Integrated Configuration and Optimization of Electric Vehicle Aggregators for Charging Facilities in Power Networks With Renewables

Abstract

This paper investigates the optimal configuration and control of electric vehicle aggregators equipped with smart meters in the regional power distribution network. The mixed optimization of configuration and control is originally formulated as a mixed integer nonlinear programming (MINLP) problem with the stochastic operation model of EV aggregations. The controllable charging load is calculated by using a queuing model with non-homogeneous Poission process and integrated into the reliability constrained economic dispatch of the power system. Benders decomposition technique is adopted to lessen the dimensional and computational complexities introduced by the charging control of massive EVs. The total operating cost is minimized in the master problem, and charging plan for each individual EV is made by EVA in the sub-problems. In the case of any subproblem violations in the feasibility check of the optimal solution, Benders cuts are produced and added to the master problem. The proposed method is demonstrated in a 123-bus test network and EVA optimization results are compared with the direct control of EV aggregations for validating the effectiveness of the proposed method in the electrification of automobiles.