Railway vehicle roof high-voltage induction alarm device and method

Abstract

The invention relates to the technical field of induction alarm devices, in particular to a rail vehicle roof high-voltage induction alarm device and method, and the method comprises the steps: obtaining the magnetic field intensity of each monitoring position of a roof between adjacent carriages of a rail at each moment; obtaining an asymmetric index according to the difference of the magnetic field intensities between the symmetric monitoring positions at each moment and the distance distribution condition; dividing the same curve time period according to the asymmetric indexes between the symmetric monitoring positions at different moments, and obtaining the symmetric anomaly degree of the same curve time period by combining the time length and the asymmetric indexes in the same curve time period; obtaining an abnormal state time period according to the difference of the asymmetric indexes of the adjacent same curve time periods and the symmetric abnormal degree; and obtaining a gap abnormity early warning result according to the symmetric abnormity degree and the asymmetric index in the abnormal state time period of the current moment. According to the invention, the situation of misjudgment of abnormity is avoided to a certain extent, and the accuracy of abnormity alarm is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of induction alarm devices, and in particular to a high-voltage induction alarm device and method for a rail vehicle roof. Background Art

[0002] With the development of rail transit electrification, high-voltage jumpers on the roofs of rail vehicles always play a key role in power transmission. By transitioning the original high-voltage jumper connection devices, including fixed connection devices, from a complex connection method to a simple connection method, and from an unstable method to a stable electrical connection method.

[0003] However, existing mechanical protection for high-voltage connections (such as insulating shields) lacks the ability to perceive the status of high-voltage jumpers in real time and provide active warnings. For example, some devices rely on physical contact to trigger alarms (such as stretch-type alarm switches), which poses a risk of response lag or false alarms. At the same time, when the train turns, the actual gap between adjacent carriages will stretch or shorten, and the corresponding high-voltage jumper will deform, and the symmetry of the magnetic field will be destroyed, resulting in distortion of the magnetic field distribution. The radial and axial magnetic induction intensity amplitudes will decrease sharply, and the distortion area will expand with the degree of deformation, and the intensity and distribution of the surrounding magnetic field will also change accordingly. These changes cause the alarm device to confuse the actual data changes of the train turning with the actual abnormal situation, resulting in misjudgment of faults, making the alarm device less accurate. Summary of the Invention

[0004] In order to solve the technical problem that the existing alarm device confuses the actual data changes of the train turning with the real abnormal situation, resulting in the situation of fault misjudgment and the low accuracy of the alarm device, the purpose of the present invention is to provide a high-voltage induction alarm device and method for the roof of a railway vehicle. The technical solution adopted is as follows:

[0005] In a first aspect, the present invention provides a rail vehicle roof high-voltage sensing alarm method, comprising:

[0006] Obtaining the magnetic field strength at each monitoring position on the roof between adjacent carriages on the track at each moment when the train is running, wherein the monitoring positions are symmetrically distributed along the jumper between adjacent carriages, and the moments include the current moment and historical moments;

[0007] According to the difference in magnetic field intensity and distance distribution between symmetrical monitoring positions at each moment, the asymmetric index between the symmetrical monitoring positions at each moment is obtained;

[0008] Based on the differences in the distribution of asymmetric indicators between symmetrical monitoring positions along the jumper at different times, the same curve time period is divided. The degree of symmetric anomaly in each same curve time period is obtained by combining the time length within the same curve time period and the overall distribution of the asymmetric indicators at each time moment.

[0009] According to the difference of asymmetric indicators between adjacent time periods of the same curve and the distribution of symmetric abnormality, the time period of the same curve is divided to obtain abnormal state time periods;

[0010] According to the symmetrical abnormality degree and the asymmetrical index in the abnormal state time period at the current moment, an abnormality warning result of the gap between adjacent cars on the track at the current moment is obtained.

[0011] Preferably, obtaining the asymmetric index between the symmetrical monitoring positions at each moment according to the difference in magnetic field strength and the distance distribution between the symmetrical monitoring positions at each moment specifically includes:

[0012] Determine each symmetrical monitoring group based on the extension direction of the jumper between adjacent carriages of the track, wherein the two monitoring positions in the symmetrical monitoring group are symmetrically distributed along the jumper;

[0013] Determining a magnetic field difference factor for each symmetrical monitoring group at each moment based on a difference in magnetic field strength between two monitoring positions within each symmetrical monitoring group at each moment;

[0014] Determine a distance distribution factor for each symmetrical monitoring group based on the distance between any monitoring position in each symmetrical monitoring group and the jumper;

[0015] Based on the product of the magnetic field difference factor and the distance distribution factor of each symmetrical monitoring group at each moment, the asymmetric index of the symmetrical monitoring group at each moment is determined.

[0016] Preferably, dividing the same curve time period according to the difference in distribution of asymmetric indicators between symmetrical monitoring positions distributed along the jumper at different times specifically includes:

[0017] Constructing an asymmetric characteristic curve at each moment according to the distribution of asymmetric indicators of all symmetric monitoring groups distributed along the jumper at each moment;

[0018] Based on the difference between the asymmetric characteristic curves of every two moments, all moments are clustered, and the consecutive moments belonging to the same cluster are regarded as a same-curve time period.

[0019] Preferably, dividing the same curve time period according to the difference in distribution of asymmetric indicators between symmetrical monitoring positions distributed along the jumper at different times specifically includes:

[0020] Constructing an asymmetric characteristic curve at each moment according to the distribution of asymmetric indicators of all symmetric monitoring groups distributed along the jumper at each moment;

[0021] Based on the difference between the asymmetric characteristic curves of every two moments, all moments are clustered, and the consecutive moments belonging to the same cluster are regarded as a same-curve time period.

[0022] Preferably, constructing an asymmetric characteristic curve at each moment according to the distribution of asymmetric indicators of all symmetric monitoring groups distributed along the jumper at each moment specifically includes:

[0023] The straight line where the jumper is located is used as the horizontal axis, and the asymmetric index of each symmetrical monitoring group at each moment is used as the vertical axis to construct the asymmetric characteristic curve at each moment.

