Distribution network cable defect detection method and equipment based on harmonic component characteristics

Abstract

The embodiment of the invention discloses a distribution network cable defect detection method and equipment based on harmonic component characteristics, belongs to the technical field of cable detection, and solves the problem that the existing cable defect detection is relatively high in labor and time cost. Based on a non-contact harmonic current sensor, a signal processing and analyzing module, a power supply module and a display module, the distribution network cable harmonic abnormal characteristic live detection device is constructed. Constructing an energy taking module according to the distribution characteristics of the magnetic field around the cable and the output voltage of the induction coil, constructing an energy management module according to the output voltage and the output power of the energy taking module, and constructing an energy taking device based on the energy taking module and the energy management module; based on the energy taking device and the distribution network cable harmonic abnormal characteristic live detection device, constructing a harmonic detection device, and carrying out harmonic data detection on the to-be-detected cable; and processing the harmonic data through a preset detection algorithm, and determining a defect type and state evaluation information corresponding to the to-be-detected cable according to the obtained harmonic characteristic data.

Description

Technical Field

[0001] The present application relates to the field of cable detection technology, and in particular to a method and device for detecting defects in distribution network cables based on harmonic component characteristics. Background Art

[0002] With the continuous advancement of urbanization, urban underground pipeline resources are becoming increasingly complex, the distribution network cable line structure is complex, the operating environment is diverse, the equipment density is increasing, and the proportion of old cables is increasing. As a result, the failure rate of distribution network cable lines remains high, and the risks of major fires or large-scale power outages are becoming increasingly severe, which is likely to seriously affect the safe operation of urban power grids.

[0003] The current means of evaluating the status of distribution network cables are mainly based on daily manual inspections and offline detection (voltage withstand, oscillation wave partial discharge, ultra-low frequency dielectric loss, etc.). Due to many restrictive factors such as cable grid construction, laying channels and strict control of planned power outages and maintenance time, as well as the risk of cumulative damage to the insulation of old cables or accelerated failures caused by offline high-voltage testing, offline high-voltage testing technology is becoming increasingly difficult to widely use.

[0004] Currently, common means of detecting live distribution network cables include high-frequency pulse partial discharge, ultrasonic partial discharge, infrared thermal imaging and other technologies. In actual application, they all have problems such as limited detection coverage, difficulty in detection implementation, few identifiable defects, inherent bottlenecks in sensitivity and anti-interference ability, etc. More importantly, it is difficult to accurately detect insulation defects of the entire distribution network cable through a single detection. Multiple measurements are required, with high labor and time costs, and great difficulties in application. Summary of the invention

[0005] The embodiments of the present application provide a distribution network cable defect detection method and device based on harmonic component characteristics, which are used to solve the following technical problems: the common distribution network cable live detection methods currently have problems in actual application, such as limited detection coverage, difficulty in detection implementation, few identifiable defects, inherent bottlenecks in sensitivity and anti-interference capabilities, and more importantly, it is difficult to accurately check the insulation defects of the entire distribution network cable through a single detection, and multiple measurements are required, with high labor and time costs, and great difficulties in application.

[0006] The present application embodiment adopts the following technical solutions:

[0007] The embodiment of the present application provides a distribution network cable defect detection method based on harmonic component characteristics. It includes: constructing a distribution network cable harmonic abnormal characteristic live detection device based on a non-contact harmonic current sensor, a signal processing and analysis module, a power module and a display module; constructing an energy acquisition module according to the magnetic field distribution characteristics around the cable and the output voltage of the induction coil, and constructing an energy management module according to the output voltage and output power of the energy acquisition module, so as to construct an energy acquisition device based on the energy acquisition module and the energy management module; constructing a harmonic detection device based on the energy acquisition device and the distribution network cable harmonic abnormal characteristic live detection device; performing harmonic data detection on the cable to be tested through the harmonic detection device; processing the harmonic data through a preset detection algorithm to obtain harmonic characteristic data, so as to determine the defect type and status evaluation information corresponding to the cable to be tested based on the harmonic characteristic data.

