A method and system for detecting the insulation performance of a cable protection tube

Abstract

The application provides a kind of cable protection pipe insulation performance detection method and system.Belongs to the technical field of electrical test and material degradation evaluation of power cable auxiliary facilities.The method comprises: carrying out global resistance preliminary measurement on the sample data of the pipe body to be measured, collecting the initial potential distribution of the pipe body surface and the environmental temperature and humidity parameters, and generating the pipe body basic state data set;According to the preset gradient, gradually apply the radial load to simulate the uneven settlement stress of soil body, synchronously collect the strain response data of the pipe body at different deformation positions, and generate the pipe body deformation distribution data set with continuous bending stress;Carry out global potential dynamic scanning on the pipe body deformation distribution data set, and generate the potential abnormal point position marking data set;Through the combination of coaxial cavity loading and four-point bending simulation working condition, insulation detection can be completed under the actual buried stress state, effectively prevent the situation that only rely on static no-load test misses micro-crack hidden danger, greatly improve the matching degree of on-site detection result and actual working condition.

Description

Technical Field

[0001] The present invention proposes an insulation performance detection method and system for cable protection pipes, belonging to the technical field of electrical tests for power cable auxiliary facilities and material degradation assessment. Background Art

[0002] In current domestic cable corridor projects, the verification of the insulation performance of cable protection pipes has long followed the conventional procedures of overall megohmmeter shaking test and laboratory static power frequency withstand voltage test, and the deviation problems exposed by such detection methods in actual engineering scenarios have not been systematically solved for a long time. Most on-site operators use the method of inserting a copper brush into the pipe orifice to complete the resistance reading of the entire pipe. However, during the stages of storage, open-air transportation and pre-burial storage of the pipes, the outer surface is extremely prone to adsorb environmental condensation. During the test, the pointer of the megohmmeter often swings violently irregularly, and most operators directly judge the insulation as qualified on the grounds that the on-site environmental humidity exceeds the standard, completely covering up the real insulation resistance drop problem caused by micro-crack carbonization on the inner wall of the pipe.

[0003] The current laboratory power frequency withstand voltage test requires the pipe body to be placed flat without any external load throughout the whole process, which is completely divorced from the long-term bending working condition caused by uneven settlement of backfill soil after burial. The subcritical partial discharge precursor formed after moisture invades the micro-cracks generated on the inner surface layer of the pipe body under the tensile effect cannot be effectively excited in the static pressurization environment, and a large number of pipes with hidden defects directly enter the laying link. Meanwhile, the existing detection evaluates the entire protection pipe as a single resistance unit, without separate insulation verification for the spigot-and-socket rubber ring and the connection area of the half coupling. More than 90% of the cable outer sheath wear accidents are rooted in the sealing failure of the protection pipe joint or the fastening bolt piercing the insulation layer, and such joint insulation hidden dangers have always been in the detection blind area. Conventional detection can only give a binary judgment of pass or fail at the current moment, cannot quantify the insulation degradation trend corresponding to the current moisture absorption rate of the pipe, and cannot provide early warning for corridor operation and maintenance before the rainy season comes. The industry has always lacked an in-situ insulation decomposition detection scheme suitable for small-diameter multi-material protection pipes. Summary of the Invention

[0004] The present invention provides an insulation performance detection method and system for cable protection pipes to solve the problems mentioned in the above background art. The insulation performance detection method for cable protection pipes proposed by the present invention includes the following steps: S1. Perform a full-domain initial resistance test on the sample data of the pipe body to be tested, collect the initial potential distribution on the pipe body surface and environmental temperature and humidity parameters, and generate a basic state data set of the pipe body. S2. Apply radial loads step by step according to the preset gradient to simulate the uneven settlement stress of the soil, and simultaneously collect strain response data at different deformation locations of the pipe to generate a deformation distribution dataset of the pipe with continuous bending stress; perform a full-domain potential dynamic scan on the deformation distribution dataset of the pipe to generate a potential anomaly point marker dataset. S3. Inject a high-frequency step pulse signal into the pipe body under test with continuous bending stress, collect the full-path reflection waveform of the pulse propagating along the pipe body, decompose the time-domain features of the reflection waveform to generate full-path impedance distribution sequence data; perform spatial coordinate conversion on the full-path impedance distribution sequence data to generate a spatial location dataset of insulation anomaly points. S4. Under the condition of maintaining a stable four-point bending load, a stepped AC test voltage is synchronously applied to the tube body, and partial discharge pulse signals at different voltage levels are continuously collected to generate a phase-resolved partial discharge spectrum dataset; feature clustering is performed on the phase-resolved partial discharge spectrum dataset to generate a discharge type classification label dataset. S5. Apply a wideband sweep test signal to the marked insulation anomaly points to generate a full-band dielectric characteristic response dataset; perform feature fitting on the full-band dielectric characteristic response dataset to obtain the moisture absorption degree and insulation degradation rate parameters at different locations of the pipe, and generate a meter-by-meter insulation state quantification dataset. S6. Input the meter-by-meter insulation state quantification dataset into the preset Weibull breakdown probability model, substitute the insulation degradation rate parameters at different locations to complete the probability iteration calculation, output the insulation degradation distribution results of the entire pipe body and the surface flashover remaining time parameters, and generate a complete cable protection pipe insulation performance test report data.

[0005] This invention proposes an insulation performance testing system for cable protection pipes, the system comprising: One or more processors; Memory, used to store one or more programs; Wherein, when the one or more programs are executed by the one or more processors, the one or more processors are made to implement the method described in any one of the above.

