Train early warning system based on vehicle health monitoring

Abstract

The invention relates to the technical field of subway early warning, in particular to a train early warning system based on vehicle health monitoring, which comprises an acquisition module, a determination module, a screening module, a judgment module, an adjustment module and an early warning module. According to the method, through multi-parameter cross monitoring and layer-by-layer screening judgment, the recognition precision of the abnormal carriage is effectively improved, non-structural wear deviation caused by natural factors such as turning and windward can be eliminated through introduction of wind speed and curvature data, the screening pertinence is improved, the pressure data is subjected to normalization and fluctuation dispersion analysis, and the accuracy of the abnormal carriage recognition is improved. The stability and adaptability of the judgment standard are ensured, the thickness and current data are used for dynamically adjusting the threshold value, so that the system can adaptively optimize the judgment standard according to the actual running condition of the vehicle, and the problems of slow early warning response and low anomaly recognition precision caused by excessive dependence on a fixed line sensor and a single data source are effectively solved.

Description

Technical Field

[0001] The present invention relates to the field of subway early warning technology, and in particular to a train early warning system based on vehicle health monitoring. Background Art

[0002] With the acceleration of urbanization, subways, as a core component of urban public transportation, are experiencing increasingly complex operating networks, with a growing number of lines and carriages. Operating at high speeds, with large passenger volumes and high density, subway systems face significant safety risks and equipment wear. The health of subway vehicles and tracks is crucial to the overall safety and operational efficiency of the system, especially during long-term, high-frequency operations. Traditional inspection methods are unable to meet the needs of real-time monitoring and early warning of subway vehicles, necessitating the need for more advanced technologies to improve safety and prevent failures.

[0003] Patent document with publication number CN104192174A discloses a train warning system and a train warning method, which system includes: multiple foreign object detection sensors, a foreign object detection processing unit, a lookout server, and at least one intelligent terminal, wherein: multiple foreign object detection sensors are configured to be installed along the railway or in a predetermined section and collect environmental information within their respective monitoring ranges; the foreign object detection processing unit is configured to receive environmental information collected from multiple foreign object detection sensors, and determine whether a foreign object is detected based on the received environmental information, wherein when no foreign object is detected, the collected environmental information can be sent directly to the lookout server at regular intervals; when a foreign object is detected, a detection result is generated and the foreign object is marked at a corresponding position in the detection result to form a foreign object alarm result and the foreign object alarm result is sent to the lookout server; the lookout server is configured to receive position and motion information, manual alarm information and main alarm information from all intelligent terminals. The smart terminal receives and displays the environmental information detected by the foreign object detection sensor within the alarm range and the observation range; when the foreign object detection sensor within the alarm range detects a foreign object, the smart terminal receives and displays the environmental information detected by the foreign object detection sensor within the observation range and the foreign object alarm result within the alarm range, and the position and motion information of other smart terminals.

[0004] It can be seen that the train warning system has the following problems: the system mainly relies on multiple foreign object detection sensors installed along the railway to collect environmental information and judge foreign objects. If the foreign object is in a blind spot or the sensor fails, it is easy to miss the detection, and the coverage and reliability are limited; the alarm range and observation range are static sections set based on the position and motion information of the smart terminal, and do not consider dynamic factors such as changes in train speed and train load differences, and there is a risk of misjudgment or omission; the system only relies on a single source (foreign object detection sensor) for judgment, and does not integrate other key parameters such as the train's own operation data or track status. The accuracy and robustness of the judgment results are not high. Summary of the Invention

[0005] To this end, the present invention provides a train warning system based on vehicle health monitoring, which is used to overcome the problems of slow warning response and low abnormality recognition accuracy in the existing technology due to over-reliance on fixed line sensors and a single data source by fusing train operation status data with on-board sensor information.

[0006] To achieve the above objectives, the present invention provides a train warning system based on vehicle health monitoring, comprising:

[0007] The acquisition module is used to collect real-time data on the left and right surface wind speeds of each monitoring car during subway operation, the curvature of the current track, the thickness of the brake pads, the car current, and the pressure at the connection between adjacent monitoring cars;

[0008] a determination module connected to the acquisition module, configured to determine a number of temporary carriages according to the thickness and a preset wear threshold;

[0009] a screening module, connected to the collection module and the determination module respectively, for screening out a number of carriages of interest based on the curvature, the left surface wind speed, and the right surface wind speed corresponding to each of the temporary carriages;

[0010] a determination module, connected to the acquisition module and the screening module respectively, for determining a number of abnormal carriages based on the total pressure between any two adjacent carriages of interest and a preset abnormality determination threshold;

[0011] an adjustment module, connected to the determination module and the determination module, respectively, for adjusting the preset wear threshold to form an adjusted wear threshold, or adjusting the preset abnormality determination threshold to form an adjusted abnormality determination threshold according to the thickness and the compartment current of each abnormal compartment within a preset adjustment time period;

[0012] An early warning module is connected to the determination module and is used to issue early warnings to all abnormal carriages that are re-determined based on the adjusted wear threshold or the adjusted abnormality determination threshold.

[0013] Furthermore, the determining module includes:

[0014] a thickness change calculation unit, configured to calculate an absolute value of a difference between two thicknesses at adjacent moments within a predetermined time period to form a plurality of thickness change values;

[0015] a variation fluctuation calculation unit connected to the thickness variation calculation unit, for calculating the standard deviation of all the thickness variation values ​​to form a thickness variation fluctuation value;

[0016] A determination unit is connected to the change fluctuation calculation unit and is used to determine that the monitored carriage is the temporary carriage when the thickness change fluctuation value is greater than the preset wear threshold, so as to determine a number of temporary carriages.

[0017] Furthermore, the screening module includes:

[0018] a duration recording unit, for recording a duration during which the curvature is greater than a preset curvature threshold within a preset recording period;

[0019] a curvature fluctuation calculation unit connected to the duration recording unit, for calculating, when the duration is greater than a preset duration threshold, the absolute value of the difference between the curvatures at adjacent moments within the duration to form a plurality of curvature differences, and calculating the standard deviation of all the curvature differences to form a curvature change fluctuation value;

[0020] a wind difference fluctuation calculation unit, configured to calculate an absolute value of a difference between the left surface wind speed and the right surface wind speed to form a wind speed difference value, and calculate a standard deviation of all wind speed differences within the duration to form a wind difference fluctuation value;

[0021] A screening unit is connected to the curvature fluctuation calculation unit and the windage fluctuation calculation unit respectively, and is used to screen out a number of focus carriages according to the curvature change fluctuation value and the windage fluctuation value.