[0008] The embodiment of the present application is aimed at the near-field live coupling technology of the harmonic component characteristics of the distribution network cable, and a harmonic characteristic live detection device is developed, and a standardized on-site detection scheme is formed. The device and the detection scheme not only meet the needs of long-term and large-scale live census, but can also be applied to temporary trend monitoring, adapt to the needs of different scales and application scenarios, and realize the practical engineering application of harmonic sensing technology. Secondly, the embodiment of the present application proposes a set of insulation defect identification and diagnosis algorithms based on harmonic abnormal fingerprints by determining the correlation characteristics between "insulation degradation / aging defects-harmonic components / amplitudes" in the distribution network cable. Combined with the evolution trend of harmonic characteristics, an evaluation method and criterion for the overall operating status of the cable are proposed, and finally relevant technical standards are formed, realizing the identification of latent defects in distribution network cable lines based on harmonic signal abnormal characteristics. It provides reliable technical support for the health monitoring, preventive maintenance and safe operation of cables in power systems.

[0009] In one implementation of the present application, an energy harvesting module is constructed according to the magnetic field distribution characteristics around the cable and the output voltage of the induction coil, specifically including: determining the output voltage of the induction coil by constructing a three-dimensional simulation model of the three-core cable; determining the layout of the energy harvesting module based on the output voltage, the coil angle position and the current phase corresponding to the three-core cable; wherein the energy harvesting module is used to collect the magnetic field energy of the three-core cable; inputting preset key structural parameters into the constructed three-core cable model, and determining the output power corresponding to the energy harvesting module under different preset key parameters; wherein the preset key structural parameters include at least one of the number of coil turns, the conductor diameter, the magnetic thickness and the magnetic core length; determining the optimal parameters based on the output power, and optimizing based on the optimal parameters to obtain the energy harvesting module.

[0010] In one implementation of the present application, before determining the output voltage of the induction coil by constructing a three-dimensional simulation model of a three-core cable, the method also includes: treating the three-phase conductor corresponding to the three-core cable as an infinitely long straight conductor and passing a three-phase industrial frequency current through it to obtain the magnetic field generated by the three-phase conductor in the surrounding space, and establishing a magnetic field calculation function; determining the distance from the field point to the center of the conductor based on the coordinates of the field point in the polar coordinate system and the distance from the center of the cable to the center of the conductor, and establishing a distance calculation function; substituting the distance calculation function into the magnetic field calculation function, and directional decomposing the magnetic field generated by the three-phase conductor at the field point to obtain a magnetic field distribution function in the space around the cable, so as to simplify the cable model.

[0011] In one implementation of the present application, the spatial magnetic field distribution function around the cable is:

[0012]

[0013] Among them, B A is the magnetic field of the surrounding space corresponding to A; B B is the magnetic field in the surrounding space corresponding to B; B C is the magnetic field in the surrounding space corresponding to C; u 0 is the vacuum permeability; I A is the A phase power frequency current; I B is the power frequency current of phase B; I C is the C-phase power frequency current; r and φ are the coordinates of the field point in the polar coordinate system; a is the distance from the cable center to the conductor center.

[0014] In one implementation of the present application, an energy management module is constructed according to the output voltage and output power of the energy acquisition module, specifically including: based on the circuit topology of the half-wave rectifier boost circuit, the output voltage of each induction coil is rectified by half-wave rectification to obtain a DC output voltage, and the output voltages of the three induction coils are superimposed in series to increase the output voltage of the energy acquisition device; based on the energy management chip, the output voltage of the half-wave rectifier boost circuit is stabilized; multiple induction coils are connected in parallel, and the parallel-connected coils are output according to the rectifier boost circuit to increase the output power of the energy acquisition module; an energy management module is constructed by a half-wave rectifier boost circuit, an energy management chip and multiple parallel-connected induction coils.