[0006] The beneficial effects of this invention are as follows: By combining coaxial cavity loading with four-point bending simulation, insulation testing can be completed under conditions closely resembling actual buried stress, effectively preventing the omission of micro-crack hazards by relying solely on static no-load testing, and significantly improving the matching degree between on-site test results and actual working conditions. Simultaneously, the combined positioning method of high-frequency TDR and partial discharge spectrum can directly pinpoint the specific location of each insulation anomaly, eliminating the need to treat the entire pipe as a single resistance unit for general judgment, significantly reducing the probability of missed detections in joint areas. By capturing water molecule polarization characteristics through broadband complex dielectric sweep frequency, the degree of hidden moisture inside the pipe can be accurately identified, effectively avoiding the problem of subjective judgment of compliance based solely on megohmmeter fluctuation data. It can provide the current meter-by-meter insulation status and predict the remaining time of subsequent flashovers, leaving sufficient handling window for maintenance. Attached Figure Description

[0007] Figure 1 This is a diagram illustrating the steps of the method described in this invention; Figure 2 This is a detailed flowchart of step S4 in this invention. Detailed Implementation

[0008] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0009] One embodiment of the present invention, such as Figure 1 As shown, a method for testing the insulation performance of a cable protection pipe includes: S1. Seal both ends of the cable protection pipe to be tested and connect it to the high-voltage electrode and the grounding electrode to form a closed coaxial test cavity between the inner cavity of the pipe and the outer metal layer. Introduce circulating water vapor with a preset humidity into the cavity to complete the uniform condensation adhesion treatment on the entire surface of the pipe and generate sample data of the pipe under test with a uniform condensation layer. Perform a preliminary global resistance measurement on the sample data of the pipe under test, collect the initial potential distribution on the surface of the pipe and the ambient temperature and humidity parameters, and generate a basic state dataset of the pipe. S2. Place the pipe sample with a uniform condensation layer on a four-point bending loading platform, apply radial loads step by step according to a preset gradient to simulate the uneven settlement stress of the soil, and simultaneously collect strain response data at different deformation locations of the pipe to generate a pipe deformation distribution dataset with continuous bending stress; perform a full-domain potential dynamic scan on the pipe deformation distribution dataset to capture the abrupt change points of the pipe surface potential under stress and generate a potential anomaly point marker dataset. S3. Inject a high-frequency step pulse signal into the pipe body under test with continuous bending stress, collect the full-path reflection waveform of the pulse propagating along the pipe body, decompose the time-domain features of the reflection waveform, extract the echo delay and amplitude parameters of different impedance abrupt change points, and generate full-path impedance distribution sequence data; perform spatial coordinate conversion on the full-path impedance distribution sequence data, map each impedance abrupt change point to the corresponding length position of the pipe body, and generate a spatial location dataset of insulation anomaly points. S4. Under the condition of maintaining a stable four-point bending load, a stepped AC test voltage is synchronously applied to the tube body, and partial discharge pulse signals at different voltage levels are continuously collected. The phase distribution and amplitude variation characteristics of the pulses are extracted to generate a phase-resolved partial discharge spectrum dataset. Feature clustering is performed on the phase-resolved partial discharge spectrum dataset to distinguish three types of signals: microcrack discharge in the tube body matrix, surface condensation discharge along the surface, and joint gap discharge, and a discharge type classification label dataset is generated. S5. Apply a wideband sweep test signal to the marked insulation anomaly points, collect the complex dielectric constant response curve in the 10Hz to 1MHz frequency band, extract the characteristic frequency points and dielectric loss peak parameters corresponding to water molecule polarization relaxation, and generate a full-band dielectric characteristic response dataset; perform feature fitting on the full-band dielectric characteristic response dataset to obtain the moisture absorption degree and insulation degradation rate parameters at different locations of the pipe, and generate a meter-by-meter insulation state quantitative dataset. S6. Input the meter-by-meter insulation state quantification dataset into the preset Weibull breakdown probability model, substitute the insulation degradation rate parameters at different locations to complete the probability iteration calculation, output the insulation degradation distribution results of the entire pipe body and the surface flashover remaining time parameters, and generate a complete cable protection pipe insulation performance test report data.

[0010] The working principle and effects of the above technical solution are as follows: By simulating soil settlement stress and combining it with condensation environment for insulation testing, the actual service conditions of the cable protection pipe can be fully reproduced, significantly improving the accuracy and authenticity of insulation performance test results and reducing the deviation caused by conventional static testing. Through the combined use of pulse waveform analysis and spatial positioning, the abnormal insulation points inside the pipe can be accurately located, effectively reducing the occurrence of missed or false detections of insulation defects. By clustering the characteristics of partial discharge signals to distinguish different discharge types, the causes of various insulation defects can be accurately identified, and surface moisture defects and structural damage to the substrate can be specifically differentiated. By combining full-band dielectric characteristic testing with probabilistic model deduction, the insulation status and deterioration progress of the pipe can be refined meter by meter, enhancing the comprehensiveness of the investigation of potential insulation hazards in cable protection pipes and avoiding the problem that a single testing dimension cannot simultaneously address surface moisture and internal structural damage, leading to one-sided test results and an inability to accurately predict the service life of the equipment.

[0011] In one embodiment of the present invention, S1 includes: S11. Clean the residual dust and oil on the inner and outer walls of the cable protection pipe at both ends of the test pipe, grind the metal contact area of ​​the pipe opening to remove the oxide film, complete the flat treatment of the electrode bonding contact surface, and output the pipe opening pre-treatment to complete the pipe body physical marking information. S12. Secure the high-voltage electrode to one end of the tube cavity, attach the grounding electrode to the metal cladding layer of the outer wall of the tube, fill the gap between the electrode and the tube wall with insulating sealant, build a coaxial closed cavity without air leakage, and output the information that the sealed cavity assembly is complete. S13. Adjust the output of the constant temperature and humidity water vapor generator to a fixed relative humidity of circulating gaseous water vapor, control the water vapor to flow through the closed cavity at a uniform speed, and continuously maintain a fixed circulation time so that a condensation film of uniform thickness is formed on the entire area of ​​the inner and outer walls of the tube, and output the full-area condensation adhesion to complete the sample data of the tube to be tested. S14. Set up an array-type surface potential acquisition probe and move it at a constant speed along the tube axis to scan, simultaneously reading the real-time temperature value inside the cavity, the air humidity value around the cavity, and the DC resistance value of the entire tube wall. Integrate all the acquired values ​​to form a complete initial potential distribution sequence of the tube surface, and output the integrated basic state dataset of the tube.

