Fig 1 - uploaded by Haoyu Wang
Content may be subject to copyright.
Source publication
To ensure railway operations safe, track geometry parameters, e.g., track gauge, are usually inspected using track geometry cars. The measurement frequency of track geometry cars is low (twice per year) due to high operational costs and track possession. An innovative way to perform track inspection at high frequency and affordable cost is using mo...
Contexts in source publication
Context 1
... tracks are regularly inspected to assess the track quality and ensure their safe operation [1][2][3]. The track quality is described by track geometry parameters normally including track gauge, alignment, longitudinal level (also referred to as surface or profile [4], as shown in Fig. 1), cross level, and twist [5][6][7]. Track geometry tends to drift away from the design geometry as the number of passed tonnage growing [8]. After collection, track geometry parameters are used for the detection of defects depending on the amplitudes. Various thresholds, i.e., the alert limit, intervention limit and immediate action ...
Context 2
... figure, we first compared the measurement results and then calculated the discrepancies (absolute value). According to [12], the 95 percentile of the discrepancies is used as an indicator evaluate the repeatability and reproducibility of the device, the limits of which are 0.5 mm and 0.8 mm, respectively. The histograms of the tests are shown in Fig. ...
Context 3
... This may mean the effect of measuring distance becomes more significant in some specific locations than others. To validate the assumption, the corresponding location of the maximal discrepancy of Test 3 and Test 4 (18.431 km and 18.408 km, respectively), as indicated by P1 in Fig. 9f and P2 in Fig. 9g, are marked on the map as shown in Fig. ...
Context 4
... shown in Fig. 11, the main discrepancy in both Test 3 and Test 4 appears in the same bridge transition zone at the bridge-embankment transition (when the train leaves the bridge) indicating that measuring distance has an effect on the track geometry measurement in track transition zones. The reason for this may be related to the severe track geometry ...
Context 5
... analyse the effect of the measuring distance in track transition zones, a theoretical study was performed using the Finite Element (FE) method. The dynamic model used is shown in Fig. 12a. The model consists of three main parts, namely two ballast tracks on the embankment and a slab track on a bridge. The ballast tracks are both 48 m long, and the bridge section is 24 m long. The total length of the model is 120 m. The vehicle model is 23 m long, moving from the one end of the track to the other end at the velocity 140 ...
Context 6
... [21]. The cross-sectional and mass properties of the UIC54 rails are used and the element length of rail is 75 mm. Spring-damper elements between rails and sleepers are used to simulate fasteners. Sleepers and ballast are modelled by three-dimensional elastic bodies which are composed of the selective reduced integrated hexahedral solid Fig. 11. Track section corresponding to maximal discrepancies, P1 from and P2 from Fig. 9. elements. The vehicle model is idealised as a multibody system consisting of one carbody, two bogies and four wheelsets. The primary and secondary suspensions are modelled by spring-damper ...
Context 7
... model the track geometry irregularity in the transition zone, a downwards differential settlement is added to the embankment, while the track on the bridge remains unchanged, as shown in Fig. 12b. The differential settlement used here is 8 mm, which is often found in field measurements e.g., in [24]. After adding the differential settlement, the track geometry is indicated in Fig. 12c (enlarged 30 times). Some sleepers next to the bridge are hanging due to the differential settlement and the constraint of the track on ...
Context 8
... in the transition zone, a downwards differential settlement is added to the embankment, while the track on the bridge remains unchanged, as shown in Fig. 12b. The differential settlement used here is 8 mm, which is often found in field measurements e.g., in [24]. After adding the differential settlement, the track geometry is indicated in Fig. 12c (enlarged 30 times). Some sleepers next to the bridge are hanging due to the differential settlement and the constraint of the track on ...
Context 9
... system is recorded, regarded as the track longitudinal level measured by the inspection system model. The model of the inspection system is placed at the close location (1.1 m) and the far location (3.1 m) respectively as per the field measurements (Section 3). The calculated track longitudinal level (D1) are compared at two locations in Fig. ...
Context 10
... shown in Fig. 13, the results are very similar in the track sections on the embankment (before 45 m and after 85 m) as well as in the section on the bridge (from 56 m to 74 m). Considering the track sections on the embankment having a lowing track stiffness and that on the bridge high, it shows that the loaded and unloaded measurements producing the ...
