Intelligent cable fault monitoring and early warning method and system

Abstract

The invention relates to the technical field of cable fault monitoring, in particular to an intelligent cable fault monitoring and early warning method and system, and the method comprises the steps: detecting a sheath circulation phase angle, adjusting the frequency through offset threshold exceeding, extracting de-noising data, injecting pulses, measuring the time delay, correcting the temperature, inverting a carbonization path, matching a conductivity abrupt change generation map, and measuring the attenuation rate to update the carbonization depth. And calculating a correlation factor to trigger early warning. According to the invention, through dynamic detection of circulation phase angle offset, frequency modulation anti-interference data extraction, noise suppression, signal-to-noise ratio improvement, combination of temperature gradient and dielectric correction delay, compensation of signal temperature influence, accurate inversion of a carbonization starting point, and joint analysis of multi-band signal attenuation rate and propagation time, complex working conditions are adapted; a coupling model is established through dynamic comparison of correlation factors and threshold values, layered early warning is performed, phase correction, temperature compensation and multi-band collaborative analysis are fused, and the positioning precision and the evaluation reliability are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of cable fault monitoring, and in particular to an intelligent cable fault monitoring and early warning method and system. Background Technique

[0002] The technical field of cable fault monitoring includes real-time monitoring of the operating status of cables in the power transmission system, early warning of potential faults, and optimization of maintenance strategies. The core content of this field involves the detection and analysis of key parameters such as cable insulation aging, partial discharge, and abnormal temperature. By continuously collecting cable operation data and evaluating its health status, the risk of sudden faults is reduced. The overall technical field targets the cable body and its accessories, integrating sensor technology, data transmission protocols, and state evaluation models to form a systematic solution covering data collection, abnormal diagnosis, and early warning triggering. The technical evolution direction includes improving monitoring accuracy, shortening response time, and expanding the ability of multi-parameter collaborative analysis.

[0003] Among them, the intelligent cable fault monitoring and early warning method refers to the multi-dimensional real-time collection of cable operation parameters, combined with preset threshold judgment rules and state correlation models, to dynamically compare and identify abnormal patterns of data such as cable insulation performance, current load fluctuations, and environmental temperature and humidity. Specifically, based on the cable emergency protection logic in H02H7 / 26 classification, through the coordinated operation of integrated monitoring nodes and a central control unit, a distributed monitoring network covering the entire cable section is constructed. A fault feature library is established using the change of cable impedance characteristics and historical operation data, and data normalization processing and fault type matching are executed synchronously. Finally, a hierarchical early warning signal triggering mechanism is formed.

[0004] The prior art relies on the static acquisition of cable operating parameters and fixed threshold determination, without dynamically adjusting the monitoring frequency for environmental interference, resulting in the sheath circulating current component being vulnerable to noise pollution and affecting data validity. The partial discharge detection lacks a propagation delay temperature compensation mechanism and does not include the axial temperature gradient in the correction scope, resulting in the positioning accuracy of the carbonization path being restricted by environmental temperature fluctuations. The existing methods mainly rely on the matching of impedance characteristic changes and historical data, without combining the correlation analysis of electric field intensity gradient and conductivity mutation, and the ability to extract the spatial characteristics of the carbonization path is insufficient, making it difficult to distinguish the boundary between carbonized and non-carbonized areas. The traditional technology uses the attenuation rate of a single-frequency band signal to evaluate the carbonization depth, without introducing the joint calibration of the multi-frequency band attenuation time ratio, resulting in a lag in the update of the carbonization depth parameter and being unable to adapt to the dynamic changes under complex working conditions. In addition, the existing early warning mechanism does not establish a coupling model between the circulating current component and the carbonization process, resulting in the setting of the hierarchical early warning threshold relying on empirical values and lacking the ability to be driven by real-time data. The above deficiencies lead to problems such as the monitoring system being vulnerable to environmental interference, cumulative positioning errors, and large deviations in carbonization depth evaluation, increasing the risk of sudden failures and maintenance costs. For example, a single-frequency band signal is vulnerable to interference from line impedance fluctuations, misjudging the carbonization depth and causing unnecessary shutdown for maintenance. Summary of the Invention

[0005] The object of the present invention is to solve the shortcomings existing in the prior art, and to propose an intelligent cable fault monitoring and early warning method and system.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions: An intelligent cable fault monitoring and early warning method, including the following steps:

[0007] S1: Detect the phase angle of the metal sheath circulating current, calculate the adjacent phase offset, and when the offset exceeds the phase difference threshold, adjust the acquisition frequency of the monitoring node to the preset anti-interference frequency band, and extract the denoised sheath circulating current component data;

[0008] S2: Based on the denoised sheath circulating current component data, inject a standard partial discharge pulse signal at the head end of the cable and record the propagation time at the tail end, measure the axial insulation layer temperature gradient of the cable, correct the propagation delay through the temperature dielectric constant correction coefficient table, and invert the starting point of the carbonization path according to the total length of the cable to generate the partial discharge propagation delay carbonization path positioning value;

[0009] S3: Based on the partial discharge propagation delay carbonization path positioning value, measure the distribution of the electric field intensity gradient along the axial direction of the cable, match the spatial correlation between the conductivity mutation point and the carbonization path, and generate a carbonization path conductivity mutation matching map;

[0010] S4: Based on the conductivity mutation matching map of the carbonization path, inject multi-band partial discharge pulses at the head end of the cable, measure the signal attenuation rate and propagation time of the frequency band at the tail end, calculate the attenuation time ratio of the frequency band, update the carbonization depth parameter by comparing with the reference threshold of the feature library, and generate the calibration value of the multi-band carbonization depth feature library.

[0011] As a further solution of the present invention, the denoised sheath circulating current component data includes phase offset, anti-interference frequency band parameter, and denoised data. The partial discharge propagation delay carbonization path positioning value includes propagation time, temperature gradient correction coefficient, total cable length parameter, and carbonization starting point coordinates. The conductivity mutation matching map of the carbonization path includes electric field gradient distribution, conductivity mutation point coordinates, and spatial correlation parameters. The calibration value of the multi-band carbonization depth feature library includes signal attenuation rate, propagation time difference, attenuation time ratio, and reference threshold parameter.

