Research on Scheduling Based on Spatial and Temporal Distribution Characteristics of Piled Shared Bicycle

Abstract

With the development of bicycles, the problem of having no piles to park and no cars to use also appears. Because the driving mechanism of bicycle travel is not understood, the operation personnel in charge of redistribution have no way to start. In this paper, from the perspective of subway station and city bus transfer system, starting from the characteristics of road network, combining with the nature of land use, the spatial-temporal characteristics of quantitative analysis is carried out to study the spatial-temporal distribution law of travel behavior. Based on the above factors and the supporting conditions of the bicycle facilities, the common least square model was established. It was found that the number of stations in the traffic district was the most important variable, indicating that the basic supporting facilities had a great impact on the number of bicycle trips. Then, the number of bicycles and piles was deeply studied. Firstly, the stations are classified according to K-means, and defined as low-frequency high flow, high-frequency low flow, low-frequency low flow, and medium-frequency low flow according to the average of clustering results. Then, the redistribution suggestions are put forward respectively for these categories. For outflow clustering with large values of O and D, R can be predicted by BP neural network to provide reference for redistribution and distribution.