Search

Online quantitative safety monitoring is the key technology for ensuring the operational safety of the automatic train protection (ATP) system for the future advanced train control system. However, the traditional formal verification method can only deal with the system of small scale since these verification algorithms exhaustively search all executable paths of formal models. Especially for the formal models with uncertain parameters, the problem of memory overflow will occur due to the state space explosion, which makes verification algorithms fail to converge. To solve this problem, an intelligent quantitative safety monitoring approach is proposed by integrating the probabilistic model checking method (PMCM) with neural network in this paper. To begin with, the instantiated continuous-time Markov Chains model is verified by PMCM offline. Then, the neural network is constructed and optimized based on the offline verification results. The quantitative safety boundaries (QSBs) for all quantitative safety levels are also computed by the designed algorithm. Furthermore, hierarchical iterative evaluation method is applied to efficiently evaluate the reliability performance of the ATP online. Finally, the quantitative safety level of ATP or its subsystem will be determined by both reliability performance and the previous QSBs or the neural network online. [ABSTRACT FROM AUTHOR]

In recent years, with the rapid development of Guangdong-Hong Kong-Macao Greater Bay Area, the passenger flow between cities in the region has increased rapidly, and the requirements for transportation efficiency of intercity railways are getting higher and higher. Moving block and automatic turn-back technology is an important development direction of signal system to improve railway transportation efficiency, which can effectively shorten the tracking interval between trains and the turn-back interval of turn-back stations, thus improving the transportation capacity and efficiency of lines. At present, moving block technology and automatic turn-back have not been applied in intercity railway. According to the technical status of moving block and automatic turn-back, the feasibility of function optimization of existing intercity CTCS2+ATO train control system in Pearl River Delta is deeply analyzed in this paper, and the overall framework of realizing moving block and automatic turn-back in the intercity CTCS2+ATO train control system in Pearl River Delta is put forward. Finally, the working principle and realization of CCS block mode switching, driving permit calculation, automatic turn-back function optimization and turn-back process are introduced emphatically. The results show that it is feasible to realize moving block and automatic turn-back under the condition that the existing intercity CTCS2+ATO train control system in the Pearl River Delta is slightly changed, which is of great significance to improve the transportation efficiency of intercity railway lines in the Pearl River Delta. [ABSTRACT FROM AUTHOR]

In order to study the effect of the velocity pulse on the dynamic response of the train-bridge system of the high-speed railway simple supported beam bridge, the velocity pulse is simulated by the trigonometric function method and superimposed with the far-field earthquake without pulse to synthesize the pulse with different pulse types, pulse periods and pulse peaks. A 10 × 32m typical high-speed railway simple supported beam bridge is considered an case illustrating study. Then, the dynamic response of train-track-bridge coupling system is calculated by train-track-bridge seismic analysis software TTBSAS. Afterwards, the influence of pulse near-field earthquakes parameters and vertical components on dynamic response of train-bridge system and the safety of the train on the bridge are discussed in detail.The new derailment evaluation index is adopted to evaluate driving safety under earthquakes. The train safety control of simply supported beam bridge under the action of near-field earthquake is studied. The results show that the impact of pulse ground motion on the dynamic response of the train-track-bridge coupling system is significantly higher than that of no-pulse ground motion, especially the impact on the bridge and rail subsystem is more significant than that of train subsystem. Under the excitation of ground motion intensity of 0.05g ∼ 0.15g, the safe speed threshold of pulse near-field ground motion is smaller than that of far-field ground motion. When the ground motion intensity is 0.20g ∼ 0.30g, the safe speed threshold of pulse near-field ground motion and far field ground motion is 200km/h. So, the pulse near-field earthquake poses a greater threat to the safety of the train on the bridge than the far-field earthquake. Therefore, the influence of pulse near-field earthquakes should be considered in the seismic design. The research results of this paper can provide theoretitcal support for the design of a high-speed railway bridge in the near-field area. [ABSTRACT FROM AUTHOR]

Automatic train operation (ATO) is a critical component of automatic train control (ATC) systems. The ATO automatically adjusts the speed of trains, ensuring the safety of trains and increasing the passenger capacity of urban railway networks. The traditional ATO models employ a linear approximation approach to fit the nonlinear control model of the train, such as proportional–integral–derivative (PID) or model predictive control (MPC). However, due to the complexity of actual train operations, these methods fail to accurately fit train models. Thus, they are unable to shorten the operational intervals of trains and meet the demand for improving system efficiency of ATC. In recent years, machine learning-based methods have been employed for deriving more accurate train control models. However, these methods require trains to offload massive amounts of data to central servers for centralized training, which is challenging in CBTC systems due to limited wireless bandwidth. In this paper, we propose using federated learning (FL)-based edge computing to train the ATO model without transmitting a large amount of data. The onboard computing devices are used to perform local federated learning when the train is idle. Additionally, we present a Tab-Transformer-based machine learning model for ATO policy to improve the prediction accuracy of federally trained models. Our extensive simulation results demonstrate that the proposed federated learning scheme improves the efficiency of model training, and the Tab-Transformer-based ATO model achieves better driving performance. [ABSTRACT FROM AUTHOR]

The pulse effect of near-fault ground motion has asevere threat to high-speed railway train running on bridge. Current research generally relies on numerical simulation and lacks physical test data. In this paper, a scale indoor physical test of high-speed train running on bridge is conducted and dynamic characteristics of high-speed railway train–track–bridge system under near-fault pulse-type ground motion are explored. The bridge of the physical test is excited by shaking table with lateral inputs of near-fault pulse-type ground motion, pulse-free ground motion and far-field ground motion. A corresponding numerical model is established and validated by the test data. Then, the influence regularities of pulse effect on system dynamic response and running safety indexes are explored under different peak ground acceleration (PGA) and train speeds using data of the physical tests and the numerical simulation comprehensively. Results suggest that compared to the near-fault pulse-free ground motion and the far-field ground motion, the near-fault pulse-type ground motion leads to larger lateral beam displacement, lateral train acceleration, and running safety indexes in different degrees. And from the data results of this paper, its lateral wheel–rail force and derailment coefficient are the first to break through the limit values. By studying the parameter sensitivity of lateral train acceleration, it was found that under the near-fault pulse-type ground motion, the lateral train acceleration is significantly more sensitive to changes of train speed and PGA. From the perspective of energy, the near-fault pulse-type ground motion results in spectrum intensity (SI) indexes which are larger and improve faster with the increase of PGA than other ground motions. This indicates that it inputs significantly more energy into the high-speed railway train–track–bridge system, although it may not excite larger lateral beam acceleration. [ABSTRACT FROM AUTHOR]

