A catenary icing monitoring method based on multi-sensor data fusion

Abstract

The application discloses a kind of contact net icing monitoring method based on multi-sensor data fusion, it is related to data fusion technical field, for solving the problem that contact net asymmetric stress and dance exception is difficult to find in time, by multi-sensor data fusion, contact net is monitored in real time, contact line with icing is marked by identifying existence, detect the support stress data of suspension arm to marked contact line, identify potential stress anomaly, utilize laser ranging sensor to collect along line sag data and calculate sag distribution characteristics, by fusing support stress data and sag distribution characteristics, calculate asymmetric index to judge the symmetry state of marked contact line, when symmetry state is asymmetry, voltage fluctuation data is collected when train passes through marked contact line and divided detection area, obtain the contact pressure data and vibration data of each divided area by pantograph, evaluate dance characteristics and identify local anomaly, improve the stability of bow net contact and power supply safety.

Description

Technical Field

[0001] This invention relates to the field of data fusion technology, and more specifically, to a method for monitoring icing of overhead contact lines based on multi-sensor data fusion. Background Technology

[0002] The overhead contact line is a crucial component of the railway electrification system, and its stability directly impacts train power supply safety and operational efficiency. In winter and low-temperature environments, the contact line is prone to icing, leading to unstable contact between the pantograph and the contact wire, resulting in problems such as galloping, abnormal stress, and voltage fluctuations. Existing contact line inspection methods primarily rely on manual inspections or single-sensor monitoring, such as strain sensors measuring stress, voltage sensors monitoring fluctuations, or laser rangefinders measuring sag.

[0003] The existing technology has the following shortcomings:

[0004] Currently, single monitoring methods cannot fully reflect the stress state, sag characteristics, and galloping of the overhead contact system. Manual inspections are time-consuming and lack real-time performance, making it impossible to dynamically assess the status of the overhead contact system during train operation. Furthermore, the lack of multi-dimensional data fusion analysis makes it difficult to identify asymmetric stress areas and local galloping offsets caused by icing. At the same time, existing methods cannot provide timely warnings of potential risks before anomalies occur, making it difficult to detect asymmetric stress and galloping anomalies in the overhead contact system in a timely manner. Consequently, power supply safety and train operation stability cannot be effectively guaranteed. Therefore, a monitoring method for overhead contact system icing based on multi-sensor data fusion is proposed.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a contact wire icing monitoring method based on multi-sensor data fusion. This method utilizes multiple sensors, such as strain sensors, laser rangefinders, voltage sensors, and vibration sensors, to collect real-time data on the force, sag, voltage, and vibration of the contact wire in the contact wire. Based on a multi-sensor data fusion algorithm, the method comprehensively analyzes the symmetry, galloping characteristics, and local anomalies of the contact wire in the contact wire, thereby achieving the identification of asymmetric stress areas, galloping offset assessment, and anomaly early warning to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for monitoring icing of overhead contact lines based on multi-sensor data fusion, comprising the following steps:

[0008] Step S1: Mark the contact line with ice in the area to be tested, set the analysis time, detect the support force data of the suspension arm on the marked contact line through the strain sensor, and calculate the support imbalance coefficient based on the support force data;

[0009] Step S2: Detect the sag data along the marked contact line using a laser rangefinder, evaluate the sag distribution characteristics based on the sag data, and analyze the symmetry state of the marked contact line by fusing the support imbalance coefficient and the sag distribution characteristics.

[0010] Step S3: When the symmetrical state is asymmetrical, the voltage fluctuation data of the marked contact line is collected during the process of the train passing through the marked contact line. The frequency of the arc pulse is calculated using the voltage fluctuation data and the marked contact line is divided into regions.

[0011] Step S4: Detect the contact pressure data of the train's pantograph to each divided area and the vibration data of the marked contact wire. Calculate the number of pressure triggers based on the contact pressure data, and evaluate the galloping characteristics of the divided areas in combination with the vibration data to determine whether there are any abnormalities in the divided areas.

[0012] In a preferred embodiment, in step S1, the transverse diameter of each contact line cross-section in the horizontal direction in the test area is collected by a laser ranging sensor, and the transverse diameter is compared with a preset diameter threshold to filter and mark the contact lines.

