Multi-stage stochastic framework for simultaneous energy management of slow and fast charge electric vehicles in a restructured smart parking lot

Abstract A widespread appeal of electric vehicles entails, among other elements, the provision of fast charging stations to reduce the range anxiety. We present a novel structure and operating mechanism for EV parking lots where, traditionally, parked vehicles were charged in level-1 or level-2 charging modes. Due to ubiquitous presence of these parking lots in all urban and residential areas, the proposed approach actually leads to creation of just as many fast charging stations each accommodating a modest number of fast charging vehicles in proportion to the base capacity of the parking lot. A three-stage scheduling framework based on stochastic programming and MPC is introduced, consisting of day-ahead, periodic real-time and intra-period real-time planning for joint scheduling task of inflexible loads, slow- and fast-charged vehicles. The comprehensive fast charge allocation algorithm incorporates V2V concept, M/G/∞ queuing theory and a resource priority stack based on Most Laxity First concept. By providing energy to EVs of different charging classes, the proposed model can be considered a unified approach to prevent or at least reduce the need for separate fast charging stations in urban areas. Comprehensive examples demonstrate the effectiveness of the proposed approach in handling charging load of different EV classes.