No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
A systematic review of acceleration-based estimation of railway track quality
Track and preventive maintenance are necessary for the safe and comfortable operation of railways. Track displacement measured by track inspection vehicles or trolleys has been primarily used for track management. Thus, vibration data measured in in-service vehicles have not been extensively used for track management. In this study, we propose a new technique for estimating track irregularities from measured car body vibration for track management. The correlation between track irregularity and car body vibration was analysed using a multibody dynamics simulation of travelling rail vehicles. Gaussian process regression (GPR) was applied to the track irregularity and car body vibration data obtained from the simulation, and a method was proposed to estimate the track irregularities from the constructed regression model. The longitudinal-level, alignment, and cross-level irregularities were estimated from the measured car body vibrations and travelling speeds on a regional railway, and the results were compared with the actual track irregularity data. The results showed that the proposed method is applicable for track irregularity management in regional railways.
A comprehensive methodology is presented for the estimation of track vertical irregularities of railway bridges from vehicle responses, based on a novel and lightweight multi-layer-perceptron (MLP) deep learning architecture. Firstly, a vehicle–track–bridge interaction (VTBI) model is established for the generation of the datasets of deep learning networks. Secondly, the lightweight deep learning architecture is meticulously designed to identify the track's vertical irregularities. Then, the effectiveness of the proposed technique is validated through examples of a single-span simple bridge and a three-span bridge. Further, the sensitivity of the present method against various factors is investigated, including different combinations of vehicle responses, measurement noise, vehicle speed and different classes of track irregularity. It is confirmed that the identified track irregularities of the railway bridges, regardless of irregularity class and noise level, are in excellent agreement with the ground-truth ones. The increase in vehicle speed to some extent reduces the estimation accuracy of the irregularity. The proposed method has higher identification accuracy and efficiency for longer irregularity sequences of multi-span railway bridges.
Track irregularities directly affect the quality and safety of railway vehicle operations. Quantitative detection and real-time monitoring of track irregularities are of great importance. However, due to the frequent variable vehicle speed, vehicle operation is a typical non-stationary process. The traditional signal analysis methods are unsuitable for non-stationary processes, making the quantitative detection of the wavelength and amplitude of track irregularities difficult. To solve the above problems, this paper proposes a quantitative detection method of track irregularities under non-stationary conditions with variable vehicle speed by order tracking analysis for the first time. Firstly, a simplified wheel–rail dynamic model is established to derive the quantitative relationship between the axle-box vertical vibration and the track vertical irregularities. Secondly, the Simpson double integration method is proposed to calculate the axle-box vertical displacement based on the axle-box vertical acceleration, and the process error is optimized. Thirdly, based on the order tracking analysis theory, the angular domain resampling is performed on the axle-box vertical displacement time-domain signal in combination with the wheel rotation speed signals, and the quantitative detection of the track irregularities is achieved. Finally, the proposed method is validated based on simulation and field test analysis cases. We provide theoretical support and method reference for the quantitative detection method of track irregularities.
Track smoothness has become an important factor in the safe operation of high-speed trains. In order to ensure the safety of high-speed operations, studies on track smoothness detection methods are constantly improving. This paper presents a track irregularity identification method based on CNN-Bi-LSTM and predicts track irregularity through car body acceleration detection, which is easy to collect and can be obtained by passenger trains, so the model proposed in this paper provides an idea for the development of track irregularity identification method based on conventional vehicles. The first step is construction of the data set required for model training. The model input is the car body acceleration detection sequence, and the output is the irregularity sequence of the same length. The fluctuation trend of the irregularity data is extracted by the HP filtering (Hodrick Prescott Filter) algorithm as the prediction target. The second is a prediction model based on the CNN-Bi-LSTM network, extracting features from the car body acceleration data and realizing the point-by-point prediction of irregularities. Meanwhile, this paper proposes an exponential weighted mean square error with priority inner fitting (EIF-MSE) as the loss function, improving the accuracy of big value data prediction, and reducing the risk of false alarms. In conclusion, the model is verified based on the simulation data and the real data measured by the high-speed railway comprehensive inspection train.
Track geometry is one of the critical indicators of railway tracks' condition which requires continuous monitoring and maintenance over time. In this paper, a novel artificial intelligence (AI) based framework is proposed for railway track geometry inspection using vibration data collected from a dedicated measuring high-speed train. This AI-based anomaly track detection approach consists of two main stages. Firstly, a subset of features that best characterizes the track condition is defined. Several dynamic features from time domain data are extracted and importance scores are assigned to them, to determine the most effective subset for the purpose of track condition monitoring. Secondly, a data-driven based anomaly detection approach is developed to assess and identify track geometrical defects. In this stage, the acceleration responses collected from an in-service train traversing on a healthy track zone are employed as input into a One-Class Support Vector Machine (OCSVM) algorithm. The proposed algorithm defines the anomalies as relative changes to the historical behaviour. A comprehensive dataset from field measurements using a Société Nationale des Chemins de Fer Français (SNCF) Réseau IRIS320 highspeed train is used in this paper to implement the proposed approach. In addition, the impact of using different features and different locations/directions of the sensors on the accuracy of detecting geometrical defects is investigated. It is also shown that the OCSVM approach outperforms other algorithms based on Isolation Forest (IF), Local Outlier Factor (LOF), and Robust Mahalanobis Distance (RMD) in terms of recall, precision, and F1-score. The proposed anomaly detection approach has demonstrated a 12% increase in defect detection accuracy compared to the direct utilization of the raw acceleration response, which can facilitate track monitoring using in-service trains while providing cost-efficient maintenance in the future.