Context 11
... 75 m to 85 m). Especially, the results of the unloaded measurement account for 83.6% of the loaded results, which means more than 15% value lost. The bottom values of the loaded and unloaded case are 4.23 mm and 5.06 mm respectively. The numerical results correlate strongly with those in the field as shown by the comparison of discrepancies in Fig. ...
Context 12
... shown in Fig. 14, discrepancies can be found both before and after the bridge in the transition zone, wherein the one in the bridgeembankment transition (train moving off the bridge) is larger. It can also be seen that the discrepancy curve of the measurements taken in the field fluctuates more compared to that of the simulation. This proves that the ...
Context 13
... has been measured monthly using the mobile track inspection system. The whole track has been measured five times, including August 2018, September 2018, October 2018, November 2018, and January 2019. December 2018 is unfortunately not measured due to practical reason. The track longitudinal level (D1) in a representative location is analysed in Fig. ...
Context 14
... seen in Fig. 15a, the longitudinal levels of the 5 runs are similar to each other before 20.5 km and after 23 km but between these values is a relatively large variation. The largest change appears at 21.85 km, which can be seen in Fig. 15b; the amplitudes increase from September 2018 reaching the maximum level in November 2018, while reduces in ...
Context 15
... seen in Fig. 15a, the longitudinal levels of the 5 runs are similar to each other before 20.5 km and after 23 km but between these values is a relatively large variation. The largest change appears at 21.85 km, which can be seen in Fig. 15b; the amplitudes increase from September 2018 reaching the maximum level in November 2018, while reduces in January 2019. The standard deviation of the longitudinal level is calculated using a 25 m sliding window (100 samples) following the method used in [25], as shown in Fig. ...
Context 16
... The largest change appears at 21.85 km, which can be seen in Fig. 15b; the amplitudes increase from September 2018 reaching the maximum level in November 2018, while reduces in January 2019. The standard deviation of the longitudinal level is calculated using a 25 m sliding window (100 samples) following the method used in [25], as shown in Fig. ...
Context 17
... to Fig. 15a, Fig. 16a shows that the standard deviation is higher at 20.7 km, 21.8 km and 24.2 km than at other locations. Fig. 16b shows the detailed change of the standard deviation overtime at 21.8 km, revealing the local degradation process of the track quality. The standard deviation is reduced in January 2019, indicating that maintenance was performed ...
Context 18
... to Fig. 15a, Fig. 16a shows that the standard deviation is higher at 20.7 km, 21.8 km and 24.2 km than at other locations. Fig. 16b shows the detailed change of the standard deviation overtime at 21.8 km, revealing the local degradation process of the track quality. The standard deviation is reduced in January 2019, indicating that maintenance was performed between November 2018 and January 2019. After consulting with the railway owner (ProRail), it was confirmed ...
Context 19
... a track maintenance was performed in December 2018. It should be noted that the quality of the track for those locations with a high standard deviation is poor and remains poor even after maintenance, confirming the 'memory effect' of track [26,3,27]. The reason for the fast degradation is due to poor subgrade confirmed by the maintenance staff. Fig. 17 shows the average standard deviation of the longitudinal level D1 over the entire track section. The increase rate can be calculated (indicated by the red solid line), reflecting the rate of track degradation. Moreover, the effect of maintenance can be seen by the vertical dotted line. This information can provide a deeper ...
Similar publications
Plant turnaround shutdown is a fundamental asset management in capital intensive process-based industries. Offshore platform is like a ‘town’ in the middle of the sea with a facility for well drilling to explore, process for petroleum and also natural gas, to extract and store. The platform also has facilities for workforce as well. Shutting do...
Citations
... Rights reserved. a smoothed vertical position (Wang et al., 2021). On the basis of this measurement the fractal values can be computed, via appropriate filtering and processing steps. ...
