Key Techniques Of Public Bicycle System Optimization In Beijing

With the rapid development of the bicycle sharing system,it presents a number of problems in the process of rapid development.The previous researches on traditional public bicycle are based on the small sample survey and the main content is planning,but fewer researches on optimization method of public bicycle system.Therefore,the study of historical travel patterns is very important for the network structure optimization of the public bicycle system.The paper is organized as follow: firstly,made analysis construction status,operation characteristics,travel characteristics and user characteristics through the data from survey and public bicycle system.The time and space characteristics of public bicycles are excavated deeply.Based on that,public bicycle travel OD network is generated by GIS and eight kinds of typical structures of single station are extracted including scattered structure,single-line structure,isolated point structure,edge radiation structure,internal radiation structure,ring-shaped radiation structure,uniform network structure,and strong ring-shaped radiation structure.In order to better quantify these structures,three structural characteristic attributes are established including expected line density,radiation intensity of critical stations and radiation angle of critical stations.Meanwhile,a typical structure simulation and identification method is established based on ant colony algorithm.Then,the 23 types of factors affecting public bicycle travels were determined through investigation and screening.Finally,a model for predicting the public bicycle travel structure based on the BP neural network is constructed through studying the relationship between the influence factors and the characteristic attributes,which takes 23 types of factors as input and takes three kinds of typical characteristics as output,By changing the input variables——influencing factors,to get different output variables —— travel structure features,then make travel structure optimization of bicycle system by changing the influencing factors.This paper provides the key technical support and important theoretical basis for the optimization of the public bicycle system.