A machine learning-integrated multi-criteria decision-making approach based on consensus for selection of energy storage locations

Abstract

In this research, the location of energy storage systems (ESS) is decided by comparing and contrasting multi-criteria decision-making (MCDM) methods and machine learning (ML) techniques. MCDM methods are better than mathematical methods because they can take into account more than one criterion and give a clearer indication of preference. Furthermore, the integration of ML methods into the MCDM methods can enhance decision-making capabilities. Based on the geographic coordinates of wind power plants, K-means++ and elbow methods are used to determine the alternative locations for ESS. TOPSIS, ARAS, EDAS, and MOORA are chosen to rank the alternatives according to the defined criteria. BORDA approach is used to combine the rankings, and a consensus is established on the overall rating. The findings of this research suggest that the combination of ML and MCDM techniques accurately identifies and prioritizes potential locations for ESS by reducing the size of the alternative set. Therefore, the complicated structure that was present during the decision phases is simplified, allowing the decision-makers to work more quickly and with less effort. The results indicate a reference for energy providers to select an appropriate location for ESS according to specific criteria.