[0013] The analysis time is preset and divided into multiple analysis moments. Multiple suspension arms along the marked contact line are detected. The bending strain value of the suspension arm to the marked contact line is collected by strain sensors installed on each suspension arm.

[0014] The average value of the bending strain of the same cantilever at each analysis time is used to obtain the support stress data;

[0015] The ratio of the force deviation to the average support force is used as the support imbalance coefficient.

[0016] In a preferred embodiment, in step S2, a laser is emitted to multiple measuring points of the marked contact line by a laser range sensor. The reflected signals from the measuring points are received by the laser range sensor and converted into the sag height value of the corresponding measuring point.

[0017] Combine the sag height values ​​of all measuring points into sag data along the line according to the order of the marked contact line;

[0018] The sag deviation is obtained by taking the absolute value of the difference between the sag height values ​​of adjacent measuring points in the sag data along the line.

[0019] In a preferred embodiment, in step S2, the historical database is accessed to obtain the sag distribution benchmark, which is a reference value for statistically analyzing the sag deviation under normal contact wire conditions.

[0020] Compare the sag deviation with the sag distribution benchmark:

[0021] If the sag deviation is greater than the sag distribution reference, then the sag deviation is marked.

[0022] Conversely, no sag deviation is marked.

[0023] The number of sag deviations is statistically marked as the number of abnormal sags, and the ratio of the number of abnormal sags to the total number of sag deviations is used as the sag distribution characteristic.

[0024] In a preferred embodiment, in step S2, the support imbalance coefficient and the sag distribution characteristics are standardized to obtain the support imbalance factor and the sag distribution factor, respectively.

[0025] The asymmetric index is calculated by integrating the support imbalance factor and the sag distribution factor.

[0026] If the asymmetry index is greater than the preset asymmetry threshold, then the symmetry state of the marked contact line is determined to be asymmetric.

[0027] Conversely, the symmetry of the marked contact line is determined to be symmetry.

[0028] In a preferred embodiment, in step S3, when the symmetry state of the marked contact line is asymmetric, a monitoring cycle is set during the process of the train running to the marked contact line area, and the length of the monitoring cycle is the result of dividing the structural feature length of the contact line by the train running speed.

[0029] During the monitoring period, voltage sampling sensors installed at both ends and the middle section of the contact wire continuously collect the instantaneous voltage value of the marked contact wire at a preset sampling frequency, forming voltage fluctuation data.

[0030] In a preferred embodiment, in step S3, the voltage fluctuation data is processed by the first time derivative to obtain the voltage change rate.

[0031] When the rate of change of voltage is greater than the arc pulse determination threshold, an arc pulse event is determined to have occurred at the corresponding sampling time.

[0032] The frequency of arc pulses is calculated by statistically analyzing the total number of arc pulse events occurring within a monitoring period and normalizing the time duration.

[0033] The coordinates of voltage sampling points along the contact line are correlated with the corresponding arc pulse frequencies. Based on the spatial gradient of the arc pulse frequency, the locations of local extreme points are identified, and the marked contact line is divided into multiple regions using the locations of these local extreme points as boundaries.

[0034] In a preferred embodiment, in step S4, as the train passes through the marked contact line zones, the contact pressure applied by the pantograph to the contact line is collected in real time by a contact pressure sensor installed on the pantograph slide.

[0035] Accelerometer-type vibration sensors are installed in the corresponding designated areas to synchronously collect the vibration acceleration of the contact wire when the train passes.

[0036] When the contact pressure is greater than the preset contact pressure threshold, it is determined that an effective contact has been formed between the pantograph and the contact wire at the corresponding sampling time.

[0037] The number of valid contact events occurring in each divided area is statistically analyzed, and the number of pressure triggers in each divided area is calculated.

[0038] In a preferred embodiment, in step S4, the vibration acceleration is subjected to a fast Fourier transform to obtain the vibration spectrum, and the vibration energy spectral density is calculated.

[0039] The pressure triggering number and vibration energy spectral density were standardized to obtain the pressure triggering factor and oscillation energy factor.

