To advance the energy conservation of subway stations, this study focused on the optimization of control strategies for ventilation and airconditioning (VAC) systems. In this study, Beijing, Shanghai, and Guangzhou were selected as representative cities for the cold, hotsummer coldwinter, and hotsummer warmwinter regions, respectively. In addition, a model of a subway station with a platform screen door system was established using a transient system simulation program (TRNSYS). The performances of automation and optimized timetable control were compared in terms of station environment, energy performance, and renovation costs. The results show that both automation and optimized timetable control can effectively improve the station air temperature during the cooling season, thereby achieving a similar station environment. Automatic control can reduce the annual energy consumption of the VAC system by 41%49% compared with the current conventional timetable control, whereas optimized timetable control can decrease the annual energy use by 38%48%. For renovation costs, optimized timetable control merely requires improvement in management without supplementing devices, whereas automatic control requires higher equipment and maintenance costs. Therefore, we recommend the adoption of optimized timetable control for subway stations to achieve energy savings.

To improve the efficiency and accuracy of urban railtransit security inspection systems, this paper designs a novel patternrecognition mode to effectively combine artificial intelligence (AI) image recognition technology with manual centralized pattern recognition. First, based on the current liquid inspection, a liquid detection algorithm is introduced to avoid the open inspection of safe liquids. Second, security products are classified according to their risk levels. Finally, AI confidence judgment, manual sampling, or necessary inspection charts are combined to determine the pattern recognition mode, which can flexibly adjust the depth of the AI intervention according to the accuracy of the AI image recognition and the requirements for pattern recognition at different stages. With the continuous improvement in the accuracy of AI image recognition, it gradually changes from an AIassisted manualbased pattern recognition mode to an AIbased manualassisted pattern recognition mode and finally achieves a fully intelligent pattern recognition mode. A case analysis reveals that the judgment graph model can further achieve rapid security inspection, cost reduction, and efficiency increase without reducing the safety inspection level of urban rail transit stations.

In order to address the issue of longterm dependence caused by the extended time span of wheel wear data and improve the prediction accuracy, an improved BiLSTM metro wheel wear prediction model is proposed by optimizing Bidirectional long shortterm memory network (BiLSTM) with Sparrow search algorithm (SSA). Firstly, the hyperparameters of the BiLSTM algorithm, such as the number of neurons, iteration count, input batch size, and learning rate, are optimized using the SSA. This optimization process is conducted within a specified range to obtain the optimal values of these hyperparameters. This optimization process aims to obtain the optimal parameter values. Subsequently, the SSABiLSTM network model is constructed using these optimal parameter values to predict and analyze wheel wear. Tread wear and flange wear are taken as the research objects, and the measured historical wear data of wheel No.1 of the metro’s carriage # 1 are used as inputs to metro and validate the model, and compare the prediction results with those of MLP, LSTM, BiLSTM and SSA-LSTM models. The results show that the improved bidirectional long short-term memory network model has higher wear prediction accuracy, and the mean absolute percentage error (MAPE) of tread wear is reduced by 13.28%, 10.32%, and 1.47%, and flange wear by 9.5%, 0.46%, and 0.02%. The wear of the No. 1 wheel of the same metro No.2 and No.4 cars is predicted and compared with the measured wear data. The average absolute percentage error of tread wear is 1.34% and 1.42%, respectively, and the average absolute percentage error of rim wear is 0.18% and 0.19%, respectively. The results confirm that the model exhibits strong generalization capabilities. The wheel wear prediction model based on improved BiLSTM network (SSA-Bi-LSTM) has high prediction accuracy and good generalization, which provides theoretical support for the intelligent management of metro wheelsets and prolongs wheel service life.

Rapid urbanization and population growth have led to a continuous increase in passenger flow in urban rail transit, which presents significant challenges to the safety, comfort, and stability of rail transit operations. To solve the problem of excessive load rate of urban rail transit during peak hours, we propose a cooperative passenger flow control method for urban rail transit based on deep reinforcement learning. This method uses the full load rate between intervals as its state, a flow restriction strategy as its action, and the passenger flow experience as its reward. It generates an optimal flow restriction scheme through multiround reinforcement learning. We validated the effectiveness of this method by constructing simulation experiments using data from the Beijing subway network. The simulation results show that the cooperative passenger flow control method can effectively reduce passenger flow in a section, relieve congestion during peak hours, and improve passenger travel comfort.

The current evaluation system for the sustainable development of urban rail transit often employs a simple hierarchical weighted summation method. However, this method does not account for the interactions among indicators at the same level or the multiple effects of lowerlevel indicators on upperlevel indicators, thus failing to reflect the actual sustainable development status systematically, comprehensively, and accurately. To address this issue, this study introduces a new evaluation approachthe ESE (EconomicSocialEnvironmental) evaluation theory and method. This approach builds on the new understanding that data impacts the three dimensions of economy, society, and environment. An ESE spatial Cartesian coordinate system is constructed to represent all operational scenarios of urban rail transit. Within this threedimensional ESE space, the sustainable development status of urban rail transit is classified into "eight types and four categories," allowing for the quantification of sustainability levels and enabling a threedimensional visual representation.