[0012] The working principle and effects of the above technical solution are as follows: By meticulously cleaning and polishing the cable protection pipe opening, a tight and uniform fit between the electrode and the pipe body is ensured, reducing interference from contact impurities and oxide layers in the test data. Constructing a sealed, leak-free coaxial test chamber maintains a stable, sealed testing environment, preventing external airflow and humidity fluctuations from interfering with condensation adhesion. Uniform water vapor circulation control ensures a uniform condensation layer across the entire pipe body, guaranteeing a consistent sample state in the early stages of testing and avoiding test deviations caused by uneven local condensation. Full-area scanning of potential, temperature, humidity, and resistance parameters by the array probe captures the initial electrical state of the pipe body comprehensively, ensuring the comprehensiveness and consistency of basic test data. This effectively improves the overall accuracy of subsequent insulation test data and avoids distortion of overall test results due to improper sample preparation and basic data acquisition.

[0013] In one embodiment of the present invention, S2 includes: S21. Raise the four-point bending loading platform support bracket to the standard horizontal height, calibrate the coaxiality and horizontality of the four support rollers of the platform, remove debris and debris from the platform bearing area, and output the loading platform calibration ready working condition record information. S22. The tube to be tested, with a uniform condensation layer on the transfer belt, is placed stably in the middle bearing area of ​​the four support rollers. The limiting blocks clamp the two ends of the tube to prevent axial slippage during the load application process. The tube is then fixed in place to complete the physical positioning information. S23. Drive the loading head to slowly press down from zero load starting point according to equal differential gradient. After each single load is stabilized for a fixed time, read the output values ​​of the strain gauges pasted on the top, bottom, left and right sides of the pipe body. Continuously record the strain values ​​of the entire pipe section corresponding to each load level throughout the process, and output a complete dataset of pipe deformation distribution with continuous bending stress. S24. The mobile multi-channel potential acquisition module traverses the outer wall of the pipe segmentally at a uniform speed to collect potential values. It compares the potential difference between adjacent acquisition points to filter out points that exceed the stable fluctuation threshold. It records the axial length coordinates and radial orientation of the corresponding points and outputs the point coordinate labeling to complete the potential abnormal point marking dataset.

[0014] The working principle and effects of the above technical solution are as follows: Precise calibration and debris removal of the bending loading platform ensures stable and uniform load application, reducing test errors caused by equipment assembly deviations. Fixing the pipe position using a limiting structure prevents pipe slippage and swaying during testing, maintaining overall test stability. Gradual, step-by-step pressurization and stable pressure collection of strain data closely matches the stress characteristics of uneven soil settlement, improving the accuracy and relevance of pipe deformation data. Multi-channel module segmented collection of potential data comprehensively covers all test areas of the pipe while accurately identifying specific points of potential anomalies, enhancing the precision of anomaly location identification. This effectively avoids distortion of deformation data caused by uneven load application and pipe displacement, making the correlation between stress and potential changes more consistent with actual working conditions.

[0015] In one embodiment of the present invention, step S3 includes: S31. The pulse signal transmitting terminal is connected to the high voltage electrode at one end of the tube body. The signal acquisition terminal is connected to the grounded metal layer on the outer wall of the tube body. The shielded cable wraps the entire signal transmission line to isolate external electromagnetic noise. The output pulse signal transceiver circuit is built and the working condition is recorded. S32. The pulse generator outputs a high-frequency step pulse signal with a fixed rise rate and introduces it into the coaxial cavity of the tube. The high-speed waveform acquisition card continuously captures the continuous waveform curve formed by the pulse reflecting back and forth inside the insulation layer of the tube wall, and outputs the original pulse full-path reflection waveform data for storage. S33. Decompose each independent echo pulse unit in the original reflected waveform, extract the time interval and voltage peak interval corresponding to a single pulse, and count the time difference and voltage peak value of each echo unit from transmission to return reception. Summarize all unit values ​​to form an ordered full path impedance distribution sequence data. S34. The length conversion is completed by matching the pulse propagation fixed wave velocity value with the echo delay value. The severity level of impedance change is distinguished according to the echo amplitude. The axial length coordinate obtained by the conversion is simultaneously bound with the radial azimuth coordinate recorded in the previous stage. The coordinate level binding is output to complete the spatial positioning dataset of insulation abnormality point.

[0016] The working principle and effects of the above technical solution are as follows: By constructing a pulse signal transceiver circuit with shielded lines, external electromagnetic interference can be isolated, improving the purity of pulse signal transmission and reducing the adverse effects of noise signals on waveform acquisition. Through high-frequency step pulse penetration detection, the reflected waveform inside the pipe insulation layer can be completely captured, allowing for comprehensive acquisition of insulation impedance change information. By breaking down and refining the time domain and amplitude parameters of each echo pulse, the details of impedance changes can be refined, improving the regularity and accuracy of the impedance data sequence. Through a positioning method combining wave velocity conversion and azimuth binding, precise spatial positioning of insulation anomaly points can be achieved, and the severity of different defects can be distinguished, effectively improving defect positioning accuracy. This avoids the problem that traditional detection methods can only roughly determine the fault area and cannot accurately locate specific points, making insulation defect investigation more detailed and comprehensive.

