Multi-attribute Optimization of Electric Vehicle Fleet Charging Scheduling

Abstract

Electric vehicles become an important resource to achieve more reliable and sustainable power system. However, for the optimal integration of electric cars in power network, several objectives have to be fulfilled simultaneously: the minimization of the driver discomfort, maximization of the revenues coming from possible services to the grid and the maximization of the vehicle fleet charging station efficiency. This problem is solved using the multi attribute utility theory that allows different decision maker’s attitude toward risk. The final output is the decision algorithm for a day-ahead schedule of electric cars usage offering only secondary frequency regulation, from the fleet owner perspective. There are generally two approaches for the generation of solution sets for the multi objective optimization: scalarization and nonscalarization methods. Scalarization methods explicitly use a scalarization function to convert the problem into a single objective program. In our approach, we used multiplicative aggregation as one of the well-known scalarization technique. The optimization problem is set as a MINLP problem. The optimization function is nonlinear, with linear constraints.