不同时空出租车供求匹配程度

Taxi Resource Allocation in the“Internet+”Era

打车软件实现了乘客于出租车司机之间的信息互通,在一定程度上提高了出租车资源的供求匹配程度,缓解了打车难问题。但对于不同时空来说,时间与空间的不同都会影响出租车资源供求匹配程度。本文基于“苍穹——滴滴快的大数据平台”,以成都市为例,以成都市各地区对出租车的需求程度的不同,对出租车运行经纬度进行K-mean聚类,将成都分为四个城区;以乘客与司机间的关系,确立出租车供求匹配指标:未成单率a't和空车率。选择一个城区为研究对象,以小时为单位,定义距离最近原则和等可能到达原则,对时间加权平均得到每个城区的a',b',求出在成都市不同时空的出租车资源供求匹配程度:3好城区的匹配程度最好,4号城区的匹配程度最差;7点(早高峰)和17点(晚高峰)匹配程度最好。分析当时打车软件的补贴情况,由于价格补贴会影响供给量、需求量、可行域范围,构建关系图,发现三者成循环的关系:需求量和供给量相互依赖。并给出使用博弈论研究补贴方案的方向。

The taxi software realizes the information exchange between passengers and taxi drivers,which improves the matching of supply and demand of taxi resources to a certain extent,and alleviates the difficulty of taxiing.However,for different time and space,the difference between time and space will affect the matching of supply and demand of taxi resources.This article is based on"Cangwu-Drips Big Data Platform",taking Chengdu as an example,taking the degree of demand for taxis in various regions of Chengdu,and performing K-mean clustering on the latitude and longitude of taxis,which will be divided into Chengdu.For the four urban areas;the relationship between passengers and drivers,the matching indicators of taxi supply and demand:unsuccessful and empty rate.Select a city as the research object,define the distance from the nearest principle and the equal-reaching principle in hours,and obtain the time-weighted average of each urban area to find the matching of the supply and demand of taxi resources in different time and space in Chengdu:3 Good City The match is the best,and the matching degree in the 4th city is the worst;the 7 points(early peak)and 17 points(late peak)match the best.