Railway tracks must be managed appropriately because their conditions significantly affect railway safety. Safety is ensured through inspections by track maintenance staff and maintenance based on measurements using dedicated track geometry cars. However, maintaining regional railway tracks using conventional methods is becoming difficult because of their poor financial condition and lack of manpower. Therefore, a track condition diagnostic system is developed, wherein onboard sensing devices are installed on in-service vehicles, and the vibration acceleration of the car body is measured to monitor the condition of the track. In this study, we conduct long-term measurements using the system and evaluate changes in the track conditions over time using car-body vibration data. Filed test results showed that sections with degraded tracks were identified using car-body vibration data. The track degradation trend can be constructed using the results obtained. Furthermore, this study demonstrated that the track maintenance effect could be confirmed. A method for improving train position using the yaw angular velocity is proposed. The track irregularity position can be shown more clearly by monitoring the track condition using position-corrected data using the proposed method. It is also shown that the time-frequency analysis of measured car-body vertical acceleration is effective for evaluating the track condition more clearly.
The aim of this work is to develop a model-based methodology for monitoring lateral track irregularities based on the use of inertial sensors mounted on an in-service train. To this end, a gyroscope is used to measure the wheelset yaw angular velocity and two accelerometers are used to measure lateral acceleration of the wheelset and the bogie frame. The main contribution of the present work is the development of a very efficient Kalman-based monitoring strategy to estimate the lateral track irregularities. The Kalman filter is based on a highly simplified linear bogie model that is able to capture the most relevant dynamic behaviour of the vehicle. The behaviour of the designed filter is assessed through the use of a detailed multibody model of an in-service vehicle running on a straight track with realistic irregularities. The model output is used to generate virtual measurements that are subsequently used to run the filter and validate the proposed estimator. In addition, the equivalent parameters of the simplified model are identified based on these simulations. In order to prove the robustness of the proposed technique, a systematic parametric analysis has been performed. The results obtained with the proposed method are promising, showing high accuracy and robustness for monitoring lateral alignment on straight tracks, with a very low computational cost.
Track irregularity detection serves as one of the most essential technologies to interpret the in-service condition of the high-speed railway (HSR) and ensure the comfortability and safety in HSR service. This study newly proposed a novel on-board detection technique integrated with newly developed algorithm for identifying the longitudinal track irregularity to obtain the condition monitoring for further condition based maintenance (CBM). Such on-board detection technique includes data acquisition unit, spatial-time synchronous calibration unit and data processing unit, which can be directly installed in commercial high-speed trains. Via inertial reference method, an algorithm is developed to derive longitudinal irregularity from the acceleration measured in axle box of high-speed train by combining multiple digital filters. A prototype is developed and is installed on a comprehensive inspection train to verify the feasibility and test on multiple HSR lines eventually. The comparison affirms that the detection system has high accuracy and sound repeatability, and certain on-board detection technique enables the real-time online monitoring of track condition.
Tracks are critical and expensive railroad asset, requiring frequent maintenance. The stress from heavy car axle loads increases the risk of deviations from uniform track geometry. Irregularities in track geometry, such as track warping, can cause an excessive harmonic rocking condition that can lead to derailments, traffic delays, and associated financial losses. This paper presents an approach to enhance the location identification accuracy of track geometry irregularities by combining measurements from sensors aboard Hi-Rail vehicles. However, speed variations, position recording errors, low GPS update rates, and the non-uniform sampling rates of inertial sensors pose significant challenges for signal processing, feature extraction, and signal combination. This study introduces a method of extracting features from the fused data of inertial sensors and GPS receivers with multiple traversals to locate and characterize irregularities of track geometry. The proposed method provides robust detection and enhanced accuracy in the localization of irregularities within spatial windows along the track segment. Tradeoff analysis found that the optimal spatial window size is 5-meter.
Proper maintenance of railway tracks is essential to ensure railway safety. A track condition monitoring system was developed for the preventive maintenance of the track by installing the on-board sensing device in the in-service vehicles and monitoring the track condition by measuring the car-body acceleration. This paper describes the application of time–frequency analysis for the condition monitoring of railway tracks. Car-body acceleration simulated by a 10-DOF vehicle model, with a faulty track, was used to identify the track faults using Hilbert–Huang transform (HHT). Field tests were carried out using the developed track condition monitoring system to show the effectiveness of the time–frequency analysis using HHT. Simulation studies and field test results showed that HHT can yield a good time–frequency resolution and that intrinsic mode functions (IMFs) can provide the detailed information of track faults.
The condition of railway infrastructure is currently assessed by track recording cars, wayside equipment, onboard monitoring techniques and visual inspections. These data sources deliver valuable information for infrastructure managers on the asset’s condition but are mostly carried out in time-based intervals. This paper examines the potential of fibre optic cables, which are already installed in cable troughs alongside railway tracks, to monitor railway infrastructure conditions. The sensing technique, known as distributed acoustic/vibration sensing (DAS/DVS), relies on the effect of Rayleigh scattering and transforms the optical fibre into an array of “virtual microphones” in the thousands. This sensing method has the ability to be used over long distances and thus provide information about the events taking place in the proximity of the monitored asset in real-time. This study outlines the potential of DAS for the identification of different track conditions and isolated track defects. The results are linked to asset data of the infrastructure manager to identify the root cause of the detected signal anomalies and pattern. A methodology such as this allows for condition-based and component-specific maintenance planning and execution and avoids the installation of additional sensors. DAS can pave the way toward a permanent and holistic assessment of railway tracks.