In the 1850s, Portugal initiated an ambitious public works program spearheaded by railroads. In this article, I analyze how policymakers (congressmen, government officials, engineers, and military officers) debated and managed the construction and operation of the Portuguese railroad network using the theoretical framework of critical infrastructure. I show that railroads were deemed critical to modernize the country, draw it closer to the European core nations and further from the periphery, and to attract traffic to the Portuguese harbors. The absence of railroads was considered a vulnerability that could jeopardize Portugal as a nation and it fostered a sense of urgency that motivated policymakers to act, to a large extent, hastily. The construction of the network brought about other vulnerabilities that, for different motives, could threaten the future of the country. In a parallel way, the implementation of the system was marked by vulnerabilities that originated within or surrounding it, which limited its technical potential. Balancing between a literature review and the use of primary sources (parliamentary debates, technical reports, the press, and photography), I argue that the criticality associated with railroads, as a sociotechnical construction, was central to motivate their construction, but different vulnerabilities, inherent or external to the system, limited their influence and the advantages touted by its promoters. [ABSTRACT FROM AUTHOR]

The rail rolling stock undercarriage condition monitoring is proposed by using an automated measuring system, located on the railway track and measuring the specified parameters of the wheels directly, while the train is moving. Regular undercarriage condition monitoring reduces the costs of preventive maintenance of rolling stock without compromising the traffic safety. An algorithm has been developed for the operation of a special software package for visualizing and assessing monitoring data on the condition of the undercarriage of rail rolling stock. The software package consists of separate software modules that can be used independently of each other. It is possible to make short- or long-term predictions of the behavior of any of the monitored parameters using an proposed automated measuring system. [ABSTRACT FROM AUTHOR]

Interpreting transcriptome data is an important yet challenging aspect of bioinformatic analysis. While gene set enrichment analysis is a standard tool for interpreting regulatory changes, we utilize deep learning techniques, specifically autoencoder architectures, to learn latent variables that drive transcriptome signals. We investigate whether simple, variational autoencoder (VAE), and beta-weighted VAE are capable of learning reduced representations of transcriptomes that retain critical biological information. We propose a novel VAE that utilizes priors from biological data to direct the network to learn a representation of the transcriptome that is based on understandable biological concepts. After benchmarking five different autoencoder architectures, we found that each succeeded in reducing the transcriptomes to 50 latent dimensions, which captured enough variation for accurate reconstruction. The simple, fully connected autoencoder, performs best across the benchmarks, but lacks the characteristic of having directly interpretable latent dimensions. The beta-weighted, prior-informed VAE implementation is able to solve the benchmarking tasks, and provide semantically accurate latent features equating to biological pathways. This study opens a new direction for differential pathway analysis in transcriptomics with increased transparency and interpretability. Author summary: The ability to measure the human transcriptome has been a critical tool in studying health and disease. However, transcriptome datasets are too large and complex for direct human interpretation. Deep learning techniques such as autoencoders are capable of distilling high-level features from complex data. However, even if deep learning models find patterns, these patterns are not necessarily represented in a way that humans can easily understand. By bringing in the prior knowledge of biological pathways, we have trained the model to "speak the language" of the biologist, and represent complex transcrtomes, in simpler concepts that are already familiar to biologists. We can then apply the tool to compare for example samples from lung cancer cells to healthy cells, and show which biological processes are perturbed. [ABSTRACT FROM AUTHOR]

To address the inefficiencies, poor real-time capabilities, and challenges in achieving comprehensive dynamic inspections of existing railway external environmental hazard monitoring technologies, this paper investigated a satellite remote sensing intelligent monitoring technology system suitable for railway external environmental monitoring. The system was tested and applied in the management of external environments along the Shanghai-Kunming High-speed Railway segment from K563 to K641. Initially, highresolution satellite images with approximately 500m width on either side of the initial trial segment were obtained. Using remote sensing interpretation systems, hazards such as hard and light floating objects in the railway external environment were extracted, establishing a precise hazard database covering 108 locations. Subsequently, a highly automated management system centered on GIS technology and a mobile intelligent inspection application (APP) were developed for daily management and dynamic inspections of the railway external environment archive. Leveraging this system, three phases of high-resolution satellite imagery were used to monitor changes in the trial segment, identifying a total of 63 hazard changes, with 50 confirmed through field verification, achieving an inclusion rate of approximately 80%. The study results demonstrate that the proposed satellite remote sensing intelligent monitoring technology comprehensively reduces labor costs, alleviates manpower shortages, enhances railway operational safety management efficiency, reduces safety risks, and ensures the safe transportation of railways. [ABSTRACT FROM AUTHOR]

In order to solve the problems of low level of information technology, poor real-time performance, and insufficient operation records in traditional online operation methods of railways, this paper conducted research on railway online operation safety control methods based on mobile spatiotemporal positioning trajectories. Continuous real-time positioning data of online operators and large equipment were collected through mobile positioning terminals. By utilizing the operational railway line equipment and facility ledger, an accurate range of electronic fence control for online operations was established. Based on spatial analysis technology, real-time continuous monitoring of personnel and large equipment operation status was achieved, accurately obtaining line intervals, mileage, and operation deviation distance. Overcrossing and overtime operation behaviors were detected in a timely manner, achieving the goal of accurate and real-time warning. According to the requirements of management, the functional design and software development of the regulatory GIS platform carried out which dynamically visualizes the current status of operations along the line in real time. The conclusion drawn is that the safety control methods and technical processes proposed in this article are deeply integrated with railway operation safety protection business, which can achieve automatic identification of illegal operation behavior and supervision assistance for multi-line operations, effectively avoiding safety risks in railway online operations. [ABSTRACT FROM AUTHOR]

