Train Doors Research Articles

A platform noise dataset of various types of subway stations during all operating hours

The subjective perceptions of passengers are intricately linked to the noise levels in subway stations of urban rail transit systems. In this study, comprehensive measurement was conducted on the noise environment on the platforms during operational hours at four distinct types of subway stations within an urban metro network. To ensure precise data collection, four different noise and sound pressure measurement points were strategically positioned in the middle and at both ends of each station platform. Subsequently, the platform noise data was continuously monitored throughout the operation period, yielding a time series of original data on sound pressure levels for standardized continuous measurement of the platform. The collected results spanning the entire operation period were organized into hourly sets, encompassing information on noise sources such as vehicles entering and leaving the station, the opening and closing of train doors, and platform announcements. This dataset provides original data on noise sources at subway platforms throughout the all-day train operation period, aiding researchers in understanding the sources and levels of noise at subway station platforms. Additionally, this research supports noise control on subway platforms and provides a scientific basis for predicting the acoustic environment of rail transit stations and improving the platform acoustic environment.

Wavelet-Based Analysis of Motor Current Signals for Detecting Obstacles in Train Doors

Trains used in urban mass passenger transit often have side entrance doors through which passengers can rapidly enter and exit the train. These doors are typically electrically powered and automated. Many incidents have occurred in which a passenger is trapped and injured while passing through the doors as they are closing. Existing solutions rely on sensitive-edge sensors and current signal peak detection in the time domain to detect door obstructions. However, these methods have notable limitations: sensors are expensive, and sensor failure can result in safety risks, while time-domain signal analysis is prone to noise, potentially leading to false peak detection. The proposed efficient and cost-effective method enhances safety by implementing the torque control of a DC motor which limits the door closing force to prevent potential injuries. In addition, it reduces reliance on traditional edge sensors, which are prone to failure and may result in undetected obstructions. By using a robust time–frequency domain approach, the system ensures more accurate detection, minimizing potential injury risks. An obstruction of the door causes a corresponding change in the motor current. These changes can be detected by using the discrete wavelet transform to decompose the current signal. The norm and peak of the current are used as obstacle detection features, and appropriate threshold values are obtained from a simulation. The simulation results were validated through an experiment. The proposed novel system effectively detects forces between 100 and 200 N (indicating the presence of an object) within 0.3 s and complies with EN14752 safety standard. It can also differentiate between soft and hard objects trapped in train doors.

The Effect of Seat Layout on the Interaction of Passengers Inside the Train Carriage: An Experimental Approach for Urban Services

The platform–train interface (PTI) is one of the most complex spaces in metro stations. At the PTI, the interaction of passengers boarding and alighting reaches high density, affecting the boarding and alighting time, among other variables related to safety and efficiency. Developing research was performed to study the effect of seat layout on the interaction of passengers by means of experiments in a controlled environment. The laboratory facility included a mock-up of a train carriage and its adjacent platform. The train was representative of urban services in the Valparaiso Metro (Chile). The results showed that seat layout changed the patterns of interaction of passengers inside the train carriage. If seats were parallel to the movement of the train, then wider corridors inside the train were generated, and therefore, the number of passengers using this space could increase up to three times. However, in urban services, passengers were located closer to the train doors to be prepared for alighting, and therefore, the passenger numbers at the central hall remained the same with the seat layout. In addition, most passengers always used seats even if they were in a different position due to the aforementioned reasons. Further research will include passengers with reduced mobility and remaining inside the train while others are alighting to identify the effect of the space used on the interaction of passengers inside the train.

A Study on Health Indicator for Rivet Breakage in Train Door System Using Motor Current Signal

In this study, health indicator is developed to diagnose the degradation state against the rivet breakage, which is one of the critical failures in the train door system. For this purpose, motor current signals are acquired during the routine maintenance for the train doors of new, 13 and 15 years old, in which the rivet breakages are found in some of those in the 15 years. Features are extracted for decomposed signals after performing the wavelet packet decomposition. Useful features are selected that show monotonicity with respect to the door age and rivet failure, while those with redundancy are removed. Health indicator model is constructed using the selected features by regression to assess the current health against the rivet failure. As a result, it is found that the health indicator increases as the door ages and reaches the maximum value for those with rivet failures, which may be valuable information to the maintenance engineer.