The determination of the precise track irregularity with unfavorable wavelength, which shall induce vehicle’s violent vibration in terms of the vehicle’s speeds, still challenges the researchers. This study proposes a feasible study of assessing the track irregularity by using the transfer function and the measured carriage-body acceleration by combining the ARX model with state space model. The ARX model and state space model are constructed using system identification to obtain the transfer relation between the track irregularity and the carriage-body acceleration, respectively. The model’s parameters are estimated by the measured data from the high-speed China Railway Comprehensive Inspection Train (CRCIT). The correlation value between the predicted and measured carriage-body acceleration shows that both models can effectively represent the transfer characteristics between the track irregularity and the carriage-body acceleration. Furthermore, the models can help assess the proportion of the vibration caused by track irregularity with the specific wavelengths and determine the track irregularity with unfavorable wavelength.
A cab-based track monitoring system has been developed which makes use of the existing on-board GSM-R cab radio present in the majority of trains operating in the UK. With the addition of a low-cost sensor, type, location and severity of the track defects are reported using the system. The system improves safety and network performance by efficiently directing maintenance crews to the location of defects, minimising time spent on maintenance and inspection. Initially, vehicle dynamic simulation was used to test the feasibility of the system for defect monitoring and to develop compensation factors for vehicle type and operating speed. Novel on-board signal processing techniques are also presented through comparison of vibration response from sites with known defects and outputs from Network Rail’s (NR) New Measurement Train (NMT). Good agreement was reported for track faults in relation to vertical and lateral alignment and dip faults. Statistically, good agreement has been demonstrated, suggesting that the data acquired could be used to provide an indication of track quality thereby improving network performance, reducing rough ride and leading to improved passenger comfort. Improvements in the measured and statistical correlation are anticipated through the use, of multi-train / multi-journey and machine learning methods..
The worldwide increase in frequency of traffic for passenger trains and the rise of freight trains over the recent years necessitate the more intense deployment of track monitoring and rail inspection procedures. The wheel-rail contact forces, induced by the static axle loads of the vehicle and the dynamic effects of ground-borne vibration coming from the track superstructure, have been a significant factor contributing to the degradation of the railway track system. Measurements of track irregularities have been applied since the early days of railway engineering to reveal the current condition and quality of railway lines. Track geometry is a term used to collectively refer to the measurable parameters including the faults of railway tracks and rails. This paper is aiming to review the characteristics of compact inertial measurement systems (IMUs), their components, installation, the basic measures of the quality of the track using motion sensors, like accelerometers, gyroscopes and other sensing devices mounted on different places of the vehicle. Additionally, the paper briefly discusses the fundamentals of inertial navigation, the kinematics of the translational and rotational train motions to obtain orientation, velocity and position information.
A track condition monitoring system that uses a compact on-board sensing device has been developed and applied for track condition monitoring of regional railway lines in Japan. Monitoring examples show that the system is effective for regional railway operators. A classifier for track faults has been developed to detect track fault automatically. Simulation studies using SIMPACK and field tests were carried out to detect and isolate the track faults from car-body vibration. The results show that the feature of track faults is extracted from car-body vibration and classified from proposed feature space using machine learning techniques.
This article performs an extensive review on condition monitoring techniques for rail vehicle dynamics. In particular, the review focuses on applications of model-based approaches for on-board condition monitoring systems. The article covers condition monitoring schemes, fault detection strategies as well as theoretical aspects of different techniques. Case studies and experimental applications are also summarized. All the mentioned issues are discussed with the goal of providing a detailed overview on condition monitoring in railway vehicle dynamics.
In this paper, a methodology to predict the track longitudinal level using bogie vertical acceleration from in-service vehicles is proposed. To account for the effect of vehicle speed, the acceleration levels are double integrated on-board the vehicle. Synthetic indicators like the RMS are then computed over predefined track sections of 100 m, to reduce the amount of data to be stored and analysed. Then, a linear regression model between the double integrated indicators and the direct track geometry measurements collected by a TRV is built, to verify the degree of correlation of the two quantities. To this end, data collected during a long-term monitoring campaign along the Italian railway network are considered. The regression model is finally adopted to predict the RMS of the longitudinal level using the signals collected on-board the vehicle. The comparison between the predicted and measured data is shown to be promising towards the possibility of condition monitoring of the track geometry both on high-speed and conventional lines.
Monitoring railway tracks through drive-by vibration data collected by in-service trains offers a cost-effective and adaptable solution for inspecting multiple railway lines. However, numerous existing drive-by monitoring methods rely on supervised learning models, necessitating extensive labelled data for each line. In this paper, a novel framework is proposed based on Unsupervised Domain Adaptation (UDA) concept which facilitates the transfer of a geometric defects diagnosis model learned from one line to a new line without the need for any labelled data from the new line. The proposed framework learns the dynamic-based features that are sensitive to damage and also invariant to different railway tracks. It comprises three components: data pre-processing, UDA implementation, and damage diagnosis. The framework uses the data from the source domain, including corresponding labels, as well as the unlabelled data from the target domain as input. The outputs of the framework consist of the predicted labels for the target domain. The performance of the proposed framework is evaluated using a comprehensive dataset of field measurements of a high-speed train passing 4 different lines within the French high-speed rail network. The proposed UDA framework is implemented using four common UDA algorithms including Information-Theoretical Learning (ITL), Geodesic Flow Kernel (GFK), Transfer Component Analysis (TCA), and Subspace Alignment (SA). The results show that the proposed framework has a 14% increase in the anomaly detection accuracy compared to traditional unsupervised learning methods in which UDA is not used. Furthermore, this study investigates the impact of incorporating a percentage of target data labels during training (semi-supervised domain adaptation), along with various sensor layouts and different tuning parameters, on the accuracy of the proposed approach. The results show that the proposed framework can significantly facilitate the monitoring of railway track conditions using the data collected by in-service trains which could be great interest of railway owners.
