A subway emergency management and intelligent command system and method

Abstract

The present application relates to the technical field of safety monitoring and intelligent management of rail transit systems, and specifically discloses a subway emergency management and intelligent command system and method, which uses temperature sensors and vibration sensors to respectively collect temperature data of electrical connection points and track vibration data; based on Haar wavelet transform, temperature abnormal fluctuation characteristic values are calculated to identify abnormal states of the electrical connection points, while fast Fourier transform is combined with mean clustering analysis to extract track vibration amplitude and frequency characteristics, a comprehensive characteristic vector is constructed as input, a random forest model is used for risk prediction, according to the risk score output by the model, the future running state of the track is divided into three levels of high risk, general risk and low risk, a detailed inspection and maintenance plan is automatically triggered for the high-risk area, and the train scheduling strategy is optimized to ensure operation safety, key information is displayed through data visualization technology, and the response speed and accuracy of decision makers are improved.

Description

Technical Field

[0001] This invention relates to the field of safety monitoring and intelligent management technology for rail transit systems, specifically to a subway emergency management and intelligent command system and method. Background Technology

[0002] With the rapid development of urban rail transit systems, ensuring their safety and reliability has become a key challenge in operation and management. As a crucial component of modern urban public transportation, the monitoring and maintenance of subway operation is vital for ensuring passenger safety and improving operational efficiency. Traditional monitoring methods primarily rely on periodic manual inspections and simple sensor data collection, such as monitoring the temperature and vibration of track electrical connection points. However, these methods often fail to achieve real-time, accurate risk assessment, especially in the face of the complex and ever-changing urban rail transit environment. Therefore, there is an urgent need for a method integrating advanced data analysis technologies to achieve comprehensive monitoring and intelligent management of track health.

[0003] The existing technology has the following shortcomings:

[0004] While existing technologies can provide basic information about track status to some extent, they suffer from several significant shortcomings. First, traditional methods typically rely on threshold alarm mechanisms, which can only identify obvious faults that have already occurred, but struggle to provide early warnings of potential risks. This often results in reactive rather than preventative maintenance. Second, current data analysis methods are relatively limited, lacking the ability to comprehensively analyze various types of data and failing to capture deeper changes in the track system. Furthermore, existing technologies are inefficient when processing large-scale datasets, making it difficult to support real-time monitoring needs. Finally, the lack of effective data visualization tools hinders decision-makers from making quick and accurate judgments when faced with complex monitoring data, further limiting the effectiveness of emergency management and command systems. These problems highlight the importance of developing a more intelligent and efficient track health monitoring and management system. Summary of the Invention

[0005] The purpose of this invention is to provide a subway emergency management and intelligent command system and method to solve the problems mentioned above.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] A subway emergency management and intelligent command method includes the following steps:

[0008] S1: During subway operation, the temperature data of the track electrical connection points is monitored in real time, and the abnormal fluctuation of the temperature data is used to determine whether there is any abnormality in the working status of the electrical connection points;

[0009] S2: Based on the judgment result, the working status of the electrical connection point is divided into normal and abnormal;

[0010] S3: Provide early warning based on abnormal electrical connection points;

[0011] S4: Based on normal electrical connection points, obtain vibration data of the corresponding track, process the vibration data, extract vibration amplitude features and vibration frequency features, and construct a track risk prediction model based on the extracted vibration amplitude features and vibration frequency features to predict the degree of track risk.

[0012] S5: Based on the model output, the risk level of future orbital operation status is divided into three levels: high risk, moderate risk and low risk. Early warning processing is carried out for high-risk orbital operation status.

[0013] As a further aspect of the present invention: the determination of whether the working state of the electrical connection point is abnormal specifically includes:

[0014] According to the time series, the temperature data of each electrical connection point is collected in real time. Based on the abnormal fluctuation amplitude of the temperature data, the abnormal temperature fluctuation characteristic value is calculated. It is determined whether the abnormal temperature fluctuation characteristic value of each electrical connection point is greater than or equal to the preset threshold. If it is, it is recorded as an abnormal electrical connection point; otherwise, it is recorded as a normal electrical connection point.

[0015] As a further aspect of the present invention: the process for obtaining the temperature anomaly fluctuation characteristic value is as follows:

[0016] According to the time series, the temperature data of each electrical connection point is collected in real time. The difference between the temperature data collected at adjacent time points is calculated and the absolute value is taken to obtain the temperature data difference sequence.