Intercity Rail Transit Platform Anomaly Detection Using Door Tracking-Based Key Frame Extraction and AnoDet Network

The anomalies in the risk space between platform doors and train will threaten the safety of intercity rail transit. However, performing anomaly detection tasks with an accurate grasp of the time gap between the train doors closed and the train departure is challenging. The scarcity of anomalous samples and the lack of prior knowledge of anomalies also pose considerable obstacles to anomaly detection. To solve the aforementioned problems, a method based on key frames extraction and a two-stage Transformer network is proposed. The switch statuses of the train door are extracted by the door tracking algorithm, which effectively provides key frames for subsequent anomaly detection. An end-to-end two-stage Transformer, which is named AnoDet, is proposed for unsupervised anomaly localization and supervised multi-classification. The AnoDet is based on an image masking strategy, dual-channel network, knowledge migration base and anomaly object classification set prediction loss. The unsupervised localization model first utilizes the masking strategy to remove a large number of image patches that may contain anomalies in the inputs. Then, it trains a dual-channel network based on the remaining normal patches to inpaint anomalous inputs into normal images. The supervised classifier proposes a knowledge migration base to provide prior knowledge of anomalies and alleviate the scarcity of anomalous samples. It also introduces a set prediction loss to achieve a natural connection with the unsupervised model. In the intercity rail station anomaly dataset, our method achieves 97.8% precision and 96.6% recall for anomaly localization and obtains 88.9% mAP for multi-classification, outperforming the state-of-the-arts.

System for Security and Prevention of Terrorism Attacks at MRT Lebak Bulus, Jakarta

Public transportation systems such as the Lebak Bulus MRT are likely to become prime targets for terrorists due to the potential for disruption, destruction, and possible escape of perpetrators, due to size, openness, accessibility, lack of identification of passengers, and the number of people carried by the transportation system. Risk prevention and preparedness in the Lebak Bulus MRT transportation system is critical to internal security and requires proper analysis of asset vulnerabilities, clear awareness of criticality, possible countermeasures, and adequate methods to design, scale and optimize protection. This study uses a literature study research method. The author uses previous research sources that are relevant to the topic under study, especially related to the security system and prevention of terrorism attacks at MRT Lebak Bulus Jakarta. CPTED is a complex strategy that will include a variety of interventions and behavioral responses from different groups of users. CPTED involves territoriality, surveillance, access control, activity support, image/management, and target setting. In the Mass Rapid Transportation (MRT) Lebak Bulus, the vulnerability to terrorist attacks can be identified based on physical security theory. The main vulnerability is the MRT central depot. All MRT control is in the central depot, not on the mechanics. The MRT driver is only in charge of opening and closing the train doors, so that the control of the MRT, which is located at the Lebak Bulus central depot, places the vulnerability to hijackings by terrorists.

Deep reinforcement learning based train door adaptive control in metro tunnel evacuation optimization

Passenger evacuation in a metro tunnel is affected by distinct factors in the environment; thus, it is a huge challenge for both designers and managers throughout the metro system’s life cycle. A lateral evacuation platform is usually designed in a modern metro tunnel, and its evacuation efficiency is directly affected by train door opening strategy. By comparing the strategies of opening one, two, three, and four train doors, it is found that opening more doors improves evacuation efficiency. However, when two or more doors are accessed, this can lead to congestion on the evacuation platform near the train door. To reduce the congestion and improve evacuation efficiency, an adaptive train door control policy is proposed in this study, which features an adaptive dynamic programming method, namely, deep q network (DQN). The information of the environment and the state of the train doors are selected as the input and output of the control policy. By simulating scenarios under random, sequence, and DQN policies, the evacuation efficiencies are detailed. Results show that the proposed adaptive control method effectively improves evacuation efficiency. This method helps overcome the disadvantages associated with the fixed train door opening strategy and provides an optimal train door control policy during tunnel evacuation.

Research on Generating Method of On-board Equipment Test Sequence in ATO System of High Speed Railway

With the development of railway technology, the Train control system of Beijing-Zhangjiakou high-speed railway, which was opened in 2019, applies CTCS-3+ATO (Automatic Train Operation) technology to realize automatic driving at 350km/h, including automatic train opening, automatic and precise parking, and the linkage between train doors and platform doors. To ensure the safety and reliability of train control system, it is necessary to test the system. In this paper, the test cases used in the testing work are concatenated into test sub-sequences according to certain principles, and the concatenation problems of test sub-sequences are converted into TSP (Travelling Salesman Problem) for solving. PSO (Particle Swarm Optimization) algorithm was used to obtain the test sequence of ATO system in high-speed railway, which provided guidance for the test work.

