Ev charging management in a real-time optimization framework considering operational constraints

Abstract

The electrification of transportation plays a central role in the decarbonization of energy systems. Although electric vehicles (EVs) are expected to reduce energy related CO2 emissions, the increasing demand imposed by large scale EV adoption presents serious challenges for distribution systems (DSs), which were not originally designed to accommodate such loads. This study proposes a mixed integer quadratically constrained programming (MIQCP) framework to optimize the operation of an EV parking lot (EVPL) under DS constraints. The model compares three widely adopted objective functions: minimization of active power loss, charging cost, and uncontrolled charging impact, which is represented by minimizing the total charging time. The optimization is based on an AC power flow formulation that explicitly captures voltage limits, load factor, and active and reactive power constraints. A rolling horizon based real time optimization strategy is employed to manage forecast uncertainties in EV behavior and photovoltaic generation. Real world conditions are reflected by incorporating two actual DS topologies from T & uuml;rkiye and ten EV types with different technical specifications, evaluated at fifteen minute resolution. The results show that cost oriented EV charging strategies lead to the most adverse effects on system operation, including a %23.06 increase in distribution line losses and a %38.56 reduction in load factor compared to a base scenario without EVs. These findings highlight the critical need for objective aware planning by distribution system operators and aggregators, particularly in the context of growing EV penetration and uncertain renewable integration.