[0040] The difference between the oscillation energy factor and the pressure triggering factor is used as the dancing feature for dividing the region;

[0041] When the dancing feature exceeds the preset dancing threshold, it is determined that there is an anomaly in the corresponding segmented area;

[0042] When the dancing feature is less than or equal to the preset dancing threshold, it is determined that there is no abnormality in the corresponding division area, that is, the dancing state is within the normal range.

[0043] The technical effects and advantages of this invention are as follows:

[0044] This invention uses multi-sensor data fusion to monitor the overhead contact line in real time, identifies and marks contact lines with icing, detects the support force data of the cantilever arm on the marked contact lines, identifies potential stress anomalies, and uses a laser rangefinder to collect sag data along the line to calculate sag distribution characteristics. By fusing support force data and sag distribution characteristics, an asymmetry index is calculated to determine the symmetry state of the marked contact lines. When the symmetry state is asymmetric, voltage fluctuation data is collected and detection areas are divided when the train passes the marked contact lines. Contact pressure data and vibration data of the pantograph on each divided area are obtained, galloping characteristics are evaluated, and local anomalies are identified. This invention can effectively identify asymmetric stress areas and galloping offsets caused by icing, improving pantograph-catenary contact stability and power supply safety. Attached Figure Description

[0045] Figure 1This is a flowchart illustrating the process of a contact wire icing monitoring method based on multi-sensor data fusion according to the present invention.

[0046] Figure 2 This is an interactive flowchart of a contact wire icing monitoring method based on multi-sensor data fusion according to the present invention. Detailed Implementation

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

[0048] This invention uses multi-sensor data fusion to monitor the overhead contact line in real time, identifies and marks contact lines with icing, detects the support force data of the cantilever arm on the marked contact lines, identifies potential stress anomalies, uses a laser rangefinder to collect sag data along the line to calculate sag distribution characteristics, and calculates an asymmetry index by fusing support force data and sag distribution characteristics to determine the symmetry state of the marked contact lines. When the symmetry state is asymmetric, voltage fluctuation data is collected and detection areas are divided when the train passes the marked contact lines, and contact pressure data and vibration data of the pantograph on each divided area are obtained to evaluate galloping characteristics and identify local anomalies. This invention can effectively identify asymmetric stress areas and galloping offsets caused by icing.

[0049] Example 1

[0050] Please see Figures 1 to 2 A method for monitoring icing of overhead contact lines based on multi-sensor data fusion includes the following steps:

[0051] Step S1: Mark the contact line with ice in the area to be tested, set the analysis time, detect the support force data of the suspension arm on the marked contact line through the strain sensor, and calculate the support imbalance coefficient based on the support force data;

[0052] Step S2: Detect the sag data along the marked contact line using a laser rangefinder, evaluate the sag distribution characteristics based on the sag data, and analyze the symmetry state of the marked contact line by fusing the support imbalance coefficient and the sag distribution characteristics.

[0053] Step S3: When the symmetrical state is asymmetrical, the voltage fluctuation data of the marked contact line is collected during the process of the train passing through the marked contact line. The frequency of the arc pulse is calculated using the voltage fluctuation data and the marked contact line is divided into regions.

[0054] Step S4: Detect the contact pressure data of the train's pantograph to each divided area and the vibration data of the marked contact wire. Calculate the number of pressure triggers based on the contact pressure data, and evaluate the galloping characteristics of the divided areas in combination with the vibration data to determine whether there are any abnormalities in the divided areas.

[0055] The specific implementation is as follows:

[0056] In step S1, ice and other deposits may appear on the contact wire during operation, and the contact wire is prone to structural asymmetry. Under wind-induced excitation, the asymmetry will cause the contact wire's galloping response to deviate in direction, which will affect the stability of the pantograph-catenary contact and the safety of power supply.

[0057] The transverse diameter of each contact line cross section in the test area is collected by a laser rangefinder, and the transverse diameter is compared with a preset diameter threshold to filter and mark the contact lines.

[0058] If the transverse diameter of the contact wire is greater than the preset diameter threshold, the contact wire is marked.

[0059] Conversely, the contact line is not marked.

[0060] It should be explained that a laser rangefinder is a detection element that measures spatial distance by emitting a laser beam and receiving the reflected light from the target; the preset diameter threshold can be set according to the bare wire diameter of the contact wire, the upper limit of allowable wear, and the outer diameter change under typical icing conditions.

