基于多元Logit模型的城市公交准时可靠性分析  

Bus On-time Performance Evaluation Based on Multinomial Logit Model

作　　者：曹眉舒 刘刚[2] 廖明军[3,4] 王凯英[3] 

机构地区：[1]山东高速科技发展集团有限公司,山东济南250031 [2]福州大学土木工程学院,福建福州350116 [3]北华大学汽车与建筑工程学院,吉林吉林132013 [4]道路灾变防治及交通安全教育部工程研究中心,湖南长沙410076

出　　处：《北华大学学报（自然科学版）》2016年第2期261-265,共5页Journal of Beihua University（Natural Science）

基　　金：国家自然科学基金项目(51338008);长沙理工大学交通运输工程及测绘科学与技术领域研究基地开放基金项目(kfj150401);吉林市杰出青年科技引导项目(201467009)

摘　　要：以多维度大数据为基础,分析不同时间、不同天气条件下的公交准时可靠性.结果表明:工作日和晚高峰期间的准时比例明显低于其他时段,周六和周日午夜的准时比例明显低于工作日午夜的准时比例;利用多元Logit模型预测了公交早到、晚到和准时与环境变量、时间变量及运行特性等之间的定量关系.结果表明:公交准时性在晚高峰、线路末端和恶劣天气条件下降低.研究结果可为出行者制定出行计划提供依据,减少等待时间;同时可为管理者优化公交发车间隔,提高公交服务水平提供有效的决策支持.Based on multi-source big data,this study first analyzed the temporal bus on-time performance under different weather conditions. The results showed that the on-time performance during weekday and peak hours was worse than it in other time periods. The on-time performance in the midnight during weekend was also worse. A multinomial logit model was developed and estimated,which relates early,late and on-time bus arrivals to environmental,temporal and operating characteristics. The model results showed that the probability of on-time failures increases during afternoon peak periods,as buses progressed further along their routes and under adverse weather conditions. An understanding of bus service reliability was necessary to help traveler make reasonable travel plan to decrease waiting time,and to develop strategies that optimize transit operation and help transit agencies provide better services.

关 键 词：公交服务可靠性 准时性 多源数据 多元LOGIT模型 

分 类 号：U121[交通运输工程]