[0017] In one embodiment of the present invention, S33 includes: Read the original pulse full-path reflection waveform storage data, perform full-domain segmentation processing on the continuous waveform data according to the pulse waveform baseline threshold, remove waveform base noise and invalid clutter data, separate all independent and complete echo pulse units within the waveform period, and generate pure independent echo pulse unit set data; For each independent echo pulse unit in the ensemble data, the time domain interval is truncated, the start acquisition time of the single pulse signal and the time when the pulse signal decays to zero are locked, the complete time coverage interval corresponding to a single pulse is calibrated, and pulse time domain interval labeling data is generated. The amplitude range of each echo pulse unit that has been annotated in the time domain is truncated, the lowest voltage baseline value and the highest peak voltage value of the single pulse waveform are captured, the voltage amplitude fluctuation range corresponding to a single pulse is locked, and pulse amplitude range annotation data is generated. The time difference between the initial signal transmission time and the signal return reception time corresponding to each pulse unit is calculated. The peak voltage amplitude value corresponding to each pulse unit is recorded synchronously. The time delay parameters and amplitude parameters of all pulse units are collected in batches to generate pulse characteristic parameter collection data. All collected feature parameters are arranged and organized in an orderly manner according to the pulse propagation time sequence, and the parameter storage format and arrangement dimension are unified. All pulse feature parameters are integrated to form a full-path impedance distribution sequence data with unified dimension and coherent time sequence.

[0018] The working principle and effects of the above technical solution are as follows: By performing global noise reduction and segmentation processing on the original pulse waveform, it is possible to filter out background noise and invalid signals from the detection environment, improve the purity of the echo pulse unit, and reduce the interference of noise data on subsequent parameter analysis. By precisely trunculating and calibrating the pulse time domain and amplitude range one by one, the signal details corresponding to each impedance change can be completely preserved, making the acquisition of pulse characteristic parameters more detailed and comprehensive. By batch aggregating the core parameters of pulse delay and amplitude, it is possible to organize scattered signal data and improve the regularity of parameter statistics. By uniformly arranging various characteristic parameters in time sequence, it is possible to ensure the temporal continuity of impedance distribution data and maintain the uniformity of the overall data dimension, effectively avoiding the insulation impedance analysis deviation caused by waveform data disorder and feature loss, making the results of subsequent defect location and grade determination more accurate and reliable.

[0019] In one embodiment of the present invention, S34 includes: Retrieve the time delay parameters of each echo unit stored in the full path impedance distribution sequence data, match the fixed propagation wave velocity parameters of the pulse in the insulation medium of the cable protection pipe, perform medium propagation distance conversion calculation, obtain the axial length position parameters corresponding to each impedance change point, and generate the original data of the axial coordinates of the insulation abnormal point. Read the voltage amplitude parameters of each echo unit inside the full path impedance distribution sequence data, and complete the hierarchical comparison and division by combining the threshold range of impedance change amplitude of the insulating medium. Classify the impedance change points of different degrees of anomaly and generate impedance change severity level labeling data. Retrieve the radial orientation data of the pipe body retained in the previous testing process, and dimensionally associate and bind the axial length parameters and radial orientation parameters of each impedance change point to form a single point all-round spatial parameter combination data, and generate multi-dimensional point spatial parameter set data. The multi-dimensional spatial parameter set data is standardized and verified, invalid point data with excessive parameter matching deviation are removed, the spatial coordinate arrangement order and grade labeling format of all valid points are unified, and a standardized spatial location dataset of insulation anomaly points is generated.

[0020] The working principle and effects of the above technical solution are as follows: By combining pulse delay with medium wave velocity for distance conversion, the axial position of insulation anomaly points can be accurately calculated, improving the accuracy of longitudinal positioning of pipe fault points. Through voltage amplitude parameter comparison and grading, the severity of various impedance abrupt changes can be distinguished, achieving quantitative differentiation of insulation defects. By binding axial parameters with radial orientation data, complete spatial point information can be constructed, allowing fault location determination to no longer be limited to a single dimension. By performing regularization verification and invalid data removal on multi-dimensional spatial parameters, detection errors caused by parameter matching deviations can be reduced, improving the effectiveness and standardization of overall positioning data. This entire processing method can achieve precise three-dimensional positioning of insulation defects while ensuring data standardization, avoiding problems such as inaccurate fault judgment and excessive positioning deviation caused by messy original signal parameters and missing dimensions, continuously improving the overall accuracy of insulation defect detection.

[0021] One embodiment of the present invention, such as Figure 2 As shown, S4 includes: S41. Lock the current downward stroke of the four-point bending loading platform head to maintain constant load operation, disconnect the pulse signal transceiver circuit, reconnect the frequency converter AC voltage output terminal to the electrodes at both ends of the pipe body, and record the information after the output steady-state load voltage test circuit is switched. S42. The AC voltage source gradually increases the output amplitude from the reference low voltage starting point according to a fixed voltage difference. Each voltage level is continuously stabilized for a fixed acquisition time. The high-frequency partial discharge acquisition equipment continuously captures the discharge pulse waveform released by the pipe wall insulation layer and outputs multi-voltage level raw partial discharge pulse storage data. S43. Extract the AC phase interval and pulse peak value variation amplitude corresponding to the discharge pulse in each AC voltage cycle. Plot a continuous spectrum with phase value as the horizontal axis and pulse amplitude as the vertical axis. Integrate the spectrum content of all voltage levels to form a complete phase-resolved partial discharge spectrum dataset. S44. Extract three core features from the pulse clusters within the spectrum: the phase concentration interval, the pulse amplitude fluctuation interval, and the pulse repetition frequency. Group the clusters by feature similarity to complete the cluster classification. Match the exclusive feature intervals of the three types of discharge phenomena to complete the signal classification. Output the category label to complete the discharge type classification and labeling dataset.

