Two-Level Trip Selection and Price Incentive Scheduling in Electric Vehicle-Sharing System

The rebalance operations have been an essential problem in car-sharing service. In this paper, a two-level price incentive trip selection process is proposed to mitigate the imbalance issue in an electric vehicle-sharing (EVS) system. Specifically, at the perspective of customers, a trip price plan is made based on the adoption rate incorporating stochastic utility function in the first-level trip selection. The second-level selection adopts part of customers kept in the first-level selection, which brings less reposition cost happened in the scheduling operations in the EVS service. In the two-level trip selection process, the uncertain parameters, i.e., customers’ price expectation, potential travel demand, are assumed as random variables with known statistical measures, e.g., marginal moments, obtained from the real world. And the corresponding worst-case chance constraints combined with these random variables are further approximated as the convex optimization. In a real-world case study, the computational results demonstrate several economic and environmental benefits of our two-level selection program in the EVS system.