Optimized allocation of microgrids’ distributed generations and electric vehicle charging stations considering system uncertainties by clustering algorithms

• Published in: IET renewable power generation 2024-08, Vol.18 (11), p.1798-1818
• Main Authors: Yaghoubi‐Nia, Mohammad‐Reza, Hashemi‐Dezaki, Hamed, Niasar, Abolfazl Halvaei
• Format: Article
• Language: English
• Subjects: Algorithms; Alternative energy sources; Analysis; Battery chargers; electric vehicle charging stations; Electric vehicles; Monte Carlo method; optimisation; photovoltaic power systems; reliability; renewable energy sources; Simulation methods; smart power grids
• ISSN: 1752-1416; 1752-1424
• DOI: 10.1049/rpg2.13038

The reliability‐oriented optimized sizing and placement of electric vehicle (EV) charging stations (EVCSs) has received less attention. In addition, the literature review shows that a research gap exists regarding a clustering‐based method to optimize the allocation of DGs and EVCSs, considering the system uncertainties. This article tries to fill such a knowledge gap by proposing a new clustering‐based method to optimize the allocation of DGs and EVs simultaneously, considering the uncertainties of EV behaviours and stochastic behaviours of renewable DGs. Developing a new stochastic model for EVs using the clustering algorithm is one of the essential contributions. The uncertain parameters, e.g. EV charging loads based on EV owners’ behaviours (arrival time, departure time, and driving distance) and renewable DGs, would be clustered. The proposed method could solve the execution time challenges of Monte Carlo simulation‐based approaches to concern the stochastic behaviours of smart grids. The simultaneously reliability‐oriented optimal allocation of EVCSs, DGs, and protection equipment, using the proposed clustering‐based algorithm is another main contribution. The IEEE 33‐bus test system is studied to examine the introduced method. Simulation results imply that a 1.45% accuracy improvement could be obtained compared to available analytical ones, while its execution time is appropriate. The allocation of renewable/non‐renewable distributed generations and electric vehicle charging stations are optimized. Considering system uncertainties by clustering algorithms is the main novelty of this research.