Partially Observable Markov Decision Processes (POMDPs) can model complex sequential decision-making problems under stochastic and uncertain environments. A main reason hindering their broad adoption in real-world applications is the unavailability of a suitable POMDP model or a simulator thereof. Available solution algorithms, such as Reinforcement Learning (RL), typically benefit from the knowledge of the transition dynamics and the observation generating process, which are often unknown and non-trivial to infer. In this work, we propose a combined framework for inference and robust solution of POMDPs via deep RL. First, all transition and observation model parameters are jointly inferred via Markov Chain Monte Carlo sampling of a hidden Markov model, which is conditioned on actions, in order to recover full posterior distributions from the available data. The POMDP with uncertain parameters is then solved via deep RL techniques with the parameter distributions incorporated into the solution via domain randomization, in order to develop solutions that are robust to model uncertainty. As a further contribution, we compare the use of Transformers and long short-term memory networks, which constitute model-free RL solutions and work directly on the observation space, with an approach termed the belief-input method, which works on the belief space by exploiting the learned POMDP model for belief inference. We apply these methods to the real-world problem of optimal maintenance planning for railway assets and compare the results with the current real-life policy. We show that the RL policy learned by the belief-input method is able to outperform the real-life policy by yielding significantly reduced life-cycle costs.
... The structural and ballast degradation measurements were carried out outside the area where the ballast was removed and replaced. For geometric measurements, a Track Recording Car (TRC) model EM100 [36] was used, as it is commonly employed for the continuous assessment of geometric quality in railway networks [28,[37][38][39][40][41][42]. This vehicle can measure all geometric parameters at every 0.25 m, while running at the operational speed of the track. ...
The tamping process is crucial for correcting geometric irregularities on railway tracks, especially on newly constructed or renovated tracks that are more susceptible to accumulating defects. Recent studies have demonstrated that multiple insertions of tamping tines lead to greater ballast compaction and improved lateral resistance of the track. However, further evaluation is necessary to assess their impact on other important track parameters. This study evaluated single and multiple insertions to determine which would promote more effective maintenance for newly built or renovated track conditions, focusing on geometry correction, track structural behaviour, and potential ballast degradation. These tamping methods were tested in a heavy haul line, showing that multiple insertions improved geometric quality, increased lateral resistance, and enhanced vertical track stiffness without causing significant ballast degradation. Therefore, using multiple insertions is recommended for newly built or renovated tracks to enhance geometric quality and structural behaviour, providing a more effective maintenance solution.
... Springs, dampers, rods and other connections are implemented using elastic and elastic-viscous power elements. In 131 this paper, the rheological models of Hooke and Voigt [15], respectively, are used to describe elastic and elastic-viscous 132 force elements (see Figure 4). The mathematical model of force F and moment M in the coordinate system of the elastic element are determined by 151 the following expression in matrix form: ...
The Author analyze the transitional mode of movement of a passenger train caused by sudden braking by the driver. To study the possibility of derailment of passenger cars, a simulation model of the movement of a train consisting of a locomotive and twenty passenger cars was developed in the Universal Mechanism software environment, designed to study the dynamics and kinematics of mechanical systems, which include railway rolling stock. The developed model allows taking into account the longitudinal, transverse and vertical vibrations of all cars and locomotive. All bodies in the model are assumed to be absolutely rigid. The assumption of non-deformability of the bodies is based on the fact that the stiffness of the spring suspension and elastic connections is significantly less than the structural rigidity of the bogie frames and the bolster structure, and the frequency of elastic vibrations of these bodies is much higher than the frequency of their vibrations on the spring suspension. Passengers and cargo in a wagon are considered to be non-deformable and integral with the wagon body, similarly in the case of a locomotive body. Between the elements of the system, connections are involved that regulate certain relative movements of these elements.
... For example, the track quality and ride comfort at transition zones have been analyzed using acceleration measured by smartphones [14]. In [15], track vertical displacement due to different loading conditions in transition zones was measured using a mobile system called the rail infrastructure alignment acquisition (RILA) system. In our previous work [16], key performance indicators (KPIs) for monitoring transition zones were proposed from wavelet-based features of axle box acceleration (ABA) signals. ...
This study presents a measuring framework for railway transition zones using a case study on the Swedish line between Boden and Murjek. The final goal is to better understand the vertical dynamics of transition zones using hammer tests, falling weight measurements, and axle box acceleration (ABA) measurements. Frequency response functions (FRFs) from hammer tests indicate two track resonances, for which the FRF magnitudes on the plain track are at least 30% lower than those at the abutment. The falling weight measurements indicate that the track on the bridge has a much higher deflection than the track on the embankment. Two features from ABA signals, the dominant spatial frequency and the scale average wavelet power, show variation along the transition zone. These variations indicate differences in track conditions per location. Finally, the ABA features in the range of 1.05–2.86 m⁻¹ are found to be related to the track resonance in the range of 30–60 Hz. The findings in this paper provide additional support for physically interpreting train-borne measurements for monitoring transition zones.