[0017] The difference sequence of temperature data is processed using Haar wavelet transform, which decomposes the difference sequence of temperature data into approximation coefficients and detail coefficients.

[0018] For each level of detail coefficient, calculate the energy of the corresponding level of detail coefficient;

[0019] The characteristic value of temperature anomaly fluctuation is obtained by summing the squares of the ratio of the energy of each detail factor to the total energy of all detail factors.

[0020] As a further aspect of the present invention: the processing of vibration data specifically includes:

[0021] Based on normal electrical connection points, vibration data of the corresponding track is obtained. According to the degree of abnormal change in the vibration amplitude characteristics of the track within the monitoring period, the abnormal characteristic value of the vibration amplitude of the corresponding track is calculated. According to the degree of abnormal fluctuation in the vibration frequency of the track within the monitoring period, the abnormal characteristic value of the vibration frequency of the corresponding track is calculated.

[0022] As a further aspect of the present invention: the process for obtaining the abnormal vibration amplitude characteristic value is as follows:

[0023] Vibration data of the track is acquired in real time according to the time series. The acquired vibration data is then converted to the frequency domain using Fast Fourier Transform to obtain the complex spectrum. ;

[0024] Calculate the energy of each frequency component, calculate the total energy of all frequency components, calculate the proportion of each frequency component relative to the total energy, calculate the standard deviation of the energy proportions of all frequency components, and obtain the abnormal characteristic value of vibration amplitude.

[0025] As a further aspect of the present invention: the process for obtaining the abnormal vibration frequency characteristic value is as follows:

[0026] pass Mean value clustering analysis is used to analyze vibration data obtained through Fast Fourier Transform (FFT). Specifically, it includes determining the number of clusters. Random selection We have 1 initial cluster center, which represents the typical amplitude level of different frequency bands; for each frequency component, we assign it to the nearest cluster center, i.e., find the minimum Euclidean distance; update the position of each cluster center to the average of its members, and repeat the assignment and update steps until the predetermined number of iterations is reached.

[0027] Calculate the absolute value of the difference between each frequency component and its cluster center to obtain the distance between each frequency component and its cluster center. Calculate the average of the distances of all frequency components to obtain the vibration frequency anomaly characteristic value.

[0028] As a further aspect of the present invention: the construction of the orbital risk prediction model specifically includes:

[0029] Anomaly features of vibration amplitude and vibration frequency are combined to form a comprehensive feature vector, which is used as the input to the risk prediction model. The model is trained with the goal of minimizing the error between the risk score of the future trajectory and the risk score of the actual trajectory. The output of the model is the risk score of the future trajectory. The risk prediction model is a random forest model.

[0030] As a further aspect of the present invention: based on the model output, the risk level of the future orbital operating state is divided into three levels, specifically including:

[0031] Based on the model output, the risk score of the future trajectory is compared with a preset first threshold. If the risk score is greater than or equal to the preset first threshold, it is recorded as high risk. If the risk score is less than the preset first threshold, the risk score is compared with a preset second threshold. If the risk score is less than or equal to the preset second threshold, it is recorded as low risk. If the risk score is greater than the preset second threshold and less than the preset first threshold, it is recorded as general risk.

[0032] As a further aspect of the present invention: Management of the aforementioned high-risk orbital operating state specifically includes:

[0033] Based on the predicted high-risk track sections, track adjustments and replacement of damaged components are carried out during non-operational time; and real-time monitoring of the current section continues, with the monitoring frequency increased until the next track operation status assessment meets the low-risk condition, at which point the monitoring frequency is reduced.

[0034] A subway emergency management and intelligent command system includes:

[0035] The data acquisition module collects temperature data of the track within the monitoring section using a temperature sensor and vibration data of the track within the monitoring section using a vibration sensor.

[0036] An electrical connection point anomaly identification module determines whether there is an anomaly in the working status of the electrical connection point based on the degree of abnormal fluctuation in temperature data.

[0037] An electrical connection point classification module, which classifies the working state of electrical connection points into normal and abnormal based on the judgment result;

[0038] An electrical connection point early warning module, which performs early warning processing based on abnormal electrical connection points;

[0039] The track risk prediction module acquires vibration data of the corresponding track based on normal electrical connection points, processes the vibration data, extracts vibration amplitude and frequency features, and constructs a track risk prediction model based on the extracted vibration amplitude and frequency features to predict the degree of track risk.