Track defects are gradually emerging with the development of urban rail transits. However, there is rare research implemented to diagnose track conditions in real time. Although some intelligent data-driven methods seem to have the potential to achieve the track condition diagnosis, it’s hard to acquire sufficient labeled data in actual applications. This study proposes a dynamics simulation-assisted transfer learning (TL) method for label-scarce track condition diagnosis. Firstly, a dynamics model of axle-box bearings considering the service environment is established. Based on this model, a large amount of axle-box vibration signals corresponding to healthy/defective track conditions is simulated. Wavelet transform is performed for these signals to characterize their time-frequency energy distribution modes in the format of time-frequency maps, which are considered source-domain data. Similarly, the time-frequency maps of the collected signals during vehicle operation are served as the target-domain data. Subsequently, a sub-domain alignment TL network is constructed to map the data from the source and target domain into a deep feature space. In this network, unlabeled target-domain data are classified to obtain their pseudo labels. Finally, Wasserstein distance measure and multiple domain discriminators are employed to achieve label alignment between two domains for each corresponding category. A feature centroid-driven loss function is applied to further reduce the intra-class variations, ultimately realizing accurate knowledge transfer from simulated signals to collected signals. A two-level sliding window algorithm is designed to detect abnormal axle-box vibration signal parts which are then diagnosed through the well-trained network. The proposed method is validated through a transfer diagnosis experiment using simulated signals and collected signals. This study provides a promising solution to diagnose different track conditions, which is of great significance for ensuring running safety in urban rail transits.
Purpose
This article aims to predict the rapid track geometry change in the short term with a higher detection frequency, and realize the monitoring and maintenance of the railway state.
Design/methodology/approach
Firstly, the ABA data needs to be filtered to remove the DC component to reduce the drift due to integration. Secondly, the quadratic integration in frequency domain for concern components of the vertical and lateral ABA needs to be done. Thirdly, the displacement in lateral of the wheelset to rail needs to be calculated. Then the track alignment irregularity needs to be calculated by the integration of lateral ABA and the lateral displacement of the wheelset to rail.
Findings
By comparing with a commercial track geometry measurement system, the high-speed railway application results in different conditions, after removal of the influence of LDWR, identified that the proposed method can produce a satisfactory result.
Originality/value
This article helps realize detection of track irregularity on operating vehicle, reduce equipment production, installation and maintenance costs and improve detection density.
Fault diagnosis of railway track irregularities (TI) via supervised learning algorithms is a difficult task due to the lack of suitable, labeled datasets. Class imbalance in the data poses an additional issue. While it may be possible to continuously monitor the condition of the railway track using sensors mounted on regular in-service vehicles, the presence and location of faults is unknown (unlabeled data). At the same time, most of the collected information will correspond to the nominal, non-faulty condition (imbalanced dataset), which impacts on the performance of Machine Learning classifiers. In this paper, a nobel method to generate synthetic TI using advanced Generative Adversarial Networks (GAN) is presented and used, in conjuction with a numerical model of the track-vehicle interaction, to perform data augmentation and obtain a large, labeled and balanced dataset, suitable for supervised learning classification. Inertial measurements from a vehicle-track scale model are then used as test dataset to validate the data augmentation process.
Railway systems are important components of transport networks which need to be well maintained over time. Railway track inspection is currently conducted visually or using special vehicles called track geometry vehicles (TRVs). These methods are normally labor-expensive and not always effective. In this paper, vibration data collected from in-service trains are employed for the purpose of track monitoring. The proposed approach could be more efficient approach where the railway tracks can be monitored on a daily basis at a little cost. A novel data-driven method is proposed to detect several defects of railway tracks using vertical acceleration data collected from an operational passenger train. An open access dataset from a field study is used in this study. A data cleaning process is performed to extract useful data from the whole dataset. In addition, condition-sensitive features are extracted from the raw data. Finally, a multilabel classification algorithm based on the least-square support-vector machine (LS-SVM) is used to classify the type of the defect. Results show that when an LS-SVM is used, 92% accuracy can be achieved for two types of defects, tamping and surfacing using frequency domain features. In addition, the impact of using different features extraction methods and different classification algorithms on the accuracy of the proposed approach is studied.KeywordsRail maintenanceData drivenData reductionSignal processing
The monitoring of the railway infrastructure is nowadays of the utmost importance to guarantee the reliability of this kind of transportation and the safety of people involved. To this end, diagnostic trains are regularly employed to perform monitoring activities. However, while along high-speed railway lines their runs are usually scheduled once every two/four weeks, the diagnostic runs are much less frequent along main lines. Therefore, techniques able to detect anomalies on the line through measurements taken on-board in-service vehicles have been proposed in recent years. In this work, a wireless system developed to simultaneously monitor the vehicle dynamics and identify possible issues on specific track sections is presented. The detection of track defects is made possible by analysing acceleration RMS computed in specific frequency bands, considering several travels along the same track section recorded during a long-term experimental campaign. The proposed method would allow carrying out continuous monitoring campaigns, providing freight wagon with a cheap and easy to install measuring setup. In the end, the designed system could support the maintenance strategy along conventional lines, where the runs of the diagnostic vehicles are scheduled at long time periods one apart the other.KeywordsCondition MonitoringRailway InfrastructureFreight WagonOn-board MeasurementsWireless SensorsTrack Defects