[0024] Preferably, the combination of the time length within the same curve time period and the overall distribution of the asymmetry index at each moment to obtain the degree of symmetry abnormality for each same curve time period specifically includes:

[0025] Obtaining the average value of the asymmetric indexes between all monitoring positions at each moment in each same curve time period to determine the asymmetric balance factor for each same curve time period;

[0026] The degree of symmetric abnormality of each same-curve time period is determined based on the product of the negative correlation coefficient of the time length of each same-curve time period and the asymmetric balancing factor.

[0027] Preferably, the method of dividing the same curve time period into abnormal state time periods according to the difference in asymmetric indicators between adjacent same curve time periods and the distribution of symmetric abnormality degrees specifically includes:

[0028] Cluster the same curve time period to obtain clusters contained in different clustering results;

[0029] According to the difference of asymmetric index between adjacent same-curve time periods in each cluster in each clustering result and the degree of symmetric abnormality between adjacent same-curve time periods, the clustering objective function value of each clustering result is constructed;

[0030] The continuous time periods of the same curve in each cluster in the clustering results corresponding to the maximum value of the clustering objective function are regarded as abnormal state time periods.

[0031] Preferably, the clustering objective function value of each clustering result is constructed based on the difference in asymmetry index between adjacent same-curve time periods in each cluster cluster in each clustering result and the degree of symmetry abnormality between adjacent same-curve time periods, specifically including:

[0032] For any clustering result, in each cluster, the mean difference of the asymmetric index corresponding to the same monitoring position between each two adjacent time periods of the same curve is calculated to obtain the first characteristic coefficient corresponding to each two adjacent time periods of the same curve;

[0033] Calculating the product of the symmetric abnormality degree of each two adjacent time periods of the same curve to obtain the second characteristic coefficient corresponding to each two adjacent time periods of the same curve;

[0034] The cumulative result of the first characteristic coefficient and the second characteristic coefficient corresponding to all two adjacent time periods of the same curve in all clusters is determined as the clustering objective function value of any clustering result.

[0035] Preferably, obtaining the abnormal gap warning result between adjacent cars on the track at the current moment based on the symmetric abnormality degree and the asymmetric index in the abnormal state time period at the current moment specifically includes:

[0036] Obtain the first mean of the asymmetry index between all monitoring positions at all times within the abnormal state time period at the current moment, and the second mean of the symmetry abnormality degree of all time periods with the same curve within the abnormal state time period at the current moment; normalize the product of the first mean and the second mean to obtain the abnormality degree of the gap between adjacent carriages on the track at the current moment; determine the gap abnormality warning result based on the abnormality degree of the gap between adjacent carriages on the track at the current moment.

[0037] Preferably, determining the gap abnormality warning result according to the abnormality degree of the gap between adjacent cars on the track at the current moment specifically includes:

[0038] If the abnormality of the gap between adjacent carriages on the track at the current moment is greater than or equal to the preset abnormality threshold, an abnormality warning will be issued; if the abnormality of the gap between adjacent carriages on the track at the current moment is less than the preset abnormality threshold, no abnormality warning will be issued.

[0039] In a second aspect, the present invention provides a high-voltage sensing alarm device for a rail vehicle roof, which is used to implement the steps of a high-voltage sensing alarm method for a rail vehicle roof. The high-voltage sensing alarm device for a rail vehicle roof specifically includes:

[0040] A data acquisition module is used to obtain the magnetic field strength at each monitoring position on the roof of adjacent carriages on the track at each moment when the train is running, wherein the monitoring positions are symmetrically distributed along the jumper between adjacent carriages, and the moments include the current moment and historical moments;

[0041] A first feature analysis module is used to obtain an asymmetric index between the symmetrical monitoring positions at each moment based on the difference in magnetic field strength and distance distribution between the symmetrical monitoring positions at each moment;

[0042] The second feature analysis module is used to divide the same curve time period according to the differences in the distribution of asymmetric indicators between symmetrical monitoring positions distributed along the jumper at different times. The degree of symmetry anomaly in each same curve time period is obtained by combining the time length within the same curve time period and the overall distribution of the asymmetric indicators at each time point.

[0043] The abnormal state judgment module is used to divide the same curve time period into abnormal state time periods based on the difference in asymmetric indicators and the distribution of symmetric abnormality between adjacent same curve time periods;

[0044] The abnormality warning module is used to obtain an abnormality warning result of the gap between adjacent cars on the track at the current moment according to the symmetrical abnormality degree and the asymmetrical index in the abnormal state time period at the current moment.

[0045] The embodiments of the present invention have at least the following beneficial effects:

[0046] The present invention first sets monitoring positions around the high-voltage jumper between adjacent carriages to monitor the changes in magnetic field intensity in real time, providing a data basis for the subsequent analysis of the symmetry characteristics of the magnetic field intensity. Then, through the difference in magnetic field intensity and the distance distribution between the symmetrically distributed monitoring positions, a preliminary analysis is made of the asymmetry of the magnetic field intensity between the symmetrically distributed monitoring positions around the high-voltage jumper, which is also the asymmetry index. Secondly, the asymmetry index is used to divide the time period. Through the asymmetry characteristics shown in the continuous time period, the time points that may appear at the same inner diameter bend angle are divided into a time period, and the degree of abnormality is quantified. The degree of symmetrical abnormality characterizes the actual degree of abnormality, such as the gap that exists when the original symmetry characteristics are destroyed. Furthermore, by further dividing the time period of the same curve, the interference of the difference in magnetic field intensity fluctuations caused by different inner diameter bend angles on the abnormal magnetic field intensity fluctuations is avoided. Finally, through the specific abnormal performance in the continuous abnormal time period at the current moment, the situation of misjudging the occasional normal phenomenon of abnormal magnetic field intensity at a single time point or time period as abnormal is reduced, thereby improving the accuracy of abnormal alarm. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0048] Figure 1This is a flowchart of the steps of a high-voltage sensing alarm method for a rail vehicle roof provided by the present invention;

[0049] Figure 2 is a schematic diagram of the distribution of monitoring locations provided by the present invention;

[0050] Figure 3 is a schematic diagram of the connection of high-voltage jumpers between adjacent carriages provided by the present invention;

[0051] Figure 4 This is a flowchart of specific implementation steps for obtaining the asymmetric index between symmetrical monitoring positions at each moment provided by the present invention;

[0052] Figure 5 is a schematic diagram of the distribution of magnetic field intensity along the jumper in space provided by the present invention;

[0053] Figure 6 is a flowchart of the specific implementation sub-steps of step S300 provided by the present invention;

[0054] Figure 7 This is a flowchart of the steps of dividing the same curve time period to obtain the abnormal state time period provided by the present invention;

[0055] Figure 8 This is a schematic diagram of the module structure of a high-voltage induction alarm device on the roof of a rail vehicle provided by the present invention;

[0056] in, Figure 3 The numbers are: A, in the front car, B, in the rear car, C, train travel direction, 1, insulation shield, 2, inductive sensor probe, 3, insulation rubber layer, 4, radiator. DETAILED DESCRIPTION

[0057] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the following, in combination with the accompanying drawings and preferred embodiments, describes in detail a rail vehicle roof high-voltage sensing alarm device and method proposed by the present invention, its specific implementation method, structure, characteristics and effects.