[0040] The risk classification and emergency response module classifies the risk level of future orbital operation status into three levels based on the model output: high risk, moderate risk, and low risk, and provides early warning processing for high-risk orbital operation status.

[0041] The beneficial effects of this invention are:

[0042] (1) This invention innovatively achieves in-depth analysis of temperature data and vibration characteristics of track electrical connection points by integrating real-time monitoring with advanced data analysis technologies, such as Haar wavelet transform, fast Fourier transform, and mean clustering analysis, and accurately identifies potential risk factors. For electrical connection points, Haar wavelet transform is used to calculate the characteristic values ​​of abnormal temperature fluctuations, enabling early warning of their working status; for track vibration, fast Fourier transform combined with mean clustering analysis is used to extract vibration amplitude and frequency characteristics, constructing a comprehensive risk prediction model to effectively capture subtle changes and predict future risks. This method not only enables maintenance teams to take targeted measures in advance to prevent failures, thereby significantly improving the safety and operational stability of the rail transit system, but also significantly reduces service interruptions and maintenance costs caused by sudden failures by optimizing maintenance plans and resource allocation, and extends the service life of equipment. In addition, the combination of data visualization technology to display key information further enhances the accuracy and efficiency of the decision-making process, providing strong technical support for subway emergency management and intelligent command, and ensuring the efficient and stable operation of the rail transit system.

[0043] (2) Based on the constructed random forest risk prediction model, this invention realizes an accurate assessment and graded early warning mechanism for the risk level of future track operation status. Through the deep integration of historical data and real-time monitoring information, it automatically triggers detailed inspection and maintenance plans for high-risk areas and optimizes train scheduling strategies to ensure operational safety. Specifically, the model uses advanced technologies such as Haar wavelet transform, fast Fourier transform, and mean clustering analysis to extract key features through in-depth analysis of electrical connection point temperature data and track vibration characteristics, thereby generating a comprehensive risk score. For specific sections predicted to be high-risk, the system can not only issue timely alarms, but also clearly display the location of abnormal points, temperature change trends, vibration modes, and historical data with the help of data visualization technology, greatly improving the response speed and decision accuracy of decision-makers. This comprehensive risk management framework not only enhances the efficiency and accuracy of subway emergency management and intelligent command, but also promotes the scientific rationalization of resource allocation and maintenance plans, fundamentally ensuring the efficient and stable operation of the rail transit system and significantly reducing the safety risks and economic losses caused by potential faults. Finally, this invention provides a comprehensive, dynamic, and forward-looking solution, laying a solid foundation for improving the safety and reliability of urban rail transit. Attached Figure Description

[0044] The invention will now be further described with reference to the accompanying drawings.

[0045] Figure 1 This is a flowchart of a subway emergency management and intelligent command method according to the present invention;

[0046] Figure 2This is a flowchart of a subway emergency management and intelligent command system according to the present invention. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] Please see Figure 1 As shown, this invention is a subway emergency management and intelligent command method, comprising the following steps:

[0049] S1: During subway operation, the temperature data of the track electrical connection points is monitored in real time, and the abnormal fluctuation of the temperature data is used to determine whether there is any abnormality in the working status of the electrical connection points;

[0050] S2: Based on the judgment result, the working status of the electrical connection point is divided into normal and abnormal;

[0051] S3: Provide early warning based on abnormal electrical connection points;

[0052] S4: Based on normal electrical connection points, obtain vibration data of the corresponding track, process the vibration data, extract vibration amplitude features and vibration frequency features, and construct a track risk prediction model based on the extracted vibration amplitude features and vibration frequency features to predict the degree of track risk.

[0053] S5: Based on the model output, the risk level of future orbital operation status is divided into three levels: high risk, moderate risk and low risk. Early warning processing is carried out for high-risk orbital operation status.

[0054] In S1, during subway operation, the temperature data of the track electrical connection points is monitored in real time. Based on the degree of abnormal fluctuation in the temperature data, it is determined whether there is any abnormality in the working status of the electrical connection points. Specifically, this includes:

[0055] According to the time series, the temperature data of each electrical connection point is collected in real time. Based on the abnormal fluctuation amplitude of the temperature data, the abnormal temperature fluctuation characteristic value is calculated. It is determined whether the abnormal temperature fluctuation characteristic value of each electrical connection point is greater than or equal to the preset threshold. If it is, it is recorded as an abnormal electrical connection point; otherwise, it is recorded as a normal electrical connection point.

