New energy bound-based model for optimal charging of electric vehicles with solar photovoltaic considering low-voltage network's constraints

Abstract

This paper introduces a linear programming (LP)-based optimisation method of charging electric vehicles (EVs) in a decentralised fashion. It exploits the available photovoltaic (PV) power to charge EV batteries while maintaining the low-voltage (LV) network within its operational limits. A new energy-bound model is implemented in order to meet the connected EVs energy requirements. This model highlights two main aspects: first, the proposed formulation seeks to compute both the upper and lower energy boundaries from the arrival energy and no from zero. Second, the charging power is dynamically adjusted by combining a fixed and variable charging rate to assure the technical limits of the network. This means maximising power delivered to all EVs for a given period by optimising the charging rate of each EV connected. Besides, a network sensitivity analysis technique is developed to manage voltage and loading constraints. The accuracy of the proposed linear approximation was tested simulating two cases (moderate and high penetration level of PVs and EVs) on a real LV feeder. Results over a set of simulations for winter and summer seasons demonstrate that this method can be effectively implemented as a charging strategy and for energy planning studies.