[0022] Furthermore, the screening unit includes:

[0023] A first curve drawing subunit is configured to draw a change curve of all the curvature change fluctuation values ​​within the duration to form a first curve when the wind difference fluctuation value is greater than a preset wind difference fluctuation threshold;

[0024] A second curve drawing subunit is configured to draw a change curve of all the wind difference fluctuation values ​​within the duration to form a second curve when the wind difference fluctuation value is greater than the preset wind difference fluctuation threshold;

[0025] a synchronization degree calculation subunit, connected to the first curve drawing subunit and the second curve drawing subunit respectively, for calculating the cosine similarity between the first curve and the second curve to form a change synchronization degree;

[0026] The screening subunit is connected to the synchronization degree calculation subunit and is used to determine that the temporary car is the car of interest when the change synchronization degree is less than a preset standard synchronization degree, so as to screen out several cars of interest.

[0027] Furthermore, the determination module includes:

[0028] a section marking unit, configured to mark the two focused carriages and all the monitored carriages between the two focused carriages as carriage sections, thereby forming a plurality of carriage sections;

[0029] a pressure fluctuation calculation unit connected to the section marking unit, for calculating the standard deviation of all the pressures in each compartment section to form a plurality of pressure fluctuation values;

[0030] A determination unit is connected to the pressure fluctuation calculation unit and is used to determine a number of abnormal carriages according to the pressure fluctuation value and the preset abnormality determination threshold.

[0031] Furthermore, the determination unit includes:

[0032] a pressure normalization subunit, configured to perform normalization processing on each of the pressure fluctuation values ​​to form a plurality of pressure normalized fluctuation values;

[0033] a fluctuation change calculation subunit, connected to the normalization subunit, for calculating the absolute value of the difference between the pressure fluctuation values ​​in any two adjacent compartment sections to form a plurality of fluctuation change values;

[0034] a dispersion calculation subunit, connected to the fluctuation change calculation subunit, for calculating the standard deviation of all the fluctuation change values ​​to form a fluctuation dispersion;

[0035] The determination subunit is connected to the dispersion calculation subunit and is used to determine that the focus carriages in the corresponding two carriage sections are all abnormal carriages when the fluctuation dispersion is less than the preset abnormality determination threshold, forming a plurality of abnormal carriages.

[0036] Furthermore, the adjustment module includes:

[0037] an adjustment change calculation unit, configured to calculate an absolute value of a difference between two thicknesses at adjacent moments within the preset adjustment time period to form a plurality of adjustment change values;

[0038] an adjustment fluctuation calculation unit connected to the adjustment change calculation unit and configured to calculate a standard deviation of all the adjustment change values ​​to form an adjustment fluctuation value;

[0039] a current fluctuation calculation unit, configured to calculate a standard deviation of the currents of all the carriages within the preset adjustment time period to form a current fluctuation value;

[0040] An adjustment unit is connected to the adjustment fluctuation calculation unit and the current fluctuation calculation unit respectively, and is used to adjust the preset wear threshold according to the adjustment fluctuation value and the current fluctuation value to form the adjustment wear threshold, or adjust the preset abnormality determination threshold to form the adjustment abnormality determination threshold.

[0041] Furthermore, the adjustment unit includes:

[0042] a fluctuation normalization subunit, configured to normalize the adjusted fluctuation value to form an adjusted normalized fluctuation value, and to normalize the current fluctuation value to form a current normalized fluctuation value;

[0043] a coefficient calculation subunit, connected to the fluctuation normalization subunit, for calculating a correlation coefficient between the adjusted normalized fluctuation value and the current normalized fluctuation value;

[0044] An adjustment subunit is connected to the coefficient calculation subunit, and is used to adjust the preset wear threshold according to the correlation coefficient and the preset correlation range to form the adjusted wear threshold, or to adjust the preset abnormality judgment threshold according to the correlation coefficient, the preset correlation range, the adjusted normalized fluctuation value and the current normalized fluctuation value to form the adjusted abnormality judgment threshold.

[0045] Furthermore, the adjusting subunit is configured to adjust the preset wear threshold according to the correlation coefficient and the preset correlation range, and adjusting the wear threshold includes:

[0046] The adjustment subunit is used to determine that the wear is natural wear when the correlation coefficient is greater than the maximum value of the preset correlation range, and to reduce the preset wear threshold according to the relative deviation between the correlation coefficient and the maximum value of the preset correlation range to form the adjusted wear threshold.

[0047] Furthermore, the adjustment subunit is configured to adjust the preset abnormality determination threshold according to the correlation coefficient, the preset correlation range, the adjusted normalized fluctuation value, and the current normalized fluctuation value, and forming the adjusted abnormality determination threshold includes:

[0048] The adjustment subunit is configured to calculate, when the correlation coefficient is less than the minimum value of the preset correlation range, the current normalized fluctuation value is less than the preset current normalized fluctuation threshold, and the adjusted normalized fluctuation value is less than the preset adjusted normalized fluctuation threshold, a product of a relative deviation between the preset current normalized fluctuation threshold and the current normalized fluctuation value and a preset current weight to obtain a current product, and to calculate a product of a relative deviation between the preset adjusted normalized fluctuation threshold and the adjusted normalized fluctuation value and a preset adjustment weight to obtain an adjusted product;

[0049] The adjustment subunit is further configured to calculate the sum of the current product and the adjustment product to obtain an adjustment factor;

[0050] The integral subunit is further configured to increase the preset abnormality determination threshold according to the adjustment factor to form the adjusted abnormality determination threshold.

[0051] Compared with the existing technology, the beneficial effect of the present invention is that it effectively improves the recognition accuracy of abnormal carriages through multi-parameter cross-monitoring and layer-by-layer screening and judgment. The introduction of wind speed and curvature data can eliminate non-structural wear deviations caused by natural factors such as turning and headwind, thereby improving the targeted screening; pressure data is normalized and subjected to fluctuation dispersion analysis to ensure the stability and adaptability of the judgment criteria; thickness and current data are used to dynamically adjust the threshold, enabling the system to adaptively optimize the judgment criteria according to the actual operating conditions of the vehicle, avoiding false alarms or missed alarms due to the different speeds of wear changes in individual carriages, improving the system's perception depth and response accuracy of the train's health status, and effectively solving the problems of slow warning response and low abnormality recognition accuracy caused by over-reliance on fixed sensors along the line and a single data source.

[0052] Furthermore, using the thickness change fluctuation value as an indicator to reflect the uneven wear and abnormal wear trend of the brake pad can eliminate the accidental fluctuations caused by single measurement errors and reflect the long-term wear status through statistical standard deviation. The logical correlation is that a large standard deviation of continuous thickness changes means that the wear rate is unstable, which may indicate abnormal wear of the brake pad; therefore, comparing this indicator with the preset wear threshold can help to accurately identify carriages with abnormal wear, improve the accuracy and reliability of the early warning system, and then optimize maintenance strategies and ensure safe train operation.