[0015] In one implementation of the present application, before processing the harmonic data through a preset detection algorithm to obtain the harmonic characteristic data, the method also includes: taking the sample data of the proportion of 2-11 harmonic components as input, and constructing a covariance matrix according to the standardized data set; determining the eigenvalues ​​and corresponding eigenvectors of the covariance matrix; based on the eigenvalues ​​and eigenvectors, determining the contribution rate and cumulative contribution rate corresponding to each principal component; eliminating harmonics based on the contribution rate and cumulative contribution rate; clustering the remaining harmonic component data after elimination, and obtaining the influence relationship between different operating conditions and harmonic characteristics based on the clustering results; wherein the different operating conditions factors include at least one of metal shielding, hot and cold states, and load current.

[0016] In one implementation of the present application, harmonic data is processed by a preset detection algorithm to obtain harmonic characteristic data, so as to determine the defect type and status assessment information corresponding to the cable to be tested based on the harmonic characteristic data, specifically including: processing the harmonic data by a self-organizing mapping algorithm to obtain first harmonic characteristic data; constructing a cable failure probability model and an operation risk assessment model through LASSO regression and Weibull distribution, and processing the harmonic data through the model to obtain second harmonic characteristic data; comparing the first harmonic characteristic data and the second harmonic characteristic data with a preset cable defect harmonic characteristic information table, respectively, to obtain the defect type corresponding to the cable to be tested and the status assessment information corresponding to the cable to be tested.

[0017] In one implementation of the present application, before processing the harmonic data through a self-organizing mapping algorithm to obtain the first harmonic characteristic data, the method also includes: acquiring sample harmonic data and preprocessing the sample harmonic data; using the preprocessed sample harmonic data as input samples to construct and train a self-organizing mapping model to obtain a trained self-organizing mapping model; processing the harmonic data through principal component analysis and a self-organizing mapping algorithm to obtain first harmonic characteristic data, specifically including: inputting the harmonic data into the trained self-organizing mapping model, mapping the harmonic data to neurons in a grid through the trained self-organizing mapping model; obtaining the first harmonic characteristic data based on the distribution and adjacent relationship of the neurons, so as to determine the defect type of the cable to be tested based on the first harmonic characteristic data.

[0018] In one implementation of the present application, a cable failure probability model and an operation risk assessment model are constructed through LASSO regression and Weibull distribution, specifically including: using LASSO regression, fitting the data by minimizing the residual sum of squares, and preventing overfitting by adding regularization terms, to construct a cable failure probability model to quantitatively analyze the degradation trend of distribution network cable defects; wherein the distribution network cable defects include at least one of semi-conductive suspension defects, metal tip defects, metal suspension defects on the insulation surface, and residual defects of the insulation tape on the surface of the crimping tube; constructing a fault risk probability model based on a healthy state through a preset exponential function; determining model parameters based on Weibull distribution and fitting a bathtub curve, and constructing a fault risk probability model based on operating years based on the determined model parameters; obtaining an operation risk assessment model based on the fault risk probability model based on a healthy state and the fault risk probability model based on operating years.

[0019] An embodiment of the present application provides a distribution network cable defect detection device with harmonic component characteristics, including: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can: construct a distribution network cable harmonic abnormal characteristic live detection device based on a non-contact harmonic current sensor, a signal processing and analysis module, a power supply module and a display module; construct an energy acquisition module according to the magnetic field distribution characteristics around the cable and the output voltage of the induction coil, and construct an energy management module according to the output voltage and output power of the energy acquisition module, so as to construct an energy acquisition device based on the energy acquisition module and the energy management module; construct a harmonic detection device based on the energy acquisition device and the distribution network cable harmonic abnormal characteristic live detection device; perform harmonic data detection on the cable to be tested through the harmonic detection device; process the harmonic data through a preset detection algorithm to obtain harmonic characteristic data, so as to determine the defect type and status evaluation information corresponding to the cable to be tested based on the harmonic characteristic data.