To ensure the safety of ancient buildings during subway shield tunneling, a general technical approach involving initial assessment, reinforcement, construction monitoring, subsequent evaluation, and additional reinforcement has been established. This study built a 1: 3.52 physical model of ancient buildings and conducted shaking table tests to obtain the damping ratio of structures Ansys was used to build a threedimensional model. Using structural and modal analysis for vulnerable structures for strengthening and prevention, this study provides a reference for the followup vibration monitoring and shield construction. The construction parameters were adjusted according to the construction monitoring data in the test area. The results showed that the maximum displacement of the structure was approximately 14.01 mm under static analysis, which verified the validity of the threedimensional model. The basic frequency of the structural model was 0.590 46 Hz and the most vulnerable position was sandalwood. During shield tunneling, the acceleration of ancient buildings was less than 0.003 m/s², the vibration frequency was primarily concentrated at 0.6–20.5 Hz, and the vibration frequency was in the normal range. The paper proposes that the overall solution of subway shield tunneling through the ancient buildings involves a uniform driving speed, avoiding the operation peak period, using the grouting reinforcement between the strata in the affected area, supporting the vulnerable structure of the ancient buildings, and adopting a vibration isolation and vibration reduction scheme during the construction. This provides a reference for similar subsequent projects.

Mediumand lowvolume rail transit, as a supplement to metro networks in super or mega megacities and as the backbone of large and mediumsized cities, is essential for solving urban traffic problems. Based on a survey of more than 2600 mediumand lowvolume rail transit operation lines in 496 cities in 65 countries, a database of mediumand lowvolume rail transit lines worldwide was constructed. Based on this database, 9 types of mediumand lowvolume rail transit systems (gear rail, tram, suspended monorail, automated people mover systems (APM), straddle monorails, linear motor systems, mediumand lowspeed maglev, electronic guidance rubbertired system, and beamguiding rubbertired system) have been obtained worldwide, with a total length of 18 744.84 km. And statistics were conducted on the distribution of mediumand lowvolume rail transit in various continents and countries. The results showed that it is mainly distributed in 32 countries in Europe, including Germany and Russia, with a total length of 13 256.0 km, accounting for 70.7%. Statistics show that the distribution of various rail transit systems in different countries/cities, with different urban characteristics and functional positioning of different lines shows that mediumand lowvolume rail transit systems mainly serve cities with a population of less than 3 million. 76.25% of tram lines, 47.82% of straddle monorail lines, and 42.11% of linear motor system lines are backbone lines, 84.62% of gear rail lines are tourist lines, and 70.73% of APM lines are airport dedicated lines. Additionally, the origin and development of each system are summarized, and the technical characteristics, parameters, and applicability of each system are analyzed. The results provide a reference for the selection of urban medium and lowvolume rail transit systems in China.

To provide maintenance advice regarding a scenario in which floating slab tracks are soaked in water, we established a numerical model of a submerged steel spring floating slab track vibration based on the fluidsolid coupling theory. The vibration reduction of the floating slab track with different accumulation water depths under a running train considering the track, water, and tunnel base was analyzed. The results showed that the water under the floating slab influences the vibration reduction performance. When the depth of the accumulated water was lower than the 1/2 height of the slabside space, the effect on the vertical vibration of the tunnel increased by approximately 5 dB at a peak frequency of 63 Hz. When the water filled with the entire height of the slab side space, the vertical vibration level of the tunnel at its peak frequency of 63 Hz increased to approximately 13.4 dB, an increase of 30.7%. The insertion loss of the tunnel vibration level also increased by approximately 13.36 and 13.67 dB at 63 and 80 Hz, respectively. Therefore, when the height of the water under the floating slab was over the 1/2 height of the slab side space, which means that the vertical vibrations of the slab and tunnel at their peak value frequency were larger than 5 and 10 dB, respectively, than under normal conditions without water, maintenance work to drain away the water should be performed immediately.

A switch machine is a core equipment of a signal system with the largest maintenance volume, strongest maintenance difficulty, and a wide range of equipment failures. With the increase in intelligent operation and maintenance research in the urban rail transit industry, research on intelligent operation and maintenance of switch machines has attracted the interest of industry scholars. It is the foundation for achieving intelligent maintenance of urban rail signals and intelligent operation and maintenance production organization modes. This article summarizes the limitations of traditional microcomputer monitoring systems and proposes an intelligent system based on the maintenance difficulties and pain points of switch machines in the industry. The system includes critical state perception, intelligent fault diagnosis, critical state warning, and health assessment, and proposes specific implementation methods. The system was successfully applied to Nanning Metro Lines 4 and 5, and significant benefits were obtained.

In view of the lack of research on rescue scenarios in the operation of metro railway and the inability of the electric multiple unit (EMU) to fully meet the timeliness requirements of rescue, the runaway risk and braking strategy of metro train rescue conditions are studied. Firstly, a longitudinal dynamic model is established for the coupling of multi formation trains and then Beijing New Airport Line is token as an example to analyze the potential risk of train slip caused by the existing urban rail transit operation under rescue scenario. Finally, based on relevant standards, this study analyzes the comfort of passengers under different braking modes during the rescue process. The results show that when coupled at a speed of 5 km/h on a 33‰ slope, the acceleration and jerk rate of the train will reach 10.5 m/s² and 9.9 m/s³ while at holding brake mode, and slip will occur. However, the acceleration and jerk rate of the train at emergency braking mode decrease by 63.1% and 54.7%, respectively, and no slip occurrs. When the coupling speed is reduced to 3 km/h, the maximum acceleration and maximum impact rate are reduced to 2.1 m/s² and 2.4 m/s³, respectively, which means passenger comfort significantly improves. When the train is parked on a slope, due to the braketraction switching process, the holding brake force should reach to 60% or more of the maximum service braking force to ensure that the vehicle will not slide on the maximum slope.