[0053] Furthermore, by calculating the standard deviation of the fluctuations of the curvature and wind speed difference through statistical methods, it is possible to objectively quantify the local state fluctuations caused by track curvature or uneven wind speed during the dynamic operation of the carriage, thereby distinguishing between natural fluctuations and abnormal conditions. It can not only filter out misjudgments caused by short-term noise or instantaneous fluctuations, but also highlight the statistical signals caused by continuous abnormalities, thereby accurately identifying potential risk carriages, improving the accuracy and reliability of the early warning system, and providing data support for subsequent threshold adjustment and fault repair.

[0054] Furthermore, by plotting the curvature and wind difference fluctuations separately in the time series and performing cosine similarity calculations, a quantitative analysis of the synchronization of the changing trends of the two key dynamic parameters is achieved, which can effectively distinguish between parameter differences caused by environmental factors or normal operation fluctuations and asynchrony under true abnormal conditions; the logical correlation between such data ensures that attention will only be triggered when the fluctuation trends of the two parameters are obviously inconsistent, thereby improving the accuracy of screening, reducing the misjudgment rate, and providing a reliable basis for the subsequent system to dynamically adjust the abnormal judgment threshold and accurate early warning.

[0055] Furthermore, by statistically processing the standard deviation of pressure data within the car section, the degree of discreteness of pressure fluctuations at the connections between cars can be objectively quantified, so that structural instability or abnormal stress problems reflected by local pressure anomalies can be identified using preset thresholds, ensuring that abnormal judgments are made only when pressure fluctuations deviate significantly from normal levels, effectively reducing the risk of false alarms caused by short-term fluctuations or noise, and at the same time improving the sensitivity and accuracy of the early warning system to potential faults.

[0056] Furthermore, through normalization processing and subsequent difference calculation, the consistency of pressure fluctuations in adjacent car sections is quantified, and the standard deviation is used to reflect the degree of dispersion of the fluctuations, thereby ensuring that only when the pressure fluctuations are extremely consistent (that is, the dispersion is low) are they judged as abnormal. Such a design helps to filter out errors caused by short-term random fluctuations or measurement noise, and enhance the sensitivity and accuracy of the early warning system to real structural abnormalities or uneven force problems. At the same time, it is reflected in the data logic that when the dispersion of pressure fluctuations is lower than the preset abnormality judgment threshold, it indicates that the pressure changes in each section tend to be consistent, and there may be an overall abnormality risk, thereby providing a reliable basis for timely operation and maintenance.

[0057] Furthermore, through statistical analysis of thickness changes and car current data, using standard deviation as a fluctuation indicator, the dynamic trends of brake pad wear and current load changes can be accurately reflected, making the dynamic adjustment of the threshold more in line with the actual operating status. In the face of long-term wear trends or sudden status changes, the adaptability and judgment accuracy of the early warning system are improved, the risk of false alarms is reduced, and at the same time, maintenance and safety measures can be implemented at the best time. The fluctuation of thickness and the current fluctuation value jointly reflect the relationship between equipment load and wear status. When the fluctuation value is large, it means that the load change has a greater impact on wear, and the threshold needs to be adjusted to match the current operating conditions.

[0058] Furthermore, through normalization processing and correlation coefficient calculation, the intrinsic correlation between the thickness change fluctuation and the car current fluctuation is quantified. When the correlation between the two is strong, it means that the wear change is significantly affected by the load. The system adjusts the preset threshold to more accurately reflect the actual working conditions, thereby improving the accuracy of abnormal identification and the adaptability of the system. The thickness fluctuation and current fluctuation data are considered together, and the correlation between the data is fully utilized based on the preset relevant range, which not only reduces the risk of false alarms, but also ensures that the early warning can be triggered in a timely and accurate manner under critical conditions.

[0059] Furthermore, by identifying the high correlation between wear and current, it is possible to accurately identify natural wear situations, and on this basis reduce the threshold for wear judgment, increase the system's sensitivity to natural wear, prevent misjudgment as abnormal conditions, and make the system threshold dynamically adapt to the actual operating status, thereby improving the accuracy and intelligence level of wear identification.

[0060] Furthermore, when data fluctuations are small and correlations are weak, the tolerance for anomaly judgment is automatically increased to avoid misjudging slight fluctuations or occasional differences as anomalies, effectively reducing the false alarm rate. By fusing multi-dimensional normalized fluctuations with their relative deviations and introducing weight design, it can ensure that the degree of influence of each indicator on anomaly judgment matches the actual operating status, thereby achieving a more robust and accurate anomaly identification strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 Schematic diagram of a train warning system based on vehicle health monitoring according to this embodiment;

[0062] Figure 2 This is a decision logic diagram for determining a temporary carriage by the determination unit of this embodiment;

[0063] Figure 3 A logic decision diagram for selecting the carriage of interest by the selection subunit of this embodiment;

[0064] Figure 4 This is the logic determination diagram for the determination subunit of this embodiment to determine an abnormal carriage. DETAILED DESCRIPTION

[0065] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention.

[0066] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0067] See also Figure 1As shown, it is a schematic diagram of the train warning system based on vehicle health monitoring in this embodiment;

[0068] This embodiment provides a train warning system based on vehicle health monitoring, including:

[0069] The acquisition module is used to collect real-time data on the left and right surface wind speeds of each monitoring car during subway operation, the curvature of the current track, the thickness of the brake pads, the car current, and the pressure at the connection between adjacent monitoring cars;

[0070] a determination module connected to the acquisition module, configured to determine a number of temporary carriages according to the thickness and a preset wear threshold;

[0071] a screening module, connected to the collection module and the determination module respectively, for screening out a number of carriages of interest based on the curvature, the left surface wind speed, and the right surface wind speed corresponding to each of the temporary carriages;

[0072] a determination module, connected to the acquisition module and the screening module respectively, for determining a number of abnormal carriages based on the total pressure between any two adjacent carriages of interest and a preset abnormality determination threshold;

[0073] an adjustment module, connected to the determination module and the determination module, respectively, for adjusting the preset wear threshold to form an adjusted wear threshold, or adjusting the preset abnormality determination threshold to form an adjusted abnormality determination threshold according to the thickness and the compartment current of each abnormal compartment within a preset adjustment time period;

[0074] An early warning module is connected to the determination module and is used to issue early warnings to all abnormal carriages that are re-determined based on the adjusted wear threshold or the adjusted abnormality determination threshold.

[0075] Any two adjacent focus carriages refer to a single focus carriage and another focus carriage closest to it, and these two focus carriages may be connected to each other or not.

[0076] The acquisition module includes multiple sensor units installed in each monitoring car of the train, specifically including: wind speed sensors installed on the outer surfaces of the left and right sides of the car, respectively, for real-time collection of the left and right surface wind speeds generated by the interaction of the car body with the external airflow during operation; a curvature sensor installed at the bottom of the train, for obtaining the curvature information of the track corresponding to the current position of the train; a thickness sensor installed in the brake system, for detecting the real-time remaining thickness of the brake pad; a current sensor installed in the car circuit system, for collecting the current value when the car is running; and a pressure sensor arranged on the connecting device of adjacent cars, for obtaining the connection pressure value between adjacent monitoring cars.