[0022] The working principle and effects of the above technical solution are as follows: By maintaining a constant bending load during voltage test circuit switching, the stress state of the pipe can be kept consistent, avoiding interference from changes in operating conditions on partial discharge signal acquisition and ensuring uniform and stable test conditions. By progressively increasing the AC voltage and stabilizing it for signal acquisition, the changes in discharge pulses under different voltage conditions can be fully captured, enriching the coverage of partial discharge data and reducing signal omissions. By combining phase and amplitude parameters to draw partial discharge maps, the distribution pattern of discharge signals can be intuitively presented, improving the intuitiveness and completeness of discharge feature display. Through multi-dimensional pulse feature clustering and classification, discharge signals of different causes can be accurately distinguished, and the fault types of insulation defects can be refined, effectively reducing the misjudgment of faults caused by the confusion of different discharge signals and significantly improving the accuracy of identifying insulation defect types in cable protection pipes.

[0023] In one embodiment of the present invention, S44 includes: Read the complete phase-resolved partial discharge spectrum dataset, perform global feature picking on all discharge pulse clusters distributed within the spectrum, batch collect the phase concentration interval, amplitude fluctuation interval and pulse repetition frequency corresponding to each pulse cluster, and generate a multi-dimensional original feature dataset of partial discharge pulse clusters. Feature normalization processing is performed on the multidimensional original feature dataset of partial discharge pulse clusters to unify the quantization standard of pulse features under different voltage levels, eliminate the feature offset problem caused by the difference in test conditions, and generate a standardized pulse cluster feature dataset. A global similarity comparison is performed on the standardized pulse cluster feature dataset, and the pulse clusters are automatically grouped and divided according to the degree of feature overlap, generating multiple groups of pulse cluster grouping data with the same feature origin; The system retrieves the specific characteristic reference intervals of various insulation discharge phenomena stored in the system, compares and matches the characteristic data of each group of pulse clusters with the reference intervals across the entire domain, completes the classification and assignment of different pulse signals, and generates preliminary classification data of discharge signals. The preliminary classification data of discharge signals are verified and standardized as a whole, the category labeling format of all discharge signals is unified, the classification labeling of all pulse signals is solidified, and a classification labeling dataset of discharge type is generated.

[0024] The working principle and effects of the above technical solution are as follows: By capturing multi-dimensional features of partial discharge pulse clusters across the entire domain, key information of various discharge signals can be comprehensively retained, reducing analytical biases caused by incomplete feature acquisition. Normalization of the original features smooths out feature differences caused by different test voltages, ensuring consistent comparison standards for various pulse features and improving the fairness and reliability of data comparison. Grouping pulse clusters through feature similarity comparison enables automatic aggregation of discharge signals from the same source, improving the regularity of signal classification. Full-domain matching and identification with discharge reference feature intervals accurately distinguishes discharge signals corresponding to different insulation defects, reducing confusion between different discharge types. Final data verification and format standardization solidify accurate classification results and unify data output standards, avoiding errors in defect judgment caused by scattered and disordered feature data, effectively improving the accuracy and stability of discharge fault identification in cable protection pipes.

[0025] In one embodiment of the present invention, step S5 includes: S51. The mobile broadband impedance analysis probe is sequentially attached to each insulation abnormality point marked with potential and impedance. The space between the probe and the pipe wall is filled with conductive coupling medium to eliminate air gaps. The single-point test probe attachment ready status record information is output. S52. The impedance analysis equipment starts from a low frequency of 10Hz and moves upward at a constant speed to complete the full-band linear frequency sweep output. It simultaneously reads the real part and imaginary part of the complex dielectric constant corresponding to each frequency node, records the continuous change curve of the entire frequency band, and outputs the original dielectric constant change curve of a single point for storage. S53. Screen the frequency characteristic nodes generated by the polarization motion of water molecules adsorbed on the pipe wall inside the curve, record the frequency values ​​corresponding to the nodes and the maximum peak value of dielectric loss at the concave part of the curve, and summarize all abnormal point frequency band parameters to form a complete dielectric characteristic full-frequency band response dataset. S54. A nonlinear curve fitting algorithm is used to process the numerical values ​​of the full-band response curve. The trend of the fitted curve is used to obtain the total amount of water accumulation inside the pipe wall at the corresponding point and the change rate of molecular chain breakage of the insulation material. The two types of values ​​are summarized in segments according to each meter of pipe length, and the segmented numerical summaries are output to complete the meter-by-meter insulation state quantification dataset.

[0026] The working principle and effects of the above technical solution are as follows: By filling the probe with a conductive coupling medium to eliminate the air gap between the probe and the pipe wall, the frequency signal attenuation caused by the air layer can be reduced, improving the accuracy of dielectric parameter acquisition. Through a linear continuous frequency sweep from 10 Hz to 1 MHz, the dielectric response changes of the insulating material over a wide frequency range can be fully captured, reducing feature loss caused by single-frequency point testing. By screening the frequency peak nodes corresponding to water molecule polarization, the insulation degradation characteristics caused by moisture can be extracted separately, distinguishing between material loss itself and loss changes caused by water vapor. By nonlinear fitting to convert the total moisture content and material aging rate, and then summarizing the values ​​by meter segment, the degree of insulation degradation at a single point can be quantified, and the overall insulation change trend of the pipe can be visually displayed. This avoids the problem that narrow-frequency testing can only qualitatively determine the degree of moisture and cannot quantify the aging progress, significantly improving the precision of insulation condition assessment.