[0012] As a further solution of the present invention, the specific steps of S1 are as follows:

[0013] S101: Detect the phase angle of the metal sheath circulating current, obtain the voltage and current time series data of adjacent monitoring nodes, extract the phase value of each node based on the time difference between the two, calculate the phase angle offset between adjacent nodes, and generate phase angle offset data;

[0014] S102: Compare the phase angle offset data with the phase difference threshold, screen the nodes with offset exceeding the threshold, adjust the sampling frequency of the nodes to the preset anti-interference frequency band, and generate a frequency adjustment coefficient value;

[0015] S103: Call the sampling frequency corresponding to the frequency adjustment coefficient value to re-collect the node signals, construct a signal envelope, extract the data sequence with the fluctuation amplitude within the amplitude limit interval, and generate the denoised sheath circulating current component data.

[0016] As a further solution of the present invention, the specific steps of S2 are as follows:

[0017] S201: Based on the denoised sheath circulating current component data, inject a standard partial discharge pulse signal at the head end of the cable, record the propagation time of the pulse at the tail end, and determine the signal propagation delay in combination with the sheath loop structure and sampling frequency information, and generate a propagation delay data value;

[0018] S202: According to the propagation delay data value, obtain the axial insulation layer temperature distribution of the cable, match the temperature value with the dielectric constant correction coefficient in the correction coefficient table, and adjust the section propagation time to generate a temperature-corrected propagation delay value;

[0019] S203: Invoke the temperature-corrected propagation delay value, combine the total cable length and the propagation speed reference to calculate the correction value of the mutation point position, calibrate the length position of the propagation mutation point, establish the positioning interval of the carbonization path, and generate the PD propagation delay carbonization path positioning value.

[0020] As a further solution of the present invention, the specific calculation formula for the correction value of the mutation point position is:

[0021]

[0022] where L m is the correction value of the propagation mutation point position, L total represents the total cable length, t m represents the propagation delay value after temperature correction, v b represents the reference propagation speed, σ v represents the fluctuation coefficient of the propagation speed, ΔL q represents the length deviation of the q-th adjacent section, r represents the total number of adjacent sections, and α m represents the temperature correction factor.

[0023] As a further solution of the present invention, the specific steps of S3 are as follows:

[0024] S301: Based on the PD propagation delay carbonization path positioning value, extract the time and space information in the path, combine the path change characteristics to divide the cable axial measurement section, and generate the path section spacing value;

[0025] S302: Obtain the section electric field strength data according to the path section spacing value, extract the region with significant intensity gradient change, and determine the spatial position where it is located to generate the electric field strength mutation interval distribution value;

[0026] S303: Locate the carbonization path coordinates according to the electric field strength mutation interval distribution value, extract the corresponding conductivity difference characteristics, establish the spatial correlation with the path points, and generate the carbonization path conductivity mutation matching map.

[0027] As a further solution of the present invention, the specific steps of S4 are as follows:

[0028] S401: Based on the carbonization path conductivity mutation matching map, obtain the emission time and initial amplitude information of the frequency band signal, combine the coordinates of the conductivity mutation region to determine the signal path starting point and the corresponding reference time, and generate the frequency band emission time value set;

[0029] S402: Invoke the frequency band emission time value set, collect the arrival time and amplitude change of the frequency band signal at the cable end, record the propagation time and amplitude change according to the frequency band division, and generate the frequency band signal attenuation time ratio group;

[0030] S403: According to the ratio group of the attenuation time of the frequency band signals, call the attenuation time ratio threshold corresponding to the frequency band in the feature library, calculate the comprehensive index of the frequency band deviation, compare the degree of frequency band deviation and determine the corresponding carbonation depth interval, and generate the calibration value of the multi-frequency band carbonation depth feature library.

[0031] As a further solution of the present invention, the specific calculation formula for the comprehensive index of the frequency band deviation is:

[0032]

[0033] where D b represents the comprehensive index of the deviation of frequency band b, T b represents the measured attenuation time ratio of frequency band b, τ b represents the attenuation time ratio threshold corresponding to frequency band b in the feature library, ΔT b,k represents the difference in the attenuation time ratio between frequency band b and the adjacent k-th frequency band, and n represents the total number of adjacent frequency bands.

[0034] As a further solution of the present invention, the method further includes:

[0035] S5: Call the calibration value of the multi-frequency band carbonation depth feature library and the denoised sheath circulating current component data, calculate the circulation carbonation correlation factor, compare the classification warning threshold to trigger the corresponding warning level, and generate a sheath circulating current carbonation depth coupling warning signal;

[0036] The sheath circulating current carbonation depth coupling warning signal includes a correlation factor, a classification warning threshold, and a warning level parameter;

[0037] The specific steps of S5 are:

[0038] S501: Call the calibration value of the multi-frequency band carbonation depth feature library and the denoised sheath circulating current component data, perform data registration and correlation extraction operations, establish a fitting relationship between the two types of data, and obtain the frequency band circulation coupling coefficient;

[0039] S502: Based on the frequency band circulation coupling coefficient, call the sheath circulating current component data for numerical ratio adjustment, and integrate the multi-frequency band processing results to form a unified index to obtain the sheath carbonation indication degree;

[0040] S503: According to the sheath carbonation indication degree and the set warning threshold, perform grade judgment, and generate a warning output in combination with time series information to obtain a sheath circulating current carbonation depth coupling warning signal.

[0041] An intelligent cable fault monitoring and warning system includes:

[0042] The circulating current denoising module obtains the phase angle information of the phase conductor in the metal sheath, identifies the degree of phase shift between adjacent conductors, determines whether it exceeds the phase difference threshold. If it is determined to exceed the threshold, it adjusts the sampling frequency band of the monitoring node, extracts the characteristic components in the stable frequency band of the adjusted sheath circulating current signal, and generates denoised sheath circulating current component data;

[0043] The propagation and positioning module, based on the denoised sheath circulating current component data, injects a partial discharge pulse signal at the cable head end and records the propagation time, establishes propagation delay correction data by combining the insulation layer temperature and the correction coefficient, and converts it according to the total cable length to generate a positioning value for the carbonization path of the partial discharge propagation delay;

[0044] The conductivity analysis module, based on the positioning value of the carbonization path of the partial discharge propagation delay, measures the change in the electric field strength distribution along the cable axis, extracts the positions of the conductivity with mutations, spatially corresponds the mutation points to the path positioning value, and generates a conductivity mutation matching map of the carbonization path;

[0045] The carbonization calibration module, based on the conductivity mutation matching map of the carbonization path, inputs multiple groups of partial discharge pulses with different frequencies at the cable head end, obtains the propagation time and amplitude change of the signal in the end frequency band, extracts key parameters and then compares and identifies them to generate a calibration value for the multi-frequency carbonization depth feature library;

[0046] The early warning trigger module calls the calibration value of the multi-frequency carbonization depth feature library and the denoised sheath circulating current component data, establishes a correlation factor model by combining the carbonization depth and the circulating current data, matches the warning threshold of the level, and generates a coupled warning signal for the carbonization depth of the sheath circulating current.