[0077] The early warning module automatically generates corresponding early warning signals by receiving all abnormal carriage information re-determined by the judgment module based on the adjusted wear threshold or abnormal judgment threshold, and issues alarms in a multi-channel form through the on-board display terminal, the dispatching center monitoring system and the operation and maintenance platform. Specifically, it displays the specific abnormal carriage number and abnormality type on the on-board terminal, pops up an alarm prompt window in the dispatching center and simultaneously pushes it to the maintenance personnel's mobile device, and records the time of the abnormality, carriage number and corresponding data parameters to ensure the visualization of train operation safety risks, traceability of information and timely operation and maintenance response.

[0078] The preset wear threshold refers to the minimum thickness limit used to determine whether the brake pad has abnormal wear. It depends on the material of the brake pad, service life standards, braking frequency and vehicle operating environment, and is usually set between 8mm and 12mm. In this embodiment, the preset wear threshold is set to 10mm. This can not only identify risks in time before the brake pad enters the accelerated wear stage, but also reduce false alarms caused by normal wear of the equipment, ensuring that the determined data fluctuations reflect the actual abnormal wear trend.

[0079] The preset abnormality judgment threshold is the standard deviation limit value used to determine whether the fluctuation dispersion is abnormal. It depends on the historical standard deviation statistics of the pressure changes at the connections between adjacent carriages during normal train operation. It is generally set between 1.0 and 3.0. In this embodiment, it is set to 1.8, which can effectively identify abnormal fluctuations caused by loose connections or structural deviations, while avoiding false alarms caused by normal curvature changes.

[0080] The preset adjustment time is the time window used by the system to observe the changing trends of key parameters of abnormal carriages. It depends on a comprehensive evaluation of factors such as the train speed, maintenance response cycle, and parameter change rate. It is generally set between 5 and 30 minutes. In this embodiment, it is set to 15 minutes, which can take into account both response timeliness and data stability, and is conducive to dynamically adjusting the threshold to achieve more reasonable early warning judgment.

[0081] The acquisition module captures key operational data in real time, including wind speed on the left and right sides of each car, track curvature, brake pad thickness, car current, and joint pressure. The system first identifies temporary cars based on brake pad thickness and preset wear thresholds. It then combines wind speed differences with track curvature to identify cars of concern that may be experiencing abnormal force or operating posture. Next, the judgment module analyzes pressure changes at the joints between cars of concern and compares them with preset abnormality thresholds to identify abnormal cars. After identifying abnormal cars, the adjustment module tracks the thickness and current trends within a preset adjustment period, dynamically adjusting the wear threshold or abnormality threshold to better reflect actual operating conditions. Finally, the early warning module reassesses all cars of concern based on the adjusted thresholds and issues early warning signals for cars with true abnormalities.

[0082] Through multi-parameter cross-monitoring and layer-by-layer screening and judgment, the accuracy of identifying abnormal carriages has been effectively improved. The introduction of wind speed and curvature data can eliminate non-structural wear deviations caused by natural factors such as turning and headwind, improving the targetedness of screening. Pressure data is normalized and subjected to fluctuation dispersion analysis to ensure the stability and adaptability of the judgment criteria. Thickness and current data are used to dynamically adjust the threshold, enabling the system to adaptively optimize the judgment criteria based on the actual operating conditions of the vehicle, avoiding false alarms or missed alarms due to the varying speeds of wear changes in individual carriages. This improves the system's perception depth and response accuracy of the train's health status, effectively solving the problems of slow warning response and low anomaly identification accuracy caused by over-reliance on fixed sensors along the line and a single data source.

[0083] Please continue reading Figure 2 As shown, it is a decision logic diagram of the determination unit of this embodiment for determining a temporary carriage;

[0084] The determination module includes:

[0085] a thickness change calculation unit, configured to calculate an absolute value of a difference between two thicknesses at adjacent moments within a predetermined time period to form a plurality of thickness change values;

[0086] a variation fluctuation calculation unit connected to the thickness variation calculation unit, for calculating the standard deviation of all the thickness variation values ​​to form a thickness variation fluctuation value;

[0087] A determination unit is connected to the change fluctuation calculation unit and is used to determine that the monitored carriage is the temporary carriage when the thickness change fluctuation value is greater than the preset wear threshold, so as to determine a number of temporary carriages.

[0088] The preset determination time refers to the time window used to collect and count the brake pad thickness change data during the initial screening stage, which depends on the train running speed, braking cycle and wear change rate, and is usually set between 5 minutes and 30 minutes; in this embodiment, the preset determination time is set to 10 minutes, which is sufficient to capture the continuous change trend of the brake pad thickness, ensure the stability of the statistical data, and timely reflect the real-time wear status.

[0089] First, the thickness change calculation unit is used to calculate the absolute value of the difference in brake pad thickness between two consecutive moments within a preset time period, thereby forming a series of thickness change values; next, the change fluctuation calculation unit is used to calculate the standard deviation of these thickness change values ​​to obtain a thickness change fluctuation value; finally, the determination unit determines that when the thickness change fluctuation value exceeds the preset wear threshold based on the comparison result of this thickness change fluctuation value and the preset wear threshold, the monitored car is determined to be a temporary car, thereby screening out the cars that need special attention.

[0090] Using the thickness change fluctuation value as an indicator to reflect the uneven wear and abnormal wear trend of the brake pad can eliminate the accidental fluctuations caused by single measurement errors and reflect the long-term wear status through statistical standard deviation. The logical correlation is that a large standard deviation of continuous thickness changes means that the wear rate is unstable, which may indicate abnormal wear of the brake pad; therefore, comparing this indicator with the preset wear threshold helps to accurately identify carriages with abnormal wear, improve the accuracy and reliability of the early warning system, and then optimize maintenance strategies and ensure safe train operation.

[0091] Specifically, the screening module includes:

[0092] a duration recording unit, for recording a duration during which the curvature is greater than a preset curvature threshold within a preset recording period;

[0093] a curvature fluctuation calculation unit connected to the duration recording unit, for calculating, when the duration is greater than a preset duration threshold, the absolute value of the difference between the curvatures at adjacent moments within the duration to form a plurality of curvature differences, and calculating the standard deviation of all the curvature differences to form a curvature change fluctuation value;

[0094] a wind difference fluctuation calculation unit, configured to calculate an absolute value of a difference between the left surface wind speed and the right surface wind speed to form a wind speed difference value, and calculate a standard deviation of all wind speed differences within the duration to form a wind difference fluctuation value;

[0095] A screening unit is connected to the curvature fluctuation calculation unit and the windage fluctuation calculation unit respectively, and is used to screen out a number of focus carriages according to the curvature change fluctuation value and the windage fluctuation value.