[0058] In the following description, different references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0059] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0060] The specific scheme of the high-voltage sensing alarm device and method for the roof of a railway vehicle provided by the present invention is described in detail below with reference to the accompanying drawings.

[0061] Before introducing the specific solutions provided in the embodiments of the present application, the specific implementation scenarios in the present application are explained to facilitate understanding by those skilled in the art, and are not intended to limit the uses in the present application.

[0062] According to Faraday's law of electromagnetic induction, changes in magnetic flux in a closed loop generate an induced electromotive force at both ends of the coil. In the high-voltage system on top of a rail vehicle, the electromagnetic induction effect is amplified by the strong electric and magnetic fields generated by the high voltage, and is manifested as follows:

[0063] Strong electric field coupling: A high-intensity electrostatic field is formed around the high-voltage conductor (high-voltage jumper), which induces electric field with the adjacent conductor or environmental medium.

[0064] High-frequency magnetic field interference: Rapid changes in high-voltage current (such as switching operations and arc discharges) cause high-frequency magnetic field fluctuations, which may interfere with nearby equipment.

[0065] See also Figure 1 , which shows a flowchart of a method for high-voltage sensing alarm on a rail vehicle roof provided by one embodiment of the present invention, the method comprising the following steps:

[0066] Step S100, obtaining the magnetic field strength of each monitoring position on the roof between adjacent carriages on the track at each moment when the train is running, wherein the monitoring positions are symmetrically distributed along the jumper between adjacent carriages, and the moments include the current moment and historical moments.

[0067] During the running of the train, the high-voltage system on the top of the rail vehicle is monitored in real time. After a period of running, the monitoring data of the current moment and multiple historical moments in the previous period are collected in real time. Considering that around the high-voltage jumper on the top of the car, the farther away from the jumper, the lower the magnetic field strength, and the closer to the jumper, the higher the magnetic field strength, and the magnetic field strength is symmetrically distributed along the jumper during normal straight-line running, different monitoring positions are set symmetrically along the jumper when monitoring the magnetic field strength on the top of the car.

[0068] like Figure 2 As shown, it is a plane diagram parallel to the ground. The straight lines in the figure are high-voltage jumpers between cars. Each rectangle is located at a different monitoring position. The intervals between adjacent monitoring positions are equal, so that the monitoring positions are distributed in a grid shape, and the position distribution is symmetrical. An inductive sensor is set at each monitoring position to collect the magnetic field strength of the corresponding position in real time.

[0069] In some embodiments, by installing inductive sensors around high-voltage jumpers, the spacing between the sensors and the high-voltage jumper assembly is designed in combination with the magnetic field strength attenuation characteristics. For example, for a 25kV high-voltage system, the sensor installation distance is typically 50-200mm. The specific distance setting also needs to be optimized through finite element simulation. Implementers need to determine it based on the specific implementation scenario. At the same time, multiple sensors are arranged at key nodes of the jumper assembly (such as connections and bends) to form a distributed monitoring network that covers electromagnetic field blind spots. The change in the inductive reactance of the iron core coil is used to sense the approach of a magnetic object or a change in the gap, which is suitable for insulation gap monitoring of high-voltage jumper assemblies.

[0070] like Figure 3 As shown, Figure 3 The front car A and the rear car B are adjacent cars in motion, and an insulating shielding cover 1 is fixed on the top of the high-voltage jumper on the top of the two adjacent cars to reduce the wind resistance during high-speed driving of the rail vehicle and isolate the influence of external interference or media; an insulating rubber layer 3 is installed directly above the front car A and the rear car B to adapt to the changes in the gap between the cars during the train travel, and a plurality of inductive sensor probes 2 are installed on the insulating rubber layer 3 according to the bending shape of the high-voltage jumper to monitor the magnetic field changes to determine the conductor offset; at the same time, since the maximum magnetic flux change is perpendicular to the direction of the jumper, the same gap installation method is used in the actual installation process, so that the inductive sensor probes are installed in a grid shape, so that the straight line formed by any vertical column of sensor probes is perpendicular to the direction of the jumper, which can still monitor the maximum magnetic flux change; at the same time, a radiator 4 is installed at the rear of the streamlined baffle to reduce the air temperature inside the space to avoid failures caused by high temperature.

[0071] It should be noted that Figure 2 and Figure 3 This is merely a schematic diagram of one possible implementation. In some embodiments, the monitoring locations can be distributed in a grid pattern of 7*7 or 5*5, with symmetry along the high-voltage jumper. In some embodiments, the high-voltage jumper can employ, for example, a dual-terminal spiral structure, and this embodiment does not impose any specific limitations.

[0072] Step S200: obtaining an asymmetric index between the symmetrical monitoring positions at each moment according to the difference in magnetic field strength and the distance distribution between the symmetrical monitoring positions at each moment.

[0073] As the mileage of trains increases, the lifespan of the corresponding equipment decreases and aging occurs. The aging of the roof high-voltage jumper assembly leads to a decrease in its ability to resist vibration, and the vibration causes abnormal changes in the insulation gap, which may cause arc discharge. Therefore, when conducting safety monitoring of the high-voltage jumpers of rail vehicles, it is necessary to judge the possibility of faults by monitoring the gap changes of the jumpers and to provide early warning through the alarm device.

