Filter-based feature selection for rail defect detection

Over the last few years research has been oriented toward developing a machine vision system for locating and identifying, automatically, defects on rails. Rail defects exhibit different properties and are divided into various categories related to the type and position of flaws on the rail. Several kinds of interrelated factors cause rail defects such as type of rail, construction conditions, and speed and/or frequency of trains using the rail. The aim of this paper is to present an experimental comparison among three filtering approaches, based on texture analysis of rail surfaces, to detect the presence/absence of a particular class of surface defects: corrugation.Received: 7 April 2002, Accepted: 13 April 2004, Published online: 13 July 2004     

Artificial neural network application for modeling the rail rolling process

Rail rolling process is one of the most complicated hot rolling processes. Evaluating the effects of parametric values on this complex process is only possible through modeling. In this study, the production parameters of different types of rails in the rail rolling processes were modeled with an artificial neural network (ANN), and it was aimed to obtain optimum parameter values for a different type of rail. For this purpose, the data from the Rail and Profile Rolling Mill in Kardemir Iron & Steel Works Co. (Karabük, Turkey) were used. BD1, BD2, and Tandem are three main parts of the rolling mill, and in order to obtain the force values of the 49 kg/m rail in each pass for the BD1 and BD2 sections, the force and torque values for the Tandem section, parameter values of 60, 54, 46, and 33 kg/m type rails were used. Comparing the results obtained from the ANN model and the actual field data demonstrated that force and torque values were obtained with acceptable error rates. The results of the present study demonstrated that ANN is an effective and reliable method to acquire data required for producing a new rail, and concerning the rail production process, it provides a productive way for accurate and fast decision making.     

Monitoring the state of the moving train by use of high performance systems and modern computation methods

The objective of this work has been to study the propagation of elastic waves in rails. It presents the comparison of calculations obtained by the grid-characteristic and discontinuous Galerkin methods. The propagation of elastic waves in the presence and absence of the karst inclusion in the ground under the embankment, diagnosed in these cases from the rails, are compared. The wave pictures and diagnosed signals for four types of defects of a fractured character: vertical and horizontal head layering, cross fracture in the head and cracks in the rail web are given. The grid-characteristic method on the curvilinear structural meshes and the discontinuous Galerkin method on the nonstructured triangular meshes make it possible to solve efficiently the tasks on monitoring the state of the moving train and rail, including a great number of integrity violations, dynamic interactions in the train-rail system, and obtain the full wave picture.     

Human factors in support of a successful railway: a review

One of the critical infrastructure components in most economies across the world is the rail network. In different nations rail is responsible for ensuring that there is not complete gridlock on the roads in commuter hours, and for moving both people and freight for long distances in an as efficient manner as possible. This critical role, a number of high profile accidents and proposals for new network control philosophies and systems have led to a great upsurge in human factors rail research and applications in the past few years. This paper provides a retrospective on rail human factors research covering driving, signalling and control, maintenance, incident reporting systems, passengers and the public, planning and technical systems change. This research foundation, and also current major rail human factors programmes, are placed in the context of technology, investment, competition, cultural and safety requirements and constraints. The paper concludes with an examination of where rail human factors should and will be going into the future. Rail human factors research has grown rapidly in both quantity and quality of output over the past few years. There was an early base of work at a few institutions carried out over the 1960 s and 1970 s, followed by a lull in the 1980 s and early 1990 s. The continual influences of safety concerns, new technical system opportunities, reorganisation of the business, needs to increase effective, reliable and safe use of capacity, and increased society, media and government interest have now accelerated rail human factors research programmes in several countries. In this paper we review the literature on rail human factors research, covering driving, signalling and control, maintenance, reporting systems, passenger interests, planning and technical systems change. Current major rail human factors programmes are summarised and future research needs proposed. It is asserted that general human factors models and methods are being re-assessed, and new ones developed, to meet the requirements of the railways. This paper has, with the agreement of all editors concerned, been published in substantially similar form as: Wilson J. R. and Norris B. J., 2005, Rail human factors: past, present and future. Applied Ergonomics, 36, 6, 649–660.     