[0096] The preset recording period refers to a fixed time window used to continuously collect and count curvature and wind speed data. It depends on the train speed, data collection frequency, and the system's requirements for capturing dynamic changes. It is usually set between 5 minutes and 30 minutes. In this embodiment, it is set to 10 minutes, which can not only cover sufficient dynamic change data, but also maintain the real-time and representativeness of data statistics, thereby ensuring the accuracy of subsequent fluctuation calculations.

[0097] The preset curvature threshold is used to determine whether the current track curvature is abnormal or within the critical value that needs to be monitored. It depends on the track design standards, vehicle dynamic behavior and the curvature variation range during normal operation. It is usually set at 0.1m. -1 Up to 0.3m -1 In this embodiment, it is set to 0.2m -1 , which can better distinguish normal curvature changes from track curves that may cause abnormal stress on the carriage, thereby improving the accuracy of screening carriages of concern.

[0098] The preset duration threshold refers to the shortest time requirement when the track curvature continuously exceeds the preset curvature threshold. It depends on the dynamic change characteristics of the track curvature, the train operating environment, and the system's response speed to abnormal judgment. It is usually set between 2 minutes and 5 minutes. In this embodiment, it is set to 3 minutes, which is sufficient to eliminate misjudgments caused by occasional data fluctuations and can capture real abnormal situations in a timely manner, thereby improving the sensitivity and reliability of the system.

[0099] First, the duration recording unit records the duration of time during which the curvature value of the car exceeds the preset curvature threshold within a preset recording period; when the duration exceeds the preset duration threshold, the curvature fluctuation calculation unit takes the absolute value of the difference between the curvature data of adjacent moments within this period, and calculates the standard deviation of these differences, that is, forms the curvature change fluctuation value; at the same time, the wind difference fluctuation calculation unit takes the absolute value of the difference between the wind speed on the left surface and the wind speed on the right surface of the car, and calculates its standard deviation within the same time window to form the wind difference fluctuation value; the screening unit then screens the cars that meet the abnormalities or potential risks as cars of interest based on the calculated curvature change fluctuation value and wind difference fluctuation value for further monitoring and judgment.

[0100] By calculating the standard deviation of the fluctuations in curvature and wind speed differences through statistical methods, it is possible to objectively quantify the local state fluctuations caused by track curvature or uneven wind speed during the dynamic operation of the carriage, thereby distinguishing between natural fluctuations and abnormal conditions. It can not only filter out misjudgments caused by short-term noise or instantaneous fluctuations, but also highlight the statistical signals caused by continuous abnormalities, thereby accurately identifying potentially risky carriages, improving the accuracy and reliability of the early warning system, and providing data support for subsequent threshold adjustments and fault repairs.

[0101] Please continue reading Figure 3As shown, it is a logic decision diagram of the screening subunit screening the focused carriages in this embodiment;

[0102] The screening unit comprises:

[0103] A first curve drawing subunit is configured to draw a change curve of all the curvature change fluctuation values ​​within the duration to form a first curve when the wind difference fluctuation value is greater than a preset wind difference fluctuation threshold;

[0104] A second curve drawing subunit is configured to draw a change curve of all the wind difference fluctuation values ​​within the duration to form a second curve when the wind difference fluctuation value is greater than the preset wind difference fluctuation threshold;

[0105] a synchronization degree calculation subunit, connected to the first curve drawing subunit and the second curve drawing subunit respectively, for calculating the cosine similarity between the first curve and the second curve to form a change synchronization degree;

[0106] The screening subunit is connected to the synchronization degree calculation subunit and is used to determine that the temporary car is the car of interest when the change synchronization degree is less than a preset standard synchronization degree, so as to screen out several cars of interest.

[0107] The preset wind difference fluctuation threshold is the standard limit for determining whether the fluctuation of the surface wind speed difference on both sides of the car is abnormal. It depends on the natural fluctuation range of the wind speed on the left and right sides during normal operation of the train and the environmental aerodynamic characteristics. It is usually set between 0.5m / s and 2.0m / s. In this embodiment, it is set to 1.2m / s. It can effectively distinguish between normal wind speed fluctuations caused by environmental factors and abnormal fluctuations caused by abnormal car structure or abnormal operating status, thereby providing a reliable basis for the screening of cars of concern.

[0108] The preset standard synchronization degree refers to the cosine similarity threshold used to measure the similarity between the curvature change fluctuation curve and the wind difference fluctuation curve. It depends on the system's sensitivity requirements for the synchronous change trend and is usually set between 0.85 and 0.95. In this embodiment, the preset standard synchronization degree is set to 0.9, which helps to accurately screen the carriages of interest while ensuring that the change trends are obviously different, avoid misjudging normal carriages due to weak disturbances, and improve the accuracy and stability of screening.

[0109] First, the first curve drawing subunit detects when the wind speed difference fluctuation value exceeds the preset wind speed difference fluctuation threshold. It then plots all curvature change fluctuation values ​​collected within the preset recording period in a time series as a first curve. Simultaneously, the second curve drawing subunit plots all wind speed difference fluctuation values ​​collected within the same time period as a second curve. Next, the synchronization calculation subunit calculates the cosine similarity of these two curves to obtain a change synchronization index. Finally, the screening subunit compares this change synchronization index with a preset standard synchronization index. If the change synchronization index falls below the preset standard synchronization index, the corresponding temporary carriage is identified as a carriage of concern, thereby screening several carriages of concern with potential abnormal risks for subsequent accurate judgment and early warning.

[0110] By plotting the curvature and wind difference fluctuations separately in the time series and calculating the cosine similarity, a quantitative analysis of the synchronization of the changing trends of the two key dynamic parameters is achieved, which can effectively distinguish between parameter differences caused by environmental factors or normal operation fluctuations and asynchrony under true abnormal conditions; the logical correlation between the data ensures that attention will only be triggered when the fluctuation trends of the two parameters are obviously inconsistent, thereby improving the accuracy of screening, reducing the misjudgment rate, and providing a reliable basis for the subsequent system to dynamically adjust the abnormal judgment threshold and accurate early warning.

[0111] Specifically, the determination module includes:

[0112] a section marking unit, configured to mark the two focused carriages and all the monitored carriages between the two focused carriages as carriage sections, thereby forming a plurality of carriage sections;

[0113] a pressure fluctuation calculation unit connected to the section marking unit, for calculating the standard deviation of all the pressures in each compartment section to form a plurality of pressure fluctuation values;

[0114] A determination unit is connected to the pressure fluctuation calculation unit and is used to determine a number of abnormal carriages according to the pressure fluctuation value and the preset abnormality determination threshold.

[0115] First, the section marking unit divides the two focus carriages and all monitored carriages between them into several continuous carriage sections; then, the pressure fluctuation calculation unit counts the pressure data collected at the connection points in each carriage section, calculates the standard deviation of these pressure values, and forms the corresponding pressure fluctuation value; then, the judgment unit compares the pressure fluctuation value of each section with the preset abnormal judgment threshold, and determines the abnormal carriage based on whether it exceeds the threshold.