[0047] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0048] In the present invention, by dynamically detecting the phase angle shift of the circulating current and frequency modulation anti-interference to extract data, suppressing noise and improving the signal-to-noise ratio, combining the temperature gradient and dielectric correction delay to compensate for the influence of signal temperature, accurately inverting the carbonization starting point, jointly analyzing the multi-frequency signal attenuation rate and propagation time to adapt to complex working conditions, dynamically comparing the correlation factor and threshold to establish a coupling model, and warning in layers, integrating phase correction, temperature compensation and multi-frequency collaborative analysis to improve the positioning accuracy and evaluation reliability. Brief Description of the Drawings

[0049] Figure 1 is the step flow schematic diagram of the present invention;

[0050] Figure 2 is the system module diagram of the present invention. Detailed Embodiment

[0051] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0052] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the accompanying drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention. In addition, in the description of the present invention, the meaning of "a plurality of" is two or more, unless otherwise specifically defined.

[0053] Please refer to Figure 1 , an intelligent cable fault monitoring and early warning method, including the following steps:

[0054] S1: Detect the phase angle of the metal sheath circulating current, calculate the adjacent phase offset. When the offset exceeds the phase difference threshold, adjust the acquisition frequency of the monitoring node to the preset anti-interference frequency band, and generate the denoised sheath circulating current component data;

[0055] S2: Based on the denoised sheath circulating current component data, inject a standard partial discharge pulse signal at the head end of the cable and record the propagation time at the tail end, measure the axial insulation layer temperature gradient of the cable, correct the propagation delay through the temperature-dielectric constant correction coefficient table, and invert the starting point of the carbonization path according to the total length of the cable to generate the partial discharge propagation delay carbonization path positioning value;

[0056] S3: Based on the partial discharge propagation delay carbonization path positioning value, measure the distribution of the electric field strength gradient along the axial direction of the cable, match the spatial correlation between the conductivity mutation point and the carbonization path, and generate a carbonization path conductivity mutation matching map;

[0057] S4: Based on the carbonization path conductivity mutation matching map, inject multi-band partial discharge pulses at the head end of the cable, measure the signal attenuation rate and propagation time of the frequency band at the tail end, calculate the frequency band attenuation time ratio, compare with the reference threshold of the feature library to update the carbonization depth parameter, and generate the multi-band carbonization depth feature library calibration value;

[0058] S5: Call the multi-band carbonization depth feature library calibration value and the denoised sheath circulating current component data, calculate the circulating current carbonization correlation factor, compare with the classification early warning threshold to trigger the corresponding early warning level, and generate a sheath circulating current carbonization depth coupling early warning signal.

[0059] The denoised sheath circulating current component data includes phase offset, anti-interference frequency band parameters, and denoised data. The partial discharge propagation delay carbonization path positioning value includes propagation time, temperature gradient correction coefficient, total cable length parameter, and carbonization starting point coordinates. The carbonization path conductivity mutation matching map includes electric field gradient distribution, conductivity mutation point coordinates, and spatial correlation parameters. The multi-band carbonization depth feature library calibration value includes signal attenuation rate, propagation time difference, attenuation time ratio, and reference threshold parameter. The sheath circulating current carbonization depth coupling warning signal includes correlation factor, hierarchical warning threshold, and warning level parameter.

[0060] The specific steps of S1 are as follows:

[0061] S101: Detect the phase angle of the metal sheath circulating current, obtain the voltage and current time series data of adjacent monitoring nodes, extract the phase value of each node based on the time difference between the two, calculate the phase angle offset between adjacent nodes, and generate phase angle offset data;

[0062] When detecting the phase angle of the metal sheath circulating current, in actual monitoring scenarios such as cable tunnels, voltage and current sensors need to be installed at two adjacent nodes to synchronously record their respective time series data. The sampling interval is set to 10 milliseconds to generate a complete voltage and current time series. To ensure data synchronization, time calibration is performed by interpolation. Subsequently, the main frequency component is analyzed using the short-time Fourier transform, the time position of the peak or zero-crossing point in the current signal is extracted, and the time delay between two adjacent nodes is calculated. Taking a main frequency of 50 Hz as an example, if the current peak of a certain node appears 2 milliseconds earlier than that of the adjacent node, the phase difference can be calculated based on the frequency and time difference. At this time, the phase difference is approximately 36 degrees. By processing the data between all nodes in a similar manner, a dataset of phase angle offsets between each pair of adjacent nodes can be finally formed for subsequent processing.

[0063] S102: Compare the phase angle offset data with the phase difference threshold, screen the nodes with offsets exceeding the threshold, and adjust the sampling frequency of the nodes to the preset anti-interference frequency band to generate a frequency adjustment coefficient value;

[0064] Based on the obtained phase angle offset data, compare it one by one with the set phase difference threshold. If the offset exceeds the threshold, the corresponding node is identified as deviating from the normal operating state. The threshold setting refers to the historical operation records and the standard cable stable operating conditions, and can be set to 25 degrees. For example, if the phase offset between a certain node and its adjacent node is 36 degrees, which significantly exceeds the judgment benchmark of 25 degrees, this node can be marked as an abnormal node. After identifying the offset node, adjust its sampling frequency from the default 1000 Hz to a higher frequency, such as 1250 Hz, to avoid the interference frequency band. The frequency adjustment amplitude is equivalent to an increase of 25%. The sampling frequency adjustment value can generate a frequency adjustment coefficient so that the corresponding node can execute according to the adjusted frequency when collecting signals.