Cellular vehicle-to-everything (C-V2X) enables short-distance communication between vehicles and other users to improve road safety through data sharing. Conventional research on C-V2X typically assumes that vehicles travel at the same speed with a fixed inter-vehicle distance ( D i s i n t e r ). However, this assumption does not reflect the real driving environment or promote road traffic efficiency. Conversely, assigning different speeds to vehicles without a structured approach inevitably increases the collision risk. Therefore, determining appropriate speeds for each vehicle in the C-V2X framework is crucial. To this end, considering the road environment and mobility, this study introduces an adaptive speed mechanism based on the congestion level of a zone and D i s i n t e r . First, the given scenario is divided into several zones. Subsequently, based on the congestion level of a zone and the D i s i n t e r level, an adaptive speed is defined for each vehicle. This approach ensured that vehicles adopt lower speeds in congested situations to reduce the collision probability and higher speeds in sparse traffic cases to improve traffic efficiency. The performance of the proposed adaptive speed scheme is compared with that of conventional fixed-speed settings. The results show that the adaptive speed control scheme outperforms conventional fixed-speed schemes in terms of the packet reception ratio (PRR) and collision ratio (CR). Specifically, the proposed mechanism can reduce the CR to 0 and ensure that the PRR is higher than 0.98 in low-density scenarios. [ABSTRACT FROM AUTHOR]

This paper is based on the proximity engineering project of the Baishiyi tunnel group passing under the Chongqing West Station track group. Considering the train load and the spatial relationship of the tunnel and track groups, the settlement patterns, horizontal displacement, and differential settlement of the tunnel–strata–tracks system during the excavation process are studied through theoretical calculations and numerical simulation methods. The results indicate that the tunnel vault, strata, and track settlement deformation patterns are similar. Throughout the tunnel construction process, the tracks underwent uplift, settlement, and eventually stabilized. The settlement trough formed by the excavation of the three tunnels below the track group has an impact range of 25–145 m. Between 35 and 75 m, the differential settlement of the double track gradually increases with excavation. As the tunnel face reaches 75 m, the track differential settlement gradually converges and tends to stabilize. To minimize the impact of underpass tunnel construction on track groups, it is recommended to use a combination of full-section hole grouting and surface reinforcement grouting for ground reinforcement. Additionally, optimizing the construction parameters, including the step length and primary support closure time, and strengthening the locking anchor can further reduce the impact. [ABSTRACT FROM AUTHOR]

The safety of the high-speed train traveling through the stationary thunderstorm downburst wind was studied. First, a thunderstorm wind test device was used to simulate the stationary thunderstorm downburst wind. Based on the rigid model pressure measurement tests, the aerodynamic forces of the train traveling along different paths through the stationary thunderstorm downburst wind were measured. The influence of the radial distance of the crossing path on the aerodynamic force coefficients of the train was investigated. On this basis, an unsteady aerodynamic model of the high-speed train crossing through the stationary thunderstorm downburst wind was established, and dynamic response analysis was carried out using the SIMPACK multibody dynamics simulation software to explore further the safety of the high-speed train crossing through the stationary thunderstorm downburst wind. The research showed that the stationary thunderstorm downburst wind field has significant spatial variation characteristics compared with the atmospheric boundary layer wind field. When the train passes through the thunderstorm downburst wind, the radial wind speed and wind yaw angle experienced by the train constantly change, and the change curve shows a symmetrical distribution. The aerodynamic force of the train will undergo sudden loading and unloading processes, and the lateral force coefficient of the train on different paths shows a “pulse-type” variation. Moreover, the lateral force coefficient increases with the increase of wind yaw angle. Under the influence of the thunderstorm downburst wind, the variation trend of the aerodynamic force coefficients of the train is consistent with that under crosswind. However, there are significant differences in the numerical values. Therefore, it is impossible to simply use the formula for calculating the aerodynamic force coefficients of the train under crosswinds to predict the aerodynamic force coefficients of the train under the thunderstorm downburst wind. While passing through the thunderstorm downburst wind, the overturning coefficient index plays a decisive role in the safety of train operations. Train rollover is the main form of train safety accidents, while derailment accidents are not easy. The numerical results obtained in this study are significant for evaluating the operational safety while moving trains traversing the stationary thunderstorm downburst wind. [ABSTRACT FROM AUTHOR]

Summary: Real‐time estimation of crowd counting in underground metro systems, constrained by limited space, is crucial for managing heightened pedestrian volumes and responding promptly to emergencies. To address this challenge, we propose a passenger state transition‐based model, called STRmt, designed for the seamless and continuous monitoring of real‐time crowd movement within service areas of stations and trains, leveraging auto fare collection systems (AFC) as a comprehensive sensor network. Our innovation lies in modeling the dynamic movement of passengers within a metro system over time as a state transition process aligned with the train schedule. To achieve this, we introduce a spatio‐temporal deep learning framework, denoted as STnet, designed to dynamically predict these state transitions. The performance of our method is rigorously assessed through extensive experiments conducted spanning 2 years in Shenzhen, China, utilizing AFC data, train schedule data, and weather data. The results demonstrate that the proposed method surpasses baseline methods, achieving an estimation precision of 0.92. [ABSTRACT FROM AUTHOR]

Currently, the static configuration data checking of the safety critical system is realized by the existing constraint rules which generated from the professional norms. Most of these constraint rules are for single-category data and relatively simple, and only some basic data errors can be identified with which. Therefore, it is necessary to excavate more complete and comprehensive data constraint rules to further improve the performance of the data verification. Data association rules describe the correlation and dependence between data items and reflect the rules and patterns of some attributes appearing simultaneously. Based on which, the in-depth research on the methods of static data verification in the train control system is carried out. Firstly, clustering and dimension reduction methods are used to divide data items into multiple sub-range intervals, which can solve the problem that association rules of the static data are difficult to extract due to floating-point data measurement. Then, a DSRJ algorithm is proposed to judge the existence of association relations among sub-range data items, and a neural network model will be constructed in which a large number of data samples are trained to obtain the data relation function that is transformed further into an association rule. The new subset of data samples can be conducted safety verification according to this rule. [ABSTRACT FROM AUTHOR]

