A bilevel optimization approach for Balancing Markets with electric vehicle aggregators and smart charging

Abstract

Demand Response (DR) programs can help alleviate the management of the electrical distribution grid by reducing loads in specified areas. They can be enabled within the energy Balancing Market (BM). Aggregators can manage different customers providing flexibility. Recently, Electric Vehicles Aggregators (EVAs) have emerged as significant players in the BM because they can manage fleets of electric vehicles (EVs) in the distribution grid. This paper addresses a multi-objective optimization problem for a distribution power grid that includes EVs and smart charging parks. At the higher level, the Distribution System Operator (DSO) considers the characteristics of each BM actor to minimize costs. Meanwhile, EVAs focus on controlling EV charging at the lower level to maximize their profit. The optimization problems of EVAs and other actors are replaced by KKT (Karush–Kuhn–Tucker)​ conditions, which are embedded as constraints in the DSO decision problem. Moreover, the resulting bilinear terms (in the optimization problem constraints) are linearized to fasten the finding of an optimal solution. The overall optimization problem is a mixed-integer quadratic programming (MIQP) and has been applied to the IEEE 13-bus test benchmark. The results demonstrate a reduction of about 6% of power loss in the grid achieved by the developed model. Besides, the linearized model can afford a more discretized model due to the reduction of computational effort.