Comparative Realistic Objectives Oriented Optimization Framework for EV Charging Scheduling in a Distribution System

Abstract

The integration of large-scale electric vehicles (EVs) into the distribution system has emerged as a critical topic of research with the proliferation of EVs over the years. To mitigate the negative effects of EVs on the distribution system (DS), in this study, the optimal operation of an EVPL is investigated with a model in the form of mixed-integer quadratic constrained programming (MIQCP) that aims to minimize a variety of realistic objectives including active power losses, charging cost or voltage deviations while taking DS constraints into account. Also, uncertain behavior of the EVPL has been considered via machine-learning based forecasting by using historic data. The effectiveness of the proposed model has been evaluated using a 33-bus test system with 15-minute time granularity and compared to models that had various objective functions.