[0027] In one embodiment of the present invention, step S6 includes: S61. Remove out-of-limit abnormal values ​​generated during the internal acquisition process of the meter-by-meter insulation state quantization dataset, rearrange the remaining valid values ​​in ascending order of the axial length of the pipe body, and output the cleaned and sorted standardized segmented quantization data sequence. S62. Retrieve the Weibull breakdown probability basic model pre-stored in the detection system, and sequentially import the total moisture absorption value and insulation degradation rate value of each meter interval in the standardized segmented quantized data sequence into the model input port, and output the intermediate value of the initial breakdown probability calculation for a single meter interval. S63. Execute multiple rounds of probability iteration to update the breakdown probability value of each meter interval, integrate all interval probability values ​​to draw a continuous distribution curve of the axial deterioration degree of the pipe body, combine the condensation surface discharge characteristics to deduce the remaining available time value of the flashover phenomenon on the pipe wall surface, and output the combined result data of the global deterioration distribution curve and flashover time. S64. Integrate all the contents of the previous electrode assembly records, load deformation data, impedance location points, partial discharge classification marks, full-band dielectric response curves, full-domain degradation distribution curves and flashover remaining time, and unify and standardize them into a standardized test document format. Output a complete cable protection pipe insulation performance test report data with integrated layout.

[0028] The working principle and effects of the above technical solution are as follows: By cleaning, screening, and standardizing segmented quantified data, invalid and abnormal data generated during the testing process can be removed, improving the data quality of the input model and reducing calculation deviations caused by abnormal data. Calculating the parameters of the pipe section meter by meter using the Weibull breakdown probability model refines the insulation breakdown risk of each pipe section, improving the detail of insulation risk assessment. Updating the breakdown probability value and extrapolating the remaining flashover duration through multiple rounds of iterative calculations can align with the gradual pattern of insulation degradation, improving the accuracy of lifespan prediction results. Integrating the entire process of testing data to generate a standardized report not only completely preserves various testing information such as deformation, partial discharge, and dielectric characteristics, but also unifies the presentation format of testing results, avoiding the biased assessment caused by a single data dimension. This effectively improves the completeness and reference value of the cable protection pipe insulation testing results, providing comprehensive and reliable data support for equipment operation and maintenance judgments.

[0029] In one embodiment of the present invention, S63 includes: Read the intermediate value of the initial breakdown probability calculation for a single-meter interval, set a fixed iteration convergence threshold and an upper limit for the number of iterations, perform multiple rounds of iterative calculations on the breakdown probability parameters of each segment interval, continuously update the dynamic breakdown probability values ​​of each segment interval, and generate segment breakdown probability sequence data after iterative updates. The breakdown probability values ​​of all pipe segments are collected and iterated. The probability values ​​are continuously fitted according to the axial extension direction of the pipe to generate continuous distribution curve data of insulation degradation covering the entire length of the pipe. By retrieving the condensation adhesion state parameters and surface discharge characteristic parameters of the pipe body, and combining the insulation failure law corresponding to different degrees of degradation, the decay trend of insulation withstand performance in each segment is progressively extrapolated to generate data on the flashover development process of insulation in each segment. By combining data on the development of insulation flashover with real-time environmental operating parameters, the remaining working time for each tube segment to maintain normal insulation performance is calculated, and array data of the remaining flashover time of the entire domain is generated. The continuous distribution curve data of insulation degradation degree is integrated and collected with the array data of remaining flashover time in the whole domain to generate a complete combination result data of global degradation distribution curve and flashover time.

[0030] The working principle and effects of the above technical solution are as follows: By updating the segmented breakdown probability values ​​through multiple rounds of iterative calculations, the deviation of the initial calculation results can be continuously corrected, making the breakdown probability of each pipe segment more consistent with the actual insulation state and improving the accuracy of probability calculation. By continuously fitting the probability values ​​of each segment along the axial direction, the discreteness of segmented data can be eliminated, forming a coherent global degradation distribution trend and fully demonstrating the overall insulation decay state of the pipe. By combining condensation and discharge characteristics to deduce the insulation failure process, the degradation law under actual service environment can be matched, weakening the difference between theoretical calculations and field conditions. By comprehensively converting the remaining working time of the equipment using multiple parameters, the insulation service capacity of each pipe segment can be accurately determined, and a global equipment life assessment can be achieved. This avoids the problem that a single probability parameter can only determine the risk level but cannot predict the service life, effectively improving the comprehensiveness and practical value of cable protection pipe insulation degradation assessment.

[0031] According to one embodiment of the present invention, an insulation performance testing system for cable protection pipes is provided, the system comprising: One or more processors; Memory, used to store one or more programs; Wherein, when the one or more programs are executed by the one or more processors, the one or more processors are made to implement the method described in any one of the above.

[0032] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for testing the insulation performance of a cable protection pipe, characterized in that, The method includes: S1. Perform a preliminary global resistance measurement on the sample data of the tube body to be tested, collect the initial potential distribution on the surface of the tube body and the ambient temperature and humidity parameters, and generate a basic state dataset of the tube body. S2. Apply radial loads step by step according to the preset gradient to simulate the uneven settlement stress of the soil, and simultaneously collect strain response data at different deformation locations of the pipe to generate a deformation distribution dataset of the pipe with continuous bending stress; perform a full-domain potential dynamic scan on the deformation distribution dataset of the pipe to generate a potential anomaly point marker dataset. S3. Inject a high-frequency step pulse signal into the pipe body under test with continuous bending stress, collect the full-path reflection waveform of the pulse propagating along the pipe body, decompose the time-domain features of the reflection waveform to generate full-path impedance distribution sequence data; perform spatial coordinate conversion on the full-path impedance distribution sequence data to generate a spatial location dataset of insulation anomaly points. S4. Under the condition of maintaining a stable four-point bending load, a stepped AC test voltage is synchronously applied to the tube body, and partial discharge pulse signals at different voltage levels are continuously collected to generate a phase-resolved partial discharge spectrum dataset; feature clustering is performed on the phase-resolved partial discharge spectrum dataset to generate a discharge type classification label dataset. S5. Apply a wideband sweep test signal to the marked insulation anomaly points to generate a full-band dielectric characteristic response dataset; perform feature fitting on the full-band dielectric characteristic response dataset to obtain the moisture absorption degree and insulation degradation rate parameters at different locations of the pipe, and generate a meter-by-meter insulation state quantification dataset. S6. Input the meter-by-meter insulation state quantification dataset into the preset Weibull breakdown probability model, substitute the insulation degradation rate parameters at different locations to complete the probability iteration calculation, output the insulation degradation distribution results of the entire pipe body and the surface flashover remaining time parameters, and generate a complete cable protection pipe insulation performance test report data.