[0074] Furthermore, when a train turns, the actual gap between adjacent cars stretches or shrinks. This causes the corresponding high-voltage jumpers to deform, disrupting the symmetry of the magnetic field and causing a distortion in the magnetic field distribution. The radial and axial magnetic induction intensities decrease dramatically, and the distorted area expands with the degree of deformation, altering the intensity and distribution of the surrounding magnetic field. These changes can be confused with actual anomalies, leading to misjudgments of faults and reduced alarm accuracy. Based on this characteristic, by analyzing the distance distribution and changes in magnetic field intensity between two monitoring locations exhibiting symmetrical distributions, it is possible to quantify whether the symmetry between these symmetrically distributed monitoring locations has been disrupted.

[0075] Specifically, at each moment, the difference in magnetic field strength between two monitoring positions that present a symmetrical distribution and the distance between the two monitoring positions are obtained respectively to determine the asymmetry index between all monitoring positions that present a symmetrical distribution at each moment, that is, the asymmetry index characterizes the magnitude of the asymmetric characteristics exhibited due to the anomaly between two monitoring positions that actually have a symmetrical distribution.

[0076] Step S300, according to the difference in the distribution of asymmetric indicators between the symmetrical monitoring positions distributed along the jumper at different times, the same curve time period is divided, and the degree of symmetric abnormality of each same curve time period is obtained by combining the time length in the same curve time period and the overall distribution of the asymmetric indicators at each moment.

[0077] Since a certain degree of vibration is generated during the movement of the train, this vibration will cause a small deformation of the jumper wire, which in turn will cause small local fluctuations in the symmetry between the grid-distributed monitoring positions. At this time, the vibration will make it possible to misjudge the actual abnormal gap state of the jumper wire. Therefore, when analyzing the abnormal gap between adjacent carriages, it is necessary to first eliminate the normal fluctuations in symmetry caused by the influence of vibration.

[0078] First, when the symmetrical fluctuations of the monitoring positions occur due to turning operations during train operation, the asymmetric indicators corresponding to all symmetrically distributed monitoring positions along the axial distribution of the high-voltage jumper at each moment can show a certain trend of change, for example, they can be quantified by curve fitting or by constructing a sequence distribution.

[0079] Then, by classifying the differences in the change trends at each moment, moments with the same or similar change trends can be grouped together, thereby determining the same-curve time period. It can be understood that moments within the same-curve time period represent moments with the same or similar symmetrical fluctuations, and a same-curve time period may represent the entire process of a turning operation.

[0080] Finally, for each time period with the same curve, the duration and overall asymmetry distribution are combined to measure the possible symmetric anomalies in each time period with the same curve, that is, the degree of symmetric anomaly. The longer the duration of the time period with the same curve, the more likely it is that the time period belongs to a normal train turning process. The destruction of the symmetric characteristics at this time is a normal phenomenon, rather than a symmetric anomaly with the presence of a gap anomaly. Therefore, the smaller the value of the degree of symmetric anomaly at this time, the duration of the time period with the same curve is negatively correlated with the degree of symmetric anomaly in the time period with the same curve.

[0081] The degree of symmetric anomaly in each same-curve time period represents the possibility of an abnormal situation in which the magnetic field intensity at the symmetrically distributed monitoring positions is destroyed due to the destruction of symmetry at each moment in the corresponding same-curve time period.

[0082] Step S400 : dividing the same-curve time period into abnormal state time periods based on the differences in asymmetric indicators and the distribution of symmetric abnormality degrees between adjacent same-curve time periods.

[0083] The time periods with the same inner diameter bending angle and high-frequency and large-amplitude fluctuations in magnetic field intensity are further divided into a cluster, so as to screen and obtain the local symmetric anomaly probability of each time period including the real-time sampling time point. This operation avoids the interference of the magnetic field intensity fluctuation differences caused by different inner diameter bending angles on the abnormal magnetic field intensity fluctuations, reduces the misjudgment of the occasional normal phenomenon of abnormal magnetic field intensity at a single time point or time period as an anomaly, and improves the accuracy of abnormal alarms.

[0084] Based on this feature, on the basis of preliminarily dividing the same curve time periods with relatively close symmetry characteristics, time periods with longer duration are further divided, and certain division conditions need to be met during the division to distinguish the abnormal symmetry changes around the high-voltage jumper caused by normal curve driving behavior, and the abnormal symmetry changes around the high-voltage jumper caused by the actual fault situation of the driving gap during train driving.

[0085] Specifically, a second clustering is performed on all time periods with the same curve, resulting in multiple different secondary clustering results. The optimal clustering result is then selected from these clustering results to identify the clustering results that meet the required objective function values. The objective function is set to maximize the degree of symmetry between adjacent time periods with the same curve, and to maximize the difference in asymmetry between adjacent time periods with the same curve. By selecting the clustering results with the largest objective function values, the interference of abnormal symmetry characteristics of normal curves can be effectively eliminated, thereby enhancing the expression of abnormal characteristics of the actual abnormal situation.

[0086] Step S500: obtaining an abnormal gap warning result between adjacent cars on the track at the current moment according to the symmetrical abnormality degree and the asymmetrical index in the abnormal state time period at the current moment.

[0087] The results of the second clustering show that the time periods belonging to the same cluster represent consecutive moments with relatively similar symmetric anomaly characteristics. Furthermore, the possible anomalies at the current moment can be quantitatively measured by analyzing the anomaly characteristics of all moments in the cluster where the current moment is located and all time periods with the same curve.

[0088] As a specific example, the first mean of the asymmetry index between all monitoring locations at all times within the abnormal state time period at the current moment is obtained, as well as the second mean of the degree of symmetry abnormality for all time periods within the same curve within the abnormal state time period at the current moment. The product of the first mean and the second mean is normalized to obtain the degree of abnormal clearance between adjacent cars on the track at the current moment. The normalization method is well known and will not be further explained here.

[0089] The degree of gap anomaly indicates the possibility that the abnormal change in magnetic field strength around the high-voltage jumper is caused by gap anomaly at the current moment. The greater the possibility, the greater the possibility of gap anomaly between adjacent vehicles on the track at the current moment; the smaller the possibility, the smaller the possibility of gap anomaly between adjacent vehicles on the track at the current moment.

