A static bike repositioning model in a hub-and-spoke network framework

This paper addresses a static bike repositioning problem by embedding a short-term demand forecasting process, the Random Forest (RF) model, to account for the demand dynamics in the daytime. To tackle the heterogeneous repositioning fleets, a novel repositioning operation strategy constructed on the hub-and-spoke network framework is proposed. The repositioning optimization model is formulated using mixed-integer programming. An artificial bee colony algorithm, integrated with a commercial solver, is applied to address computational complexity. Experimental results show that the RF can achieve a high forecasting accuracy, and the proposed repositioning strategy can efficiently decrease the users' dissatisfaction.