Classification of Livebus arrivals user behavior

With the increasing use of Intelligent Transport Systems, large amounts of data are created. Innovative information services are introduced and new forms of data are available, which could be used to understand the behavior of travelers and the dynamics of people flows. This work analyzes the requests for real-time arrivals of bus routes at stops in London made by travelers using Transport for London's LiveBus Arrivals system. The available dataset consists of about one million requests for real-time arrivals for each of the 28 days under observation. These data are analyzed for different purposes. LiveBus Arrivals users are classified based on a set of features and using K-Means, Expectation Maximization, Logistic regression, One-level decision tree, Decision Tree, Random Forest, and Support Vector Machine (SVM) by Sequential Minimal Optimization (SMO). The results of the study indicate that the LiveBus Arrivals requests can be classified into six main behaviors. It was found that the classification-based approaches produce better results than the clustering-based ones. The most accurate results were obtained with the SVM-SMO methodology (Precision of 97%). Furthermore, the behavior within the six classes of users is analyzed to better understand how users take advantage of the LiveBus Arrivals service. It was found that the 37% of users can be classified as interchange users. This classification could form the basis of a more personalized LiveBus Arrivals application in future, which could support management and planning by revealing how public transport and related services are actually used or update information on commuters.     

A mathematical programming technique for matching time-stamped records in logistics and transportation systems

Time-stamped data for transportation and logistics are essential for estimating times on transportation legs and times between successive stages in logistic processes. Often these data are subject to recording errors and omissions. Matches must then be inferred from the time stamps alone because identifying keys are unavailable, suppressed to preserve confidentiality, or ambiguous because of missing observations. We present an integer programming (IP) model developed for matching successive events in such situations and illustrate its application in three problem settings involving (a) airline operations at an airport, (b) taxi service between an airport and a train station, and (c) taxi services from an airport. With data from the third setting (where a matching key was available), we illustrate the robustness of estimates for median and mean times between events under different random rates for “failure to record”, different screening criteria for outliers, and different target times used in the IP objective. The IP model proves to be a tractable and informative tool for data matching and data cleaning, with a wide range of potential applications.     

Mining missing train logs from Smart Card data

This paper shows how to recover the arrival times of trains from the gate times of metro passengers from Smart Card data. Such technique is essential when a log, the set of records indicating the actual arrival and departure time of each bus or train at each station and also a critical component in reliability analysis of a transportation system, is missing partially or entirely. The procedure reconstructs each train as a sequence of the earliest exit times, called S-epochs, among its alighting passengers at each stations. The procedure first constructs a set of passengers, also known as reference passengers, whose routing choices are easily identifiable. The procedure then computes, from the exit times of the reference passengers, a set of tentative S-epochs based on a detection measure whose validity relies on an extreme-value characteristic of the platform-to-gate movement of alighting passengers. The tentative S-epochs are then finalized to be a true one, or rejected, based on their consistencies with bounds and/or interpolation from prescribed S-epochs of adjacent trains and stations. Tested on 12 daily sets of trains, with varying degrees of missing logs, from three entire metro lines, the method restored the arrival times of 95% of trains within the error of 24 s even when 100% of logs was missing. The mining procedure can also be applied to trains operating under special strategies such as short-turning and skip-stop. The recovered log seems precise enough for the current reliability analysis performed by the city of Seoul.     

From Twitter to detector: Real-time traffic incident detection using social media data