[0061] The analysis time is preset and divided into multiple analysis moments. Multiple suspension arms along the marked contact line are detected. The bending strain value of the suspension arm to the marked contact line is collected by strain sensors installed on each suspension arm.

[0062] Bending strain value refers to the relative tensile or compressive deformation of the cantilever when it is bent under stress, reflecting the magnitude of the supporting force of the cantilever on the contact line;

[0063] The average bending strain of the same cantilever at each analysis time is used to obtain the support force data, which reflects the support state of the cantilever on the marked contact line.

[0064] The force deviation value is obtained by subtracting the maximum and minimum values ​​from the support force data of each cantilever arm; the average support force is obtained by taking the average value of the support force data.

[0065] The ratio of the force deviation to the average support force is used as the support imbalance coefficient;

[0066] The support imbalance coefficient is used to quantify the unevenness of the force on the cantilever arms along the marked contact line. It reflects whether there is a force deviation on the marked contact line. The smaller the support imbalance coefficient, the more uniform the force distribution of each cantilever arm is, the more balanced the force on the marked contact line is, and the lower the risk of galloping. The larger the support imbalance coefficient, the more uneven the force on the cantilever arms is, which may cause the marked contact line to gallop and deviate.

[0067] It should be noted that the preset analysis time can be set according to the length of the marked contact line and the rate of change of current; the strain sensor is a mechanical sensor used to measure the minute deformation of structural materials. In this embodiment, it is used to collect the bending strain value of the cantilever arm on the marked contact line in real time.

[0068] When the contact wire becomes icy during operation, it can easily lead to structural asymmetry, causing a deviation in the galloping direction and affecting the stability of the pantograph-catenary contact and power supply safety. This step involves acquiring contact wires with attached icing, deploying strain sensors along the wire to collect the bending strain values ​​of the cantilever arms in real time, calculating the support force data and support imbalance coefficient of each cantilever arm, reflecting whether there is a force deviation in the marked contact wire, and promptly identifying contact wires with abnormal support force, which helps improve the operational safety, stability, and power supply reliability of the contact network.

[0069] In step S2, a laser is emitted to multiple measuring points of the marked contact line through a laser range sensor. The reflected signals from the measuring points are received by the laser range sensor and converted into the sag height value of the corresponding measuring point.

[0070] Combine the sag height values ​​of all measuring points into sag data along the line according to the order of the marked contact line;

[0071] The sag deviation is obtained by taking the absolute value of the difference between the sag height values ​​of adjacent measuring points in the sag data along the line.

[0072] Access the historical database to obtain the sag distribution benchmark. The sag distribution benchmark is a reference value for statistically analyzing the sag deviation distribution characteristics under normal conditions of the contact wire. Compare the sag deviation with the sag distribution benchmark:

[0073] If the sag deviation is greater than the sag distribution reference, then the sag deviation is marked.

[0074] Conversely, no sag deviation is marked.

[0075] The number of marked sag deviations is taken as the number of abnormal sags, and the ratio of the number of abnormal sags to the total number of sag deviations is taken as the sag distribution characteristic, reflecting the influence of sag distribution when there is icing along the structure of the marked contact line.

[0076] The larger the sag distribution, the more pronounced the sag area along the marked contact line, which may be caused by the force deviation of the marked contact line due to icing, increasing the risk of the marked contact line galloping and offsetting. The smaller the sag distribution, the more uniform the sag along the contact line and the better the structural symmetry.

[0077] After standardizing the support imbalance coefficient and sag distribution characteristics respectively, we obtain the support imbalance factor and sag distribution factor.

[0078] The asymmetric index is calculated by combining the support imbalance factor and the sag distribution factor. ,in, It is an asymmetric exponent. To support the imbalance factors, The sag distribution factor is , and Preset adjustment weights;

[0079] The asymmetry index is compared with a preset asymmetry threshold to determine the symmetry state of the marked contact line.

[0080] If the asymmetry index is greater than the preset asymmetry threshold, then the symmetric state is judged to be asymmetric.

[0081] Conversely, if the state is symmetrical, then it is determined to be symmetrical.