2. The method for testing the insulation performance of a cable protection pipe according to claim 1, characterized in that, S1 includes: S11. Clean the residual dust and oil on the inner and outer walls of the cable protection pipe at both ends of the test pipe, grind the metal contact area of ​​the pipe opening to remove the oxide film, complete the flat treatment of the electrode bonding contact surface, and output the pipe opening pre-treatment to complete the pipe body physical marking information. S12. Secure the high-voltage electrode to one end of the tube cavity, attach the grounding electrode to the metal cladding layer of the outer wall of the tube, fill the gap between the electrode and the tube wall with insulating sealant, build a coaxial closed cavity without air leakage, and output the information that the sealed cavity assembly is complete. S13. Adjust the output of the constant temperature and humidity water vapor generator to a fixed relative humidity of circulating gaseous water vapor, control the water vapor to flow through the closed cavity at a uniform speed, and continuously maintain a fixed circulation time so that a condensation film of uniform thickness is formed on the entire area of ​​the inner and outer walls of the tube, and output the full-area condensation adhesion to complete the sample data of the tube to be tested. S14. Set up an array-type surface potential acquisition probe and move it at a constant speed along the tube axis to scan, simultaneously reading the real-time temperature value inside the cavity, the air humidity value around the cavity, and the DC resistance value of the entire tube wall. Integrate all the acquired values ​​to form a complete initial potential distribution sequence of the tube surface, and output the integrated basic state dataset of the tube.

3. The method for testing the insulation performance of a cable protection pipe according to claim 1, characterized in that, S2 includes: S21. Raise the four-point bending loading platform support bracket to the standard horizontal height, calibrate the coaxiality and horizontality of the four support rollers of the platform, remove debris and debris from the platform bearing area, and output the loading platform calibration ready working condition record information. S22. The tube to be tested, with a uniform condensation layer on the transfer belt, is placed stably in the middle bearing area of ​​the four support rollers. The limiting blocks clamp the two ends of the tube to prevent axial slippage during the load application process. The tube is then fixed in place to complete the physical positioning information. S23. Drive the loading head to slowly press down from zero load starting point according to equal differential gradient. After each single load is stabilized for a fixed time, read the output values ​​of the strain gauges pasted on the top, bottom, left and right sides of the pipe body. Continuously record the strain values ​​of the entire pipe section corresponding to each load level throughout the process, and output a complete dataset of pipe deformation distribution with continuous bending stress. S24. The mobile multi-channel potential acquisition module traverses the outer wall of the pipe segmentally at a uniform speed to collect potential values. It compares the potential difference between adjacent acquisition points to filter out points that exceed the stable fluctuation threshold. It records the axial length coordinates and radial orientation of the corresponding points and outputs the point coordinate labeling to complete the potential abnormal point marking dataset.

4. The method for testing the insulation performance of a cable protection pipe according to claim 1, characterized in that, The S3 includes: S31. The pulse signal transmitting terminal is connected to the high voltage electrode at one end of the tube body. The signal acquisition terminal is connected to the grounded metal layer on the outer wall of the tube body. The shielded cable wraps the entire signal transmission line to isolate external electromagnetic noise. The output pulse signal transceiver circuit is built and the working condition is recorded. S32. The pulse generator outputs a high-frequency step pulse signal with a fixed rise rate and introduces it into the coaxial cavity of the tube. The high-speed waveform acquisition card continuously captures the continuous waveform curve formed by the pulse reflecting back and forth inside the insulation layer of the tube wall, and outputs the original pulse full-path reflection waveform data for storage. S33. Decompose each independent echo pulse unit in the original reflected waveform, extract the time interval and voltage peak interval corresponding to a single pulse, and count the time difference and voltage peak value of each echo unit from transmission to return reception. Summarize all unit values ​​to form an ordered full path impedance distribution sequence data. S34. The length conversion is completed by matching the pulse propagation fixed wave velocity value with the echo delay value. The severity level of impedance change is distinguished according to the echo amplitude. The axial length coordinate obtained by the conversion is simultaneously bound with the radial azimuth coordinate recorded in the previous stage. The coordinate level binding is output to complete the spatial positioning dataset of insulation abnormality point.

5. The method for testing the insulation performance of a cable protection pipe according to claim 4, characterized in that, S33 includes: Read the original pulse full-path reflection waveform storage data, perform full-domain segmentation processing on the continuous waveform data according to the pulse waveform baseline threshold, remove waveform base noise and invalid clutter data, separate all independent and complete echo pulse units within the waveform period, and generate pure independent echo pulse unit set data; For each independent echo pulse unit in the ensemble data, the time domain interval is truncated, the start acquisition time of the single pulse signal and the time when the pulse signal decays to zero are locked, the complete time coverage interval corresponding to a single pulse is calibrated, and pulse time domain interval labeling data is generated. The amplitude range of each echo pulse unit that has been annotated in the time domain is truncated, the lowest voltage baseline value and the highest peak voltage value of the single pulse waveform are captured, the voltage amplitude fluctuation range corresponding to a single pulse is locked, and pulse amplitude range annotation data is generated. The time difference between the initial signal transmission time and the signal return reception time corresponding to each pulse unit is calculated. The peak voltage amplitude value corresponding to each pulse unit is recorded synchronously. The time delay parameters and amplitude parameters of all pulse units are collected in batches to generate pulse characteristic parameter collection data. All collected feature parameters are arranged and organized in an orderly manner according to the pulse propagation time sequence, and the parameter storage format and arrangement dimension are unified. All pulse feature parameters are integrated to form a full-path impedance distribution sequence data with unified dimension and coherent time sequence.