In light of the rapid development of electrified railways, the safety and stability of train operations, as well as the catenary's interaction with current quality, have garnered widespread attention. Electrified train operation with additional track irregularities serves as a principal excitation source within the vehicle–bridge–catenary system, significantly influencing the vibration characteristics of the system. Addressing the aforementioned issues, we first established the vehicle–track dynamics model and the bridge–catenary finite element model based on the principles of coupled dynamics of the vehicle–track system. These models are interconnected using dynamic forces between the wheel and rail. Subsequently, within the vehicle–track coupled system, track random irregularities are introduced as input excitations for the coupled model, and the dynamic response of the system is simulated and solved. Then, the obtained wheel–rail forces are applied to the bridge–catenary coupled system finite element model in the form of time-varying moving load forces. Finally, the dynamic response characteristics of the catenary portal structure under different conditions are determined. Meanwhile, a study on the vibration characteristics of the truss-type pillar portal structure was conducted, and the results were compared with those of existing models. The results indicate that the vertical and lateral forces between the vehicle and track are positively correlated with the speed and irregularity amplitude. Response values such as the derailment coefficient and wheel load reduction rate are within the specified range of relevant standards. The low-order natural resonant frequency of the truss-type pillar structure has, on average, increased by 0.86 compared to the existing pillar structure, which signifies improved stability. Furthermore, under various conditions, the average reductions in maximum displacement and stress response of this pillar structure are 13.2% and 14.19%, respectively, in comparison to the existing pillar structure, rendering it more suitable for practical engineering applications. [ABSTRACT FROM AUTHOR]

The precise identification of railway subgrade defects remains a significant challenge for the railway industry globally. Due to the limitations of individual monitoring techniques, comprehensive information on subgrade defects cannot be obtained. In fact, the presence of subgrade defects can significantly increase the risk of traffic accidents during high-speed train operations, which may affect the safety of train operations and economic development. The monitoring of subgrade health status is used as a pre-disaster planning method that is urgently required to avoid accidents and guide the maintenance strategy. Therefore, a novel "integrated" holistic monitoring approach for subgrade structures is presented based on satellite remote sensing, a comprehensive inspection vehicle, and a ground-based testing technique. Additionally, the monitoring content is more clearly defined during the service life of the subgrade. The method is used to investigate the location, development trend, and the cause of subgrade defects on the Shanghai–Nanjing high-speed railway. Some new viewpoints are put forward: First, the monitoring content for assessing the health status of the subgrade should encompass the foundation settlement, the track geometry status, and the monitoring of deformation and defects within the subgrade. Second, the mileage points K235 and K299 of the subgrade, as well as K236 and K237 of the bridge–subgrade transition sections, are estimated to be locations with potential defects based on the differential InSAR and track quality index. Third, the result of settlement monitoring and ground-penetrating radar analysis illustrates that sections K235 +540 to +680 and K299 +680 to +750 are diagnosed as defect positions triggered by the rapid drop of water level and engineering activity, respectively. Fourth, the "integrated" holistic monitoring technique for subgrade service status might be expected to be a promising method that can be useful in developing maintenance plans and implementing fault recovery for railway infrastructure. [ABSTRACT FROM AUTHOR]

The growth of the global construction market has attracted international companies to participate in overseas projects. Overseas projects are extremely dynamic with numerous uncertainties, raising the need to collect information about construction in host countries. Due to the vast amounts of text data in the construction industry, an automated method, specifically information retrieval, is required to find the necessary information. Previous studies have suggested automated methods to review various construction documents. However, these studies required substantial manual effort and mainly focused on only one language, resulting in loss of vital information because it is buried in documents written in the host country's language. To address these limitations, this study proposes a cross-lingual information retrieval (CLIR) framework using pretrained Bidirectional Encoder Representations from Transformers (BERT) models to retrieve information from multilingual construction documents. The proposed framework employs language models (i.e., monolingual, multilingual, and cross-lingual) and trains these models on a construction data set to enhance their ability in construction-specific text. The framework achieved reliable performance of retrieval, even with minimal additional training using domain-specific data. The results indicate that training on the domain data set raises the level of retrieval, increasing the mean reciprocal rank of a specific task by up to 0.2128. With the employment of a monolingual model with machine translation, CLIR in a specific domain could be performed effectively without the need for a labeled data set. The suggested CLIR framework offers a practical alternative for dealing with construction documents in overseas projects, reducing time and cost while improving risk identification and mitigation. [ABSTRACT FROM AUTHOR]

This paper presents an Event-B model of a speed control system, a part of the case study provided in the ABZ2020 conference. The case study describes how the system regulates the current speed of a car according to a set of criteria like the driver's desired speed, the position of a possible preceding vehicle, but also a given speed limit that the driver must not exceed. For that purpose, this controller reads different information from the available sensors (key state, desired speed) and takes adequate actions by acting on the actuators of the car's speed according to the information read. To formally model this system, we adopt a stepwise refinement approach with the Event-B method. We consider most of the features of the case study. All proof obligations of the invariant properties have been discharged using the Rodin provers. Our model has been validated using ProB by applying the different provided scenarios. This validation has permitted us to point out and correct some mistakes, ambiguities and oversights contained in the first versions of the case study. [ABSTRACT FROM AUTHOR]