[0082] It should be noted that the laser rangefinder is a non-contact displacement measurement sensor, deployed beside or above the marked contact line, used to collect the sag height value of the marked contact line measuring points. The position of the measuring points in the marked contact line can be set according to the contact line length, suspension spacing, and galloping sensitive sections. For example, a measuring point can be placed every 1 to 2 meters to ensure that the sag changes along the line are fully measured. The historical database is used to manage the sag distribution benchmark of the contact line in a non-iced and non-attached state. The standardization processing methods include, but are not limited to, standard linear transformation based on interval scaling, Z-Score standardization based on statistics, or normalization method based on nonlinear mapping functions. The application methods of standardization processing will not be elaborated here. The preset adjustment weight can be set according to the contact line stress balance and statistical analysis of abnormal sag distribution data. For example, when the sag distribution abnormality is large, the adjustment weight corresponding to the sag distribution factor can be increased. The preset asymmetric threshold can be set according to the asymmetric exponential distribution of the contact line in a normal symmetrical state.

[0083] By acquiring sag data along the marked contact line, the sag distribution characteristics are calculated. Combined with the support imbalance coefficient, the asymmetry index is calculated and compared with a preset symmetry threshold to determine the symmetry state. When the asymmetry index is greater than the threshold, it indicates that there are uneven stress or abnormal sag areas caused by icing along the marked contact line, and the structure is in an asymmetric state, increasing the risk of galloping and offset. Conversely, when the asymmetry index is low, the marked contact line structure has good symmetry. This step can identify asymmetric stress areas caused by icing in real time, enhancing the stability of pantograph-catenary contact and the safety of power supply.

[0084] In step S3, when the symmetry of the marked contact line is asymmetrical, the voltage fluctuation monitoring process during the train's passage over the marked contact line is initiated. Specifically, during the train's journey to the marked contact line area, a monitoring cycle is set based on the train's speed and the structural characteristic length of the contact line. Specifically, the length of the contact line's structural characteristic length is divided by the train's speed as the duration of the monitoring cycle. Within the monitoring cycle, voltage sampling sensors located at both ends and the middle section of the contact line continuously collect the instantaneous voltage value of the marked contact line at a preset sampling frequency, forming voltage fluctuation data to reflect the changes in the instantaneous contact state between the train's pantograph and the contact line.

[0085] The structural characteristic length refers to the horizontal span length between the marked contact line and two adjacent suspension support structures.

[0086] It should be noted that the voltage sampling sensor is a voltage detection element installed on the conductive part of the contact wire. Its working principle is based on potential difference measurement. It acquires the instantaneous voltage signal of the contact wire in real time during operation through a non-contact voltage sampling probe. The preset sampling frequency is used to determine the time resolution of the voltage sampling sensor in acquiring the instantaneous voltage of the contact wire within the monitoring period. In order to ensure the complete capture of high-frequency transient signals such as arc pulses, the sampling frequency satisfies the Nyquist sampling theorem.

[0087] After obtaining voltage fluctuation data, its first-order time derivative is processed to obtain the voltage change rate. An arc pulse detection threshold is used as the criterion; when the voltage change rate exceeds the threshold, an arc pulse event is determined to have occurred at the corresponding sampling time. The arc pulse frequency is calculated by statistically analyzing the total number of arc pulse events within the monitoring period and normalizing the time duration. The mathematical expression for this frequency is:

[0088] ;

[0089] in, The frequency of the electric arc pulse. For the duration of the monitoring period, The instantaneous voltage value of the contact wire at sampling time t is given, where t is the sampling time within the monitoring period. The rate of change of voltage. The threshold for determining the electric arc pulse. For Heaviside step function, It is the derivative of t, used to assign a value of 1 when the threshold judgment condition is met, and a value of 0 when the threshold judgment condition is not met.

[0090] It should be noted that the first-order time derivative is used to describe the rate of change of the voltage signal over time, and its physical meaning is the amplitude of voltage change per unit time; the Heaviside step function is a mathematical discriminant function used to implement threshold logic judgment in signal processing; the arc pulse judgment threshold is used to distinguish between normal voltage fluctuations and violent voltage changes caused by arc discharge. By collecting contact line voltage fluctuation data under stable environmental conditions, normal contact line condition and no arc discharge, calculating its voltage change rate, obtaining the statistical distribution of the voltage change rate under normal operating conditions, and calculating its mean and standard deviation, the result of adding a certain number of times the standard deviation to the mean is used as the arc pulse judgment threshold.