... The data given in Tables 1 and 2 and Figures 6 and 7 show that the stability factor against wheel derailment when the locomotive passes turnouts on its side up to a speed of 50 km/h and in a straight direction up to 90 km/h are in acceptable limits, i.e., they take values of at least 1.4. The results show that the dynamic performance of the diesel locomotive when passing turnouts meets the requirements of the "Norms for Permissible Speeds of Locomotives and Wagons on Railway Tracks of Gauge 1520 (1524) mm of Railway Transport of the Republic of Kazakhstan" (hereinafter referred to as the norms for permissible speeds) [17,18]. Consequently, the speed of movement on the turnouts in accordance with the norms for permissible speeds of movement according to the assessment of dynamic indicators is limited to 40 km/h when moving on the turnouts on the side and the design indicators when driving on the turnouts in the forward direction [19]. ...
... Experimentally obtained values of the vertical dynamics coefficients of the first and second stages of the spring suspension of the tested diesel locomotive CKD6E-2108 did not exceed the permissible limits, and the maximum values for the vertical dynamics coefficients were 0.11 for the first stage and 0.25 for the second stage. In this case, it should be noted that the obtained limit value of 0.25 of the vertical dynamics coefficient of the second stage was recorded when moving at a speed of 90 km/h, which is 12.5% higher than the designated speed [18]; - ...
In the interaction of the rolling stock and the upper structure of the railway track, intense dynamic loads occur. They have a destructive effect both on the parts of the rolling stock and on the elements of the superstructure of the track. In order to develop a durable, rational and reliably functioning design of cars and locomotives with good dynamic properties and good indicators of the impact of rolling stock on the railway track, along with theoretical computational studies, experimental studies are also required, which are usually the final stage in the design and implementation of rolling stock or the modernization of existing ones, such as locomotives and wagons, in order to improve their strength and dynamic performance. This article presents the results of field tests to determine the dynamic performance of the type CKD6e diesel locomotive. The description of the preparation of the CKD6e shunting locomotive for testing is given. An analysis of the dynamic performance of a diesel locomotive during the passage of turnouts, on a straight section of the track and in a curve with a radius of 400 m, was carried out. The studies performed showed that the minimum value of the stability factor against wheel derailment on a straight section of the track is significantly higher than the standard value. The experimentally obtained ratio of frame forces to the static load from the wheelset on the rails, the coefficients of vertical dynamics of the first and the second stages of suspension and the coefficient of stability against derailment of the wheel from the rail were registered on the track section in a curve with a radius of 400 m meet the current requirements. A calculation scheme and equations of vertical oscillations are proposed, an analysis is carried out according to the graphs of movements of bogies and a locomotive body when moving along irregularities of different lengths at different speeds.
... The average monthly tonnage is 3.8 MGT with a standard deviation of less than 0.3 MGT, which can be regarded as constant train loads. The measurement system is RILA (Rail Infrastructure Alignment Acquisition system) [48], developed by Fugro, is mounted on any passenger train using GNSS (Global Navigation Satellite System), IMU (Inertial Measurement Unit), LiDAR scanner, video cameras, and laser vision technologies (as shown in figure 2) to measure track geometry data. For safety reasons, it is usually attached to the rear of a passenger train via a custom-built adaptor. ...
... Measurement system for track geometry. Reprinted from[48], Copyright (2021), with permission from Elsevier. ...
... Sensor configuration. Reprinted from[48], Copyright (2021), with permission from Elsevier. ...
With the fast development of computer science, many prediction models based on machine learning methods have been used in the railway industry, which can better predict the random characteristics in track degradation, plan maintenance activities, and eventually meet the requirement of railway transportation. However, the applicability and generality of these models are unclear and the comparative analysis of these models on the random railway track is rare, which makes it hard for railway engineers to choose the most suitable prediction models in practice. In this paper, the track longitudinal level of a section of a railway which measured monthly by Rail Infrastructure Alignment Acquisition system (RILA) for 1.5 years in the Netherlands has been analysed using multiple mathematical methods. After that, three machine learning-based prediction models were developed to predict the future development of the track longitudinal level, using Support Vector Machine (SVM), Grey Model (GM) and Deep Neural Network (DNN). The prediction performance of different prediction models is compared and discussed. Recommendations for choosing prediction models and further development are provided.