[0056] The process for obtaining the temperature anomaly fluctuation characteristic value is as follows:

[0057] According to the time series, the temperature data of each electrical connection point is collected in real time. The difference between the temperature data collected at adjacent time points is calculated and the absolute value is taken to obtain the temperature data difference sequence.

[0058] The difference sequence of temperature data is processed using Haar wavelet transform, which decomposes the difference sequence of temperature data into approximation coefficients and detail coefficients.

[0059] For each level of detail coefficient, calculate the energy of the corresponding level of detail coefficient. The calculation expression is as follows: In the formula, This indicates the number of layers in the Haar wavelet transform. Indicates the first Layer detail factor, Indicates the first Energy of layer detail factor Indicates the first A detailed coefficient, This represents the total number of detail coefficients;

[0060] The characteristic value of temperature anomaly fluctuation is obtained by summing the squares of the ratio of the energy of each detail factor to the total energy of all detail factors.

[0061] It should be noted that by combining differential sequence calculation and Haar wavelet transform, the characteristic value of abnormal temperature fluctuation at electrical connection points can be accurately calculated. The larger the characteristic value of abnormal temperature fluctuation, the higher the degree of abnormality of the corresponding electrical connection point. By comparing the characteristic value of abnormal temperature fluctuation with the threshold, it can be directly used to guide the decision-making process in emergency management and intelligent command systems.

[0062] In S3, early warning processing is performed based on abnormal electrical connection points, specifically including:

[0063] Based on the abnormal electrical connection point, the system automatically triggers an alarm, highlighting the location of the abnormal point in red on the monitoring interface and labeling the abnormality type as temperature abnormality. The location of the abnormal electrical connection point is precisely located on the track structure diagram, and through data visualization technology, information such as the temperature change trend, historical data, and abnormal fluctuation characteristics of the abnormal point are displayed in chart form to help decision-makers fully understand the abnormal situation. The nearest maintenance team is notified to rush to the site, increase the frequency of inspections, and arrange planned maintenance.

[0064] In S4, based on normal electrical connection points, vibration data of the corresponding track is acquired. This vibration data is processed to extract vibration amplitude and frequency features. Based on these extracted features, a track risk prediction model is constructed, specifically including:

[0065] Based on normal electrical connection points, vibration data of the corresponding track is obtained. According to the degree of abnormal change in the vibration amplitude characteristics of the track within the monitoring period, the abnormal characteristic value of the vibration amplitude of the corresponding track is calculated. According to the degree of abnormal fluctuation in the vibration frequency of the track within the monitoring period, the abnormal characteristic value of the vibration frequency of the corresponding track is calculated.

[0066] The process for obtaining the abnormal vibration amplitude characteristic value is as follows:

[0067] Vibration data of the track is acquired in real time according to the time series. The acquired vibration data is then converted to the frequency domain using Fast Fourier Transform to obtain the complex spectrum. ;in, This represents the complex spectrum of vibration data converted to the frequency domain. Indicates the total number of frequency components. Indicates the first Due to the symmetry of the real input signal, we only need to consider the complex values ​​of the frequency components. One frequency component;

[0068] The energy of each frequency component is calculated using the following expression: In the formula, Indicates the first The energy of each frequency component Indicates the first The complex values ​​of each frequency component Indicates the first One frequency component;

[0069] Calculate the total energy of all frequency components, calculate the proportion of each frequency component relative to the total energy, calculate the standard deviation of the energy proportions of all frequency components, and obtain the abnormal characteristic value of vibration amplitude.

[0070] The process for obtaining the abnormal vibration frequency characteristic value is as follows:

[0071] pass Mean value clustering analysis is used to analyze vibration data obtained through Fast Fourier Transform (FFT). Specifically, it includes determining the optimal number of clusters. Random selection We have 10 initial cluster centers, each representing a typical amplitude level across different frequency bands. For each frequency component, we assign it to the nearest cluster center, i.e., we find the minimum Euclidean distance. We then update the position of each cluster center to the average value of its members, calculated as follows: ,in, Indicates the first The central location of each cluster This represents the total number of cluster centers. express The energy of each frequency component Indicates the first The set of all data points in each cluster; repeat the assignment and update steps until the predetermined number of iterations is reached;

[0072] Calculate the absolute value of the difference between each frequency component and its cluster center to obtain the distance between each frequency component and its cluster center. Calculate the average of the distances of all frequency components to obtain the vibration frequency anomaly characteristic value.