[0091] The frequency of arc pulses reflects the density of arc discharge between the pantograph and the contact wire during train operation. When the contact wire has structural asymmetry, the spatial distribution of arc pulses often shows a local clustering phenomenon, that is, the frequency of arc pulses is significantly higher in the asymmetric force area than in other areas.

[0092] The coordinates of the voltage sampling points laid out along the contact line are associated with the corresponding arc pulse frequencies. Based on the spatial gradient of the arc pulse frequency, the location of local extreme points is identified. Using these as boundaries, the marked contact line is divided into multiple regions, each region corresponding to a uniform arc pulse distribution segment.

[0093] In step S4, the pantograph contact pressure data and contact wire vibration data are jointly monitored and analyzed for each divided area. Specifically, as the train passes through each divided area of ​​the marked contact wire, the contact pressure sensor installed on the pantograph slide plate collects the contact pressure applied by the pantograph to the contact wire in real time. The contact pressure reflects the strength of the mechanical contact between the pantograph and the contact wire during operation and characterizes the contact stability between the pantograph and the contact wire.

[0094] Meanwhile, acceleration-type vibration sensors are set up in the corresponding divided areas to synchronously collect the vibration acceleration of the contact wire when the train passes. Vibration acceleration represents the instantaneous acceleration change of the contact wire during the train's passage due to the force of the pantograph and the dynamic response of its own structure. Its physical meaning is to describe the change in the motion rate of the contact wire under dynamic load. The amplitude of vibration acceleration reflects the galloping amplitude and force intensity of the contact wire, while its frequency characteristics reflect the dynamic response mode of the structure and the existence of asymmetry or local flexibility differences.

[0095] It should be noted that the contact pressure sensor is a sensor used to measure the mechanical force between the pantograph slide and the contact wire in real time. It converts the contact pressure signal into an acquireable electrical signal output to reflect the instantaneous pressure applied by the pantograph to the contact wire during train operation. The acceleration vibration sensor is a sensor used to measure the vibration acceleration of the contact wire. Its function is to convert the vibration acceleration signal generated by the contact wire under the action of external force into an acquireable electrical signal, which can respond to the transient vibration and periodic movement of the contact wire during the passage of the train.

[0096] After collecting contact pressure data, a contact pressure threshold is set. When the contact pressure exceeds the threshold, a valid contact is determined to have occurred between the pantograph and the contact wire at the corresponding sampling time. The number of valid contact events in each defined area is statistically analyzed to calculate the pressure trigger count for each area. The calculation formula is as follows:

[0097] ;

[0098] in, For the number of stress triggers, For the duration of the monitoring period, Let be the contact pressure of the j-th segment at sampling time t. Here, t represents the contact pressure threshold, t is the sampling time within the monitoring period, and j is the index value of the divided region. For Heaviside step function, It is the derivative of t, used to assign a value of 1 when the threshold judgment condition is met, and a value of 0 when the threshold judgment condition is not met.

[0099] It should be noted that the contact pressure threshold is a critical pressure value used to determine whether the pantograph and the contact wire have formed effective contact during train operation. Under the condition that the contact wire is in normal condition and there are no attachments or structural abnormalities, the contact pressure data of the pantograph passing through each divided area are collected, and its statistical characteristics, including the average pressure and standard deviation, are calculated. The result of adding the average pressure and standard deviation is used as the contact pressure threshold.

[0100] The number of pressure triggers is used to reflect the effective contact continuity between the pantograph and the contact wire, and to reflect whether the mechanical action on the marked contact wire is stable during train operation. The fewer the number of pressure triggers, the less continuous the pantograph-catenary bonding state is in the corresponding area. The marked contact wire may experience greater galloping due to asymmetry and wind-induced excitation, resulting in unstable contact. Conversely, a higher number of pressure triggers indicates that the contact wire is subjected to uniform force and the interface bonding is stable.