Dedicated sensor and classifier of rail head defects

This paper presents an original system based on a specialized eddy-current sensor for the inspection of railway tracks. The device aims at the detection of broken rails and large head spalls of the rail, and it has been designed to be mounted underneath automatic driving railway vehicles. In fact, less important types of defects can also be detected. Another use of the sensor would be for predictive management of the rail. All these applications need a procedure for parametrization of the output signals. Then a classification procedure is performed by a set of neural networks which is able to assign each ‘defect’ into one particular class.     

Bond graph modeling of a railway truck on curved track

Rail vehicles become unstable beyond a critical speed because of speed dependant creep forces at the rail-wheel contact surface. There is a constant demand for improved rail vehicle model for design of high-speed vehicles, and track. Works available in the literature are based on models of trucks with various simplifying assumptions such as reduced degree of freedom, small displacements and without kinematic nonlinearities and inclination of the contact surface. In this paper an integrated bond graph model of a truck is developed for the first time without the aforementioned assumptions. It is found that truck dynamical behavior is significantly different from those found in the literature in some cases.     

基于模糊层次分析法的铁路运价自动定价方法研究

通过分析影响铁路运价定价的各种因素,选取了其中的运营成本、国家价格策略、货主市场承受能力、平行运行径路的竞争、附加成本、盈利水平、供求关系等7个影响因素,运用模糊层次分析方法,确定了各运价定价影响因素的指标权重,从而为铁路企业在新建铁路运价定价过程中提供决策依据和参考标准.     

An inverse shape design method for railway wheel profiles

Optimisation of a railway vehicle–track system is a complex process. The paper presents a procedure for optimal design of a wheel profile based on geometrical wheel/rail contact characteristics such as the rolling radii difference (RRD). The procedure uses optimality criteria based on an RRD function. The criteria account for stability of wheelset, cost efficiency, minimum wear of wheels and rails as well as safety requirements. The shape of the wheel profile approximated by a piecewise cubic Hermite interpolating polynomial is varied during the optimisation process to satisfy the optimality criteria. A numerical technique called multipoint approximations based on response surface fitting (MARS) has been chosen as an optimisation method. The proposed optimum design procedure has been applied to improve the performance of metro trains in Rotterdam (RET), The Netherlands. The trains were suffering from severe wheel tread wear and as a result of that from lateral vibrations (hunting). Using the proposed procedure, a new wheel profile has been obtained and applied to the RET metro trains. The results of the optimisation have shown that the performance of a railway vehicle can be improved by improving the contact properties of the wheel and rail. After the application of the optimised wheel profile, the instability of the metro trains has been eliminated and the lifetime of the wheels has been increased from 15,000 to 120,000 km.     

Surface polishing of quartz-based microfluidic channels using CO<Subscript>2</Subscript> laser

Recently, microgrinding using a polycrystalline diamond tool has been introduced to fabricate microchannels and structures from quartz (fused silica). Compared to wet or dry etching processes, the grinding process is very simple and time-efficient for prototyping. However, the roughness of the machined surface remains an issue, because the surface is covered with many small cracks. Poor surface roughness can affect fluid flow in the microfluidic channels. To reduce the surface roughness of microchannels generated by a grinding process, this study presents the laser polishing of quartz and investigates the effects of the translational speed and pitch of a laser spot on the surface roughness and shape accuracy of microchannels.     

Integrity-directed sequential state estimation: Assessing high reliability requirements via safe confidence intervals

This study deals with the problem of dynamic state estimation of continuous-time systems from discrete-time measurements in the context of high-integrity applications. The objective of integrity-directed estimation is to provide confidence intervals for the state with an extremely low risk of error. We suppose that the process noise can be modelled by Gaussian sequences, and that the measurement noise is Gaussian in the normal operating mode of the sensors. The evolution of the posterior probability distribution of the system’s state is deduced from recursive linear MMSE estimation. The estimation scheme presented here is equivalent to the Kalman filter, with the difference that the data is not processed directly, but collected in sets in preparation for an ulterior, slightly delayed, grouped processing. This strategy is particularly suitable for fault detection, because the estimator naturally takes into account the cross-correlations of close-in-time measurements and the decisions can be based on more data. Next, we introduce dynamic tools for detecting faults and sensor failures. A full Bayesian modelling of the sensors leads to the derivation of a dynamic multiple-model estimator performing the linear MMSE state estimation under various fault hypotheses. This estimator provides estimates of the posterior density function of the state, on which safe confidence intervals certifying very high-integrity levels can be fixed. In practice, the exponential growth of the complexity of the multiple-model estimator requires the simplification of the posterior mixture distributions. A new method for limiting the complexity of the posterior distributions in an integrity-oriented context is presented. Unlike the known mixture simplification strategies (GPB, IMM), the present method has the property to quantify and minimize the loss in integrity of the process of simplification of the distributions. Finally, the estimator is tested on a typical rail navigation problem.     