[0020] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects: The embodiments of the present application are aimed at the near-field live coupling technology of the harmonic component characteristics of the distribution network cable, and a harmonic characteristic live detection device is developed, and a standardized on-site detection scheme is formed. The device and the detection scheme not only meet the needs of long-term and large-scale live surveys, but can also be applied to temporary trend monitoring, adapt to the needs of different scales and application scenarios, and realize the practical engineering application of harmonic sensing technology. Secondly, the embodiments of the present application propose a set of insulation defect identification and diagnosis algorithms based on harmonic abnormal fingerprints by determining the correlation characteristics between "insulation degradation / aging defects-harmonic components / amplitudes" in the distribution network cables. Combined with the evolution trend of harmonic characteristics, an evaluation method and criterion for the overall operating status of the cable are proposed, and finally relevant technical standards are formed, realizing the identification of latent defects in distribution network cable lines based on harmonic signal abnormal characteristics. It provides reliable technical support for the health monitoring, preventive maintenance and safe operation of cables in power systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. In the drawings:

[0022] Figure 1 A flow chart of a method for detecting defects in distribution network cables based on harmonic component characteristics provided in an embodiment of the present application;

[0023] Figure 2 A schematic diagram of a data collection module provided in an embodiment of the present application;

[0024] Figure 3 A physical picture of a harmonic abnormality charged detection device provided in an embodiment of the present application;

[0025] Figure 4 A schematic diagram of a three-core cable parallel straight conductor model provided in an embodiment of the present application;

[0026] Figure 5 A schematic diagram of the layout of an energy harvesting coil provided in an embodiment of the present application;

[0027] Figure 6 A schematic diagram of a circuit model of an induction coil provided in an embodiment of the present application;

[0028] Figure 7 A schematic diagram of a half-wave rectification and boost circuit provided in an embodiment of the present application;

[0029] Figure 8 A schematic diagram of a coil parallel circuit model provided in an embodiment of the present application;

[0030] Fig. 9 A LASSO regression cross-validation diagram provided in an embodiment of the present application;

[0031] Fig.10 A LASSO regression fitting coefficient trajectory diagram provided in an embodiment of the present application;

[0032] Fig.11 A schematic diagram of a bathtub curve of a failure rate of an electric power device provided in an embodiment of the present application;

[0033] Fig.12 A schematic diagram of a Weibull distribution probability density curve provided in an embodiment of the present application;

[0034] Fig.13 A schematic diagram of a Weibull cumulative distribution function curve provided in an embodiment of the present application;

[0035] Fig.14 A schematic diagram of a failure rate function curve provided in an embodiment of the present application;

[0036] Fig.15 A schematic diagram of the structure of a distribution network cable defect detection device based on harmonic component characteristics provided in an embodiment of the present application. DETAILED DESCRIPTION

[0037] The embodiments of the present application provide a method and device for detecting defects in distribution network cables based on harmonic component characteristics.

[0038] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0039] The technical solution proposed in the embodiment of the present invention is described in detail below with reference to the accompanying drawings.

[0040] In one embodiment of the present application, a simple cable model is first established to simulate and analyze the cable flux containing different insulation defects, and a qualitative analysis is performed on the cable model magnetic field, focusing on typical defects such as insulation bubbles, water droplets, semi-conductive impurities, metal impurities, scratches, etc., and a corresponding relationship between typical defects and harmonic current characteristics is established. Cable defect models of different shapes are established, and different material types are added to defects of the same shape to simulate the cable flux. When simulating the interface defects of cable accessories, the 300mm2 10kV cable joint model is used for both two-dimensional and three-dimensional modeling. On this basis, different types of defects are added, the defect type, defect location and size are changed, and the flux density distribution at different positions of the cable joint and the proportion of cable grounding current harmonics are calculated.