[0090] Furthermore, the result of the gap abnormality warning is determined based on the current abnormality level between adjacent cars on the track. Specifically, if the current abnormality level between adjacent cars on the track is greater than or equal to a preset abnormality threshold, an abnormality warning is issued; if the current abnormality level between adjacent cars on the track is less than the preset abnormality threshold, no abnormality warning is issued. The abnormality threshold is set to 0.8 and can be adjusted by the implementer based on the specific implementation scenario.

[0091] When the degree of abnormality in the gap between adjacent cars on the track at the current moment is greater than or equal to the abnormality threshold, it indicates that the possibility of a gap abnormality is greater at this time, and an abnormality warning is required. At the same time, an alarm device is used for early warning. When an abnormality occurs during the above monitoring process, the alarm device is triggered and the train driver is reminded through an audible and visual alarm. The abnormal sensor data is also saved to facilitate subsequent abnormality inspections of the high-voltage jumper. When the degree of abnormality in the gap between adjacent cars on the track at the current moment is less than the abnormality threshold, it indicates that there is a certain degree of fluctuation in the magnetic field strength around the high-voltage jumper between the cars at this time, but this is a normal phenomenon and no warning is required. You can continue to pay attention to the driving status of the train.

[0092] The embodiment of the present invention divides adjacent time points that may occur at the same inner diameter bend angle into a time period, thereby avoiding the influence of magnetic field changes with different characteristics generated when the train copes with different bends on subsequent abnormal state judgment, and improving the accuracy of high-voltage jumper abnormal alarm; further divides the time period with the same inner diameter bend angle and the high-frequency and large-amplitude fluctuations of the magnetic field intensity into a cluster, thereby screening and obtaining the local symmetric abnormality probability of each time period including the real-time sampling time point. This operation avoids the interference of the magnetic field intensity fluctuation difference generated by different inner diameter bend angles on the abnormal magnetic field intensity fluctuation, reduces the misjudgment of the occasional normal phenomenon of abnormal magnetic field intensity at a single time point or time period as an abnormality, and improves the accuracy of abnormal alarm.

[0093] In some embodiments, the specific implementation process for obtaining the asymmetric index between the symmetrical monitoring positions at each moment in step S200 is as follows: Figure 4 As shown, it includes steps S201 to S204.

[0094] It should be noted that the magnetic field strength generated by the high-voltage jumper decreases in the vertical direction of the jumper current in space as the distance from the jumper increases; therefore, according to the grid distribution method of the designed inductive sensor, the change of magnetic field strength with distance can be obtained, such as Figure 5 As shown, the dotted line represents the influence range of the magnetic field strength, the black circle represents the vertical section of the high-voltage jumper, and the rectangle represents the inductive sensor; it can be understood that Figure 5 The perspective is a side view of the top of the rail vehicle in the plane where the train head is located.

[0095] Step S201: determining each symmetrical monitoring group based on the extension direction of the jumper between adjacent carriages on the track, wherein the two monitoring positions in the symmetrical monitoring group are symmetrically distributed along the jumper.

[0096] As a specific example, Figure 2As shown, on the extension line of the high-voltage jumper, each column of monitoring positions is perpendicular to the high-voltage jumper, and there are two groups of symmetrically distributed monitoring positions in the same column, that is, Figure 2 The monitoring position in the first row and first column and the monitoring position in the fifth row and first column belong to the same symmetrical monitoring group, and the monitoring position in the second row and first column and the monitoring position in the fourth row and first column belong to the same symmetrical monitoring group.

[0097] Step S202 : determining a magnetic field difference factor of each symmetrical monitoring group at each moment based on the difference in magnetic field strength between two monitoring positions in each symmetrical monitoring group at each moment.

[0098] Specifically, the absolute value of the difference in magnetic field strength between two monitoring positions in each symmetrical monitoring group at each moment is used as the magnetic field difference factor of each symmetrical monitoring group at each moment, reflecting the difference in magnetic field strength between symmetrically distributed monitoring positions at each moment.

[0099] Step S203: determining a distance distribution factor of each symmetrical monitoring group based on the distance between any monitoring position in each symmetrical monitoring group and the jumper.

[0100] Considering that the straight-line distance from the same monitoring location to the high-voltage jumper is equal, the straight-line distance between any monitoring location and the jumper within each symmetrical monitoring group can be directly obtained and normalized to form the distance distribution factor for each symmetrical monitoring group. The distance distribution factor reflects the relative distance between the monitoring locations and the jumper within the symmetrical monitoring group. The normalization method is well known and will not be further explained here.

[0101] Step S204 : determining an asymmetric index of each symmetrical monitoring group at each moment based on the product of the magnetic field difference factor and the distance distribution factor of each symmetrical monitoring group at each moment.

[0102] Taking into account that the farther the monitoring position is from the high-voltage jumper, the greater the degree of change in magnetic field strength is affected by the deformation of the high-voltage jumper, and the closer the monitoring position is to the high-voltage jumper, the smaller the degree of change in magnetic field strength is affected by the deformation of the high-voltage jumper, based on this feature, the relative position performance of each symmetrical monitoring group is used as a weight, and the magnetic field strength change corresponding to each symmetrical monitoring group is weighted to obtain the characteristic performance of the comprehensive change.

[0103] Specifically, considering that multiple symmetrical monitoring groups exist within the same column along the radial direction of the high-voltage jumper, it is necessary to synthesize the comprehensive differences among all symmetrical monitoring groups within each column to reflect the current changes in magnetic field intensity at different symmetrical positions. More specifically, at any given moment, the product of the magnetic field difference factor and the distance distribution factor is calculated for each symmetrical monitoring group. These products are then accumulated for all symmetrical monitoring groups within the same column within the grid-like distribution of monitoring positions to obtain the asymmetry index for the symmetrical monitoring group at that moment.

[0104] It can be understood that under the grid distribution of monitoring positions, different symmetrical monitoring groups located in the same column belong to the same symmetrical relationship or the same symmetrical position, that is, the connecting line between two monitoring positions in different symmetrical groups located in the same column is distributed radially and is perpendicular to the jumper symmetry axis. The asymmetry index characterizes the change in magnetic field strength between monitoring positions belonging to the same symmetrical relationship at each moment, and reflects the degree of symmetry of the magnetic field strength of the monitoring position at each same symmetrical position. The larger the value of the asymmetry index, the lower the symmetry of the magnetic field strength between the monitoring positions at the corresponding symmetrical position of each column.