[0073] It should be noted that by determining the optimal number of clusters, randomly initializing cluster centers to represent typical amplitude levels in different frequency bands, classifying each frequency component, and iteratively updating the cluster centers until convergence, and finally calculating the distance between each frequency component and its respective cluster center to obtain vibration frequency anomaly characteristic values, this technology not only achieves accurate extraction of deep-level features of vibration signals but also effectively identifies subtle changes in vibration modes and potential fault warning signals. This improves the foresight and accuracy of rail transit system maintenance, thereby significantly reducing safety risks and maintenance costs caused by equipment failures. This technology is highly effective, providing strong data support and decision-making basis for subway emergency management and intelligent command.

[0074] The construction of the orbital risk prediction model specifically includes:

[0075] An abnormal vibration amplitude feature value and an abnormal vibration frequency feature value are used to construct a comprehensive feature vector, which is used as the input of the risk prediction model. The model is trained with the goal of minimizing the error between the risk score of the future track and the risk score of the actual track. The output of the model is the risk score of the future track. The risk prediction model is a random forest model.

[0076] Training the model specifically includes:

[0077] Using a comprehensive feature vector from historical data and its corresponding actual trajectory risk scores as the training set, a random forest model is trained with the goal of minimizing the error between the risk scores of future trajectories and the actual risk scores of trajectories. Random forest is an ensemble learning method that improves the accuracy and robustness of the model by constructing multiple decision trees and aggregating their predictions. During training, different subsets are generated by resampling the dataset, and a decision tree is trained for each subset. Finally, a voting mechanism determines the final risk score. This process effectively utilizes information from a large amount of historical data, improving the accuracy of future risk assessment.

[0078] It's important to note that a well-trained random forest model can accept new, comprehensive feature vectors as input and output a risk score for the track within a given time period. These scores can be used to predict the potential risk level of the track in the future, guiding maintenance planning and resource allocation. When a risk is predicted for a particular area, inspections and maintenance can be scheduled in advance to prevent malfunctions. This method significantly improves the safety, reliability, and operational efficiency of rail transit systems while reducing costs and losses caused by unforeseen events.

[0079] In S5, based on the model output, the risk level of future orbital operating states is divided into three levels: high risk, moderate risk, and low risk. Early warning processing is implemented for high-risk orbital operating states, specifically including:

[0080] Based on the model output, the risk score of the future orbit is compared with a preset first threshold. If the risk score is greater than or equal to the preset first threshold, it indicates that the operational risk level of the corresponding orbit is high and is recorded as high risk. If the risk score is less than the preset first threshold, the risk score is compared with a preset second threshold. If the risk score is less than or equal to the preset second threshold, it indicates that the operational risk level of the corresponding orbit is low and is recorded as low risk. If the risk score is greater than the preset second threshold and less than the preset first threshold, it indicates that the operational risk level of the corresponding orbit is moderate and is recorded as general risk.

[0081] Based on the anticipated high-risk track operation conditions, and using high-risk scoring areas from risk prediction models, detailed inspection and maintenance plans are planned in advance. Necessary repairs, such as track adjustments and replacement of damaged components, are carried out during non-operational times to prevent potential malfunctions. Simultaneously, real-time monitoring of these areas is strengthened. Continuous vibration tracking is employed to monitor changes in track temperature and vibration data in real time, and risk assessments are conducted using risk prediction models. Any anomalies detected are immediately alerted, and emergency plans are activated. Furthermore, train schedules are optimized by combining historical data and real-time information, appropriately reducing train density and speed in high-risk sections to ensure operational safety. Through this comprehensive risk management approach, the probability of accidents can be effectively reduced, ensuring the stable operation of the rail transit system.

[0082] Please see Figure 2 As shown, a subway emergency management and intelligent command system includes:

[0083] The data acquisition module collects temperature data of the track within the monitoring section using a temperature sensor and vibration data of the track within the monitoring section using a vibration sensor.

[0084] An electrical connection point anomaly identification module determines whether there is an anomaly in the working status of the electrical connection point based on the degree of abnormal fluctuation in temperature data.