The increase of rail traffic in the last decades requires a continuous improvement of railway lines monitoring techniques, to provide higher levels of infrastructure safety and to properly manage effective maintenance plans. In this respect, the possibility to rely on in-service vehicles equipped with a simpler set of sensors (e.g., accelerometers) could increase data availability and support the maintenance strategy, that normally relies on special purpose diagnostic trains to periodically inspect the railway line. In this paper, data coming from vertical accelerometers installed on bogies of a commercial vehicle have been considered to monitor the track longitudinal level, that is the most important track geometry parameter that drives maintenance operations along high-speed lines. The proposed strategy relies on a multiple linear regression model that allows estimating the track longitudinal level, considering as input different predictors computed from the available acceleration data. The adoption of the pre-built regression model and the vehicle dynamic data allows to estimate the track geometry parameter along different sections of the line. These results can represent a useful tool to develop a methodology for track condition-based maintenance based on acceleration data from commercial vehicles.KeywordsRailway InfrastructureCondition-Based MaintenanceTrack GeometryRail Vehicle DynamicsIn-Service Vehicle
Track irregularities induce potential risks to the safety and stability of railway track systems. This paper proposes a novel methodology to identify vertical and lateral track irregularities. The method involves measuring system-based attitude calculation and a model-based unknown input observer estimator, based on the dynamic responses of distributed multi-sensors on the vehicle and bogie. First, a mechanical model of wheel-rail contacts is built with dynamic methods. The model considers the different directions of motion for a railway vehicle and consists of two bogies and four wheelsets. Based on the multi-sensor acceleration measurement, the vertical and lateral acceleration signals of the vehicle and bogies are integrated into the displacement signal. Then a state-space description of the vehicle suspension model is established for inverse dynamical analysis to extract the input signals. A suitable unknown input observer is constructed to estimate the track irregularities by transforming the state space equations of the vehicle into an augmented system that can monitor the track irregularities in-service. This method provides an opportunity to reduce the costs of the monitoring infrastructure and provide quicker and more reliable information about the status of a track.
Condition Monitoring (CM) describes the continuous observation of a dynamical system or process in order to track the evolution of its state and to detect signs of faults in an early stage. It constitutes the core block for performing fault diagnosis and prognosis, allowing a condition-based assessment of the system and supporting the implementation of an intelligent, cost-effective maintenance policy. This study focuses on the vehicle-based monitoring of railway infrastructure. First, the main concepts related to intelligent maintenance systems and strategies are depicted in a general framework. Later on, the specific application, the railway track-vehicle interaction system, is introduced. Railway vehicle instrumentation for track condition monitoring is analysed, with a special focus on the inertial measurement systems. A review of the processing algorithms used for railway monitoring is done and a taxonomy is proposed, based on the methodology approach: model-based, data-driven or hybrid. An analysis on the monitoring algorithms according to the geographical region is also made. It has been found that the railway vocation of each individual region determines the monitoring objectives pursued, as well as the methodological approach and the specific algorithms used. Finally, current trends and research gaps in railway monitoring are identified and outlined.
Differential subgrade settlement plays a key role in the formation of track geometry and significantly affects the running performance of a moving vehicle. This work therefore contributes to the on-line inversion of the track irregularities and the differential subgrade settlement hidden in the track irregularities is further excavated. To achieve the above goal, a vertically vehicle-track-subgrade coupled dynamics model with high accuracy and efficiency is first established by introducing the Green's function method. The track irregularities are then generated by a probabilistic model and the rail deflections caused by the settlement are also accounted for. The vertical accelerations of the vehicle excited by track irregularities and various vehicle speeds are subsequently derived. On this basis, a 1-D fully convolutional encoder-decoder network is constructed to predict the track irregularities by treating the acceleration data as the network inputs. A total of seven scenarios involving different input variables are investigated and the results show that when the wheelset, bogie and car body accelerations are simultaneously considered in network training, the prediction performance achieves the optimum. Meanwhile, the network robustness with respect to various vehicle speeds and degradation levels of track irregularities is also demonstrated. Finally, a time–frequency unification method is employed to identify the settlement locations and wavelengths from the predicted track irregularities. Two cases are conducted to further illustrate the effectiveness of the presented settlement identification method.
The paper describes a methodology for condition monitoring of rail track, which exploits acceleration measurements recorded by in-service high-speed vehicles. Estimates of the vertical track alignment are computed from bogie vertical acceleration with suitable linear regression models, relating synthetic RMS indicators representative of vehicle dynamics to track geometry parameters recorded by a diagnostic train. Two different linear models have been introduced, specifically devoted to the investigation of the overall track quality (by means of the RMS of the geometry parameter) and to the identification of isolated defects that may require maintenance intervention (through the track geometry peak value). The proposed solution has been specifically designed for high-speed applications, where trains travel at constant maximum speed for most of the journey. The methodology has been tested and validated against direct measurements taken by a diagnostic train. It allows to correctly reproduce the evolution with time of the railway line defectiveness, both in case of energy content and peak value. This result poses the basis for the development of methodologies for track condition-based maintenance.