[0041] By comparing the magnetic flux density of the cross section and the insulation interface at the three-junction point, it can be seen that at different positions of the cable, water droplets, water films, and water tree defects will cause different degrees of magnetic field distortion. Due to the existence of defects, the magnetic flux density model will be distorted when passing through the defects, and the law of distortion is roughly the same as the law of the insulation interface, but by comparing the change in magnetic flux density at the cable sheath, it can be seen that when the overall moisture is defective, the change in magnetic flux density of the sheath part is more obvious. Similarly, the embodiment of the present application also adds other defects to the cross-section model at the three-junction point, and simulates the magnetic field and current. It can be seen that the changes in magnetic flux density caused by different defect materials are also different, and different degrees of degradation of defects will affect the magnetic flux density. Secondly, the embodiments of the present application use electromagnetic field simulation analysis of distribution cables and accessories to study and analyze the flux disturbance characteristics caused by nonlinear volt-ampere characteristics of inhomogeneous media, such as typical insulation defects and structural defects. Based on the defect flux disturbance, the defective grounding current is analyzed, the grounding current waveform is derived, the harmonic proportion analysis of the current is performed, the proportion of harmonic components of the grounding current under different defects is analyzed, the magnetic field is combined with the current, and the relationship between defects and current is established from a theoretical perspective.

[0042] Specifically, the magnetic flux characteristics and harmonic current fluctuations of four types of defective cables, namely bubbles, water drops, semi-conductive impurities and metal impurities, were simulated to compare the electromagnetic changes caused by different defects. The simulation results show that the harmonic proportions caused by different defect types in cable insulation vary significantly, and the current harmonic distortion rates caused by defects of the same type are different. The current distortion rates of bubbles and water films are not much different. The defect structure of water trees and the resulting magnetic flux fluctuations are large, resulting in the harmonic distortion rate of cable joints caused by water trees being significantly higher than the other two defects. Under the condition of overall moisture, the current of cable joints caused by a large number of water tree defects is more obviously distorted in the waveform. Water film and water drop defects also cause different degrees of current distortion. Different types of defects cause different current harmonic distortion rates. The defect structure of water trees and the resulting magnetic flux fluctuations are large, resulting in the harmonic distortion rate of cable joints caused by water trees being significantly higher than the other two defects.

[0043] The embodiment of the present application also verifies the mechanism of harmonic anomalies of insulation defects of distribution network cables and accessories. Specifically, 16 typical defects of distribution network cables, including internal water defects, external water ingress defects, metal tips, metal suspension on the insulation surface, semi-conductive suspension on the insulation surface, residual insulation tape on the surface of the crimping tube, conductor eccentricity, body bending, loose crimping tube offset, crimping tube burrs, malposition of accessories, body nail damage, semi-conductive tips, longitudinal scratches on insulation, ring-cut scratches on insulation, and normal defect-free cables, are designed, and the harmonic current detection of typical defects is performed using the distribution cable and accessories harmonic current experimental platform.

[0044] The propagation characteristics of harmonic abnormality signals of insulation defects of distribution network cables and accessories were also verified. Specifically, the ground current in the cable accessory model was obtained through simulation software, the amplitude of typical harmonics was extracted through Fourier transform, and the attenuation distance of harmonic current was calculated according to the calculation formulas of various cable parameters and the propagation constant formula. When there are defects in cable accessories, the propagation attenuation characteristics of harmonic currents generated by defects in different parts in the distribution network cable line, as well as the effective range and influencing factors of harmonic component detection technology were analyzed through simulation, and finally the harmonic abnormality of insulation defects of distribution network cables and accessories was verified.