[0116] By statistically processing the standard deviation of pressure data within the carriage section, the degree of discreteness of pressure fluctuations at the connections between carriages can be objectively quantified, thereby using preset thresholds to identify structural instability or abnormal stress problems reflected by local pressure anomalies, ensuring that abnormal judgments are made only when pressure fluctuations significantly deviate from normal levels, effectively reducing the risk of false alarms caused by short-term fluctuations or noise, and at the same time improving the sensitivity and accuracy of the early warning system to potential faults.

[0117] Please continue reading Figure 4 As shown, it is a logic determination diagram of the determination subunit of this embodiment for determining an abnormal carriage;

[0118] The determination unit includes:

[0119] a pressure normalization subunit, configured to perform normalization processing on each of the pressure fluctuation values ​​to form a plurality of pressure normalized fluctuation values;

[0120] a fluctuation change calculation subunit, connected to the normalization subunit, for calculating the absolute value of the difference between the pressure fluctuation values ​​in any two adjacent compartment sections to form a plurality of fluctuation change values;

[0121] a dispersion calculation subunit, connected to the fluctuation change calculation subunit, for calculating the standard deviation of all the fluctuation change values ​​to form a fluctuation dispersion;

[0122] The determination subunit is connected to the dispersion calculation subunit and is used to determine that the focus carriages in the corresponding two carriage sections are all abnormal carriages when the fluctuation dispersion is less than the preset abnormality determination threshold, forming a plurality of abnormal carriages.

[0123] First, the pressure normalization subunit normalizes the pressure fluctuation values ​​collected in each car section to generate standardized pressure normalized fluctuation values; then, the fluctuation change calculation subunit calculates the absolute value of the difference between the pressure fluctuation values ​​in any two adjacent car sections to form a set of fluctuation change values; then, the dispersion calculation subunit performs statistical processing on these fluctuation change values, calculates their standard deviation, and forms an overall fluctuation dispersion; finally, the judgment subunit compares the calculated fluctuation dispersion with the preset abnormality judgment threshold. When the dispersion is lower than the threshold, it is determined that the cars of interest in the corresponding car section are all abnormal cars, thereby achieving accurate abnormality judgment.

[0124] Through normalization processing and subsequent difference calculation, the consistency of pressure fluctuations in adjacent car sections is quantified, and the standard deviation is used to reflect the degree of dispersion of the fluctuations, thereby ensuring that abnormalities are only judged when the pressure fluctuations are extremely consistent (that is, the dispersion is low). This design helps to filter out errors caused by short-term random fluctuations or measurement noise, and enhance the sensitivity and accuracy of the early warning system to real structural abnormalities or uneven force problems. At the same time, it is reflected in the data logic that when the dispersion of pressure fluctuations is lower than the preset abnormality judgment threshold, it indicates that the pressure changes in each section tend to be consistent, and there may be an overall abnormality risk, thus providing a reliable basis for timely operation and maintenance.

[0125] Specifically, the adjustment module includes:

[0126] an adjustment change calculation unit, configured to calculate an absolute value of a difference between two thicknesses at adjacent moments within the preset adjustment time period to form a plurality of adjustment change values;

[0127] an adjustment fluctuation calculation unit connected to the adjustment change calculation unit and configured to calculate a standard deviation of all the adjustment change values ​​to form an adjustment fluctuation value;

[0128] a current fluctuation calculation unit, configured to calculate a standard deviation of the currents of all the carriages within the preset adjustment time period to form a current fluctuation value;

[0129] An adjustment unit is connected to the adjustment fluctuation calculation unit and the current fluctuation calculation unit respectively, and is used to adjust the preset wear threshold according to the adjustment fluctuation value and the current fluctuation value to form the adjustment wear threshold, or adjust the preset abnormality determination threshold to form the adjustment abnormality determination threshold.

[0130] In this embodiment, the adjustment module first calculates the absolute value of the thickness difference between adjacent moments of the continuously collected brake pad thickness data within a preset adjustment time through the adjustment change calculation unit to form a series of adjustment change values; then, the adjustment fluctuation calculation unit calculates the standard deviation of these change values ​​to obtain the adjustment fluctuation value, thereby quantifying the volatility of the thickness change; at the same time, the current fluctuation calculation unit calculates the standard deviation of the car current data collected within the same time window to form the current fluctuation value; finally, the adjustment unit takes the adjustment fluctuation value and the current fluctuation value as input, and dynamically corrects the preset wear threshold or the preset abnormality judgment threshold according to the relative change between the two and the preset adjustment weight, to form the adjusted wear threshold and the adjusted abnormality judgment threshold, respectively.

[0131] Through statistical analysis of thickness changes and car current data, and using standard deviation as a fluctuation indicator, the dynamic trends of brake pad wear and current load changes can be accurately reflected, making the dynamic adjustment of the threshold more in line with the actual operating status. In the face of long-term wear trends or sudden status changes, the adaptability and judgment accuracy of the early warning system are improved, the risk of false alarms is reduced, and maintenance and safety measures can be implemented at the best time. The fluctuation of thickness and the current fluctuation value jointly reflect the relationship between equipment load and wear status. When the fluctuation value is large, it means that the load change has a greater impact on wear, and the threshold needs to be adjusted to match the current operating conditions.

[0132] Specifically, the adjustment unit includes:

[0133] a fluctuation normalization subunit, configured to normalize the adjusted fluctuation value to form an adjusted normalized fluctuation value, and to normalize the current fluctuation value to form a current normalized fluctuation value;

[0134] a coefficient calculation subunit, connected to the fluctuation normalization subunit, for calculating a correlation coefficient between the adjusted normalized fluctuation value and the current normalized fluctuation value;

[0135] An adjustment subunit is connected to the coefficient calculation subunit, and is used to adjust the preset wear threshold according to the correlation coefficient and the preset correlation range to form the adjusted wear threshold, or to adjust the preset abnormality judgment threshold according to the correlation coefficient, the preset correlation range, the adjusted normalized fluctuation value and the current normalized fluctuation value to form the adjusted abnormality judgment threshold.

[0136] The preset correlation range is a numerical interval used to judge the strength of the correlation between thickness fluctuations and current fluctuations, which usually depends on the system's sensitivity requirements for the degree of fluctuation linkage and the statistical analysis results of historical data; it is generally set between 0.6 and 0.9 to ensure that the correlation is actually representative and avoid excessive adjustment; in this embodiment, it is set to 0.75 to 0.85, which can not only ensure that the threshold is adjusted in time when the thickness fluctuation is highly correlated with the current fluctuation, but also avoid misadjustment in the case of weak correlation, thereby improving the system adaptability and judgment accuracy.