Condition monitoring of track geometry irregularities from onboard measurements is a cost-effective method for daily surveillance of track quality. The monitoring of Alignment Level (AL) and Cross Level (CL) track irregularities is challenging due to the nonlinearities of the contact between wheels and rails. Recently, the authors proposed a signal-based method in combination with a machine learning (ML) fault classifier to monitor AL and CL track irregularities based on bogie frame accelerations. The authors concluded that the Support Vector Machine (SVM) fault classifier outperformed other traditional ML classifiers. Thus, an important question arises: Is the previously reported decision boundary an optimal boundary? The objective of this research investigation is to obtain an optimal decision boundary according to theory of probabilistic classification and compare the same against the SVM decision boundary. In this investigation, the classifiers are trained with results of numerical simulations and validated with measurements acquired by a diagnostic vehicle on straight track sections of a high-speed line (300 km/h). A fault classifier based on Maximum A Posterior Naïve Bayes (MAP-NB) classification is developed. It is shown that the MAP-NB classifier generates an optimal decision boundary and outperforms other classifiers in the validation phase with classification accuracy of 95.9 ± 0.2%and kappa value of 80.4 ± 0.6%. Moreover, the Linear SVM (L SVM) and Gaussian-SVM (G SVM) classifiers give similar performance with slightly lower accuracy and kappa value. The decision boundaries of previously reported SVM based fault classifiers are very close to the optimal MAP-NB decision boundary. Thus, this further strengthens the idea of implementing statistical fault classifiers to monitor the track irregularities based on dynamics in the lateral plane via in-service vehicles. The proposed method contributes towards digitalization of rail networks through condition-based and predictive maintenance.
At present, the detection of subway track irregularities is mainly carried out by track inspection vehicles and track inspection trolleys. Such detections are restricted by subway service time, so they can only be carried out once every few months. This study explored the possibility of using the vibration of the vehicle body collected by a novel portable detector to detect track geometry irregularities. It makes a particular contribution to the dynamic detection of track conditions and the reduction of maintenance costs. Based on the data collected by the portable detector, wavelet transform was used to analyze the vibration of the vehicle body. The results confirmed that this method was effective in enhancing the correlation between vibration accelerations and track irregularities. Second, a data set processed by wavelet transform was resampled by a hybrid sampling method which uses clustering methods and considers data imbalance within each category. In this way, the imbalance ratio of the data set was found to be reduced without changing the original data set structure. Finally, the random forest algorithm and the gradient boost decision tree algorithm were adopted for classifying track regularity and irregularity data. The results showed that both two algorithms, especially the random forest algorithm, performed well for the longitudinal level track irregularity and the alignment track irregularity.
The increasing demand in mobility forms a major challenge for modern cities, even more so when examined under the prism of transition from traditional to CO2-free mobility. Railway infrastructure forms a main carrier for the mobility of people and goods and a salient component of critical infrastructures. The increased traffic frequency in urban transport imposes higher capacity demands and leads to more frequent damage and more severe deterioration and associated disruptions to service and availability. Aligning with the spirit of smart cities, and data-driven decision support, infrastructure operators require timely information regarding the current (diagnosis) and future (prognosis) condition of their assets in order to sensibly decide on maintenance and renewal actions. Railway condition assessment has traditionally heavily relied on-site visual inspections. Main measurement parameters for railway tracks are obtained since the 1960s. Quality, accuracy, and precision of measurements heavily evolved since then, including aspects such as storage, analysis, and interpretation of data. In recent years, specialized monitoring vehicles offer an automated means for relaying essential information on condition, obtained from diverse measurements including laser measurements, vibration, image, and ultrasonic information. Powered by this information diagnostic vehicles have shifted assessment from a reactive to a predictive mode. More recently, in-service vehicles equipped with low-cost on-board monitoring (OBM) measuring devices, such as accelerometers, have been introduced on railroad networks, traversing the network at higher frequencies than the specialized diagnostic vehicles. The collected information includes position, acceleration, and in some cases force measurements. The measured data require interpretation into quantifiable track-quality indicators, before it can be meaningfully incorporated in asset management tools. These indicators form the basis for real-time forecasting of condition evolution and asset management, which are essential traits of a transport infrastructure that fits the vision of smart cities. This chapter explores the state of the art of OBM for railway infrastructure condition assessment, conducting a thorough review of data-processing methodologies, which is further complemented with application examples.
Monitoring rail roughness in the railway network allows directing grinding actions to where they are needed to reduce rolling noise and large wheel/rail forces. To be able to measure rail roughness on a large scale, indirect measurements onboard railway vehicles have to be carried out. Existing methods use either axle box acceleration (ABA) or under-coach noise measurements to monitor the rail roughness indirectly. The two main challenges with rail roughness estimation from vibroacoustic signals measured onboard vehicles are to separate wheel and rail roughness and to take into account varying track dynamics in the railway network. Both questions have not yet been addressed sufficiently.
In this paper, an enhanced method for estimating rail roughness from ABA is presented. In contrast to all existing methods in the literature, the presented method operates in the time domain. A time-domain method has the advantage that the spatial variations of roughness become visible and paves the way for the detection of localized defects such as squats or deteriorated welds. The method is based on a previously developed time-domain model for high-frequency wheel/rail interaction and estimates the time series of the roughness from the time series of ABA. In a first step, the time series of the contact force is calculated from the axle box acceleration using a Least Mean Square algorithm for source identification. In a second step, the combined wheel/rail roughness is obtained from the contact force based on a non-linear Hertzian contact model and a convolutional approach to determine wheel and rail displacement. Separation of wheel and rail roughness is possible by cycle-averaging the contact force over a distance corresponding to the wheel perimeter and performing the second step separately for the part of the contact force originating from the wheel and the rail roughness, respectively.