[0105] In some embodiments, the specific implementation method sub-steps of step S300 are as follows: Figure 6 As shown, the process includes steps S301 to S304.

[0106] Step S301 : constructing an asymmetric characteristic curve at each moment according to the distribution of asymmetric indicators of all symmetric monitoring groups distributed along the jumper at each moment.

[0107] Specifically, the straight line where the jumper is located is used as the horizontal axis, and the asymmetric index of each symmetrical monitoring group at each moment is used as the vertical coordinate to construct the asymmetric characteristic curve at each moment.

[0108] It is understandable that in some embodiments, the monitoring positions are distributed in a grid pattern, and the symmetrical monitoring groups belonging to the same column have the same symmetrical relationship. At each moment, each symmetrical monitoring group in each column corresponds to the value of an asymmetric indicator. At any moment, the horizontal coordinate of the data point on the asymmetric characteristic curve is the position of each symmetrical monitoring group in each column on the high-voltage jumper, and the vertical coordinate is the value of the asymmetric indicator corresponding to each symmetrical monitoring group in each column. Based on this, the asymmetric characteristic curve represents the abnormal symmetry between different monitoring positions at each moment along the axial distribution of the high-voltage jumper.

[0109] In step S302 , all moments are clustered based on the difference between the asymmetric characteristic curves of every two moments, and consecutive moments belonging to the same cluster are regarded as a same-curve time period.

[0110] When a train turns, the magnetic field strength collected by each monitoring position usually becomes asymmetric. At this time, the magnetic field strength collected between the symmetrical monitoring positions is different, resulting in a change in the longitudinal magnetic field asymmetry within the symmetrical monitoring group belonging to the same column.

[0111] At the same time, considering that when a train turns a bend with the same inner diameter, the change trend of the asymmetric characteristic curve obtained at each moment is the same or similar, different moments can be classified based on this feature, and the time intervals corresponding to the train turns for bends with the same inner diameter can be divided.

[0112] Specifically, for any two moments, the DTW distance between the asymmetric characteristic curves of the two moments is used as the corresponding classification metric distance between the two moments. Then, the DBSCAN clustering algorithm is used to cluster each moment to obtain the clustering result. In the clustering result, the abnormal changes in the symmetric characteristics of all moments in the same cluster have similar trends in position distribution. Therefore, the time period composed of continuously distributed moments in the same cluster is regarded as a same-curve time period.

[0113] Each time period of the same curve represents adjacent time points that may occur at the same inner diameter curve angle, avoiding the influence of magnetic field changes with different characteristics generated when the train copes with different curves on subsequent abnormal state judgment, and improving the accuracy of subsequent high-voltage jumper abnormality alarms.

[0114] Step S303: Obtain the average value of the asymmetry indexes between all monitoring positions at each moment in each same curve time period and determine it as the asymmetry balance factor for each same curve time period.

[0115] More specifically, for any same-curve time period, the mean of the asymmetric indicators corresponding to all symmetrical monitoring groups at all times is the asymmetric balance factor of the same-curve time period, reflecting the overall distribution of asymmetric characteristics at all symmetrical positions at all times within the same same-curve time period.

[0116] Step S304 : determining the degree of symmetric abnormality of each same-curve time period based on the product of the negative correlation coefficient of the time length of each same-curve time period and the asymmetric balancing factor.

[0117] Long-term vibration can cause fatigue fracture of the mechanical components of the jumper system, aggravate gap anomalies, and lead to abnormal changes in magnetic field strength for a long time. At this time, the impact of vibration on the magnetic field strength will be amplified, and the symmetry of the magnetic field strength will fluctuate at high frequency and large amplitude, thereby making the asymmetric characteristics more pronounced.

[0118] As a specific example, the product of the inverse of the length of each same-curve time period and the asymmetric balance factor is used as the degree of symmetric anomaly of each same-curve time period. Based on this, when the value of the asymmetric balance factor of a certain same-curve time period is larger and the value of the time length is smaller, it means that the symmetric characteristic fluctuation of the current same-curve time period is larger, and the duration of the symmetric characteristic is shorter, which further indicates that the degree of symmetric anomaly of the high-voltage jumper at consecutive moments in the same-curve time period may be higher. It can be understood from this that the degree of symmetric anomaly of the same-curve time period represents the possibility of anomaly in the symmetry of the magnetic field strength between adjacent carriages at consecutive moments in the time period.

[0119] In some embodiments, as Figure 7 As shown, the specific step of dividing the same curve time period to obtain the abnormal state time period in step S400 can be implemented by steps S401 to S403.

[0120] Step S401 : clustering the same curve time period to obtain clusters contained in different clustering results.

[0121] By traversing all classification methods for the same curve time period, a variety of different clustering results are obtained, and each clustering result contains multiple clusters.

[0122] Step S402 : constructing a clustering objective function value for each clustering result based on the difference in asymmetric indicators between adjacent same-curve time periods in each cluster cluster in each clustering result and the degree of symmetric abnormality between adjacent same-curve time periods.

[0123] In this embodiment, any clustering result is taken as an example for explanation. For any clustering result, in each clustering cluster, the average difference of the asymmetric index corresponding to the same monitoring position between each two adjacent same-curve time periods is calculated to obtain the first characteristic coefficient corresponding to each two adjacent same-curve time periods; the product of the degree of symmetry abnormality of each two adjacent same-curve time periods is calculated to obtain the second characteristic coefficient corresponding to each two adjacent same-curve time periods; the cumulative result of the first characteristic coefficient and the second characteristic coefficient corresponding to all two adjacent same-curve time periods in all clustering clusters is determined as the clustering objective function value of the any clustering result.

[0124] As a specific example, the calculation method of the clustering objective function value of any clustering result can be expressed as:

[0125]

[0126] Among them, f represents the clustering objective function value of the clustering result, N0 represents the number of clusters contained in the clustering result, G n,mIndicates the degree of symmetric abnormality of the mth time period of the same curve contained in the nth cluster in the clustering result, G n,m+1 Indicates the degree of symmetric abnormality of the m+1th time period of the same curve contained in the nth cluster in the clustering result. Indicates the first characteristic coefficient between the mth time period with the same curve and the m+1th time period with the same curve contained in the nth cluster in the clustering result.