[0085] An electrical connection point classification module, which classifies the working state of electrical connection points into normal and abnormal based on the judgment result;

[0086] An electrical connection point early warning module, which performs early warning processing based on abnormal electrical connection points;

[0087] The track risk prediction module acquires vibration data of the corresponding track based on normal electrical connection points, processes the vibration data, extracts vibration amplitude and frequency features, and constructs a track risk prediction model based on the extracted vibration amplitude and frequency features to predict the degree of track risk.

[0088] The risk classification and emergency response module classifies the risk level of future orbital operation status into three levels based on the model output: high risk, moderate risk, and low risk, and provides early warning processing for high-risk orbital operation status.

[0089] The working principle of this invention is as follows: By real-time monitoring and analysis of temperature data and track vibration data at electrical connection points, the safety and reliability of the rail transit system are improved. During subway operation, temperature sensors collect temperature data from each electrical connection point in real time, and Haar wavelet transform is used to calculate abnormal temperature fluctuation characteristics to determine whether any abnormalities exist at the electrical connection points, thus achieving effective monitoring of the electrical connection point status. For electrical connection points identified as normal, corresponding track vibration data is further obtained through vibration sensors, and fast Fourier transform and mean clustering analysis are applied to extract vibration amplitude and frequency characteristics. A comprehensive feature vector is constructed as input, and a random forest model is used for track risk prediction. Based on the risk score output by the model, the future operating status of the track is divided into three levels: high risk, general risk, and low risk. Especially for areas predicted to be high risk, detailed inspection and maintenance plans are planned in advance, and measures such as enhanced real-time monitoring and optimized train schedules are taken to prevent potential faults. The system also includes a complete data processing flow, including a data acquisition module, an electrical connection point anomaly identification module, a classification module, an early warning module, a track risk prediction module, and a risk classification and emergency handling module, working together to ensure the efficient and stable operation of the rail transit system. This solution not only improves the ability to identify potential risks on the track, but also effectively reduces safety risks and maintenance costs caused by equipment failures, greatly enhancing the intelligence level of subway emergency management.

