ObjectiveAddressing the problems of insufficient fault localization accuracy，lagging handling response，and high reliance of fault handling on the experience of operations and maintenance personnel in FAO （fully automatic operation） and CBTC （communication-based train control） systems，it is necessary to research an emergency braking handling model for urban rail transit signaling systems based on decision trees. MethodRelying on the line emergency braking fault dataset and on-site handling records，a cause analysis is conducted to classify faults into three categories: signaling-dominated faults，vehicle-associated faults，and external interference and human factors，sorting out the triggers for each type of fault. Combining the coupling of fault factors and chronological logic，integrating industry expert experience，and adhering to four design criteria，a three-level decision tree braking model is constructed. With DMI （driver-machine interface） alarm information as input，the primary branch distinguishes FAO and CBTC systems; the secondary branches trace back layer by layer around ATP （automatic train protection） systems，wayside ATP systems，ATO （automatic train operation） systems，and non-signaling linkage faults; and the tertiary nodes output standardized handling operations. Result & Conclusion The three-level tree structure enables multi-dimensional fault tracing，with all nodes adopting binary logic，making the reasoning process fully interpretable. The proposed emergency braking handling model can establish a mapping relationship between alarm information and handling strategies; all types of fault triggers can be mapped to leaf nodes for precise localization，solidifying expert rules and forming a closed-loop process from fault tracing to handling instruction output.
ObjectiveTrain body vibration acceleration is a core indicator for evaluating the operational safety and stability of trains. Accurately predicting the lateral and vertical vibration accelerations of the train body is the key to realizing active safety early warning for trains. Therefore，it is necessary to develop a method for predicting train body vibration status that integrates multi-source data. MethodA train body vibration acceleration prediction model based on an LSTM （long- and short-term memory） neural network is proposed. First，a dataset covering multi-dimensional features such as operating speed，track irregularities，curve parameters，wheel profile parameters，and suspension system parameters is constructed. A genetic algorithm is utilized to obtain the optimal values of the hyperparameters for the LSTM neural network. A high-precision train body vibration status prediction model is established，taking multi-dimensional features as inputs and train body vibration acceleration as the output. Measured data are adopted to verify the accuracy of the prediction model. Result & Conclusion The RMSE （root mean square error） values of the proposed prediction model for the lateral and vertical vibration accelerations of the train body are 0.007 m/s2 and 0.013 m/s2，respectively. The prediction performance of the proposed model for both lateral and vertical vibration accelerations of the train body is superior to that of the comparison models. Especially in predicting the lateral vibration acceleration of the train body，the RMSE is reduced by 95.9%，which verifies the effectiveness and advanced nature of the proposed prediction model.
ObjectiveThe intelligent upgrade of railway electrical maintenance system has raised new demands for the fault diagnosis technology of turnout switch machines. Therefore，it is necessary to systematically evaluate the application and development paths of various methods in this field. MethodA multi-dimensional technical analysis of relevant methods and achievements is conducted to reveal the technical characteristics and bottlenecks of current machine learning-driven fault diagnosis systems. Addressing the current research status and taking the ZD6 and S700K switch machines as examples，common fault types of both DC （direct current） and AC （alternating current） switch machines are summarized，and the shortcomings of existing case databases regarding data comprehensiveness and fault category coverage are pointed out. In terms of feature extraction，targeting the data characteristics of switch machine action curves，the methods and challenges of fault feature extraction are reviewed，and the effectiveness of feature processing strategies across different dimensions is discussed. Furthermore，the primary applied algorithms in fault diagnosis are detailed，including typical application cases of support vector machines，neural networks，and cutting-edge algorithms，with a focus on analyzing the performance and applicable scenarios of each algorithm. Result & Conclusion Deep learning methods have strongly facilitated the improvement of fault diagnosis capabilities for turnout switch machines，leading to significant enhancements in both accuracy and diagnostic efficiency. Based on certain deficiencies in existing research，potential future research directions are proposed from the data，methodology，and engineering application levels. This provides a theoretical basis and technical pathway for the engineering deployment of intelligent diagnostic systems，thereby promoting the intellectualization and practical upgrade of fault diagnosis for turnout switch machines.
ObjectiveConventional UAV （unmanned aerial vehicle） inspections in urban rail transit control protection zones are limited by the 'take-off and return-to-base' mode，presenting issues such as significant power loss during return trips，a large number of deployed hangars，and high construction and O&M （operation and maintenance） costs. Therefore，it is necessary to study a network-level UAV inspection optimization scheme in complex urban environments. MethodComprehensively considering urban airspace restrictions，line-laying methods，and network connectivity，a configuration optimization method for UAV inspection systems based on leapfrog strategy is proposed. In combination with the actual geographical and business environments of the fully operational lines of Jinan Rail Transit，a multi-objective joint optimization mathematical model for 'hangar siting–tidal scheduling–route planning' is established，aiming to minimize the number of deployed hangars and maximize inspection coverage. This optimization model is applied proactively to the network-level holistic planning of Jinan Rail Transit. Result & Conclusion The measured data from the pilot section of Jinan Rail Transit Line 3 demonstrate that the leapfrog strategy extends the one-way continuous UAV inspection mileage from the conventional 3.25 km to 6.50 km，and maintains a safety redundancy of over 30% for the battery state of charge （SOC） upon landing，successfully running through the control closed loop of 'flight–fast charge–relay.' The simulation results of the network-level holistic planning show that by extending the single-line hangar spacing and implementing 'cross-line sharing of hangars at transfer hubs,' the total number of hangars deployed across the entire network is reduced from 72 under conventional independent planning to 38. To a certain extent，this solves the engineering pain points of 'unaffordable construction and lack of land' for hangars within the rail network.
ObjectiveAddressing the issue of abnormal fatigue damage and wear of high-speed railway catenary steady arms during service，the influence of train operating speed and contact wire stagger value on the load characteristics and vibration response of the steady arm is investigated. MethodBased on the pantograph-catenary coupling dynamics theory and the kinematic relationship of the steady arm，a pantograph-catenary dynamics coupling model considering the motion state of the steady arm is established. The validity of the proposed model is verified according to relevant pantograph-catenary dynamics standards. On this basis，operating conditions with different train operating speeds and different contact wire stagger values are set. Response indicators such as the three-directional loads of the steady arm，the vertical uplift of the registration point，and the vertical vibration acceleration are extracted. The variation laws of the steady arm dynamics response are analyzed in terms of maximum amplitude，standard deviation，and fluctuation characteristics. Result & Conclusion The model calculation results meet the requirements of relevant pantograph-catenary system dynamics and can reflect the pantograph-catenary dynamics response characteristics under the action of the steady arm. As the train operating speed increases，the maximum amplitudes and standard deviations of the three-directional loads of the steady arm，the vertical uplift of the registration point，and the vertical acceleration all increase. This indicates that an increase in train operating speed enhances the load fluctuation and vibration intensity of the steady arm，exacerbating fatigue damage and wear. As the contact wire stagger value increases，the fluctuations of the steady arm lateral force and the registration point uplift decrease slightly，but the static initial value of the normal interaction force of the contact pair，jointly formed by the lateral and vertical forces，increases significantly. Therefore，appropriately reducing the contact wire stagger value can lower the normal action level of the steady arm contact pair，thereby retarding the development of wear.