[0045] In addition, the harmonic fingerprint verification of the insulation defects of distribution network cables and accessories was also carried out. Specifically, a typical defect test system was first built. The typical defect test system passed the cable with the intermediate joint defect through the through-core transformer to form a ring loop, and the harmonic signal detection device was connected to the cable to be tested by using a sensor. The test system mainly includes: multiple 15m single-core distribution network cables and accessories with specifications of 1*300, 8.7 / 15kV, multiple through-core transformers, series resonant withstand voltage test devices, harmonic signal detection devices, voltage regulator cabinets, temperature measurement equipment, etc. In the test, the voltage regulator cabinet was controlled to apply currents of different gradients in the cable, and the series resonant withstand voltage device was used to apply voltage to the cable to be tested. The harmonic content in the cable grounding wire and the main conductor was measured by the harmonic signal detection device, and the temperature measurement equipment was used to monitor the cable conductor temperature in real time. According to the analysis, the ratio of the total harmonic component to the fundamental component of the power cable insulation defect at different defect levels can be obtained. The insulation defect status can be evaluated by detecting the proportion of harmonic components in the faulty cable. According to the harmonic types and content characteristics of the 15 cable defects, these defects can be classified to analyze their similarities and differences.

[0046] 1. The defect of significantly increased high harmonic content

[0047] Some defects can lead to a significant increase in high-order harmonics (especially the 3rd and 5th harmonics). This is common in defects where the cable insulation or conductor surface structure has changed, such as: insulation ring cut scratches, crimping tube burrs, semi-conductive tips, and conductor eccentricity. According to the data, the following types of defects have a significant effect on increasing the content of the 3rd and 5th harmonics. Among them:

[0048] (1) Insulation ring cut scratches: The average value of the 3rd harmonic is between 2.18% and 2.31%, and the 5th harmonic is between 2.12% and 2.22%. These data show that insulation ring cut scratches have a strong impact on low-order harmonics, especially in the early stage of degradation, the 5th harmonic content increases significantly.

[0049] (2) Crimped tube burr: The average value of the 3rd harmonic content of this defect is about 2.68%, and the 5th harmonic is 1.90%. The data shows that the crimped tube burr is more sensitive to the 3rd harmonic, while the 5th harmonic is slightly lower, but still remains at a high level. This phenomenon may be due to the local electric field concentration effect caused by the burr, which increases the content of low-order harmonics.

[0050] (3) Semiconducting tip: The third harmonic content of the semiconducting tip under high current conditions is 2.61%, and the fifth harmonic content is 2.03%. This high content is related to the increase in local field strength caused by the tip, which significantly increases the low-order harmonic content.

[0051] The 3rd and 5th harmonic contents of these defects increase significantly, and are particularly evident in the early stages of degradation and under high current conditions. The reason can be attributed to the electric field inhomogeneity caused by the defects, especially the tip or burr structure, which concentrates the electric field and causes a significant increase in low-order harmonics. However, there are also differences in the increase in the 3rd and 5th harmonics for different defects. For example, the burr of the crimped tube reaches 2.68% on the 3rd harmonic, while the 5th harmonic of the semi-conductive tip can reach up to 2.03%. This difference shows that the structural characteristics of each defect have different sensitivities to different harmonics.

[0052] 2. The defect of harmonic content fluctuating between multiple harmonics

[0053] This type of defect has a more complex effect on the harmonic content, often causing fluctuations in harmonics of different orders (3rd, 5th, 7th and 11th), but not concentrated on a specific harmonic. These defects usually involve defects or structural unevenness inside the insulation or conductor material, such as: internal water, external water ingress, metal tips, and longitudinal scratches on the insulation. Among them:

[0054] (4) Internal water defects: The 5th harmonic of internal water defects at high current (500A) reaches 1.95%, the 3rd harmonic is 1.04%, and the 7th and 11th harmonics are 0.33% and 0.05% respectively. The data show that internal water defects still have a certain effect on high-order harmonics (such as the 7th and 11th), although the increase is not as significant as that of low-order harmonics.

[0055] (5) Metal tip: The contents of the 3rd and 5th harmonics are 1.5% and 2.07% respectively, but the 7th harmonic can also reach 1.23%, which indicates that the metal tip has an impact on multiple harmonic orders under high current conditions.