[0101] For the vibration acceleration, a fast Fourier transform is performed to obtain the vibration spectrum, and the vibration energy spectral density is calculated:

[0102] ;

[0103] in, For vibrational energy spectral density, Let f be the highest frequency component, and f be the frequency variable, representing the expansion of vibration acceleration in the frequency domain. Let be the complex Fourier transform value of the vibration acceleration in the frequency domain for the j-th partitioned region, where j is the index value of the partitioned region. It is the derivative of the frequency variable.

[0104] It should be noted that the Fast Fourier Transform (FFT) is an efficient algorithm for converting time-domain signals into frequency-domain signals. It is used to decompose continuous or discrete vibration acceleration time series into a superposition representation of different frequency components.

[0105] Vibration energy spectral density represents the integral value of vibration energy per unit frequency band within the defined area. Its value reflects the galloping intensity and energy concentration of the contact line under dynamic force. The higher the value of vibration energy spectral density, the larger the galloping amplitude and the more concentrated the energy in the defined area. Icing may cause potential anomalies or local instability in the defined area.

[0106] The pressure triggering number and vibration energy spectral density were standardized to obtain the pressure triggering factor and oscillation energy factor.

[0107] The difference between the oscillation energy factor and the pressure trigger factor is used as the dancing feature for dividing the region. The dancing feature is then compared with a preset dancing threshold.

[0108] When the galloping characteristics exceed the preset galloping threshold, it is determined that there is an anomaly in the corresponding area, which means that the contact line in that area has asymmetrical galloping or unstable contact during train operation.

[0109] When the dancing feature is less than or equal to the preset dancing threshold, it is determined that there is no abnormality in the corresponding division area, that is, the dancing state is within the normal range.

[0110] It should be noted that the preset galloping threshold is a critical value used to determine abnormal galloping of the contact wire in the divided area. It is set based on the characteristics of the contact wire structure, train operating speed and test data. By conducting tests on the galloping response of the contact wire under different speeds, different contact wire types and different environmental conditions, the comprehensive distribution range of pressure triggering factor and oscillation energy factor is obtained. Based on engineering experience and safe operation requirements, the maximum comprehensive galloping characteristic value that does not show contact instability or arcing abnormality under normal operating conditions is selected as the galloping threshold.

[0111] Through the above processing, a quantitative assessment and anomaly determination of the galloping characteristics of each divided area of ​​the marked contact line can be achieved, providing a basis for subsequent contact line condition diagnosis and maintenance decisions.

[0112] Finally, it should be noted that in this paper, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0113] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0114] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0115] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of protection of the above specification.

[0116] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

Claims

1. A method for monitoring icing of overhead contact lines based on multi-sensor data fusion, characterized in that: Includes the following steps: Step S1: Mark the contact line with ice in the area to be tested, set the analysis time, detect the support force data of the suspension arm on the marked contact line through the strain sensor, and calculate the support imbalance coefficient based on the support force data; In step S1, the transverse diameter of each contact line cross section in the test area is collected by a laser ranging sensor in the horizontal direction. The transverse diameter is compared with a preset diameter threshold to filter and mark the contact lines. The analysis time is preset and divided into multiple analysis moments. Multiple suspension arms along the marked contact line are detected. The bending strain value of the suspension arm to the marked contact line is collected by strain sensors installed on each suspension arm. The average value of the bending strain of the same cantilever at each analysis time is used to obtain the support stress data; The force deviation value is obtained by subtracting the maximum and minimum values ​​from the support force data of each cantilever arm; the average support force is obtained by taking the average value of the support force data. The ratio of the force deviation to the average support force is used as the support imbalance coefficient; Step S2: Detect the sag data along the marked contact line using a laser rangefinder, evaluate the sag distribution characteristics based on the sag data, and analyze the symmetry state of the marked contact line by fusing the support imbalance coefficient and the sag distribution characteristics. In step S2, the support imbalance coefficient and sag distribution characteristics are standardized to obtain the support imbalance factor and sag distribution factor, respectively. The asymmetric index is calculated by combining the support imbalance factor and the sag distribution factor: ,in, It is an asymmetric exponent. To support the imbalance factors, The sag distribution factor is , and Preset adjustment weights; If the asymmetry index is greater than the preset asymmetry threshold, then the symmetry state of the marked contact line is determined to be asymmetric. Conversely, the symmetry of the marked contact line is determined to be symmetry; Step S3: When the symmetrical state is asymmetrical, the voltage fluctuation data of the marked contact line is collected during the process of the train passing through the marked contact line. The frequency of the arc pulse is calculated using the voltage fluctuation data and the marked contact line is divided into regions. Step S4: Detect the contact pressure data of the train's pantograph to each divided area and the vibration data of the marked contact wire. Calculate the number of pressure triggers based on the contact pressure data, and evaluate the galloping characteristics of the divided areas in combination with the vibration data to determine whether there are any abnormalities in the divided areas.