6. The method for testing the insulation performance of a cable protection pipe according to claim 4, characterized in that, S34 includes: Retrieve the time delay parameters of each echo unit stored in the full path impedance distribution sequence data, match the fixed propagation wave velocity parameters of the pulse in the insulation medium of the cable protection pipe, perform medium propagation distance conversion calculation, obtain the axial length position parameters corresponding to each impedance change point, and generate the original data of the axial coordinates of the insulation abnormal point. Read the voltage amplitude parameters of each echo unit inside the full path impedance distribution sequence data, and complete the hierarchical comparison and division by combining the threshold range of impedance change amplitude of the insulating medium. Classify the impedance change points of different degrees of anomaly and generate impedance change severity level labeling data. Retrieve the radial orientation data of the pipe body retained in the previous testing process, and dimensionally associate and bind the axial length parameters and radial orientation parameters of each impedance change point to form a single point all-round spatial parameter combination data, and generate multi-dimensional point spatial parameter set data. The multi-dimensional spatial parameter set data is standardized and verified, invalid point data with excessive parameter matching deviation are removed, the spatial coordinate arrangement order and grade labeling format of all valid points are unified, and a standardized spatial location dataset of insulation anomaly points is generated.

7. The method for testing the insulation performance of a cable protection pipe according to claim 1, characterized in that, The S4 includes: S41. Lock the current downward stroke of the four-point bending loading platform head to maintain constant load operation, disconnect the pulse signal transceiver circuit, reconnect the frequency converter AC voltage output terminal to the electrodes at both ends of the pipe body, and record the information after the output steady-state load voltage test circuit is switched. S42. The AC voltage source gradually increases the output amplitude from the reference low voltage starting point according to a fixed voltage difference. Each voltage level is continuously stabilized for a fixed acquisition time. The high-frequency partial discharge acquisition equipment continuously captures the discharge pulse waveform released by the pipe wall insulation layer and outputs multi-voltage level raw partial discharge pulse storage data. S43. Extract the AC phase interval and pulse peak value variation amplitude corresponding to the discharge pulse in each AC voltage cycle. Plot a continuous spectrum with phase value as the horizontal axis and pulse amplitude as the vertical axis. Integrate the spectrum content of all voltage levels to form a complete phase-resolved partial discharge spectrum dataset. S44. Extract three core features from the pulse clusters within the spectrum: the phase concentration interval, the pulse amplitude fluctuation interval, and the pulse repetition frequency. Group the clusters by feature similarity to complete the cluster classification. Match the exclusive feature intervals of the three types of discharge phenomena to complete the signal classification. Output the category label to complete the discharge type classification and labeling dataset.

8. The method for testing the insulation performance of a cable protection pipe according to claim 1, characterized in that, The S5 includes: S51. The mobile broadband impedance analysis probe is sequentially attached to each insulation abnormality point marked with potential and impedance. The space between the probe and the pipe wall is filled with conductive coupling medium to eliminate air gaps. The single-point test probe attachment ready status record information is output. S52. The impedance analysis equipment starts from a low frequency of 10Hz and moves upward at a constant speed to complete the full-band linear frequency sweep output. It simultaneously reads the real part and imaginary part of the complex dielectric constant corresponding to each frequency node, records the continuous change curve of the entire frequency band, and outputs the original dielectric constant change curve of a single point for storage. S53. Screen the frequency characteristic nodes generated by the polarization motion of water molecules adsorbed on the pipe wall inside the curve, record the frequency values ​​corresponding to the nodes and the maximum peak value of dielectric loss at the concave part of the curve, and summarize all abnormal point frequency band parameters to form a complete dielectric characteristic full-frequency band response dataset. S54. A nonlinear curve fitting algorithm is used to process the numerical values ​​of the full-band response curve. The trend of the fitted curve is used to obtain the total amount of water accumulation inside the pipe wall at the corresponding point and the change rate of molecular chain breakage of the insulation material. The two types of values ​​are summarized in segments according to each meter of pipe length, and the segmented numerical summaries are output to complete the meter-by-meter insulation state quantification dataset.

9. The method for testing the insulation performance of a cable protection pipe according to claim 1, characterized in that, The S6 includes: S61. Remove out-of-limit abnormal values ​​generated during the internal acquisition process of the meter-by-meter insulation state quantization dataset, rearrange the remaining valid values ​​in ascending order of the axial length of the pipe body, and output the cleaned and sorted standardized segmented quantization data sequence. S62. Retrieve the Weibull breakdown probability basic model pre-stored in the detection system, and sequentially import the total moisture absorption value and insulation degradation rate value of each meter interval in the standardized segmented quantized data sequence into the model input port, and output the intermediate value of the initial breakdown probability calculation for a single meter interval. S63. Execute multiple rounds of probability iteration to update the breakdown probability value of each meter interval, integrate all interval probability values ​​to draw a continuous distribution curve of the axial deterioration degree of the pipe body, combine the condensation surface discharge characteristics to deduce the remaining available time value of the flashover phenomenon on the pipe wall surface, and output the combined result data of the global deterioration distribution curve and flashover time. S64. Integrate all the contents of the previous electrode assembly records, load deformation data, impedance location points, partial discharge classification marks, full-band dielectric response curves, full-domain degradation distribution curves and flashover remaining time, and unify and standardize them into a standardized test document format. Output a complete cable protection pipe insulation performance test report data with integrated layout.

10. A system for testing the insulation performance of cable protection pipes, characterized in that, The system includes: One or more processors; Memory, used to store one or more programs; Wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the method of any one of claims 1 to 9.

Images