基于趋势预测的合乘收益研究  

The Research on Taxi Sharing Benefit Based on Tendency Estimation

作　　者：宋飞[1] 李蓉[1] 张思东[1] 张宏科[1] 

机构地区：[1]北京交通大学电子信息工程学院下一代互联网互联设备国家工程实验室,北京100044

出　　处：《电子学报》2014年第7期1353-1359,共7页Acta Electronica Sinica

基　　金：国家973重点基础研究发展计划(No.2013CB329100);国家科技重大专项(No.2012ZX03005003-04);高等学校博士学科点专项科研基金(No.20120009120005)

摘　　要：首先提出一种基于车载自组网的城市出租车合乘的方案;其次针对北京市6499辆出租车的GPS信息进行了较为全面的数据分析,获取81785个乘客上车时间及位置,从整体和局部两个方面定量地讨论了出租车合乘带来的益处,运用灰色预测和曲线拟合两种不同的方法建立数学模型预测上车点个数的变化趋势;最终得出在北京市六环内乘客采用合乘方式后,可减少1/3以上的上车点以及约20～30%的行驶里程,对缓解交通压力及构建绿色城市具有十分重要的意义.An urban taxi sharing scheme based on VANET is proposed firstly. We elaborates the necessary processing of the data derived from 6499 GPS sets of Beijing taxis in operation and mainly analyzes the data from both macro and micro scopes. The location and timing of 81785 picking up points are obtained as well. The feasibility and benefits of the sharing scheme are discussed. Two different prediction methods, i. e. grey prediction and curve fitting, are employed to create mathematical models to predict the trend of picking up points. The conclusion is that by adopting the taxi sharing method,more than 1/3 of the picking up points and about 20 - 30% mileage will be reduced, which will certainly contribute towards relieving the traffic pressure and creating an envi- ronmentfriendly city.

关 键 词：车载自组网 出租车 合乘方案 数据分析 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]