[0137] The adjustment fluctuation value and the current fluctuation value are normalized respectively by the fluctuation normalization subunit to generate the adjustment normalized fluctuation value and the current normalized fluctuation value respectively; then, the coefficient calculation subunit calculates the correlation coefficient of the two normalized fluctuation values ​​to reflect the correlation strength between the two parameters; finally, the adjustment subunit compares the calculated correlation coefficient with the preset correlation range, and then combines the information of the adjustment normalized fluctuation value and the current normalized fluctuation value to dynamically correct the preset wear threshold to generate the adjusted wear threshold, or adjust the preset abnormality judgment threshold to match the wear and current change status under the current operating conditions.

[0138] Through normalization processing and correlation coefficient calculation, the intrinsic correlation between thickness change fluctuations and car current fluctuations is quantified. When the correlation between the two is strong, it means that the wear change is significantly affected by the load. The system adjusts the preset threshold to more accurately reflect the actual working conditions, thereby improving the accuracy of abnormal identification and the adaptability of the system. The thickness fluctuation and current fluctuation data are considered together, and the correlation between the data is fully utilized based on the preset relevant range, which not only reduces the risk of false alarms, but also ensures that the early warning can be triggered in a timely and accurate manner under critical conditions.

[0139] Specifically, the adjusting subunit is configured to adjust the preset wear threshold according to the correlation coefficient and the preset correlation range, and adjusting the wear threshold includes:

[0140] The adjustment subunit is used to determine that the wear is natural wear when the correlation coefficient is greater than the maximum value of the preset correlation range, and to reduce the preset wear threshold according to the relative deviation between the correlation coefficient and the maximum value of the preset correlation range to form the adjusted wear threshold.

[0141] First, the correlation coefficient between thickness fluctuation and current fluctuation is calculated and compared with a preset correlation range. If the correlation coefficient exceeds the maximum value within the preset correlation range, the system determines that the wear on the vehicle body is primarily natural wear, meaning that the degree of wear is highly consistent with the current fluctuation. At this point, the adjustment subunit appropriately reduces the preset wear threshold based on the relative deviation between the correlation coefficient and the maximum value, resulting in a more refined adjustment of the wear threshold.

[0142] By recognizing the high correlation between wear and current, it is possible to accurately identify natural wear situations, and on this basis reduce the threshold for wear judgment, increase the system's sensitivity to natural wear, prevent misjudgment as abnormal conditions, and make the system threshold dynamically adapt to the actual operating status, thereby improving the accuracy and intelligence level of wear identification.

[0143] Specifically, the adjustment subunit is configured to adjust the preset abnormality determination threshold according to the correlation coefficient, the preset correlation range, the adjusted normalized fluctuation value, and the current normalized fluctuation value, and forming the adjusted abnormality determination threshold includes:

[0144] The adjustment subunit is configured to calculate, when the correlation coefficient is less than the minimum value of the preset correlation range, the current normalized fluctuation value is less than the preset current normalized fluctuation threshold, and the adjusted normalized fluctuation value is less than the preset adjusted normalized fluctuation threshold, a product of a relative deviation between the preset current normalized fluctuation threshold and the current normalized fluctuation value and a preset current weight to obtain a current product, and to calculate a product of a relative deviation between the preset adjusted normalized fluctuation threshold and the adjusted normalized fluctuation value and a preset adjustment weight to obtain an adjusted product;

[0145] The adjustment subunit is further configured to calculate the sum of the current product and the adjustment product to obtain an adjustment factor;

[0146] The integral subunit is further configured to increase the preset abnormality determination threshold according to the adjustment factor to form the adjusted abnormality determination threshold.

[0147] The preset current weight is a coefficient used to measure the importance of current fluctuations in the overall abnormality judgment. It depends on the sensitivity of current fluctuations to wear or system abnormalities and is usually set between 0.3 and 0.6. In this embodiment, it is set to 0.4, which can give a certain influence to current fluctuations but does not over-amplify them. It is suitable for the moderate sensitivity of current fluctuations in minor abnormalities and is conducive to maintaining the stability of the judgment.

[0148] The preset adjustment weight is a coefficient that measures the weight of thickness fluctuations in abnormality judgment. It depends on the ability of thickness changes to reflect wear trends. It is usually set between 0.4 and 0.7. In this embodiment, it is set to 0.6. Taking thickness as a direct indicator of wear, giving it a higher weight can enhance the system's sensitivity to real wear anomalies and ensure that the threshold adjustment logic is more in line with the actual operating status.

[0149] When the correlation coefficient is lower than the minimum value of the preset correlation range, and both the normalized current fluctuation value and the normalized thickness fluctuation value are lower than their respective preset fluctuation thresholds, this indicates that the correlation between wear and current is weak and the overall fluctuation is small, making it difficult for the system to accurately determine anomalies based on the existing fluctuation characteristics. In this case, the adjustment subunit calculates the relative deviation of the current and thickness fluctuations, multiplying them by their respective preset weights to obtain the current product and the adjustment product. These two are added together to obtain the adjustment factor. This adjustment factor is then used to increase the preset anomaly determination threshold, resulting in a more relaxed adjusted anomaly determination threshold.

[0150] When data fluctuations are small and correlations are weak, the tolerance for anomaly judgment is automatically increased to avoid misjudging slight fluctuations or occasional differences as anomalies, effectively reducing the false alarm rate. By fusing multi-dimensional normalized fluctuations with their relative deviations and introducing weight design, it can ensure that the degree of influence of each indicator on anomaly judgment matches the actual operating status, thereby achieving a more robust and accurate anomaly identification strategy.

[0151] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

Claims

1. A train warning system based on vehicle health monitoring, characterized in that: include: The acquisition module is used to collect real-time data on the left and right surface wind speeds of each monitoring car during subway operation, the curvature of the current track, the thickness of the brake pads, the car current, and the pressure at the connection between adjacent monitoring cars; a determination module connected to the acquisition module, configured to determine a number of temporary carriages according to the thickness and a preset wear threshold; a screening module, connected to the collection module and the determination module respectively, for screening out a number of carriages of interest based on the curvature, the left surface wind speed, and the right surface wind speed corresponding to each of the temporary carriages; a determination module, connected to the acquisition module and the screening module respectively, for determining a number of abnormal carriages based on the total pressure between any two adjacent carriages of interest and a preset abnormality determination threshold; an adjustment module, connected to the determination module and the determination module, respectively, for adjusting the preset wear threshold to form an adjusted wear threshold, or adjusting the preset abnormality determination threshold to form an adjusted abnormality determination threshold according to the thickness and the compartment current of each abnormal compartment within a preset adjustment time period; An early warning module is connected to the determination module and is used to issue early warnings to all abnormal carriages that are re-determined based on the adjusted wear threshold or the adjusted abnormality determination threshold.