论智慧轨道交通及其系统架构

提倡用“智慧轨道交通”来描述未来的“高效、便捷、安全、可视、可预测、环保和智慧”的现代轨道交通行业和宏观系统的发展远景视图,并提出了一个“智慧轨道交通系统架构”。该架构以“智慧轨道交通智能化地面基础设施”（SRT-IGI）和智慧列车为基础实现更透彻的感知;以“轨道交通全联网”(SRT-IoT)作为信息交换与共享平台实现更广泛的互联互通;以“与人类融合智慧管理、决策与指挥”（HCA-IMDC）应用系统为顶层结构。三部分组成结构及分布其中的“智能化”功能共同实现整个交通行业的“智慧化”。同时分别对三部分的技术内涵作了进一步的探讨。鉴于笔者尚未发现从信息技术（IT）角度全面探讨“智慧轨道交通”的文献,所以希望本框架的提出能够为从信息技术（IT）角度全面研究智慧轨道交通提供参考,进一步推动智慧轨道交通在技术上的跨越式发展,使我国轨道交通以“更透彻的感知、更广泛的互联互通、更深入的智能化处理能力”的雄姿,为国家经济建设发展和人性化地为人民服务。     

Bedside safety rails: assessment of strength requirements and the appropriateness of current designs

Bedside safety rails are used extensively throughout hospitals and residential care environments as a safety device to protect people against the risk of falling out of bed. However, several adverse incidents involving body entrapment have been linked to their use, which have resulted in serious injury to the bed occupants and, in more extreme cases, fatalities. Internationally agreed standards (i.e. European Committee for Standardisation (CEN) standards) for the design and testing of bed rails have recently been developed and are targeted at improved safety requirements (European Committee for Standardisation 19972000). The purpose of the present study was to: 1) determine whether the CEN test procedure for assessing bed rail strength and rigidity requirements is consistent with human force exertions to which bed rails may be exposed; 2) carry out loading tests on bed rails currently in use throughout the UK to determine whether they comply with the current safety recommendations proposed by the CEN standards.A laboratory study was conducted to determine the force exertions of 20 participants performing seven tasks (including activities of both bed occupants and care providers), which were considered representative of the forces to which bed rails could be exposed during normal use. Forces exerted by participants were measured using a force platform and distortions of the bed rails leading to potential entrapment zones were measured using linear displacement transducers. Static load tests were then conducted, in which incremental point loading was applied at the most adverse points on each of nine different bed rails (commonly used in UK nursing and residential care homes) in accordance with the test procedure detailed in the current CEN standards.Maximum static forces exerted by participants were found to range between 250 N and 350 N, which were within the 500 N force tolerance requirements set by the standards. Dynamic tasks involving 'rolling aggressively' against the bed rail produced the highest peak force of 722 N. None of the bed rails included in this study met the requirements of the CEN test procedures, principally due to horizontal lateral displacements being in excess of 50 mm for a 500 N applied force and/or the bed rails' inability to withstand inward horizontal forces of 500 N. Bed rails used with divan beds were often unable to withstand forces greater than 300 N due to turning moments exceeding the bed rail's restraining mechanism.Current CEN standards for assessing the strength and rigidity of bedside safety rails appear to be consistent with the human forces that might arise as a consequence of the normal movement and activities of a bed occupant or care provider. Although compliance with existing standards is not compulsory, findings from the tests carried out on the selection of bed rails in current use throughout the UK highlight the need for improvements in the design and manufacture of bed rails.     

