基于多源数据的通勤高峰期出行方式分担率预测方法研究  

Prediction Method of Travel Mode Share Rate in Commuting Peak Period Based on Multi-source Data

作　　者：宋永朝[1] 杨培 SONG Yongchao;YANG Pei(School of Traffic ＆ Transportation, Chongqing Jiaotong University, Chongqing 400074, P. R. Chin)

机构地区：[1]重庆交通大学交通运输学院,重庆400074

出　　处：《重庆交通大学学报（自然科学版）》2018年第5期84-91,共8页Journal of Chongqing Jiaotong University(Natural Science)

基　　金：长沙理工大学公路工程教育部重点实验室开放基金资助项目(kfj130102);重庆交通大学研究生教育创新基金项目(20140107)

摘　　要：鉴于大城市普遍存在交通拥堵问题,而通勤高峰期间尤为突出,科学合理开展城市交通规划是应对交通拥堵问题有效途径。传统出行方式分担率预测模型是基于人工调查数据,数据获取成本高且样本量有限,难以准确预测其分担率。采用路网数据、公交线网数据、公交站点数据、户籍数据、工作地数据等多源数据,依据交通工具的服务范围和公交站点的吸引范围,根据最短路径算法以及公共交通选择算法,对通勤高峰期居民的出行方式进行预测,从而得出不同出行方式的分担率,获取交通通勤出行分布规律。通过以重庆市主城区为例,进行城市交通通勤典型数据分析,验证了该方法的可靠性及准确性。The urban traffic congestion is a universal problem in big cities,especially in commuting peak period. Traffic planning in advance is an effective approach to deal with traffic congestion problem. The traditional prediction model of travel mode share rate was based on the artificial survey data,and it was difficult to predict the share rate accurately,due to its expensive acquisition cost and quantitative restriction of data samples. Multi-source data was utilized to analyze traffic travel share rate,including road network data,public transportation network data,public transportation stops data,household registration data,and workplace data etc. According to the service scope of the vehicle and the attraction scope of the bus station,the travel mode of residents during commuting peak period was forecasted by using the shortest path algorithm and the public transportation selection algorithm. Therefore,the share rate of different travel modes and the distribution rule of commuter trips were obtained. Taking Chongqing urban area as an example,the typical data analysis of urban commuter was carried out to verify the reliability and accuracy of the proposed method.

关 键 词：交通运输工程 通勤出行 多源数据 分担率 非集计模型 

分 类 号：U491[交通运输工程—交通运输规划与管理]