文章编号： 1672-5328（2016）01-0059-06

庄立坚，宋家骅，段仲渊，丘建栋
(深圳市城市交通规划设计研究中心有限公司，深圳市交通信息与交通工程重点实验室，广东深圳518021)

摘要： 为支持出租汽车运营评价和行业规划，科学掌握其运营现状与特点，以深圳市一个典型工作 日为例，研究浮动车数据挖掘在出租汽车运营特征研究中的应用。依托对浮动车数据的特征解析， 提出面向出租汽车运营特征分析的数据处理流程，针对空驶率、运营速度、载客运营时长、出行需 求的时空差异等方面分析深圳市出租汽车行业运营现状与特点。该研究可为出租汽车运营评价提供 新视角，从对既有数据深度挖掘的角度提升行业效益。

关键词： 智能交通系统；道路交通；浮动车数据；数据挖掘；出租汽车；运营特征

中图分类号： U491

文献标识码：A

A Floating Car Data Mining-based Taxi Operation Characteristics

Zhuang Lijian, Song Jiahua, Duan Zhongyuan, Qiu Jiandong
(Shenzhen Urban Transport Planning Center Co., Ltd., Shenzhen's Key Laboratory of Traffic Information and Traffic Engineering, Shenzhen Guangdong 518021, China)

Abstract： To improve the efficiency of taxi operation and planning, this paper aims to apply floating car data (FCD) mining technology in studying taxi's operational characteristics as well as evaluating its operational performance for Shenzhen city. Based on features mined from FCD, a detailed working procedures of analyzing taxi operational characteristics is developed. With developed method, taxi unloaded rate, operational speed and time, temporal/spatial differences of travel demands are able to extract and better understand the performance of Shenzhen taxi system. The proposed method offers a new technology of FCD mining in taxi data exploration which could be used to evaluate taxi operational performance as well as improving its operational efficiency.

Keywords： intelligent transportation systems; road traffic; floating car data (FCD); data mining; taxi; operational characteristics