Using cluster analysis and genetic algorithm for multi-objective optimisation of hybrid electricity supply systems: the case of a photovoltaic/wind/ battery grid-connected system in Yazd, Iran

Abstract

ABSTRACT In this research, a grid-connected hybrid renewable electricity system was studied with the aim of providing the required electricity to the buildings of Yazd city in Iran. All possible structures of the proposed system were simulated using the genetic algorithm and TRNSYS software. Then the final stable population was categorised into three clusters using the statistical quartiles. Based on the results, only about 1% of all investigated structures were optimum considering energy cost and CO2 emission. Investigating the best-identified structures showed that it could decrease the CO2 emission by about 92% and energy cost by about 35% compared to the case where only the grid electricity was used. Moreover, wind energy had the highest priority, and using batteries was not an appropriate solution for the study area. The results showed that identifying an optimum cluster can be used efficiently instead of as a unique solution for the studied system. Highlights Hybrid electricity supply system. Combination of Cluster Analysis and Genetic algorithm for multi-objective optimisation. Simulating the control strategies for energy systems.