A Many-Objective Investigation on Electric Vehicles’ Integration Into Low-Voltage Energy Distribution Networks With Rooftop PVs and Distributed ESSs

Abstract

Integrating electric vehicles (EVs), residential PVs, and energy storage systems (ESSs) into low-voltage distribution networks poses significant challenges, particularly regarding grid capacity and efficiency. This study addresses these challenges by developing distributed energy management systems for the EVs and the ESSs, and sizing approaches for the PVs and the ESSs, by enabling multi-agent systems and fuzzy cognitive maps. The optimization framework simultaneously considers six key objective functions which studied separately or in limited combinations: (i) ESS installation cost, (ii) low-voltage network operation parameters, (iii) active power losses, (iv) EV charging times, (v) total energy costs, and (vi) excess energy from PV systems. Using environmental conditions and energy purchase costs to define two regions, Heraklion, Greece, and Berlin, Germany, under varying EV integration rates, the proposed methodology enables the IEEE European Low Voltage feeder as a case study and Euclidean distance is used to identify optimal solutions from the Pareto frontiers. Results show that the integration of the distributed ESS contributes to a significant reduction in network’s energy consumption, while reduced solar irradiance in Berlin leads to different cost and operational patterns. Heraklion generally achieves better cost efficiency due to higher solar irradiance. Additionally, network operational parameters are significantly improved, and energy costs for residences are reduced up to 39%.

Keywords

• Electric vehicles
• energy distribution network
• distributed energy resources
• many-objective optimization
• fuzzy cognitive maps
• multi-agent systems