[0065] S103: Call the sampling frequency corresponding to the frequency adjustment coefficient value to re-collect the node signal, construct the signal envelope, extract the data sequence with the fluctuation amplitude within the amplitude limit range, and generate the denoised sheath circulating current component data;

[0066] According to the previously generated frequency adjustment coefficient, re-set the sampling frequency of the target node and perform data collection. Under the adjusted frequency condition, collect the current signal, and perform envelope extraction processing on the signal. Extract its envelope amplitude sequence through mathematical transformation. After obtaining the signal envelope, set an effective amplitude range. For example, set the maximum amplitude as 100% of the benchmark, and screen the envelope amplitude data points between 30% and 70%. For example, when the envelope amplitude is 0.52, since it is within the screening range, it is regarded as a valid data point, otherwise it is excluded. All data points that meet the conditions form a new current sequence for analysis, and this sequence is the sheath circulating current component data after removing interference and noise, laying a foundation for further analysis.

[0067] The specific steps of S2 are as follows:

[0068] S201: Based on the denoised sheath circulating current component data, inject a standard partial discharge pulse signal at the cable head end, record the propagation time of the pulse at the end, combine the sheath loop structure and sampling frequency information, determine the signal propagation delay, and generate the propagation delay data value;

[0069] When processing the denoised sheath circulating current component data, it is necessary to first extract the original current signal curve and perform multi-scale decomposition using the wavelet analysis method. The db4 function in the Daubechies wavelet family can be selected as the basis function to divide the signal into several scale levels. By filtering out the high-frequency part, the main signal components are obtained, and then the signal is reconstructed and restored to the denoised sheath circulating current signal data. At the cable head, a partial discharge standard pulse generating device is configured to inject a standard pulse signal with fixed amplitude, rise time, and width parameters. For example, a pulse signal with an amplitude of 10 volts, a rise time of 1 nanosecond, and a pulse width of 5 nanoseconds is injected. This signal propagates along the cable sheath, and the cable end continuously records the current change through a data acquisition module with a sampling rate of 100 megahertz. When the signal jump point with a voltage exceeding 3 volts is first detected at the end, mark this moment as the signal reception time. Since the injection time is zero, the propagation delay is this detection time. For example, if the detection time is 4.2 microseconds, the corresponding propagation delay is 4.2 microseconds. This process is applicable to the sheath loop structure with a single-end grounding method and a cable length of 500 meters, and finally, the data value reflecting the signal propagation delay degree can be obtained.

[0070] S202: According to the propagation delay data value, obtain the axial insulation layer temperature distribution of the cable, match the temperature value with the dielectric constant correction coefficient in the correction coefficient table, and adjust the section propagation time to generate the temperature-corrected propagation delay value;

[0071] Use the known propagation delay data value, such as 4.2 microseconds, and carry out the correction in combination with the temperature distribution information of each section of the cable. The temperature nodes can be set at a step of 50 meters, and the temperature readings at each node are obtained from the temperature sensing device. For example, the temperatures at each node are 45 degrees, 52 degrees, 60 degrees, 58 degrees, 47 degrees, and 42 degrees. Consult the preset dielectric constant correction coefficient table and extract the correction coefficient for each section according to the corresponding temperature value, such as 1.03, 1.05, 1.08, 1.07, 1.04, and 1.02 respectively. Calculate the actual propagation time for each section according to these correction coefficients. The propagation time is greatly affected by the coefficient. The higher the dielectric constant correction coefficient, the slower the propagation speed. For example, in the first section, the propagation time is close to 344 nanoseconds, and the propagation times in the subsequent sections are 353 nanoseconds, 421 nanoseconds, 378 nanoseconds, 360 nanoseconds, and 337 nanoseconds respectively. The sum of the propagation times of the six sections is about 4.59 microseconds, which is the temperature-corrected propagation delay data value. During the correction process, it is necessary to ensure that the temperature readings of each section of the cable are correctly matched to a unique correction coefficient, and at the same time, cross-section interference is avoided. After all sections are adjusted, the overall corrected propagation time can be obtained.

[0072] S203: Call the propagation delay value after temperature correction, combine the total cable length and the propagation speed reference to calculate the correction value of the mutation point position, calibrate the length position of the propagation mutation point, establish the positioning interval of the carbonization path, and generate the positioning value of the partial discharge propagation delay carbonization path;

[0073] The specific calculation formula for the correction value of the mutation point position is:

[0074]

[0075] Where, L m is the correction value of the propagation mutation point position, L total represents the total cable length, t m represents the propagation delay value after temperature correction, v b represents the reference propagation speed, σ v represents the fluctuation coefficient of the propagation speed, ΔL q represents the length deviation of the adjacent qth section, r represents the total number of adjacent sections, α m represents the temperature correction factor;

[0076] In the formula, L total is the total cable length, and the actual length measured by the laser rangefinder for the cable laying path is 1200 meters;

[0077] t m is the propagation delay value after temperature correction. According to the influence model of temperature on the signal propagation speed in IEC 60287 standard, the original propagation delay is 0.85 μs, and the correction coefficient is 1.02 when the temperature sensor detects the ambient temperature of 35 °C. Calculate t m = 0.85 × 1.02 = 0.867 μs;

[0078] v b is the reference propagation speed. According to the nominal propagation speed of cross-linked polyethylene cables in IEC 60502-2, it is taken as 1.70×10 8 m / s;

[0079] σ v is the fluctuation coefficient of the propagation speed. According to the allowable deviation of the propagation speed of ±3% in the cable manufacturing standard GB / T 12706.1-2020, calculate σ v = 0.03;

[0080] ΔL q is the length deviation of the adjacent qth section. By measuring the lengths of each section of the cable separately as 400 meters, 410 meters, and 390 meters, the deviations from the designed length of 400 meters are 0 meters, 10 meters, and -10 meters. The sum of the absolute values is |ΔL1| + |ΔL2| + |ΔL3| = 0 + 10 + 10 = 20 meters;

[0081] r is the total number of adjacent sections, taking the measured number of segments as 3;

[0082] α m is the temperature correction factor. According to the formula for the influence of temperature gradient on the propagation path in IEEE 4 - 2013, when the temperature changes by 5°C, the correction factor is taken as 0.05;

[0083] Substituting into the formula, the numerator part of the calculation is 20×0.03 2 = 0.018, and the denominator part is 3×0.03 2 + 0.05 = 0.0527. The value inside the square root is 0.018 / 0.0527 ≈ 0.3416, and the square root value ≈ 0.5845; The first - half calculation is (1200×0.867×10 -6 ) / (1.70×10 8 ) ≈ 6.12 meters;

[0084] Finally, L m = 6.12×0.5845 ≈ 3.58 meters;

[0085] This result shows that the correction value of the propagation mutation point position is 3.58 meters. Combining with the total cable length of 1200 meters, it is calibrated that the mutation point is located 3.58 meters from the starting end of the third section, generating the positioning value of the partial - discharge propagation delay carbonization path.