Rail signallers' assessments of their satisfaction with different shift work systems

This paper describes rail employee views on shift-work systems obtained from administration of the Rail Ergonomics Questionnaire (REQUEST) to a large sample of rail professionals in Great Britain, achieving a response rate of 83%. Results from signallers, the largest occupational grouping in the survey (n = 3230), are presented. ANOVA has been used to investigate the effects of different aspects of the shift patterns (the length of shift and the speed and direction of rotation of shifts) and the effects of age on the ratings of satisfaction with the shift system. The findings from the analyses indicate preferences for the 12-h shift system over the 7–9-h rotating shift system. There were no main effects of the direction of rotation of shifts, though there were significant interactions between the direction of rotation of shifts, the numbers of consecutive shifts and the age of respondents. There is a need for clear data addressing the impact on workers of key shift-related parameters such as shift length, direction of rotation of shifts, numbers of consecutive shifts and personal factors such as age. This paper presents findings of the effects of these key shift and personal parameters and their interactions on respondents' ratings of satisfaction with the shift system, using data collected with REQUEST in a national survey of rail signallers.     

Downsizing training data with weighted FCM for predicting the evolution of specific grinding energy with RNNs

Grinding plays a prominent role in modern manufacturing due to its capacity for producing parts of high accuracy and precision. Among the various grinding variables, the specific grinding energy (e_c) is key because it measures the energy required to remove the unit volume of part material and therefore it gives information about the performance of the grinding process. In addition, a measure of the specific grinding energy is also useful for estimating the power requirements of the grinding machine. Thus, a Recurrent Neural Network (RNN) is used for predicting e_c in a more realistic manner that involves the development of the e_c over time. Moreover, since performing grinding experiments is a highly time and resource consuming task, it would be very useful to downsize the required dataset to train the RNN since it could substantially reduce the time and costs involved in carrying out the experiments to generate the dataset, as well as in training the RNNs. Therefore, in this work a methodology combining Fuzzy C-Means (FCM) and RNNs for downsizing the dataset and predicting specific grinding energy is proposed. Unlike other approaches for reducing the dataset using FCM, in the current work the inputs are weighted. To achieve this, the knowledge is extracted from the weights of satisfactorily trained RNN obtained from previous work. The results show that under reduced training datasets (weighted and non-weighted FCM inputs) and non-reduced datasets (all available experiments), superior results were yielded with the RNNs obtained with the weighted approach. In fact, in some cases, for the reduced training dataset (weighted) the error is halved. Furthermore, the results show that it is more advantageous to use a reduced training dataset obtained after FCM, since this reduces the costs associated with experimental time, as well as the training time required for RNNs.     

基于虚拟仪器的铁轨损伤检测系统

本文介绍了铁质材料因裂纹或损伤而产生缺陷的定量检测原理。设计了基于虚拟仪器的铁轨损伤检测系统。详细介绍并分析了LabVIEW开发平台在数据采集和数据分析上的实现方法．实现了一台PC同时对多条铁轨进行损伤检测。并论证该方法的可行性与可靠性，得出了相关的结论。     

Rail signallers' assessments of their satisfaction with different shift work systems

This paper describes rail employee views on shift-work systems obtained from administration of the Rail Ergonomics Questionnaire (REQUEST) to a large sample of rail professionals in Great Britain, achieving a response rate of 83%. Results from signallers, the largest occupational grouping in the survey (n = 3230), are presented. ANOVA has been used to investigate the effects of different aspects of the shift patterns (the length of shift and the speed and direction of rotation of shifts) and the effects of age on the ratings of satisfaction with the shift system. The findings from the analyses indicate preferences for the 12-h shift system over the 7-9-h rotating shift system. There were no main effects of the direction of rotation of shifts, though there were significant interactions between the direction of rotation of shifts, the numbers of consecutive shifts and the age of respondents. There is a need for clear data addressing the impact on workers of key shift-related parameters such as shift length, direction of rotation of shifts, numbers of consecutive shifts and personal factors such as age. This paper presents findings of the effects of these key shift and personal parameters and their interactions on respondents' ratings of satisfaction with the shift system, using data collected with REQUEST in a national survey of rail signallers.     