This paper introduces a nonlinear PI controller for improved speed regulation in permanent magnet direct current (PMDC) motor drive systems. The nonlinearity comes from the exponential (Exp) block placed in front of the classical PI controller, which uses a tunable exponential function to map the speed error nonlinearly. Such a configuration has not been studied till now, thus meriting further investigation. We consider an exponential PI (EXP-PI) controller and to attain the best performance from this controller, its parameters are optimized offline using salp swarm algorithm (SSA), which borrows its inspiration from the way of forage and navigation of salps living in deep oceans. To indicate the credibility of SSA tuned EXP-PI controller convincingly, numerous experiments on speed regulation in PMDC motor have been implemented using DSP of TMS320F28335. The results obtained are also compared to similar results in the literature. It is shown that the proposed approach performs well in practice by ensuring tight tracking of the speed reference and superb torque disturbance rejection for the closed loop control. Furthermore, superior performance is achieved by the proposed nonlinear PI controller with respect to a fixed-gain PI controller. [ABSTRACT FROM AUTHOR]

In this paper, an in-depth study is conducted regarding the train automatic wake-up function of fully automatic driverless trains using FMEA method. The study analyzes the failure modes that may produce dangerous consequences, identifies their corresponding failure causes, and designs appropriate safety measures based on these failure modes and causes. This approach ensures that the fully automatic unmanned driving system reduces the original risks of the system, as well as the risks caused by new functions, to an acceptable degree, and achieves safe and reliable operation. [ABSTRACT FROM AUTHOR]

There is a situation on railway lines where there is no network coverage in some sections. When the train restarts due to a fault in said sections and there is no network available, it is unable to communicate with trackside equipment to obtain line data, which aff ects operational efficiency. This paper mainly studies the method of train fault recovery in sections without network coverage, using Regional Short Message Communication (RSMC) technology of BeiDou Navigation Satellite System to solve the problem of train restart due to faults and recovery in these sections. [ABSTRACT FROM AUTHOR]

As the core system of rail transit to ensure traffic safety and improve transport efficiency, the train control system plays an important role in the design and implementation of multi-network integration. This paper focuses on analyzing the characteristics and differences of the mainstream train control systems from the aspects of construction objectives, system selection, technical routes and other aspects in combination with the construction needs of the multi-network integration of the Guangdong-Hong Kong-Macao Greater Bay Area urban express line, and expounds the integration scheme and difficulties, so as to provide certain ideas and references for the selection and design of the follow-up train control systems. [ABSTRACT FROM AUTHOR]

At present, the ZPW-2000 integrated track circuits with the same standard as those used in the sections are widely used in China's high-speed railway stations, and the platform tracks in some stations adopt one-section track circuits. When the train is received at the station, the sending end of the track circuit of the platform track will face the train and send codes to provide trackside information for the train. If the train comes to a complete stop at the station and changes its direction by activating the cab on the intended end, when the starting signal is not at clear, the train may mistakenly receive the information sent by the track circuit of the adjacent line, and the train will depart from the station incorrectly, resulting in safety problems. In this paper, the method of adding an external relay circuit is proposed, and the timing of the train entering and leaving the platform track is determined by the circuit logic, and the concrete measure for applying train tail complementary code is provided, so that the double-end code sending can be realized after the train enters the track to solve the aforementioned problems. This solution only adds a small number of relays indoors, has no outdoor workload, does not need to change the train control software, and can be adapted to suit individual platform tracks of the station. The adaptation process is easy to implement and has minimal impact on transportation. This scheme can provide reference for the solution of safety protection enhancement for platform tracks equipped with integrated track circuits in high-speed railway stations. [ABSTRACT FROM AUTHOR]

The railway signal control system usually adopts a special double 2--vote--2 safety computer platform for information transmission, logic operation, object control and other functions. With the increasing varieties and numbers of external communication objects, the existing active-standby switch methods for security computer platforms may have problems such as not switching when a failure occurs or unnecessary switching when the same external failure occurs. This paper presents an active-standby switch method of double 2--vote--2 safety computer platform based on communication state judgment of communication objects. This method first defines the quantitative indicators such as switching value and priority, and designs the calculation formula of switching value. Then, from the aspects of establishing communication, judging communication interruption, and calculating the number of normal communication, the method formulates different judgment criteria for the active and the standby system. Finally, the method compares the switching values, and decides whether to switch between the active and the standby system. The method is verified in practical applications in typical scenarios. The results show that the active-standby switch method based on the communication object state judgment improves the system availability and reduces the unnecessary switching phenomena that affect the system security compared with the active-standby switch method based on the board level and the number of calculations; the designed active-standby switch method can improve the reliability and security of the system by classifying and analyzing different communication modes, different redundant structures and different control devices. [ABSTRACT FROM AUTHOR]

The transportation sector is regarded as the main culprit in greenhouse gas emission in the urban network, particularly idling vehicles waiting at signalized intersections. Although autonomous vehicles can be a promising technology to tackle vehicle idling, their environmental benefits receive little attention compared with their safety and mobility issues. This study investigated the environmental benefits of autonomous vehicles equipped with an automatic idling control function based on the queue discharge time and traffic signal information transmitted from the traffic signal controller via V2I communication using microscopic mobility and emission simulation models, VISSIM and MOVES, in Haeundae-gu in Busan, Korea. This study found that the function contributes to a significant reduction in CO2 emissions by 23.6% for all-inclusive emission and 94.3% for idling emission, respectively. Moreover, total reduced idling time accounts for 47.6% of the total travel time and 94.3% of the total idling time, respectively. Consequently, the autonomous vehicles equipped with automatic vehicle idling control function under C-ITS can play an important role in reducing greenhouse gas emissions and fuel consumption as well in the urban network. [ABSTRACT FROM AUTHOR]

The health management of railway signal equipment in the high-speed railway is a key link between intelligent operation and maintenance. Accurately predicting the health state of compensation capacitors is of great significance to ensure the reliable work of track circuits. This paper proposes an improved deep neural network algorithm focusing on the problem of long-term accurate health forecasts for compensation capacitors. First, establishing a transmission state model for degradation mechanism mining, the difference function that can quantitatively evaluate features is defined by piecewise processing cab signaling receiving voltage. Introducing the degradation model, predictive driving under both model and data is implemented. Then, the convolutional neural networks and bidirectional long–short-term memory are combined and improved to construct a novel artificial intelligence combination strategy, while parameters are optimized based on the sparrow search algorithm. Finally, facing the conditional repair of compensation capacitors, we set a reasonable threshold for the occurrence of hidden dangers to complete fault warning. This novel and practical approach effectively explores the procedure of prognosis and health management, while the refined maintenance will better utilize current monitoring information, helping the intelligence and accuracy of safety control decision-making. [ABSTRACT FROM AUTHOR]