The method was tested for simulated ABA obtained from measured wheel and rail roughness. In the relevant wavelength range from 0.5 m to 5 mm, the rail roughness could be estimated with good accuracy for known track dynamics. Overall, deviations in 1/3-octave bands between estimated and actual roughness were below 1 dB. Only for low rail roughness, higher deviations of less than 2.6 dB occurred around the pinned-pinned resonance frequency. Uncertainties in the track parameters affect the roughness estimation, where the most critical parameter is the rail pad stiffness. A deviation of 20% in rail pad stiffness leads to deviations in the rail roughness of up to 3.5 dB in single 1/3-octave bands. The results illustrate the need to extend the method for the simultaneous extraction of track parameters and roughness from measured axle box acceleration.
Track geometry monitoring is essential for track maintenance. Dedicated track inspection vehicles are scheduled to measure track geometry irregularities throughout the railway network, so cannot inspect each line frequently. It is desirable to inspect geometry much higher frequently using in-service trains. One possible way is estimating track geometry from vehicle–body accelerations because the accelerators can be easily installed in vehicle–body. However, inverting track geometry from vehicle–body acceleration is a pending issue. Up to now, most research has focused on vehicle dynamics modelling and simulation. In this paper, we solve the problem using deep learning method and realistic measurement data. The training and test data are the inspection data acquired by comprehensive inspection trains from three main high-speed railways in China. The proposed AM–CNN–GRU model combines an Attention Mechanism (AM), a Convolutional Neural Network (CNN) and a Gated Recurrent Unit (GRU). The model’s inputs are vertical and lateral accelerations, and vehicle speed. The outputs are two vertical track irregularities, with wavelengths of 3–42m and 3–120 m, respectively. We evaluated the model by comparing the estimated irregularities with the actual measurements. We discussed different models, high-speed lines, sub-rail infrastructures and speed to validate the model effectiveness.
In railway transportation, track geometry irregularity is one of the main factors in controlling train safety. At present, railway practitioners typically use the track geometry car (TGC) based on the inertial navigation system to inspect track irregularities. However, TGCs are quite expensive, and their inspection interval is relatively long. Among a variety of emerging methods, using vehicle responses to estimate track irregularities seems very promising as it enables a cheaper and more efficient solution. In this work, an extended auto-encoder (EAE) is proposed to estimate the track longitudinal irregularity through car body acceleration. The mean absolute percentage errors of the estimated results on the simulated and the real-world dataset are 2.67% and 3.75%, respectively, which is 50%–55% lower than the traditional neural network. In the frequency domain, the characteristic wavelengths of 5.4 and 32 m can be effectively identified. Besides, the Bayesian deep learning (BDL) method is introduced to improve the EAE and estimate the confidence interval of the track longitudinal amplitude. A metric (coverage width and error) for evaluating and optimizing the performance of the estimated interval is proposed. The interval estimation result in the time domain has a 98% correct coverage rate of the ground truth and 93% in the frequency domain. Within the error range of plus/minus one standard deviation, the EAE model has an estimation accuracy of 94.2% for the standard deviation of track longitudinal irregularity, and the BDL-EAE can even reach 100%. Compared with the existing methods, our proposed model only requires car body acceleration and has the potential to use ordinary in-service trains for onboard track inspection.
Accurate and timely estimation of track irregularities is the foundation for predictive maintenance and high-fidelity dynamics simulation of the railway system. Therefore, it’s of great interest to devise a real-time track irregularity estimation method based on dynamic responses of the in-service train. In this paper, a Wasserstein generative adversarial network (WGAN)-based framework is developed to estimate the track irregularities using the vehicle’s axle box acceleration (ABA) signal. The proposed WGAN is composed of a generator architected by an encoder-decoder structure and a spectral normalised (SN) critic network. The generator is supposed to capture the correlation between ABA signal and track irregularities, and then estimate the irregularities with the measured ABA signal as input; while the critic is supposed to instruct the generator’s training by optimising the calculated Wasserstein distance. We combine supervised learning and adversarial learning in the network training process, where the estimation loss and adversarial loss are jointly optimised. Optimising the estimation loss is anticipated to estimate the long-wave track irregularities while optimising the adversarial loss accounts for the short-wave track irregularities. Two numerical cases, namely vertical and spatial vehicle-track coupled dynamics simulation, are implemented to validate the accuracy and reliability of the proposed method.
The aim of this work is the development of a model-based methodology for the estimation of lateral track irregularities from measurements from different sensors mounted on an in-service vehicle: a gyroscope to measure wheelset yaw angular velocity, two accelerometers to measure lateral acceleration of the wheelset and bogie frame, and an encoder to obtain forward velocity of the vehicle. The proposed methodology is based on the Kalman filtering technique, through the use of a highly simplified linear dynamic model of a bogie, capable of capturing the most relevant lateral dynamic behaviour of the entire vehicle. The simplified dynamic model (SM) is based on a vehicle running at variable forward velocity on a track, which comprises straight, curve and transition sections. Finally, the proposed methodology has been experimentally validated through an experimental campaign carried out in a 90 m 1:10 scaled track facility at the University of Seville and an instrumented scaled vehicle. The results of the estimation of the lateral alignment are analysed in the space domain and in the space frequency domain, according to standards. These results are promising, showing a good performance for monitoring lateral alignment on straight and curve tracks, with a very low computational cost. Only in the case of very sharp curves, when continuous flange contact takes place, the estimator is not able to precisely estimate lateral alignment.