[0086] The specific steps of S3 are as follows:

[0087] S301: Based on the positioning value of the partial - discharge propagation delay carbonization path, extract the time and space information in the path, and divide the cable axial measurement section in combination with the path change characteristics to generate the path section spacing value;

[0088] Based on the PD propagation delay carbonization path positioning value, it is necessary to collect signal arrival time data from multiple PD signal receiving nodes arranged along the cable. By comparing the time sequence of receiving the same discharge signal at these nodes and combining the propagation speed of electromagnetic waves in the insulating medium, the relative position of the PD source in the axial direction of the cable can be inferred. For example, suppose three receiving nodes are located at 0 meters, 5 meters, and 10 meters of the cable, and the time they receive the same signal is 100 nanoseconds, 150 nanoseconds, and 200 nanoseconds. Based on the propagation speed of about 200 million meters per second, it can be judged that the signal source is located near the middle node. Through the joint delay calculation of multiple nodes, a spatial outline of a continuous carbonization path can be formed in the axial direction of the cable. Furthermore, according to the change trend of this path in multiple time slices, different path features such as time continuous segments, path slope mutation segments, and return segments are extracted, and the path segments are divided using the change rate threshold. The division standard can be set as when the path direction changes by more than 15 degrees between adjacent segments, it is regarded as a path boundary. For example, if the path extension angle of the previous segment is 10 degrees and the next segment is 30 degrees, then this point is determined to be a boundary point. After completing the path feature recognition, the coordinate distance between the starting point and the end point of each segment is obtained to form a path segment spacing value sequence, which is calibrated in the cable axial length coordinate system.

[0089] S302: acquiring segment electric field strength data according to the path segment spacing value, extracting the area with significant intensity gradient change, determining the spatial position thereof, and generating a distribution value of the electric field strength mutation interval;

[0090] According to the path segment spacing value generated above, electric field measurement points are set for each segment to obtain the electric field strength values ​​at several fixed positions within each segment. The number of sampling points in each segment is determined according to the segment length and sampling accuracy. For example, for a segment with a length of 5 meters, three sampling points can be set to record the electric field strength data at these points respectively. The collected electric field values ​​are arranged according to the axial position and the gradient change is analyzed, and the intensity gradient value is estimated by the difference of adjacent points. On the electric field strength change curve, the area with prominent value changes can be found. It can be defined that when the increase in electric field strength per unit length exceeds 2.5 volts per meter squared, it is considered that there is a mutation phenomenon in this segment. For example, from the first point to the end point of a segment, the electric field strength is 20 volts per meter, 25 volts per meter and 40 volts per meter respectively. The rate of change in the latter segment is much higher than that in the previous segment, which is considered a mutation segment. By identifying the mutation positions in multiple segments, extracting their spatial coordinates, marking them as electric field strength mutation areas, and finally generating a set of mutation interval coordinate sets for positioning.

[0091] S303: Locate the coordinates of the carbonization path according to the distribution value of the electric field intensity mutation interval, extract the corresponding conductivity difference characteristics, establish the spatial association with the path points, and generate the carbonization path conductivity mutation matching map;

[0092] According to the spatial distribution of the aforementioned mutation intervals, select the central position points among them as the path nodes where carbonization may exist, and obtain the conductivity information of the material at these points through local conductivity measurement means. The measurement method can adopt the alternating current impedance method or the multi-probe method, and obtain the original conductivity value according to the known measuring point positions. Compare the conductivity value differences between two adjacent measuring points. If the conductivity at a certain position is 124 microsiemens per meter and the adjacent position is 53 microsiemens, the difference between the two is 67 microsiemens, which is significantly higher than the set identification threshold of 50 microsiemens, and it can be determined that this is the position where the conductivity performance has mutated. Corresponding all the determined mutation points to the original path coordinates, establish the mapping relationship between conductivity and spatial points, and assign weights to each point according to the degree of difference. The weight value can be calculated according to the proportion of the difference to the maximum difference, and form a matching map arranged in the path order. Finally, combine the mutation points in sequence to form a conductivity mutation matching map of the carbonization path for reference in subsequent applications.

[0093] The specific steps of S4 are as follows:

[0094] S401: Based on the conductivity mutation matching map of the carbonization path, obtain the emission time and initial amplitude information of the frequency band signal, combine the coordinates of the conductivity mutation area to determine the signal path starting point and the corresponding reference time, and generate a set of frequency band emission time values;

[0095] In the initial stage of cable detection, by collecting conductivity data arranged at equal intervals along the cable length direction, construct a map containing the conductivity mutation coordinates in the carbonization path. Usually, a sampling point is arranged every 1 meter, record the conductivity value of the corresponding point. If the change value of the conductivity at a certain point compared with the adjacent front and rear points exceeds 0.02 siemens per meter, it is regarded as a conductivity mutation point, and record its position as the key coordinate in the matching map. For example, if a mutation is detected at the 35th meter, it is marked as the reference starting point coordinate. Set a partial discharge pulse injection device at the cable head end. This device can output signals of multiple frequency bands, such as 20 MHz, 50 MHz, 100 MHz, and 200 MHz. Inject a single pulse signal into each frequency band in turn. The emission device records the accurate emission time and its initial amplitude of each frequency band pulse. To ensure the accuracy of the emission time, a standard calibration time can be set, such as 2 nanoseconds, and correct it when the actually measured emission time has an offset. Collect the emission time and amplitude data of all frequency bands, and mark their corresponding frequencies, record and form a set of data values. For example, 20 MHz corresponds to 1.8 nanoseconds and an amplitude of 3.2 volts, 50 MHz corresponds to 2.0 nanoseconds and an amplitude of 3.0 volts, and so on. Combine the previously constructed conductivity mutation map to determine the pulse starting path position and the reference emission time. For example, if the path starting point corresponding to 50 MHz falls at the 35-meter position, the reference time is set to 2.0 nanoseconds. The information of all frequency bands forms a record set respectively, laying a foundation for subsequent analysis.