Design of new drugs is a challenging process: a candidate molecule should satisfy multiple conditions to act properly and make the least side‐effect—perfect candidates selectively attach to and influence only targets, leaving off‐targets intact. The amount of experimental data about various properties of molecules constantly grows, promoting data‐driven approaches. However, the applicability of typical predictive machine learning techniques can be substantially limited by a lack of experimental data about a particular target. For example, there are many known Thrombin inhibitors (acting as anticoagulants), but a very limited number of known Protein C inhibitors (coagulants). In this study, we present our approach to suggest new inhibitor candidates by building an effective representation of chemical space. For this aim, we developed a deep learning model—autoencoder, trained on a large set of molecules in the SMILES format to map the chemical space. Further, we applied different sampling strategies to generate novel coagulant candidates. Symmetrically, we tested our approach on anticoagulant candidates, where we were able to predict their inhibition towards Thrombin. We also compare our approach with MegaMolBART—another deep learning generative model, but exploiting similar principles of navigation in a chemical space. [ABSTRACT FROM AUTHOR]

As an emerging technology, virtual coupling improves the efficiency and flexibility of metro services by forming multiple trains (units) as a virtually coupled train set (VCTS) without mechanical couplers. However, to realize the desired VCTS operation in practical metro services, a significant gap to be filled is that the implicit and nonlinear safety constraints are hard to be addressed in real‐time control. Thus, this paper proposes a hierarchical control approach with a two‐layer framework. In the upper layer, a trajectory planning method is designed, which addresses the safety constraints and prescribes a reference trajectory for each unit. In the lower layer, a model predictive control approach is constructed, by which each unit accurately tracks its reference trajectory in real time. Afterward, the proposed hierarchical control approach is employed to realize VCTS inter‐station operation in field tests. The average tracking error of speed is around 8 cm/s and the stopping error is less than 12 cm. The following distance between two units in VCTS is reduced to less than 105 m at the speed of 80 km/h, which is a breakthrough in the development of VCTS. [ABSTRACT FROM AUTHOR]

A reliable, quick‐to‐assess, and automatic scoring system for cleanliness assessment in video capsule endoscopy (VCE) is presently not available. The present study proposes an approach to automatically assess the cleanliness in VCE frames as per the latest scoring system, that is, Korea‐Canada (KODA). First, a new multi‐label frame dataset containing medical scores of 28 VCE videos was generated through the proposed mobile‐based application called Artificial Intelligence‐KODA (AI‐KODA) score. The scores were saved automatically in real‐time through the application. The generated dataset was transformed into three datasets based on the scores, and each of the dataset was then randomly split into train:validate:test ratio of 60:20:20. Second, a comprehensive evaluation, interpretation, and benchmarking of the three classification tasks were performed with the help of eight transfer learning algorithms on NVIDIA RTX A5000 workstation. Thorough analysis indicates that AI‐KODA utilized with AI is reliable, quick‐to‐access, and free from observer bias. It promotes automatic scoring system for cleanliness assessment in VCE. The meta‐data is available here (link). [ABSTRACT FROM AUTHOR]

The Fourth Industrial Revolution (4IR) has brought about significant advancements in technology that have impacted various industries, including the railway sector. This paper explores the implications of the 4IR for the railway industry and identifies the essential soft skills that rail organisations need in order to thrive in this era. A systematic literature review (SLR) was conducted to establish a theoretical foundation, followed by a survey administered to professionals in the rail sector to prioritise the identified soft skills. Comparative methods were employed to analyse the survey data and to determine the relative importance of the 4IR soft skills. The findings revealed a range of key soft skills, such as adaptability, critical thinking, and change management, in which it is crucial that railway professionals excel in the 4IR era. The paper discusses the significance of these soft skills and their potential impact on rail organisations, and provides recommendations for fostering their development. This research contributes to a more nuanced understanding of the soft skills that are necessary for the railway sector’s successful adaptation to the 4IR. [ABSTRACT FROM AUTHOR]

Learning the full extent of pixel‐level instance response in a weakly supervised manner remains unsatisfactory. Peak response maps (PRMs) localizes the discriminative object regions but cannot provide complete instance information, suffering from incomplete segmentation and unreliable mask prediction by noisy proposal retrieval. This work tackles this challenging problem by mining diverse class peak responses that include more discriminative and complete object regions and retrieving more reliable proposals from noisy segment proposal galleries. First, the existing method is enhanced with two more classification branches, thus contributing to more diverse and abundant instance regions from peak response maps. The mined class peak responses from two of the branches are then merged to generate more complete peak response maps by a clustering approach in their deep feature space. Then, instance segmentation masks are retrieved from a noisy object segment proposal gallery with class confidence, which is calculated by a normal classifier to obtain cleaner mask prediction. Finally, the pseudo‐supervision can be used to train an instance segmentation network in a fully supervised manner. Experiments on the PASCAL VOC 2012 dataset and COCO dataset show that the approach works effectively and outperforms other counterparts by a margin of more than 6 %, 4%, and 3% with the mean average precision (mAP) at IoU threshold of 0.25, 0.5 and 0.75, respectively. [ABSTRACT FROM AUTHOR]