[0056] These defects usually lead to fluctuations in multiple harmonics, especially under high current conditions, where the content of multiple harmonics increases significantly. This phenomenon may be related to the inhomogeneity of the material inside the cable or on the contact interface, which causes different harmonic orders to be disturbed to different degrees. Similarly, different defects have inconsistent effects on higher harmonics. For example, internal water defects have a small effect on the 7th and 11th harmonics, while metal tip defects have a significant increase in the 7th harmonic, reaching 1.23%. These differences indicate that there are certain differences in the effects of the geometric shape and position characteristics of the defects on higher harmonics.

[0057] 3. Defects with less impact on harmonics

[0058] This type of defect has a relatively small impact on the harmonic content. The harmonic content has no significant changes at the 3rd, 5th, 7th and 11th harmonics, or only has slight changes at specific harmonics, such as: body bending, normal without defects, and accessories installed out of place. Among them:

[0059] (6) Body bending: The third and fifth harmonic contents at 500A are 0.55% and 0.64% respectively, which are significantly lower than other defect types and do not change much with the degree of insulation degradation.

[0060] (7) Normal without defects: Under 500A conditions, the contents of the 3rd and 5th harmonics are 0.47% and 0.52% respectively, while the 7th and 11th harmonics are less than 0.1%, and the changes between the degradation stages are also very small.

[0061] Generally speaking, since the physical characteristics of such defects are not obvious and the defect area is relatively smooth or uniform, these defects have little effect on harmonics, the harmonic content is low, and the fluctuation is not significant, showing a relatively stable harmonic characteristic. Such defects are usually structurally uniform or of a less severe defect type. Although in the case of extremely low degradation, normal defect-free and body bending have some fluctuations on the third harmonic, the overall fluctuation amplitude is small and it is difficult to serve as the main basis for fault diagnosis.

[0062] 4. Defects that are sensitive to insulation degradation

[0063] The harmonic characteristics of this type of defect vary significantly with the degree of insulation degradation, and usually occur when the insulation is damaged or aged, such as insulation ring cuts and scratches, crimping tube displacement and loosening, and semi-conductive suspension.

[0064] in:

[0065] (8) Insulation ring cut scratches: The content of the fifth harmonic is 2.22% in degradation stage 1 and decreases to 1.92% in degradation stage 4, indicating that the influence of insulation ring cut scratches on the fifth harmonic gradually weakens with the increase of degradation.

[0066] (9) The crimping tube is offset and loose: In different degradation stages, the third harmonic content gradually increases from the initial 1.5% to 2.0% in the later stage of degradation, showing a trend of gradually increasing with the deepening of insulation degradation. Especially at a higher degree of degradation, the increase in the third harmonic is more significant.

[0067] (10) Semiconducting suspension: The content of the fifth harmonic changes significantly with the degree of insulation degradation, increasing from 1.86% in degradation stage 1 to 2.01% in stage 4. In addition, the content of the third harmonic also increases in the later stage of degradation. Although the change is not as significant as that of the fifth harmonic, a certain growth trend can still be observed.

[0068] By summarizing and analyzing the detailed data, cable defects can be more accurately classified according to harmonic characteristics, and the similarities and differences of various defects in harmonic components can be identified. This not only provides clear indicators for fault diagnosis, but also provides a theoretical basis for cable maintenance and monitoring.

[0069] Figure 1 A flow chart of a distribution network cable defect detection method based on harmonic component characteristics provided in an embodiment of the present application is as follows: Figure 1 As shown, the distribution network cable defect detection method based on harmonic component characteristics includes the following steps:

[0070] S101. Construct a live detection device for harmonic abnormality characteristics of distribution network cables based on a non-contact harmonic current sensor, a signal processing and analysis module, a power supply module and a display module.