[0127] Further G n,m ×G n,m+1 is the second characteristic coefficient, the first characteristic coefficient The calculation method can be expressed as: Where N1 represents the number of columns of monitoring locations distributed in a grid pattern, and W n,m (z) represents the asymmetric index of the mth time period of the same curve in the zth column contained in the nth cluster, W n,m+1 (z) represents the asymmetric index of the m+1th time period of the same curve in the zth column within the nth cluster in the clustering result.

[0128] When quantifying the clustering objective function, the product is maximized to encourage time periods with high anomaly probabilities to be distributed adjacently, forming an overall high anomaly feature within the cluster. If the time periods within a cluster all have a high degree of symmetrical anomaly, this symmetrical anomaly is significantly increased through the form of a product. For example, if there are symmetrical continuous anomalies due to jumper aging in multiple consecutive time periods, the greater the possibility that they will appear in the same cluster.

[0129] By analysing the differences between adjacent time periods, we can capture sudden changes in adjacent time periods and avoid misjudging slowly changing normal operating conditions as normal, such as fluctuations on different curves. This is because real faults, such as arcing, can disrupt magnetic field asymmetry, resulting in certain sudden changes. Normal curves, however, cause relatively gradual changes, and larger differences indicate a greater likelihood of a sudden abnormal event.

[0130] The first characteristic coefficient indicates that the clustering results focus more on the longer duration of symmetry destruction, reflecting the possible existence of persistent abnormalities caused by aging of high-voltage jumpers. The second characteristic coefficient indicates that the clustering results focus more on the suddenness of symmetry destruction, reflecting the possible existence of real faults with sudden changes, such as magnetic field anomalies caused by arc discharge.

[0131] Based on this feature, in some embodiments, different attention weights can be set for the first characteristic coefficient and the second characteristic coefficient, so that during the real-time monitoring process, more emphasis is placed on the degree of abnormal attention in a certain aspect. For example, if the attention weight of the first characteristic coefficient is larger, it means that the continuous attention to the abnormality of the symmetry of the magnetic field around the high-voltage jumper is greater. In comparison, the degree of sudden attention to the abnormality of the symmetry of the magnetic field around the high-voltage jumper is smaller. For example, if the attention weight is set, it can be expressed as f = ∑ (α × A1 + β × A2), α and β represent the attention weight, α + β = 1, and the above formula indicates that the values ​​of the attention weight are equal, A1 and A2 represent the first characteristic coefficient and the second characteristic coefficient respectively. The larger the value of the attention weight, the higher the degree of attention to the characteristic performance of this dimension.

[0132] In step S403, the continuous time periods of the same curve in each cluster in the clustering result corresponding to the maximum value of the clustering objective function are combined into abnormal state time periods.

[0133] In some embodiments, the implementer may also adopt a greedy algorithm according to the specific implementation scenario, starting from a single same-curve time period as a cluster, gradually merging clusters, and each time selecting the merging operation that maximizes the increase in the objective function value to determine the final optimal clustering result.

[0134] In this embodiment, the optimal clustering result is the clustering result corresponding to the maximum clustering objective function value, and the abnormal state time period represents the situation where the magnetic field symmetry around the high-voltage jumper is destroyed due to normal turning behavior.

[0135] In some embodiments, as Figure 8 As shown, a high-voltage sensing alarm device for a rail vehicle roof is also provided, which is used to implement the steps of a high-voltage sensing alarm method for a rail vehicle roof. The high-voltage sensing alarm device for a rail vehicle roof specifically includes:

[0136] A data acquisition module is used to obtain the magnetic field strength at each monitoring position on the roof of adjacent carriages on the track at each moment when the train is running, wherein the monitoring positions are symmetrically distributed along the jumper between adjacent carriages, and the moments include the current moment and historical moments;

[0137] A first feature analysis module is used to obtain an asymmetric index between the symmetrical monitoring positions at each moment based on the difference in magnetic field strength and distance distribution between the symmetrical monitoring positions at each moment;

[0138] The second feature analysis module is used to divide the same curve time period according to the differences in the distribution of asymmetric indicators between symmetrical monitoring positions distributed along the jumper at different times. The degree of symmetry anomaly in each same curve time period is obtained by combining the time length within the same curve time period and the overall distribution of the asymmetric indicators at each time point.

[0139] The abnormal state judgment module is used to divide the same curve time period into abnormal state time periods based on the difference in asymmetric indicators and the distribution of symmetric abnormality between adjacent same curve time periods;

[0140] The abnormality warning module is used to obtain an abnormality warning result of the gap between adjacent cars on the track at the current moment according to the symmetrical abnormality degree and the asymmetrical index in the abnormal state time period at the current moment.

[0141] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A high-voltage induction alarm method for a rail vehicle roof, characterized in that: The method comprises the following steps: Obtaining the magnetic field strength at each monitoring position on the roof between adjacent carriages on the track at each moment when the train is running, wherein the monitoring positions are symmetrically distributed along the jumper between adjacent carriages, and the moments include the current moment and historical moments; According to the difference in magnetic field intensity and distance distribution between symmetrical monitoring positions at each moment, the asymmetric index between the symmetrical monitoring positions at each moment is obtained; Based on the differences in the distribution of asymmetric indicators between symmetrical monitoring positions along the jumper at different times, the same curve time period is divided. The degree of symmetric anomaly in each same curve time period is obtained by combining the time length within the same curve time period and the overall distribution of the asymmetric indicators at each time moment. According to the difference of asymmetric indicators between adjacent time periods of the same curve and the distribution of symmetric abnormality, the time period of the same curve is divided to obtain abnormal state time periods; According to the symmetrical abnormality degree and the asymmetrical index in the abnormal state time period at the current moment, an abnormality warning result of the gap between adjacent cars on the track at the current moment is obtained.

