA-II successfully converges to a Pareto front, and the best solution for the H-HES cost of electricity and efficiency for each wind REDZ is determined by compromise programming; a multi-criteria decision-making technique available in Pymoo. From the optimal cost of electricity and efficiency solutions, optimal variables are successfully obtained for optimal modelling of the H-HES for each wind REDZ.