Railhead roughness increases over time, leading to increased environmental noise and vibration. The use of axle-box acceleration (ABA) measurements on in-service railway vehicles to monitor rail roughness is potentially more cost-effective than other techniques. The measured acceleration requires signal processing to derive suitable metrics of railhead condition. A transfer function may be calibrated with direct roughness and ABA measurements made on a reference track, which may then be used to derive roughness spectra from subsequent ABA measurements. However, this approach is affected by variations in track dynamic behaviour, as well as variations in wheel roughness, which is inherently combined with rail roughness in the ABA measurement. This paper proposes an improved approach that (i) extracts the track’s dynamic stiffness parameters from the ABA measurements, enabling the derivation of the roughness-ABA transfer function for each section of track, and (ii) separates the wheel and rail roughness by synchronous averaging over several wheel revolutions. By accounting for variations in track properties and removing the influence of wheel roughness, initial modelling indicates that reliable measurements of rail roughness spectra can be obtained in practice.
Bridge frequencies and track irregularities are both the focuses of railway bridge condition assessment, which are coupled with each other in a vehicle-bridge system. Available algorithms face a great challenge when applied to simultaneously identify the natural frequencies and track irregularities of railway bridges using on-board measurement data. This paper proposes a novel algorithm for simultaneously identifying the frequencies and track irregularities of high-speed railway bridges using vehicle dynamic responses for the first time. An extended state-space model with unknown input condensation is established for time-dependent vehicle-bridge systems. We subsequently propose a new extended Kalman filter algorithm with an adaptive procedure for accelerating the convergence of estimation, which can simultaneously identify the frequencies and track irregularities of a railway bridge when a vehicle is running on it. The effectiveness of the proposed algorithm has been illustrated via numerical simulations of two real high-speed railway
bridges. The proposed algorithm provides a low-cost and high-efficient approach for identifying the natural frequencies and track irregularities of high-speed railway bridges.
In recent years, significant studies have focused on monitoring the track geometry irregularities through measurements of vehicle dynamics acquired onboard. Most of these studies analyse the vertical irregularity and the vertical vehicle dynamics since the lateral direction is much more challenging due to the non-linearities caused by the contact between the wheels and the rails. In the present work, a machine learning-based fault classifier for the condition monitoring of track irregularities in the lateral direction is proposed. The classifiers are trained with a dataset composed of numerical simulation results and validated with a dataset of measurements acquired by a diagnostic vehicle on the straight track sections of a high-speed line (300 km/h). Classifiers based on decision tree, linear and Gaussian support vector machine algorithms are developed and compared in terms of performance: good results are achieved with the three algorithms, especially with the Gaussian support vector machine. Even though classifiers are data driven, they retain the essence of lateral dynamics.
For Free Fulltext (Accepted Version) : http://urn.kb.se/resolve?urn=urn%3Anbn%3Ase%3Akth%3Adiva-269055
A track condition monitoring system that uses a compact on-board sensing device has developed and applied for track condition monitoring of regional railway lines in Japan. This study describes the application of time-frequency analysis for condition monitoring of railway tracks from car-body acceleration measured in in-service train. Simulation studies and field test results showed that Hilbert-Huang transform (HHT) gives good time-frequency resolution and intrinsic mode functions give the detail information of track faults.
This paper describes a new method to check for defects in railway tracks for improving passenger safety and comfort. The irregularities in the railway tracks are the fundamental cause of vibration, and different research projects are currently in progress for optimizing the process. External background noises causes the signals sent from the sensors to be distorted. In order to solve this issue, the Railway track Health Monitoring system uses a Dynamic differential Evolution algorithm (RHMDE) for identifying defects in railway tracks. Micro Electro Mechanical System (MEMS) accelerometers are mounted vertically and horizontally on the bogie and axle-box for sensing abnormalities. To locate the irregularities, a new method is included in the proposed RHMDE method. It automatically updates the location of an abnormality even if the signal from the Global Positioning System (GPS) is absent. Four different railway track problems were used for the experimental study, and the time and frequency domain responses were studied. The experimental setup of the proposed RHMDE is tested and compared the Chaos Particle Swarm Optimization (CPSO) and Genetic Algorithm (GA). The experiment results from the experiment prove that the proposed RHMDE method is the superior method for detecting faults in railway tracks. The RHMDE method will greatly improve the quality of railway transportation through detecting the track faults effectively and consistently.
Track geometry is an important parameter, used by railways, for routine track maintenance. Laser based Track Recording Coaches (LTRCs) are widely availed by railways, for recording track geometry defects, such as: vertical profile, alignment, gauge etc. In this paper, a novel method for monitoring vertical profile irregularities by in-service rail vehicle dynamics measurements has been proposed. The procedure takes as input, data from bogie installed inertial measurement unit (IMU) measuring vertical acceleration and pitch-rate, and encoder measuring longitudinal velocity. Extended Kalman Filter (EKF) with Rauch-Tung-Striebel (RTS) smoothing and Extended Kalman Particle Filter (EPF) are reviewed, for detecting vertical profile irregularities. Process model of nonlinear state estimation filters, EKF and EPF, has been developed by applying a novel analytical technique that approximates osculating circle of a point on the vertical profile curve. Further, measurement model has been devised by a novel procedure estimating curvatures from circles approximating trajectory sensed by vertical acceleration and pitch-rate curvature. Proposed method has been field-tested by comparing its output with LTRC’s recorded data.