YOLO-Former: Marrying YOLO and Transformer for Foreign Object Detection

The automatic detection of foreign objects between platform screen doors (PSDs) and metro train doors significantly affects personnel and property safety and maintains the train’s normal operation. However, some existing works only determine the presence of foreign objects but cannot indicate their categories. Besides, although deep-learning-based object detection algorithms can indicate the presence and categories of foreign objects, most of them only harness the information in region proposals, ignoring global contextual information. Furthermore, their performance comes at the considerable cost of computational complexity, and leading cannot be well deployed in the metro environment. To address these issues and better implement foreign object detection (FOD), we present You Only Look Once-Transformer (YOLO-Former), a simple but efficient model. YOLO-Former is accomplished based on YOLOv5 through the following procedure. First, the vision transformer (ViT) is introduced for dynamic attention and global modeling, thereby solving the problem that the original YOLOv5 only utilizes information in region proposals and has insufficient ability to capture global information. Second, the convolutional block attention module (CBAM) and Stem module are used to improve feature expression ability further and reduce floating point operations (FLOPs). Finally, we design various variants with different widths and depths to meet every need. Experiments on the foreign object detection dataset (FODD) and PASCAL VOC dataset demonstrate that YOLO-Former-x consistently outperforms other state-of-the-arts with significant margins (0.5 to 11.3 mean average precision, mAP, on FODD and 0.6 to 13.6 on PASCAL VOC dataset). Last but not least, YOLO-Former-x maintains real-time processing speed (27.32 and 28.17 frame per second, FPS, on TITAN Xp).

Evaluation of wheelchair accessibility in train stations using a spatial network

The universal design of public transportation is a crucial task worldwide, especially in aging societies. For this reason, public agencies and train operators have advocated for universal architectural design for train stations, for instance, installing elevators and designing step-free routes. However, the location of elevators is often decided on an ad-hoc basis, and wheelchair users are often forced to take unnecessary detours and collision risks. This is because few methodologies are available for making an integrated evaluation of spatial structure and wheelchair accessibility. This paper introduces an evaluation framework based on network analysis to evaluate wheelchair accessibility from the aspect of the spatial structure at train stations. Specifically, we first develop weighted network data that describe the spatial structure of train stations in detail to represent the mobility restrictions of wheelchair users and pedestrians. We next propose several indices to evaluate each station in terms of its wheelchair accessibility using these data: “reachable time indices,” which measure the time to the train doors from the ticket gate, and “level of path independence,” which describes the low risk of collisions on paths. This evaluation framework was applied to nine train stations in Tokyo, Japan, as case studies to examine the relationship between spatial structure and wheelchair accessibility. As a result, we found three important relationships. First, placing an elevator in the center of a platform narrows the range of the reachable times for wheelchair users. Second, the proximity of elevator doors to stairs or escalator entrances tends to increase the risk of collisions with pedestrians. Third, a wheelchair stairlift installed in place of elevators increases the minimum reachable time and increases the risk of collisions on designated paths. These insights will be useful for finding facilities that require improvement in the wheelchair accessibility of step-free routes. • Train-station spatial-structure network analysis reveals wheelchair accessibility. • Elevators at platform centers reduce the reachable time range of wheelchair users. • Elevator doors near stairs or escalators increase the risk of collisions. • Wheelchair stairlifts increase both the minimum reachable time and collision risk.

Feasibility study of a novel load detection system for train doors based on the acoustic characteristics of elastic tubes

Pinching accidents involving thin bags, canes used by the blind, or stroller wheels are a serious problem. To date, only a few studies have investigated the anti-pinch system of train doors. Hence, this paper proposes a novel pinching sensor that utilizes the acoustic characteristics inside an elastic tube. To investigate whether a highly sensitive sensor can be realized, the sensitivity was experimentally measured. In particular, we investigated whether a displacement of 1 mm can be detected along door rubber tubes, with a length of 2200 mm. First, we conducted load detection experiments for a conventional door rubber tube with an inner diameter of 15 mm. The sensor system comprises an earphone, a microphone, and an elastic tube. By adopting sweep signals of 0.5–4.5 kHz, power spectra inside the rubber were obtained, and the position responses were calculated via Fourier transform. Subsequently, we determined the applied loads based on the peak ratios. The subtraction method was adopted to detect small displacements (=1 mm). The obtained results indicated that displacements of 1 mm could be detected from 100 to 1400 mm. Next, load sensing experiments were conducted using five different types of door rubber tubes, and the obtained results indicated that the proposed method can be applied to both the widely used rubber tube, as well as in other types of door rubber tubes. Based on these results, we verified that the proposed system can be adopted as a highly sensitive pinching sensor system.