A typhoon disaster is a common meteorological disaster that seriously impacts natural ecology, social economy, and even human sustainable development. It is crucial to access the typhoon disaster information, and the corresponding disaster prevention and reduction strategies. However, traditional question and answering (Q&A) methods exhibit shortcomings like low information retrieval efficiency and poor interactivity. This makes it difficult to satisfy users' demands for obtaining accurate information. Consequently, this work proposes a typhoon disaster knowledge Q&A approach based on LLM (T5). This method integrates two technical paradigms of domain fine-tuning and retrieval-augmented generation (RAG) to optimize user interaction experience and improve the precision of disaster information retrieval. The process specifically includes the following steps. First, this study selects information about typhoon disasters from open-source databases, such as Baidu Encyclopedia and Wikipedia. Utilizing techniques such as slicing and masked language modeling, we generate a training set and 2204 Q&A pairs specifically focused on typhoon disaster knowledge. Second, we continuously pretrain the T5 model using the training set. This process involves encoding typhoon knowledge as parameters in the neural network's weights and fine-tuning the pretrained model with Q&A pairs to adapt the T5 model for downstream Q&A tasks. Third, when responding to user queries, we retrieve passages from external knowledge bases semantically similar to the queries to enhance the prompts. This action further improves the response quality of the fine-tuned model. Finally, we evaluate the constructed typhoon agent (Typhoon-T5) using different similarity-matching approaches. Furthermore, the method proposed in this work lays the foundation for the cross-integration of large language models with disaster information. It is expected to promote the further development of GeoAI. [ABSTRACT FROM AUTHOR]

Train traffic is a powerful source of seismic waves, with many applications for passive seismic imaging. Seismic signals were recorded a few metres away from the railway track. These records display harmonious waveforms below 15 Hz for trains driving at speeds of around 100 km hr−1. The sensors record an apparent wavelet emitted by the interaction of the axle on a few of the closest sleepers. From this, we build a simple modelling tool using shifted wavelets to simulate a train signal. The relationship involves the varying train speed, the distances between each axle, and the wavelet emitted by each axle. We propose a nonlinear deconvolution method to invert this relationship. We use a local minimization algorithm with gradients derived by the adjoint state method, and use a frequency continuation technique. A linearized picking-based inversion initializes the nonlinear inversion. On real data, we apply this automatic workflow to 300 train passages, with an excellent match between the best simulation and the data. We identify the trains decelerating as they enter a train station. We also identify the train type based on inverted wheel spacing with centimetric accuracy. The inverted wavelets are consistent with the assumption that trains emit seismic waves by bending the rail above sleepers, although the theory does not explain why the inverted wavelet is not zero phase. This automated kinematic inversion algorithm may allow for contactless railway monitoring, and be used for source characterization for subsurface monitoring below railway tracks. [ABSTRACT FROM AUTHOR]

By using a data set of 536 observations from 29 OECD countries for the 19 years from 1995 to 2013, this study investigates the factors affecting rail investment with panel data analysis. Among the pooled OLS, the fixed effects, and the random effects models, the fixed effects model is selected by statistical test. It seeks to clarify how investment is related to factors such as demand, managerial, and regulatory factors. The main results are as follows: (i) factors positively affecting investment are output as a measure of train-km, capital stock as network length, GDP per capita, and the competitiveness of passenger railways; (ii) the competitiveness of freight railways has a negative relationship with investment; (iii) the highspeed train ratio itself is not significant; (iv) a larger government debt ratio tends to weakly decrease investment; (v) accident rate has a weak positive relationship with investment, but as the government debt ratio increases, the accident rate's effect on investment decreases; (vi) among regulatory factors, investment weakly increases, as values of three indexes (the ratio of public ownership, vertical integration, and overall regulation) become larger; and (vii) deregulation of entry and greater competition in the market do not relate to investment. In conclusion, strong economic conditions such as GDP and the competitive situation in passenger rail are important for investment, but regulatory conditions are not. [ABSTRACT FROM AUTHOR]

Automatic train protection (ATP) system is essential for ensuring the safe operation of high‐speed trains. However, the existing extensive and fixed maintenance mode of the ATP system results in a waste of resources. To achieve a state of operation and maintenance that ensures both protection capability and economic efficiency, a lean method in a dynamic maintenance mode for the full life cycle of the ATP system is proposed. Firstly, reliability tests are carried out based on historical failure data. The parameter values of the possible life distribution are estimated by maximum likelihood method, and the optimal life distribution of different devices is obtained through the Kolmogorov–Smirnov hypothesis test. Secondly, a dynamic failure rate function is introduced to describe the impact of maintenance on device performance. A refined maintenance model is then established within the life cycle, and the dynamically changing preventive maintenance intervals and frequencies are obtained using a genetic algorithm. Finally, to mitigate the impact of the intermittent operation of ATP system on maintenance, the multidimensional relationships among the maintenance strategy, service time and operation mileage are revealed. The effectiveness of the proposed method is verified through an example test on a type of driver machine interface device. [ABSTRACT FROM AUTHOR]

A major railway project is a complex, giant system with multi-party participation, one characterized by complex geological conditions, long construction periods and large scale, which leads to an increased likelihood of safety risk events during construction. In order to solve the problem of scientific selection and formulation of safety risk prevention and control strategies for major railway projects, an auxiliary selection method of safety risk prevention and control strategies for major railway projects based on weighted Euclidean distance (WED) is proposed. The relevant ontology is used to conceptualize and formalize the knowledge of safety risks of major railroad projects, and combine the characteristics of major railroad projects; it refers to the prevention and control measures of historical safety risk events associated with major railroad projects, and then constructs the knowledge structure and case base around safety risks of major railroad projects and the circumstances of the case. In determining the comprehensive weights, the G1 method is used to determine the subjective weights, the anti-entropy weight method is used to determine the objective weights and game theory combines the subjective and objective weights. In comparing the array of safety risk prevention and control cases associated with major railway projects, the weighted Euclidean distance is used to calculate the similarity between these cases and the target case, which in turn assists project managers in determining the safety risk prevention and control strategies appropriate for major railway projects. This study takes Landslide No. 1 in the Tunnel A inlet planning area as an example. It utilizes the WED method to assist in selecting safety risk prevention and control strategies for major railway projects, which verifies the method's feasibility. The proposed method enriches the method of the assisted selection of safety risk prevention and control strategies for major railway projects, makes strategy formulation more scientific, has specific reference significance for the formulation of safety risk prevention and control strategies for major railway projects, and promotes the improvement of safety risk prevention and risk control for participating units. [ABSTRACT FROM AUTHOR]

