基于人流密集度的网点数量与布局预测模型——以宁波市公共自行车系统为例

针对宁波市公共自行车网点数量增加而新使用者增量和日周转率呈现下行的趋势, 对网点数量和布局进行分析和建模. 首先, 通过K-means聚类方法对不同网点按日周转率进行分类, 结果发现日周转率在4以下的低效网点高达52.86%, 低于0.5的网点占9.73%, 而在15以上的超负载网点占2.5%, 两极分化现象非常明显. 其次, 讨论了宁波市城市公共自行车专项规划中基于面积和服务人口的网点数量预测模型的不足之处. 最后, 以较为成熟的城市公交线路为基础, 通过统计各公交站点的人流密集度, 提出了基于人流密集度的网点数量与布局预测模型, 并给出了网点在空间上的布局. 同时指出各城市公共自行车网点数量与布局问题既有发展共性又有地域个性, 基于公交线路人流密集度模型具有普适性价值.

Traffic Density Based Demand Estimate and Layout Forecast for Public Bicycle Rental Stations in Ningbo

In 2014, the number of public bicycle rental stations in Ningbo was up by 46.34%, while new users and the daily turnover rate have shown an obvious decrease trend. In this paper, analysis is conducted and a model is built on both the number and layout of the rental stations. Firstly, through the K-means clustering approach, the different rental stations are classified by the daily turnover rate, and the proportion of “low-efficient rental stations” below the daily turnover rate of 4 is found to be as high as 52.86%. Also found is that the “zombie rental stations” is below 0.5, 9.73%, and the “overload rental stations” is over 15, 2.5%, showing an obvious polarization phenomenon. In addition, the number forecast model of rental stations is discussed in terms of area and service people in Ningbo Public Bicycle Plan, discovering that the model needs to be improved in more detail manner. Finally, based on the mature urban public bus routes and the traffic density, the number and layout forecast model are put forward according to the traffic density and the space layout of rental station.