[0096] S402: Call the set of frequency band emission time values, collect the arrival time and amplitude change of the frequency band signal at the end of the cable, record the propagation time and amplitude change according to the frequency band division, and generate a set of frequency band signal attenuation time ratios;

[0097] Using the set of frequency band emission time values formed in the previous stage, collect the actual arrival time and amplitude change of each frequency band signal at the end of the cable through a high-frequency signal receiving device. The device needs to be equipped with a filter module to distinguish multiple frequency band signals, and process each frequency band in turn. For example, for a 50 MHz signal, after recording its emission time at the head end as 2.0 nanoseconds and the initial amplitude as 3.0 volts, the arrival time of this frequency band signal is detected at the end as 4.8 nanoseconds, and the corresponding received amplitude is 1.2 volts. Thus, the signal propagation time is 2.8 nanoseconds. When the propagation distance is 100 meters, the corresponding propagation time change ratio is 0.028 nanoseconds per meter. At the same time, the signal amplitude drops from 3.0 volts to 1.2 volts, and the attenuation amplitude is 40%. Record the reception, time calculation, and amplitude change of all frequency band signals in this way to form a combined set of propagation time change ratios and amplitude attenuation ratios of frequency band signals. For example, the recorded value of 20 MHz is 0.032 nanoseconds per meter and the amplitude attenuation ratio is 65%. Record all frequency band data separately for subsequent calls in the carbonation depth judgment process.

[0098] S403: According to the set of frequency band signal attenuation time ratios, call the attenuation time ratio threshold corresponding to the frequency band in the feature library, calculate the comprehensive index of frequency band deviation, compare the degree of frequency band deviation, and determine the corresponding carbonation depth interval to generate the calibration value of the multi-frequency band carbonation depth feature library;

[0099] The specific calculation formula for the comprehensive index of frequency band deviation is:

[0100]

[0101] Among them, D b represents the comprehensive index of deviation of frequency band b, T b represents the measured attenuation time ratio of frequency band b, τ b represents the attenuation time ratio threshold corresponding to frequency band b in the feature library, ΔT b,k represents the difference in attenuation time ratio between frequency band b and the adjacent k-th frequency band, and n represents the total number of adjacent frequency bands;

[0102] In the formula, T b is the measured attenuation time ratio of frequency band b. The surface acoustic wave signal of the concrete specimen is collected through an electromagnetic attenuation tester, and the acoustic wave propagation time of frequency band b is measured as 1.45 μs according to ASTM C597 standard, and the reference signal propagation time is 1.00 μs. Calculate T b = 1.45 / 1.00 = 1.45;

[0103] τb is the decay time ratio threshold of frequency band b in the feature library. According to the relationship table between concrete carbonation depth and sound velocity specified in GB / T 30121-2013, the threshold τ corresponding to non-carbonated concrete is obtained b = 1.38;

[0104] ΔT b,k is the difference in decay time ratio between frequency band b and the adjacent k-th frequency band. By collecting the measured data of frequency bands b - 1 and b + 1, the decay time ratios of adjacent frequency bands are calculated to be 1.40 and 1.50 respectively, and ΔT b,1 = 1.45 - 1.40 = 0.05, ΔT b,2 = 1.45 - 1.50 = -0.03. After squaring and summing, we get 0.05 2 + (-0.03) 2 = 0.0034;

[0105] The total number of adjacent frequency bands n = 2. Substituting into the formula, the square root term is calculated as

[0106] The absolute value term is |1.45 - 1.38| / 1.38 = 0.07 / 1.38 ≈ 0.0507;

[0107] Finally, D b = 0.0507 × 0.0412 ≈ 0.00209. After taking the positive value of the square root After combining with the frequency band span correction factor, it is adjusted to 0.0707;

[0108] This result indicates that the deviation degree of frequency band b is 0.0707. According to the linear mapping relationship between carbonation depth and deviation degree in GB / T 50344-2019, the corresponding carbonation depth interval is 2.1 - 2.5 mm, and the calibration value of the feature library is generated

[0109] The specific steps of S5 are as follows:

[0110] S501: Call the calibration value of the multi-frequency band carbonation depth feature library and the denoised sheath circulating current component data, perform data registration and correlation extraction operations, establish the fitting relationship between the two types of data, and obtain the frequency band circulating current coupling coefficient;

[0111] To call the calibration values in the multi - frequency carbonation depth feature library and the processed sheath circulating current component data, it is necessary to first extract the carbonation depth reference values at multiple typical frequencies such as 50 Hz, 150 Hz, and 250 Hz, and at the same time obtain the circulating current component data collected on - site. Apply the wavelet packet threshold method to the collected data for denoising processing, align the two types of data in both the time and frequency dimensions, and uniformly process the two groups of data into the same number of time points. For example, if 1000 data points are collected per second and a total of 10 seconds of data is processed, that is 10,000 points. Use the interpolation method to ensure that the time nodes correspond. Then, perform difference correction on the circulating current data and carbonation depth data in the same frequency band, unify the reference differences through deviation subtraction operations. After registration, select a linear fitting model, use the minimum error strategy to calculate the corresponding relationship, and extract the coupling coefficient for each frequency band. For example, in the 150 - Hz frequency band, the coupling factor in the fitting result is 2.4, indicating that the proportionality factor between the change in the circulating current value and the carbonation depth in this frequency band is 2.4. After repeating this process to obtain the coupling values for all frequency bands, perform weighted combination of the coefficients of multiple frequency bands according to the preset weight ratio. For example, the weights corresponding to 50 Hz, 150 Hz, and 250 Hz are set to 20%, 30%, and 50% respectively, multiply them by the corresponding coupling coefficients of each frequency band and sum them up to obtain the final total coupling coefficient, forming a unified frequency - band circulating current coupling scalar for subsequent adjustment operations.

[0112] S502: Based on the frequency - band circulating current coupling coefficient, call the sheath circulating current component data for numerical ratio adjustment, and integrate the multi - frequency processing results to form a unified index to obtain the sheath carbonation indication degree;

[0113] After obtaining the frequency - band circulating current coupling coefficient, it is necessary to adjust the sheath circulating current component data of each frequency band in proportion. Reverse - calculate the original data according to the coupling coefficient to determine the true carbonation influence intensity. For example, in the 150 - Hz frequency band, if the collected circulating current value is 0.48 A and the corresponding coupling coefficient is 2.4, then the actual influence value is 0.2 A. After processing all frequency - band data in this way, perform unified processing using the standardization method. Subtract the historical mean value of the frequency band from each frequency - band value and then divide by the standard deviation to form a standardized index value. In this process, it can be assumed that the historical mean value of the 150 - Hz frequency band is 0.15 A and the standard deviation is 0.05 A, then the current positive deviation degree is 1. After completing the standardization conversion, superimpose the values of each frequency band according to the preset ratio. Set the adjusted weights of each frequency band to 30%, 30%, and 40% respectively, and then multiply them by the standardized results of each frequency band and sum them up to form a total value. For example, if the corresponding standardized values are 1, 0.5, and 0.2 respectively, the combined result is 0.53, which is used as the unified indication degree of the sheath carbonation degree.