[0090] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A subway emergency management and intelligent command method, characterized in that, Includes the following steps: S1: During subway operation, the temperature data of the track electrical connection points is monitored in real time, and the abnormal fluctuation of the temperature data is used to determine whether there is any abnormality in the working status of the electrical connection points; According to the time series, the temperature data of each electrical connection point is collected in real time. Based on the abnormal fluctuation amplitude of the temperature data, the abnormal temperature fluctuation characteristic value is calculated. It is determined whether the abnormal temperature fluctuation characteristic value of each electrical connection point is greater than or equal to the preset threshold. If it is, it is recorded as an abnormal electrical connection point; otherwise, it is recorded as a normal electrical connection point. The process for obtaining the temperature anomaly fluctuation characteristic value is as follows: According to the time series, the temperature data of each electrical connection point is collected in real time. The difference between the temperature data collected at adjacent time points is calculated and the absolute value is taken to obtain the temperature data difference sequence. The difference sequence of temperature data is processed using Haar wavelet transform, which decomposes the difference sequence of temperature data into approximation coefficients and detail coefficients. For each level of detail factor, calculate the energy of the corresponding level of detail factor. The calculation expression is as follows: In the formula, This indicates the number of layers in the Haar wavelet transform. Indicates the first Layer detail factor, Indicates the first Energy of layer detail factor Indicates the first A detailed coefficient, This represents the total number of detail coefficients; The characteristic value of temperature anomaly fluctuation is obtained by summing the squares of the ratio of the energy of each detail factor to the total energy of all detail factors. S2: Based on the judgment result, the working status of the electrical connection point is divided into normal and abnormal; S3: Provide early warning based on abnormal electrical connection points; S4: Based on normal electrical connection points, obtain vibration data of the corresponding track, process the vibration data, extract vibration amplitude features and vibration frequency features, and construct a track risk prediction model based on the extracted vibration amplitude features and vibration frequency features to predict the degree of track risk. Based on normal electrical connection points, obtain vibration data of the corresponding track. Calculate the abnormal characteristic value of vibration amplitude of the corresponding track according to the degree of abnormal change in the vibration amplitude characteristics of the track within the monitoring period. Calculate the abnormal characteristic value of vibration frequency of the corresponding track according to the degree of abnormal fluctuation in the vibration frequency of the track within the monitoring period. The process for obtaining the abnormal vibration amplitude characteristic value is as follows: Vibration data of the track is acquired in real time according to the time series. The acquired vibration data is then converted to the frequency domain using Fast Fourier Transform to obtain the complex spectrum. ;in, This represents the complex spectrum of vibration data converted to the frequency domain. Indicates the total number of frequency components. Indicates the first Due to the symmetry of the real input signal, we only need to consider the complex values ​​of the frequency components. One frequency component; The energy of each frequency component is calculated using the following expression: In the formula, Indicates the first The energy of each frequency component Indicates the first The complex values ​​of each frequency component Indicates the first One frequency component; Calculate the total energy of all frequency components, calculate the proportion of each frequency component relative to the total energy, calculate the standard deviation of the energy proportions of all frequency components, and obtain the abnormal characteristic value of vibration amplitude. The process for obtaining the abnormal vibration frequency characteristic value is as follows: pass Mean value clustering analysis is used to analyze vibration data obtained through Fast Fourier Transform (FFT). Specifically, it includes determining the optimal number of clusters. Random selection We have 10 initial cluster centers, each representing a typical amplitude level across different frequency bands. For each frequency component, we assign it to the nearest cluster center, i.e., we find the minimum Euclidean distance. We then update the position of each cluster center to the average value of its members, calculated as follows: ,in, Indicates the first The central location of each cluster This represents the total number of cluster centers. express The energy of each frequency component Indicates the first The set of all data points in each cluster; repeat the assignment and update steps until the predetermined number of iterations is reached; Calculate the absolute value of the difference between each frequency component and its cluster center to obtain the distance between each frequency component and its cluster center. Calculate the mean of the distances of all frequency components to obtain the vibration frequency anomaly characteristic value. The construction of the orbital risk prediction model specifically includes: An abnormal vibration amplitude feature value and an abnormal vibration frequency feature value are used to construct a comprehensive feature vector, which is used as the input of the risk prediction model. The model is trained with the goal of minimizing the error between the risk score of the future track and the risk score of the actual track. The output of the model is the risk score of the future track. The risk prediction model is a random forest model. S5: Based on the model output, the risk level of future orbital operation status is divided into three levels: high risk, moderate risk and low risk. Early warning processing is carried out for high-risk orbital operation status.

2. The subway emergency management and intelligent command method according to claim 1, characterized in that, Based on the model output, the risk level of future orbital operation is divided into three levels, specifically including: Based on the model output, the risk score of the future trajectory is compared with a preset first threshold. If the risk score is greater than or equal to the preset first threshold, it is recorded as high risk. If the risk score is less than the preset first threshold, the risk score is compared with a preset second threshold. If the risk score is less than or equal to the preset second threshold, it is recorded as low risk. If the risk score is greater than the preset second threshold and less than the preset first threshold, it is recorded as general risk.

3. The subway emergency management and intelligent command method according to claim 2, characterized in that, Management of the aforementioned high-risk orbital operating conditions includes: Based on the predicted high-risk track sections, track adjustments and replacement of damaged components are carried out during non-operational time; and real-time monitoring of the current section continues, with the monitoring frequency increased until the next track operation status assessment meets the low-risk condition, at which point the monitoring frequency is reduced.

4. A subway emergency management and intelligent command system, used in the subway emergency management and intelligent command method as described in any one of claims 1-3, characterized in that, include: The data acquisition module collects temperature data of the track within the monitoring section using a temperature sensor and vibration data of the track within the monitoring section using a vibration sensor. An electrical connection point anomaly identification module determines whether there is an anomaly in the working status of the electrical connection point based on the degree of abnormal fluctuation in temperature data. An electrical connection point classification module, which classifies the working state of electrical connection points into normal and abnormal based on the judgment result; An electrical connection point early warning module, which performs early warning processing based on abnormal electrical connection points; The track risk prediction module acquires vibration data of the corresponding track based on normal electrical connection points, processes the vibration data, extracts vibration amplitude and frequency features, and constructs a track risk prediction model based on the extracted vibration amplitude and frequency features to predict the degree of track risk. The risk classification and emergency response module classifies the risk level of future orbital operation status into three levels based on the model output: high risk, moderate risk, and low risk, and provides early warning processing for high-risk orbital operation status.

Images