2. The train warning system based on vehicle health monitoring according to claim 1 is characterized in that: The determination module includes: a thickness change calculation unit, configured to calculate an absolute value of a difference between two thicknesses at adjacent moments within a predetermined time period to form a plurality of thickness change values; a variation fluctuation calculation unit connected to the thickness variation calculation unit, for calculating the standard deviation of all the thickness variation values ​​to form a thickness variation fluctuation value; A determination unit is connected to the change fluctuation calculation unit and is used to determine that the monitored carriage is the temporary carriage when the thickness change fluctuation value is greater than the preset wear threshold, so as to determine a number of temporary carriages.

3. The train warning system based on vehicle health monitoring according to claim 2 is characterized in that: The screening module includes: a duration recording unit, for recording a duration during which the curvature is greater than a preset curvature threshold within a preset recording period; a curvature fluctuation calculation unit connected to the duration recording unit, for calculating, when the duration is greater than a preset duration threshold, the absolute value of the difference between the curvatures at adjacent moments within the duration to form a plurality of curvature differences, and calculating the standard deviation of all the curvature differences to form a curvature change fluctuation value; a wind difference fluctuation calculation unit, configured to calculate an absolute value of a difference between the left surface wind speed and the right surface wind speed to form a wind speed difference value, and calculate a standard deviation of all wind speed differences within the duration to form a wind difference fluctuation value; A screening unit is connected to the curvature fluctuation calculation unit and the windage fluctuation calculation unit respectively, and is used to screen out a number of focus carriages according to the curvature change fluctuation value and the windage fluctuation value.

4. The train warning system based on vehicle health monitoring according to claim 3 is characterized in that: The screening unit comprises: A first curve drawing subunit is configured to draw a change curve of all the curvature change fluctuation values ​​within the duration to form a first curve when the wind difference fluctuation value is greater than a preset wind difference fluctuation threshold; A second curve drawing subunit is configured to draw a change curve of all the wind difference fluctuation values ​​within the duration to form a second curve when the wind difference fluctuation value is greater than the preset wind difference fluctuation threshold; a synchronization degree calculation subunit, connected to the first curve drawing subunit and the second curve drawing subunit respectively, for calculating the cosine similarity between the first curve and the second curve to form a change synchronization degree; The screening subunit is connected to the synchronization degree calculation subunit and is used to determine that the temporary car is the car of interest when the change synchronization degree is less than a preset standard synchronization degree, so as to screen out several cars of interest.

5. The train warning system based on vehicle health monitoring according to claim 4 is characterized in that: The determination module includes: a section marking unit, configured to mark the two focused carriages and all the monitored carriages between the two focused carriages as carriage sections, thereby forming a plurality of carriage sections; a pressure fluctuation calculation unit connected to the section marking unit, for calculating the standard deviation of all the pressures in each compartment section to form a plurality of pressure fluctuation values; A determination unit is connected to the pressure fluctuation calculation unit and is used to determine a number of abnormal carriages according to the pressure fluctuation value and the preset abnormality determination threshold.

6. The train warning system based on vehicle health monitoring according to claim 5 is characterized in that: The determination unit includes: a pressure normalization subunit, configured to perform normalization processing on each of the pressure fluctuation values ​​to form a plurality of pressure normalized fluctuation values; a fluctuation change calculation subunit, connected to the normalization subunit, for calculating the absolute value of the difference between the pressure fluctuation values ​​in any two adjacent compartment sections to form a plurality of fluctuation change values; a dispersion calculation subunit, connected to the fluctuation change calculation subunit, for calculating the standard deviation of all the fluctuation change values ​​to form a fluctuation dispersion; The determination subunit is connected to the dispersion calculation subunit and is used to determine that the focus carriages in the corresponding two carriage sections are all abnormal carriages when the fluctuation dispersion is less than the preset abnormality determination threshold, forming a plurality of abnormal carriages.

7. The train warning system based on vehicle health monitoring according to claim 6 is characterized in that: The adjustment module includes: an adjustment change calculation unit, configured to calculate an absolute value of a difference between two thicknesses at adjacent moments within the preset adjustment time period to form a plurality of adjustment change values; an adjustment fluctuation calculation unit connected to the adjustment change calculation unit and configured to calculate a standard deviation of all the adjustment change values ​​to form an adjustment fluctuation value; a current fluctuation calculation unit, configured to calculate a standard deviation of the currents of all the carriages within the preset adjustment time period to form a current fluctuation value; An adjustment unit is connected to the adjustment fluctuation calculation unit and the current fluctuation calculation unit respectively, and is used to adjust the preset wear threshold according to the adjustment fluctuation value and the current fluctuation value to form the adjustment wear threshold, or adjust the preset abnormality determination threshold to form the adjustment abnormality determination threshold.

8. The train warning system based on vehicle health monitoring according to claim 7 is characterized in that: The adjustment unit includes: a fluctuation normalization subunit, configured to normalize the adjusted fluctuation value to form an adjusted normalized fluctuation value, and to normalize the current fluctuation value to form a current normalized fluctuation value; a coefficient calculation subunit, connected to the fluctuation normalization subunit, for calculating a correlation coefficient between the adjusted normalized fluctuation value and the current normalized fluctuation value; An adjustment subunit is connected to the coefficient calculation subunit, and is used to adjust the preset wear threshold according to the correlation coefficient and the preset correlation range to form the adjusted wear threshold, or to adjust the preset abnormality judgment threshold according to the correlation coefficient, the preset correlation range, the adjusted normalized fluctuation value and the current normalized fluctuation value to form the adjusted abnormality judgment threshold.

9. The train warning system based on vehicle health monitoring according to claim 8 is characterized in that: The adjusting subunit is configured to adjust the preset wear threshold according to the correlation coefficient and the preset correlation range, and the adjustment of the wear threshold includes: The adjustment subunit is used to determine that the wear is natural wear when the correlation coefficient is greater than the maximum value of the preset correlation range, and to reduce the preset wear threshold according to the relative deviation between the correlation coefficient and the maximum value of the preset correlation range to form the adjusted wear threshold.

10. The train warning system based on vehicle health monitoring according to claim 9 is characterized in that: The adjusting subunit is configured to adjust the preset abnormality determination threshold according to the correlation coefficient, the preset correlation range, the adjusted normalized fluctuation value, and the current normalized fluctuation value, and forming the adjusted abnormality determination threshold includes: The adjustment subunit is configured to calculate, when the correlation coefficient is less than the minimum value of the preset correlation range, the current normalized fluctuation value is less than the preset current normalized fluctuation threshold, and the adjusted normalized fluctuation value is less than the preset adjusted normalized fluctuation threshold, a product of a relative deviation between the preset current normalized fluctuation threshold and the current normalized fluctuation value and a preset current weight to obtain a current product, and to calculate a product of a relative deviation between the preset adjusted normalized fluctuation threshold and the adjusted normalized fluctuation value and a preset adjustment weight to obtain an adjusted product; The adjustment subunit is further configured to calculate the sum of the current product and the adjustment product to obtain an adjustment factor; The integral subunit is further configured to increase the preset abnormality determination threshold according to the adjustment factor to form the adjusted abnormality determination threshold.

Images