[0114] S503: According to the sheath carbonation indication degree and the set warning threshold, perform grade judgment, combine the time - series information to generate a warning output, and obtain the sheath circulating current carbonation depth coupling warning signal;

[0115] According to the obtained sheath carbonization indication degree, perform grade division corresponding to the set early warning threshold, and define three grades. The value lower than 0.4 is the low grade, the value between 0.4 and 0.7 is the medium grade, and the value higher than 0.7 is the high grade. For example, if the current carbonization indication degree is 0.53 and it falls within the medium grade range, then perform the corresponding grade judgment. At the same time, establish a sequential early warning model in combination with the time change, continuously monitor the index value within the set period. If the sampling period is once every 10 seconds, and 3600 consecutive samplings are taken, it is 1 hour of data. Calculate the moving average of the data in units of minutes, and each time take 60 consecutive values to average to form a window value. If more than 2 of 3 adjacent windows exceed the medium grade judgment condition, it is determined that there is a signal of rising carbonization trend during this period. For example, the average values of 3 windows are 0.51, 0.55, and 0.48 respectively. Since two of them are within the range of 0.4 to 0.7 and meet the medium grade condition, the medium-level early warning signal of the coupling of sheath circulation carbonization depth is triggered.

[0116] Please refer to Figure 2 , an intelligent cable fault monitoring and early warning system, including:

[0117] The circulation noise reduction module obtains the phase angle information of the phase conductor in the metal sheath, identifies the degree of phase shift between adjacent conductors, judges whether it exceeds the phase difference threshold. If it is judged to exceed the threshold, adjust the sampling frequency band of the monitoring node, extract the characteristic components in the stable frequency band of the adjusted sheath circulation signal, and generate the denoised sheath circulation component data;

[0118] The propagation positioning module is based on the denoised sheath circulation component data, injects a partial discharge pulse signal at the cable head end and records the propagation time, establishes propagation delay correction data in combination with the insulation layer temperature and the correction coefficient, and converts it according to the total length of the cable to generate the positioning value of the partial discharge propagation delay carbonization path;

[0119] The conductance analysis module is based on the positioning value of the partial discharge propagation delay carbonization path, measures the change of the electric field strength distribution along the cable axis, extracts the position of the conductivity with mutation, spatially corresponds the mutation point to the path positioning value, and generates the carbonization path conductivity mutation matching map;

[0120] The carbonization calibration module is based on the carbonization path conductivity mutation matching map, inputs multiple groups of partial discharge pulses with different frequencies at the cable head end, obtains the propagation time and amplitude change of the signal in the end frequency band, extracts key parameters and then compares and identifies them to generate the calibration value of the multi-frequency carbonization depth feature library;

[0121] The early warning trigger module calls the calibration value of the multi-frequency carbonization depth feature library and the denoised sheath circulation component data, establishes a correlation factor model in combination with the carbonization depth and the circulation data, matches the grade early warning threshold, and generates the coupling early warning signal of the sheath circulation carbonization depth.

[0122] The above are only the preferred embodiments of the present invention, and do not limit the present invention in other forms. Any person skilled in the art may use the technical content disclosed above to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. An intelligent cable fault monitoring and early warning method, characterized in that: The following steps are involved: S1: Detect the phase angle of the metal sheath circulation, calculate the adjacent phase offset, and when the offset exceeds the phase difference threshold, adjust the monitoring node acquisition frequency to the preset anti-interference frequency band to extract the denoised sheath circulation component data; S2: Based on the de-noised sheath circulating current component data, a standard partial discharge pulse signal is injected into the head end of the cable and the terminal propagation time is recorded, the temperature gradient of the cable axial insulation layer is measured, the propagation delay is corrected by the temperature dielectric constant correction coefficient table, the carbonization path starting point is inverted according to the total length of the cable, and the partial discharge propagation delay carbonization path positioning value is generated; S3: Based on the carbonization path positioning value of the partial discharge propagation delay, the electric field intensity gradient distribution is measured along the cable axis, the spatial correlation between the conductivity mutation point and the carbonization path is matched, and a conductivity mutation matching map of the carbonization path is generated; S4: Based on the conductivity mutation matching map of the carbonization path, a local discharge pulse is injected at the head end of the cable, and the frequency band signal attenuation rate and propagation time are measured at the end, the frequency band attenuation time ratio is calculated, and the carbonization depth parameter is updated by comparing the feature library benchmark threshold to generate a multi-band carbonization depth feature library calibration value.

2. The intelligent cable fault monitoring and early warning method according to claim 1 is characterized in that: The denoised sheath circulating current component data includes phase offset, anti-interference frequency band parameters, and denoised data; the partial discharge propagation delay carbonization path positioning value includes propagation time, temperature gradient correction coefficient, total cable length parameter, and carbonization starting point coordinates; the carbonization path conductivity mutation matching map includes electric field gradient distribution, conductivity mutation point coordinates, and spatial correlation parameters; the multi-band carbonization depth feature library calibration values ​​include signal attenuation rate, propagation time difference, attenuation time ratio, and benchmark threshold parameters.

3. The intelligent cable fault monitoring and early warning method according to claim 1 is characterized in that: The specific steps of S1 are: S101: Detect the phase angle of the metal sheath circulating current, obtain the voltage and current time series data of adjacent monitoring nodes, extract the phase value of each node based on the time difference between the two, calculate the phase angle offset between adjacent nodes, and generate phase angle offset data; S102: comparing the phase angle offset data with the phase difference threshold, screening nodes whose offset exceeds the threshold, adjusting the sampling frequency of the node to a preset anti-interference frequency band, and generating a frequency adjustment coefficient value; S103: calling the sampling frequency corresponding to the frequency adjustment coefficient value to re-collect the node signal, constructing the signal envelope, extracting the data sequence whose fluctuation amplitude is within the limit range, and generating the sheath circulation component data after denoising.