... The track longitudinal level measures the deviation in the vertical direction of the running surface on rails from the smoothed vertical position, which indicates the vertical smoothness of a track (Wang et al., 2021). ...
The Infrastructure-BIM is constantly growing in all the main European countries. The focus of this research is the digital twin creation of the Ogliastro-Sapri High-Speed rail in Italy. The entire project was carried out using Bentley Systems® BIM-based tools such as OpenRail Concept Station and OpenRail Designer. The advantages and limitations of the methodology applied to a real railway project are highlighted. As the main result, a specific interoperable asset manager is proposed for maintenance purposes.
... Springs, dampers, rods and other connections are implemented using elastic and elastic-viscous power elements. In this paper, the rheological models of Hooke and Voigt [15], respectively, are used to describe elastic and elastic-viscous force elements (see Figure 4). ...
The article analyzes the transitional mode of movement of a passenger train caused by emergency braking by the driver. To study the possibility of derailment of passenger cars, a simulation model of the movement of a train consisting of a locomotive and twenty passenger cars was developed in the Universal Mechanism software environment, designed to study the dynamics and kinematics of mechanical systems, which include railway rolling stock. The developed model allows taking into account the longitudinal, transverse and vertical vibrations of all cars and locomotive. All bodies in the model are assumed to be absolutely rigid. The assumption of non-deformability of the bodies is based on the fact that the stiffness of the spring suspension and elastic connections is significantly less than the structural rigidity of the bogie frames and the bolster structure, and the frequency of elastic vibrations of these bodies is much higher than the frequency of their vibrations on the spring suspension. Passengers and cargo in a wagon are considered to be non-deformable and integral with the wagon body, similarly in the case of a locomotive body. Between the elements of the system, connections are involved that regulate certain relative movements of these elements.
... The dynamic performance of the diesel locomotive on turnouts was determined by measurements made when the diesel locomotive was passing turnout's No. 13 of grade 1/9 and No. 4 29 of grade 1/11 on rails R65 located at the station Belle. The direction when the locomotive CKD6E-2108 moved with the 1st wheel pair forward from the station towards the entrance switches was taken as the direct course of the diesel locomotive [14]. Figure 3 shows the oscillograms of the primary measurements of dynamic processes recorded at turnouts. ...
In the interaction of the rolling stock and the upper structure of the railway track, intense dynamic loads occur. They have a destructive effect both on the parts of the rolling stock and on the elements of the superstructure of the track. In order to develop a durable, rational and reliably functioning design of cars and locomotives with good dynamic properties and good indicators of the impact of rolling stock on the railway track, along with theoretical computational studies, experimental studies are also required, which are usually the final stage in the design and implementation of rolling stock or modernization of existing ones. Locomotives and wagons in order to improve their strength and dynamic performance. The article presents the results of field tests to determine the dynamic performance of the type CKD6e diesel locomotive. The description of the preparation of the CKD6e shunting locomotive for testing is given. An analysis of the dynamic performance of a diesel locomotive during the passage of turnouts, on a straight section of the track and in a curve with a radius of 400 m, was carried out. The studies performed showed that the minimum value of the stability factor against wheel derailment on a straight section of the track is significantly higher than the standard value. The experimentally obtained ratio of frame forces to the static load from the wheelset on the rails, the coefficients of vertical dynamics of the first and second stages of suspension and the coefficient of stability against derailment of the wheel from the rail, registered on the track section in a curve with a radius of 400 m) meet the current requirements. A calculation scheme and equations of vertical oscillations are proposed, an analysis is carried out according to the graphs of movements of bogies and a locomotive body when moving along irregularities of different lengths at different speeds.
... A specific set of such condition indicators are the so-called fractal values, which are derived from the longitudinal level measurement. The longitudinal level represents the vertical smoothness of the rail and is measured via a diagnostic vehicle as the deviation of the running surface of the rail from the smoothed vertical position [35]. Fractal values are the outcome of fractal analysis. ...