2. The contact wire icing monitoring method based on multi-sensor data fusion according to claim 1, characterized in that: In step S2, a laser is emitted to multiple measuring points of the marked contact line through a laser range sensor. The reflected signals from the measuring points are received by the laser range sensor and converted into the sag height value of the corresponding measuring point. Combine the sag height values ​​of all measuring points into sag data along the line according to the order of the marked contact line; The sag deviation is obtained by taking the absolute value of the difference between the sag height values ​​of adjacent measuring points in the sag data along the line.

3. The contact wire icing monitoring method based on multi-sensor data fusion according to claim 2, characterized in that: In step S2, the historical database is accessed to obtain the sag distribution benchmark, which is a reference value for statistical sag deviation under normal contact wire conditions; Compare the sag deviation with the sag distribution benchmark: If the sag deviation is greater than the sag distribution reference, then the sag deviation is marked. Conversely, no sag deviation is marked. The number of sag deviations is statistically marked as the number of abnormal sags, and the ratio of the number of abnormal sags to the total number of sag deviations is used as the sag distribution characteristic.

4. The contact wire icing monitoring method based on multi-sensor data fusion according to claim 1, characterized in that: In step S3, when the symmetry of the marked contact line is asymmetric, a monitoring cycle is set during the process of the train running to the marked contact line area, and the length of the monitoring cycle is the result of dividing the structural feature length of the contact line by the train running speed. During the monitoring period, voltage sampling sensors installed at both ends and the middle section of the contact wire continuously collect the instantaneous voltage value of the marked contact wire at a preset sampling frequency, forming voltage fluctuation data.

5. The contact wire icing monitoring method based on multi-sensor data fusion according to claim 4, characterized in that: In step S3, the voltage fluctuation data is processed by the first time derivative to obtain the voltage change rate; When the rate of change of voltage is greater than the arc pulse determination threshold, an arc pulse event is determined to have occurred at the corresponding sampling time. The frequency of arc pulses is calculated by statistically analyzing the total number of arc pulse events occurring within a monitoring period and normalizing the time duration. The coordinates of voltage sampling points along the contact line are correlated with the corresponding arc pulse frequencies. Based on the spatial gradient of the arc pulse frequency, the locations of local extreme points are identified, and the marked contact line is divided into multiple regions using the locations of these local extreme points as boundaries.

6. The contact wire icing monitoring method based on multi-sensor data fusion according to claim 1, characterized in that: In step S4, as the train passes through the marked contact line zones, the contact pressure sensor installed on the pantograph plate collects the contact pressure applied by the pantograph to the contact line in real time. Accelerometer-type vibration sensors are installed in the corresponding designated areas to synchronously collect the vibration acceleration of the contact wire when the train passes. When the contact pressure is greater than the preset contact pressure threshold, it is determined that an effective contact has been formed between the pantograph and the contact wire at the corresponding sampling time. The number of valid contact events occurring in each divided area is statistically analyzed, and the number of pressure triggers in each divided area is calculated.

7. The contact wire icing monitoring method based on multi-sensor data fusion according to claim 6, characterized in that: In step S4, the vibration acceleration is subjected to a fast Fourier transform to obtain the vibration spectrum, and the vibration energy spectral density is calculated. The pressure triggering number and vibration energy spectral density were standardized to obtain the pressure triggering factor and oscillation energy factor. The difference between the oscillation energy factor and the pressure triggering factor is used as the dancing feature for dividing the region; When the dancing feature exceeds the preset dancing threshold, it is determined that there is an anomaly in the corresponding segmented area; When the dancing feature is less than or equal to the preset dancing threshold, it is determined that there is no abnormality in the corresponding division area, that is, the dancing state is within the normal range.

Images