The effectiveness of traditional incident detection is often limited by sparse sensor coverage, and reporting incidents to emergency response systems is labor-intensive. We propose to mine tweet texts to extract incident information on both highways and arterials as an efficient and cost-effective alternative to existing data sources. This paper presents a methodology to crawl, process and filter tweets that are accessible by the public for free. Tweets are acquired from Twitter using the REST API in real time. The process of adaptive data acquisition establishes a dictionary of important keywords and their combinations that can imply traffic incidents (TI). A tweet is then mapped into a high dimensional binary vector in a feature space formed by the dictionary, and classified into either TI related or not. All the TI tweets are then geocoded to determine their locations, and further classified into one of the five incident categories.We apply the methodology in two regions, the Pittsburgh and Philadelphia Metropolitan Areas. Overall, mining tweets holds great potentials to complement existing traffic incident data in a very cheap way. A small sample of tweets acquired from the Twitter API cover most of the incidents reported in the existing data set, and additional incidents can be identified through analyzing tweets text. Twitter also provides ample additional information with a reasonable coverage on arterials. A tweet that is related to TI and geocodable accounts for approximately 5% of all the acquired tweets. Of those geocodable TI tweets, 60–70% are posted by influential users (IU), namely public Twitter accounts mostly owned by public agencies and media, while the rest is contributed by individual users. There is more incident information provided by Twitter on weekends than on weekdays. Within the same day, both individuals and IUs tend to report incidents more frequently during the day time than at night, especially during traffic peak hours. Individual tweets are more likely to report incidents near the center of a city, and the volume of information significantly decays outwards from the center.     

软岩隧道初期支护沉降变形分析及控制方法

目前存在对软岩隧道初期支护变形的错误认识,造成施工成本的浪费、安全风险的增大、进度被严重制约等不利影响,所以非常有必要对隧道初期支护变形进行研究分析并控制。首先以“矿山法”工程实践证明软岩隧道预留变形量是不必设置的,再通过各种权威文献说明以容许变形来实现围岩自承的观念是对“新奥法”的曲解,是错误的,需要找出支护变形的真正原因并控制。通过总结分析隧道支护常见的6 种位移及其组合,得出为控制初期支护变形需要解决支护底脚地基承载力和支护结构强度2 方面问题。于是建立以“分布锚杆+钢架+喷射混凝土且系统分布锚杆模拟为具有法向和切向作用力的链杆支座的结构力学模型”,可计算出支护结构底脚应力及截面最大应力等值,由此可得出对地基承载力和支护结构强度要求的明确标准,或对改善支护底脚抗力提供数值依据。计算结果表明: 系统分布锚杆能极大地降低支护结构的弯矩、轴力内力和底脚应力,从而增强支护结构的承载力和降低基底应力,使变形和沉降得以有效控制。     

Web应用中并发控制的实现

Web应用已由原来的网站、电子商务发展成商业应用系统的一种架构-B/S架构,它已成为一个世界性的研究热点。但由于Internet网络协议固有的局限性以及Web应用中频繁的用户交互增加了在Internet上实现长事务的困难,从而,Web应用中的并发控制始终没能得到很好的解决。本文从Hibernate的乐观锁和悲观锁的实现原理出发,给出了实现悲观锁的基本思路和实现时的注意事项,在其它架构中得以推广和应用。     

站内计轴设备的工程应用

根据Azs 350U型计轴设备的特点，结合工程设计，提出站内计轴设备设计的思路，供读者参考。     

一种基于数据挖掘的GIS及在航海中的应用

根据聚类分析方法中密度凝聚的思想，提出一种新的复合聚类分析算法，进一步将这种算法用于地理信息系统的数据挖掘，并应用于船舶航线的自动设计。     

Internet中Peer-to-Peer应用流量测量与分析

Peer to Peer网络(简称p2p)作为一种新型的覆盖网络引起了越来越多研究者的兴趣。本文介绍了在我国进行的骨干互联网上p2p网络流量测量。与现有国外研究不同,本文的数据来源于核心路由器,因此克服了它们的缺陷。其研究集中在汇聚流中的3个周期性尖峰群、不同主机发送或接收流量的重尾分布、p2p流量的长相关特性以及提出了ADTE的估计方法来区分信令和数据流量。本文的研究也显示出Napster在p2p流中占大部分,这暗示着超级节点和阶层式拓扑较纯p2p结构潜在的优势。同时,观察到在我国p2p的流量仅占Internet总流量的1%弱,这个值跟国外的数据有很大区别。我们分析了其中的原因并希望该结论为我国p2p软件的发展提供参考。     

薄腹板T型梁的稳定分析

通过薄腹板T梁的稳定性理论分析,推导出列车运营荷载作用下梁整体稳定极限承载力和腹板屈曲的极限承载力,并利用ANSYS对24m高强度混凝土T型梁进行了有限元分析。