2. A rail vehicle roof high-voltage induction alarm method according to claim 1, characterized in that: The asymmetric index between the symmetrical monitoring positions at each moment is obtained based on the difference in magnetic field strength and the distance distribution between the symmetrical monitoring positions at each moment, specifically including: Determine each symmetrical monitoring group based on the extension direction of the jumper between adjacent carriages of the track, wherein the two monitoring positions in the symmetrical monitoring group are symmetrically distributed along the jumper; Determining a magnetic field difference factor for each symmetrical monitoring group at each moment based on a difference in magnetic field strength between two monitoring positions within each symmetrical monitoring group at each moment; Determine a distance distribution factor for each symmetrical monitoring group based on the distance between any monitoring position in each symmetrical monitoring group and the jumper; Based on the product of the magnetic field difference factor and the distance distribution factor of each symmetrical monitoring group at each moment, the asymmetric index of the symmetrical monitoring group at each moment is determined.

3. A rail vehicle roof high-voltage induction alarm method according to claim 2, characterized in that: The method of dividing the same curve time period according to the difference in the distribution of asymmetric indicators between the symmetrical monitoring positions distributed along the jumper at different times specifically includes: Constructing an asymmetric characteristic curve at each moment according to the distribution of asymmetric indicators of all symmetric monitoring groups distributed along the jumper at each moment; Based on the difference between the asymmetric characteristic curves of every two moments, all moments are clustered, and the consecutive moments belonging to the same cluster are regarded as a same-curve time period.

4. A rail vehicle roof high-voltage induction alarm method according to claim 3, characterized in that: The step of constructing an asymmetric characteristic curve at each moment according to the distribution of asymmetric indicators of all symmetric monitoring groups distributed along the jumper at each moment specifically includes: The straight line where the jumper is located is used as the horizontal axis, and the asymmetric index of each symmetrical monitoring group at each moment is used as the vertical axis to construct the asymmetric characteristic curve at each moment.

5. A rail vehicle roof high-voltage induction alarm method according to claim 1, characterized in that: The combination of the time length within the same curve time period and the overall distribution of the asymmetry index at each moment to obtain the degree of symmetry anomaly for each same curve time period specifically includes: Obtaining the average value of the asymmetric indexes between all monitoring positions at each moment in each same curve time period to determine the asymmetric balance factor for each same curve time period; The degree of symmetric abnormality of each same-curve time period is determined based on the product of the negative correlation coefficient of the time length of each same-curve time period and the asymmetric balancing factor.

6. A rail vehicle roof high-voltage induction alarm method according to claim 1, characterized in that: The method of dividing the same curve time period into abnormal state time periods based on the difference in asymmetric indicators between adjacent same curve time periods and the distribution of symmetric abnormality degrees specifically includes: Cluster the same curve time period to obtain clusters contained in different clustering results; According to the difference of asymmetric index between adjacent same-curve time periods in each cluster in each clustering result and the degree of symmetric abnormality between adjacent same-curve time periods, the clustering objective function value of each clustering result is constructed; The continuous time periods of the same curve in each cluster in the clustering results corresponding to the maximum value of the clustering objective function are regarded as abnormal state time periods.

7. A rail vehicle roof high-voltage induction alarm method according to claim 1, characterized in that: The clustering objective function value of each clustering result is constructed based on the difference in asymmetric indicators between adjacent same-curve time periods in each cluster cluster in each clustering result and the degree of symmetric abnormality between adjacent same-curve time periods, specifically including: For any clustering result, in each cluster, the mean difference of the asymmetric index corresponding to the same monitoring position between each two adjacent time periods of the same curve is calculated to obtain the first characteristic coefficient corresponding to each two adjacent time periods of the same curve; Calculating the product of the symmetric abnormality degree of each two adjacent time periods of the same curve to obtain the second characteristic coefficient corresponding to each two adjacent time periods of the same curve; The cumulative result of the first characteristic coefficient and the second characteristic coefficient corresponding to all two adjacent time periods of the same curve in all clusters is determined as the clustering objective function value of any clustering result.

8. The rail vehicle roof high-voltage induction alarm method according to claim 1, characterized in that: The obtaining of the abnormal gap warning result between adjacent cars on the track at the current moment according to the symmetric abnormality degree and the asymmetric index in the abnormal state time period at the current moment specifically includes: Obtain the first mean of the asymmetry index between all monitoring positions at all times within the abnormal state time period at the current moment, and the second mean of the symmetry abnormality degree of all time periods with the same curve within the abnormal state time period at the current moment; normalize the product of the first mean and the second mean to obtain the abnormality degree of the gap between adjacent carriages on the track at the current moment; determine the gap abnormality warning result based on the abnormality degree of the gap between adjacent carriages on the track at the current moment.

9. A rail vehicle roof high-voltage induction alarm method according to claim 8, characterized in that: The determining of the gap abnormality warning result according to the abnormality degree of the gap between adjacent carriages on the track at the current moment specifically includes: If the abnormality of the gap between adjacent carriages on the track at the current moment is greater than or equal to the preset abnormality threshold, an abnormality warning will be issued; if the abnormality of the gap between adjacent carriages on the track at the current moment is less than the preset abnormality threshold, no abnormality warning will be issued.

10. A high-voltage induction alarm device on the roof of a railway vehicle, characterized in that: The device is used to implement the steps of a rail vehicle roof high-voltage sensing alarm method as described in any one of claims 1 to 9, and the rail vehicle roof high-voltage sensing alarm device specifically includes: A data acquisition module is used to obtain the magnetic field strength at each monitoring position on the roof of adjacent carriages on the track at each moment when the train is running, wherein the monitoring positions are symmetrically distributed along the jumper between adjacent carriages, and the moments include the current moment and historical moments; A first feature analysis module is used to obtain an asymmetric index between the symmetrical monitoring positions at each moment based on the difference in magnetic field strength and distance distribution between the symmetrical monitoring positions at each moment; The second feature analysis module is used to divide the same curve time period according to the differences in the distribution of asymmetric indicators between symmetrical monitoring positions distributed along the jumper at different times. The degree of symmetry anomaly in each same curve time period is obtained by combining the time length within the same curve time period and the overall distribution of the asymmetric indicators at each time point. The abnormal state judgment module is used to divide the same curve time period into abnormal state time periods based on the difference in asymmetric indicators and the distribution of symmetric abnormality between adjacent same curve time periods; The abnormality warning module is used to obtain an abnormality warning result of the gap between adjacent cars on the track at the current moment according to the symmetrical abnormality degree and the asymmetrical index in the abnormal state time period at the current moment.

Images