Institute of Transportation, MOTC

       This study addresses the optimal allocation of a fleet of plug-in electric vehicles (EVs) to the stations of an EV-sharing system at the beginning of each day from the tactical planning perspective of the system operator. Stochastic demand is represented using a set of discrete scenarios with given probabilities. The objective is to maximize the expected profit for the system operator. A multi-layer time-space network is constructed to describe the movement of EVs in the system. A stochastic vehicle allocation model for EV sharing systems is then developed based on the multi-layer time-space network.
       A set of computational experiments are conducted based on the data provided by the operator of the EV-sharing system deployed in the Sun-Moon Lake National Park in Nantou, Taiwan. A C++ program integrating Gurobi Solver is developed to solve the proposed model. The results show that the proposed model is able to effectively generate optimal fleet allocations for the test instances with stochastic demand. Moreover, sensitivity analyses of the impact of the model parameters on solution performances are also conducted. This study also utilize EVPI and VSS to evaluate the performance of the stochastic model. This study provides a decision support tool that facilitates operators of EV-sharing systems in effectively determining the fleet deployment in tactical planning