Social norms as a cost-effective measure of managing transport demand: Evidence from an experiment on the London underground

Demand for public transport in cities has been and is projected to increase, putting existing transport networks under increasing strain. Nudging passengers to behave in certain ways through the creation of a salient social norm has the potential to be a cost-effective mechanism to manage transport demand. Transport for London (TfL) implemented in the second half of 2017 an experiment on one of its busiest metro train platforms. The platform surface was painted to highlight the exact location of the train doors once it comes to a full stop and to direct passengers to wait in parts of the platform that would not obstruct passengers from alighting from the train and leaving the platform. Drawing on millions of individual train waiting times, we estimate the effect of this intervention to change passenger behaviour on the platform on train waiting and delay times. We use different sets of assumptions about what the counterfactual change in waiting and delay times would have been in the absence of the intervention. The intervention has reduced train waiting times by up to 6.6%. This reduction came about mainly through reducing delay times of trains once they are delayed. The reductions tend to occur during peak traffic hours. The implied cost-savings amount to a return of £6 per £1 investment.

Air Leakage Detection of Pneumatic Train Door Subsystems Using Open Set Recognition

For pneumatically operated train doors, continuous monitoring and earlier detection of air leakages would greatly reduce train downtime and ensure both safety and high availability of the rail system. Deep learning, a powerful tool to extract useful information from a large amount of sensor data is widely used in fault detection. A large collection of known fault samples is necessary to generate reliable deep learning models. However, in real-world applications, it is very difficult to precollect all possible types of fault or anomaly conditions to train the model. When an unknown sample is fed into a traditional deep learning model, the model may make misclassification and recognize this sample as one of the known classes. To address this issue, the concept of OSR was proposed, whose goal is not only to classify known classes but also to detect unknown samples. Existing OSR methods are commonly based on complex generative models, and the testing phase is divided into multiple stages, which makes it difficult to apply these OSR methods to real-time industrial tasks. In this article, we proposed an end-to-end OSR method with less computational cost based on a streamlined and lightweight convolutional neural network. The proposed method is applied to detect different levels of air leakage on the pneumatic train door subsystem. Compared with other existing OSR methods, the proposed method dramatically shortens the running time and reduces the use of parameters in the model with negligible loss of OSR accuracy.

The influence of the density inside a train carriage on passenger boarding rate

ABSTRACT Passenger alighting and boarding time is an important factor for crowded metros with high-frequency operations. Some operators leave the train doors open until the last passenger boards or passengers voluntarily stop boarding, while others regulate the boarding process. To evaluate whether it is effective to regulate the passenger while boarding or not, this research explores how the density inside the train would influence the passenger boarding rate by conducting a laboratory experiment. The results showed that the passenger boarding rate increased as the density increased, up to approximately 2.5–3.0 passengers/m2. Beyond that point, our results suggested that the boarding rate may have decreased, but the evidence was not conclusive. It is deduced that, under the experimented density (up to around 5 passengers/m2), the density does not have any apparent negative effects on the boarding rate, which implies that there is no strong evidence to recommend passenger boarding regulation.

Intercity train dwell time estimation by using robust regression method: a study case on Surabaya-Yogyakarta line

Dwell time is important an issue to optimise not only for train operation efficiency but also punctuality. This paper proposes a model based on a field survey of intercity trains along the Surabaya-Yogyakarta line by using a robust regression analysis of 138 observations from 13 trainsets operating on the line. Assuming no differences in door or platform width, a model has been formulated to include the number of people who are alighting and boarding, the type of platform and the number of train doors on one side. There were three robust estimators applied to the model. With a 2.037% mean absolute percentage error, the robust regression using the M estimator formed the dwell time model for this study. The results show that, between the dummy variable for platform type using a binary number (0 or 1), dummy variables for high-level platforms and mid-level platforms have opposing tendencies. High-level platform use reduces dwell time by 2.014 sec based on continuous variables; in contrast, dwell time increases by 9.353 sec based on continuous variables by using mid-level platforms. Furthermore, people who are alighting need less time than those boarding. Dwell time will increase by 1.222 sec per alighting passenger and 1.314 per boarding passenger, while it will decrease by 2.727 sec per door on a one-sided trainset. Moreover, dwell time estimates with the model have reduced trainset travel time.

Passengers’ evacuation from a fire train in railway tunnel

ABSTRACTTo acquire the evacuation time and average evacuation velocity of young adults under a train fire situation in railway tunnel, experiments and numerical simulations were conducted. According to the results, if the fire train continues running, the number of passengers in adjacent compartment affects the evacuation procedure in fire compartment greatly. The average velocity decreases by 45.7% when the adjacent compartment is 40% overloaded. If the fire train stops immediately, evacuation in the fire compartment is influenced greatly by the number and location of opened train doors. The average evacuation velocity decreases by 21.6% when two doors are opened on one side other than on two sides. Also, it is advisable to set the evacuation velocities of young adult male and female to be 1.2 m/s and 1.0 m/s respectively under train fire situation in railway tunnel. The results have important implications for rail safety.