Something for nothing: management rejection of open source software in Australia’s top firms

Organisations have traditionally relied on commercial software products to support their operations. However, rising software costs and recent corporate failures have brought the provision and value of commercial software into question. Recently, open source software, as a relatively new development in the IS field, has risen in popularity as a possible panacea for these ills. If firms value low acquisition cost, ostensibly plentiful support, and source code access, why have not more firms adopted open source software? The lack of published empirical research in the area means this issue has been inadequately addressed.This paper examines why firms do not adopt open source software. This study surveyed 500 of Australia’s top firms to see why managers rejected open source software. The study found that managers rejected open source software because they could not see that it had any relevance to their operations, perceived a lack of reliable ongoing technical support of it and also appeared to see substantial learning costs or had adopted other software that they believed to be incompatible with open source software.     

Confidential Incident Reporting on the UK Railways: The ‘CIRAS’ System

The paper describes CIRAS (Confidential Incident Reporting and Analysis System), a confidential reporting system developed by the authors in collaboration with ScotRail, the Health and Safety Executive, Railtrack and the (now defunct) British Rail Board. After a two-year pilot/developmental study with ScotRail during 1995–97, the system is now subscribed to by all but one of the major train operating companies, rail infrastructure and maintenance companies with a presence in Scotland; plus a developing profile in the rest of the UK.  CIRAS gathers data in three ways: (i) from an initial report form or telephone call, (ii) from a structured follow-up telephone questionnaire, and (iii) from an in-depth interview with a researcher (telephone or face-to-face, according to priority). The interviews bring to light details of personal motive, and of intended/unintended actions, which are not commonly found on company-run databases because of their association with disciplinary procedures. Information is processed through a human factors model and fed back to the companies involved, in disidentified form, to take corrective action.  The basic structure of the human-factors model is described; data are presented on reports received to date which have been processed through this model. The system has recently been recommended by a UK Parliamentary Committee and by Railtrack Safety and Standards Directorate for extension to the UK network as a whole.     

基于卷积神经网络的铁轨路牌识别方法

轨道交通在我国综合交通运输体系中占有重要的地位,随着人工智能的发展,智能感知轨道交通周围环境的信息也变得越来越引人注目.本文结合深度学习与图像处理的方法,设计并实现了一种基于卷积神经网络的高铁轨道周边路牌数字识别的智能系统,该系统通过在高铁驾驶室内安装摄像头的方式采集运行前方的视频,并通过目标识别、语义分割等深度学习算法自动定位并识别路牌内的数字,从而解决了之前人工处理的繁琐和低效率.本算法整体系统由三个子模块构成,分别为目标检测模块、语义分割模块以及数字识别模块,其中目标检测模块基于SSD(Single shot MultiBox dector)模型,并对其进行了改进,使其更适用于轨道交通中的小目标识别;语义分割模块使用了全卷积的方式,对目标检测的结果进一步处理,准确得到路牌中的数字区域;数字识别模块的设计参考了著名的识别MNIST数据集的手写体识别系统,并针对路牌中数字的特点做了相应的改进,实现了对每个数字的准确识别.实验结果表明,本系统可适应白天、夜间情况下轨道交通的路况,识别的综合准确率为80.45%,其中,白天的平均识别准确率为87.98%,夜间的平均识别准确率为72.92%.     

城市轨道交通联调联试的组织与实施

联调联试作为城市轨道交通建设中的一个重要环节,是一项涉及多系统、多任务、技术难度大的系统工程,其组织与实施应遵循科学、安全和有序的原则进行。以南京宁和城际轨道交通一期工程为例,介绍城市轨道交通联调联试的整体流程,包括联调联试的筹划、组织与实施以及收尾,重点对联调联试的接口识别与分析进行阐述,并在此基础上创新性地提出以通信、信号、综合监控、供电四个专业横向牵头,其他专业纵向配合的调试模式,形成纵横交错的矩阵式调试组织结构,有效发挥系统优势,提高联调联试效率,为城市轨道交通联调联试的开展提供有益建议。