The article presents the concepts of control systems that use methods of automatic analysis and interpretation of signals. These issues are presented through the description of three control strategies: simple closed-loop control, control using a process of classification, and using a signal understanding technique. The systems based on the discussed concepts are illustrated with examples of controlling a milling machine and an autonomous vehicle. In addition, in more detail, a task of learning human motor activities is described. This task, due to the nature of the controlled object, which is a human being, is extremely difficult. The article shows that the advanced control process, in which the control algorithm is selected and its parameters are adapted to the current situation, may be implemented through the use of the classification process and machine learning methods in general. Changing the algorithm is also possible using signal understanding techniques. These techniques, utilizing models of the objects, allow to predict the long-term effects of the control process. The ability to build control systems that operate in the above manner is of huge practical importance. The aim of this article is to describe the methods of automatic signal interpretation used in control processes and identify the main problems related to their use. The key problems refer to the acquisition of expert knowledge by the system. In order for this knowledge to be effectively transferred, the methods used in the system should have a high level of explainability. Showing the essential nature of this feature is the main outcome of this work. [ABSTRACT FROM AUTHOR]

As the demand for enhanced railway services continues to rise, the imperative to upgrade and modernize the rail infrastructures becomes unavoidable. Automatic Train Operation (ATO) holds the potential to revolutionize railway systems by improving efficiency, safety, sustainability and operational performance. However, limited research has focused on understanding the implications of ATO implementation on project management processes. This paper thus presents an ongoing scoping review that aims to investigate the impact of ATO on project management in the railway industry. The review's methodology includes a comprehensive search strategy criteria for study selection, and data extraction procedures. The anticipated findings will contribute to bridging the gap regarding the intersection of ATO and project management in the railway industry. By understanding the impacts and challenges associated with ATO implementation, stakeholders can better manage projects involving ATO technology and promote its successful integration into railway systems, fostering advancements in the field. [ABSTRACT FROM AUTHOR]

Featured Application: Object detection method based on three-dimensional GPR data of railroad tracks using AI, trained solely on simulation data. Ground-penetrating radar (GPR) is a non-invasive technology that uses electromagnetic pulses for subsurface exploration. In the railroad sector, it is crucial to assessing soil layers and infrastructure, offering insights into soil stratification and geological features and aiding in identifying subsurface hazards. However, the automation of radargram analysis is impeded by the lack of ground truth—accurate real-world data used to validate machine learning models—thus affecting the deployment of advanced algorithms. This study focuses on generating high-quality simulated data to address the shortage of real-world data in the context of object detection along railroad tracks and presents a fully automated pipeline that includes data generation, algorithm training, and validation using real-world data. By doing so, it paves the way for significantly easing the future task of object detection algorithms in the railway sector. A simulation environment, including the digital twin of a GPR antenna, was developed for artificial data generation. The process involves pre- and post-processing techniques to transform the three-dimensional data from the multichannel GPR system into two-dimensional datasets. This ensures minimal information loss and suitability for established two-dimensional object detection algorithms like the well-known YOLO (You Only Look Once) framework. Validation involved real-world measurements on a track with predefined buried objects. The entire pipeline, encompassing data generation, processing, training, and application, was automated for efficient algorithm testing and implementation. Artificial data show promise for better performance with increased training. Future AI and sensor advancements will enhance subsurface exploration, contributing to safer and more reliable railroad operations. [ABSTRACT FROM AUTHOR]

Automated railway security systems prevent train collisions with trackside obstructions that cause accidents in high-speed railways. Rail safety is being improved and accident rates reduced through continuous research. A rapid advancement in deep learning has promoted new possibilities for research in this field. In this work, a novel deep learning-based FOD-YOLO net is proposed for detecting the fasteners faults and objects in the railway tracks. There are two basic components in the deep learning-based YOLOv8: the backbone and the head. YOLOv8 utilizes an improved version of the CSPDarknet53 network for detecting objects on the railway track. The head of YOLOv8 consists of EfficientNet with various convolutional layers with squeeze and excitation blocks for detecting any defect in the track fasteners. These layers are liable for detecting the objectness scores, bounding boxes and class probabilities structured with fully connected layers for the objects and faults in tracks. Based on the results from the Yolo network, the alert message is sent to the loco pilot to avoid accidents using fuzzy logic. The experimental fallouts of proposed FOD-YOLO net achieve higher accuracy and yields better evaluation results with 98.14% accuracy, 98.84% precision and 95.94% recall. From the experimental results, the FOD-YOLO net improves the overall accuracy range by 5.44%, 4.72%, 0.73%, and 13.18% better than Fast RCNN, YOLOv5s-VF, YOLO-GD, and 2D-SSA + Deep network respectively. [ABSTRACT FROM AUTHOR]

Automated railway security systems prevent train collisions with trackside obstructions that cause accidents in high-speed railways. Rail safety is being improved and accident rates reduced through continuous research. A rapid advancement in deep learning has promoted new possibilities for research in this field. In this work, a novel deep learning-based FOD-YOLO net is proposed for detecting the fasteners faults and objects in the railway tracks. There are two basic components in the deep learning-based YOLOv8: the backbone and the head. YOLOv8 utilizes an improved version of the CSPDarknet53 network for detecting objects on the railway track. The head of YOLOv8 consists of EfficientNet with various convolutional layers with squeeze and excitation blocks for detecting any defect in the track fasteners. These layers are liable for detecting the objectness scores, bounding boxes and class probabilities structured with fully connected layers for the objects and faults in tracks. Based on the results from the Yolo network, the alert message is sent to the loco pilot to avoid accidents using fuzzy logic. The experimental fallouts of proposed FOD-YOLO net achieve higher accuracy and yields better evaluation results with 98.14% accuracy, 98.84% precision and 95.94% recall. From the experimental results, the FOD-YOLO net improves the overall accuracy range by 5.44%, 4.72%, 0.73%, and 13.18% better than Fast RCNN, YOLOv5s-VF, YOLO-GD, and 2D-SSA + Deep network respectively. [ABSTRACT FROM AUTHOR]