The spatial and temporal variation in passenger service rate and its impact on train dwell time: A time-series clustering approach using dynamic time warping

ABSTRACTThis study aims to improve the understanding of the underlying mechanism of passenger boarding and alighting processes, as well as its potential influence on train dwell time and train operation. Empirical data collected from one of busiest metro stations in Seoul, Korea, demonstrates the spatial and temporal variation in the passenger service rate, as a result of interference between boarding, alighting, and standing passengers. This study postulates that the level of interference can be associated with the train car occupancy and the proximity of train door to entry points, as waiting passengers tend to cluster near the platform entries. A unique temporal service rate curve is derived from each door location. We introduce Dynamic Time Warping for similarity assessment and clustering. It revealed four groups of train doors showing distinct shapes of curve from each platform. The first cluster includes the train doors located closest to the platform entry points where the initial service rate is drastically impeded by severe inference among passengers. The level of interference gradually diminishes as the door is located farther away from the entry points, but the overall service rate decreases as well. A passenger service time model is derived to include the cluster variable. To test its significance, the prediction capability of the model is presented with and without the cluster variable. The results of this study may be used to guide waiting passengers along the platform to minimize interference and to avoid serious disruption during passenger service time.

Factors affecting variability in PM2.5 exposure concentrations in a metro system

The objectives of this study were to: (1) evaluate PM2.5 inflow to metro train cabins when doors open at stations; (2) assess the spatial and temporal variability in PM2.5 exposure concentration; and (3) quantify the relationship between in-cabin concentration versus outdoor and non-ambient PM2.5. We measured in-cabin PM2.5 concentrations using portable monitors at the door-side and center of a train cabin simultaneously on a Hong Kong metro line. In addition, platform and in-cabin pollutant concentrations near a train door were simultaneously measured. Short-term spikes in PM2.5 concentrations typically occur near train doors when doors open, related to inflow of ambient air aboveground and tunnel air underground. In-cabin PM2.5 exposure concentrations are typically lower away from the doors when the doors open. PM2.5 concentrations inside train cabins and on station platform operating above-ground are more influenced, compared to underground, by outdoor PM2.5. Moreover, non-ambient sources contribute approximately 50% of train in-cabin and station platform PM2.5 concentrations during underground operation. The results help more accurately quantify commuting PM2.5 exposure on a metro system, and can be used to improve population-based exposure simulation models.

한국의 빌린도시 현상의 유물화로서 옥외광고 확장을 제한하는 융복합적 탐색

Abstract To derive a check list of a borrowed city phenomenon, this discussion, through two cities subway OOH advertising in door, compared for convergence by segment analysis. First, discussions considered to be close to the topic of this discussion were arranged and summarized. Seoul and London were selected, and both sides of subway train doors in two cities selected, their states pictured and the images analyzed. As a result of comparison, in case of Seoul subway, the boundary between public information and commercial information(outdoor advertising) was very vague, so the borrowed city phenomenon was very generally analyzed. However, In case of London subway, its result was very contrary to that of Seoul. Eventually, Subway train doors of Seoul city as a borrowed city extracted check points of height, size, color, interval, space and permanency. Therefore in order to study serious discussion this contribution is necessity to limit thoughtless OOH advertising by these checklists in Korean cities.

Investigating ramp gradients for humps on railway platforms

Horizontal and vertical gaps between the train and the platform are a major safety concern for railway passengers, especially for disabled passengers. London Underground is implementing a programme to install platform humps to remove vertical differences between the train and the platform. In order to properly design platform humps, this study empirically investigated the effects of the design factors of the ramps, namely the slope and cross-fall gradients, on disabled passengers. The investigation consisted of two experiments: one where 20 participants were asked to walk on simulated slopes, and the other where 25 participants were asked to board or alight from the simulated train from or onto the slopes. The slope gradients tested were 3·0% (1:33), 5·2% (1:19) and 6·9% (1:14) with the cross-fall gradients 1·5% (1:67), 2·0% (1:50) and 2·5% (1:40). The results showed that the slope gradient does not largely affect the participants' performance of longitudinal walking on the slopes or their subjective safety evaluation, but would cause additional difficulty for them to board/alight from the train from/onto the slope. This suggests that train doors should not stop next to the ramp. There was little evidence concerning the effects of the cross-fall gradient. The results provide useful information for designing platform humps.