Depot constructions prediction model for expanding One-way Carsharing operations to a new city

Inhalt:
With the continuous improvement of people's travel demand, carsharing has gradually entered the public eye. Different usage models have different requirements for the operation of enterprises. It is extremely important for one-way carsharing (OWC) company to predict long-term demand and then provide the basis for depot location and related constructions. Currently, the studies on OWC demand prediction took order quantity of the stations as demand simply, ignoring information of multiple small sub-trips among dwell events during one trip. To address this issue, this paper aims to further classify carsharing demand into three categories according to usage patterns and establish demand forecast models separately. Based on carsharing data in Lan Zhou, the factors that affecting the demand are extensively studied in operating area, including Point of Interest (POI), operational parameters, transportation facilities, and accessibility. According to the operation data, we can analyze the deep information of the trip chains. Additionally, the demand forecast models are used to estimate demand in a new area and serve depot constructions support. The approach can predict the potential demand for new cities, which can reduce the risk of supply-demand imbalance and improve the station utilization.