4. The intelligent cable fault monitoring and early warning method according to claim 3 is characterized in that: The specific steps of S2 are: S201: Based on the de-noised sheath circulating current component data, a standard partial discharge pulse signal is injected at the head end of the cable, the propagation time of the pulse at the end is recorded, and the propagation delay of the signal is determined by combining the sheath loop structure and the sampling frequency information to generate a propagation delay data value; S202: According to the propagation delay data value, the temperature distribution of the cable axial insulation layer is obtained, the temperature value is matched with the dielectric constant correction coefficient in the correction coefficient table, the segment propagation time is adjusted, and the temperature-corrected propagation delay value is generated; S203: calling the temperature-corrected propagation delay value, combining the total cable length and the propagation speed reference, calculating the mutation point position correction value, calibrating the length position of the propagation mutation point, establishing the positioning interval of the carbonization path, and generating the partial discharge propagation delay carbonization path positioning value.

5. The intelligent cable fault monitoring and early warning method according to claim 4 is characterized in that: The specific calculation formula for calculating the mutation point position correction value is: Among them, L m Propagation mutation point position correction value, L total Represents the total length of the cable, t m represents the temperature-corrected propagation delay value, v b represents the reference propagation velocity, σ v Represents the fluctuation coefficient of propagation speed, ΔL q represents the length deviation of the adjacent qth segment, r represents the total number of adjacent segments, α m Represents the temperature correction factor.

6. The intelligent cable fault monitoring and early warning method according to claim 4 is characterized in that: The specific steps of S3 are: S301: extracting time and space information in the path based on the partial discharge propagation delay carbonization path positioning value, dividing the cable axial measurement section in combination with the path change characteristics, and generating the path section spacing value; S302: acquiring segment electric field strength data according to the path segment spacing value, extracting the area with significant intensity gradient change, determining the spatial position thereof, and generating a distribution value of the electric field strength mutation interval; S303: Locate the coordinates of the carbonization path according to the distribution value of the electric field intensity mutation interval, extract the corresponding conductivity difference characteristics, establish a spatial association with the path points, and generate a carbonization path conductivity mutation matching map.

7. The intelligent cable fault monitoring and early warning method according to claim 6 is characterized in that: The specific steps of S4 are: S401: Based on the conductivity mutation matching map of the carbonization path, the emission time and initial amplitude information of the frequency band signal are obtained, the starting point of the signal path and the corresponding reference time are determined in combination with the conductivity mutation area coordinates, and a frequency band emission time value set is generated; S402: calling the frequency band transmission time value set, collecting the arrival time and amplitude change of the frequency band signal at the end of the cable, recording the propagation time and amplitude change according to the frequency band division, and generating a frequency band signal attenuation time ratio group; S403: According to the frequency band signal attenuation time ratio group, the attenuation time ratio threshold corresponding to the frequency band in the feature library is called, the frequency band deviation comprehensive index is calculated, the frequency band deviation degree is compared and the corresponding carbonization depth interval is determined, and the multi-band carbonization depth feature library calibration value is generated.

8. The intelligent cable fault monitoring and early warning method according to claim 7 is characterized in that: The specific calculation formula for calculating the frequency band deviation comprehensive index is: Among them, D b Represents the deviation comprehensive index of frequency band b, T b represents the measured decay time ratio of frequency band b, τ b Represents the decay time ratio threshold corresponding to frequency band b in the feature library, ΔT b,k Represents the difference in attenuation time ratio between frequency band b and the adjacent kth frequency band, and n represents the total number of adjacent frequency bands.

9. The intelligent cable fault monitoring and early warning method according to claim 1 is characterized in that: The method further comprises: S5: calling the calibration value of the multi-band carbonization depth feature library and the denoised sheath circulation component data, calculating the circulation carbonization correlation factor, comparing the graded warning threshold to trigger the corresponding warning level, and generating a sheath circulation carbonization depth coupling warning signal; The sheath annular carbonization depth coupling warning signal includes a correlation factor, a graded warning threshold, and a warning level parameter; The specific steps of S5 are: S501: calling the calibration value of the multi-band carbonization depth feature library and the denoised jacket circulation component data, performing data registration and association extraction operations, establishing a fitting relationship between the two types of data, and obtaining the frequency band circulation coupling coefficient; S502: Based on the frequency band circulation coupling coefficient, the sheath circulation component data is called to adjust the numerical ratio, and the multi-band processing results are integrated to form a unified index to obtain the sheath carbonization indication degree; S503: Performing a level judgment according to the jacket carbonization indication and a set warning threshold, generating a warning output in combination with time series information, and obtaining a jacket circulation carbonization depth coupling warning signal.

10. An intelligent cable fault monitoring and early warning system, characterized in that: According to any one of claims 1 to 9, the intelligent cable fault monitoring and early warning method comprises: The circulating current denoising module obtains the phase angle information of the phase conductor in the metal sheath, identifies the phase offset degree between adjacent conductors, and determines whether it exceeds the phase difference threshold. If it is determined to exceed the threshold, the sampling frequency band of the monitoring node is adjusted, and the characteristic components in the stable frequency band of the adjusted sheath circulating current signal are extracted to generate denoised sheath circulating current component data; The propagation positioning module injects a partial discharge pulse signal at the head end of the cable based on the de-noised sheath circulating current component data and records the propagation time, establishes propagation delay correction data in combination with the insulation layer temperature and the correction coefficient, converts the data against the total cable length, and generates a partial discharge propagation delay carbonization path positioning value; The conductivity analysis module measures the change of the electric field intensity of the cable axial distribution based on the partial discharge propagation delay carbonization path positioning value, extracts the conductivity position with mutation, spatially corresponds the mutation point with the path positioning value, and generates a carbonization path conductivity mutation matching map; The carbonization calibration module inputs multiple sets of frequency-difference partial discharge pulses at the head end of the cable based on the conductivity mutation matching map of the carbonization path, obtains the propagation time and amplitude changes of the terminal frequency band signal, extracts key parameters, compares and identifies them, and generates a multi-band carbonization depth feature library calibration value; The warning trigger module calls the calibration value of the multi-band carbonization depth feature library and the denoised jacket circulation component data, combines the carbonization depth and circulation data to establish a correlation factor model, matches the level warning threshold, and generates a jacket circulation carbonization depth coupling warning signal.

Images