[0071] In one embodiment of the present application, when constructing a distribution network cable harmonic abnormal characteristic live detection device, material selection and sensor coil design are first performed. In the present application embodiment, manganese-zinc ferrite is selected as the core material of the sensor; the inductance at low frequency is obtained by coefficient fitting to complete the sensor frequency response design, and the low-frequency current magnetic field close-coupled coil design is completed based on the improvement of the quality factor Q value of the inductance. Sensor trial production is carried out according to the ferrite nonlinear material, frequency response and close-coupled coil design. Since the conductor, material selection and coil turns determine one of the main factors of inductance size, the number of coil turns, conductor material and magnetic flux area need to be adjusted according to the size of the measured cable signal during use. Harmonic signal data acquisition uses a self-developed data acquisition module, which mainly realizes the relevant content of analog input or digital input signal measurement, and can set a continuous acquisition mode or a limited number of acquisition modes at the same time. It supports up to 1MS / S sampling rate, 16-bit resolution, continuous and uninterrupted acquisition, 8RSE differential signal or 4DIFF single-ended signal analog input. Harmonic signal processing and analysis mainly use fast Fourier analysis. Figure 2 A schematic diagram of a data collection module provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, the data acquisition module includes trigger sources: software trigger and external trigger; sampling clock source: internal sampling clock is derived from external sampling clock source; FIFO buffer; analog input; analog output and communication interface.

[0072] Figure 3 A physical picture of a harmonic abnormal current detection device provided in an embodiment of the present application. Through modularization and 3D printing technology, the sensor manufacturing process is improved, the measurement accuracy of the non-contact harmonic current sensor is improved, and the sensor field perception range is optimized. High-performance integrated circuits are used to speed up the working speed of the signal processing and analysis unit. Based on the design optimization of the device component structure, the power loss of the whole set of equipment is reduced to adapt to the online energy extraction component. By developing supporting software, the operating difficulty of the detection device is reduced, and functions such as real-time large-scale storage and export of multiple sets of detection data are realized, and the human-computer interaction interface is further optimized.

[0073] S102, constructing an energy acquisition module according to the magnetic field distribution characteristics around the cable and the output voltage of the induction coil, constructing an energy management module according to the output voltage and output power of the energy acquisition module, and constructing an energy acquisition device based on the energy acquisition module and the energy management module.

[0074] In one embodiment of the present application, the output voltage of the induction coil is determined by constructing a three-core cable 3D simulation model. Based on the output voltage and the coil angle position and the current phase corresponding to the three-core cable, the layout of the energy acquisition module is determined; wherein the energy acquisition module is used to collect the magnetic field energy of the three-core cable. The preset key structural parameters are input into the constructed three-core cable 3D model, and the output power corresponding to the energy acquisition module is determined under different preset key parameters; wherein the preset key structural parameters include at least one of the number of coil turns, wire diameter, magnetic thickness and magnetic core length. Based on the output power, the optimal parameters are determined, and the energy acquisition module is obtained by optimization based on the optimal parameters.

[0075] Specifically, the embodiment of the present application proposes a simplified cable model as shown in FIG. Figure 4 A schematic diagram of a three-core cable parallel straight wire model provided in an embodiment of the present application is shown in FIG. Figure 3 As shown, the three-phase current-carrying conductor is simplified to an infinite line current located at the center of the conductor.

[0076] From the electromagnetic theory, we know that the magnetic field generated by an infinitely long parallel straight wire is:

[0077]

[0078] In the process of deriving the magnetic field in the space around a three-core cable, the three-phase conductors A, B, and C are regarded as infinitely long straight conductors, and three-phase industrial frequency currents are passed through them. The magnetic fields generated by the three-phase conductors in the surrounding space can be obtained respectively.

[0079]

[0080] Where R A , R B , R C are the distances from the field point P(r,φ) to the center of the conductor.

[0081]

[0082] Among them, r and φ are the coordinates of the field point in the polar coordinate system, and a is the distance from the center of the cable to the center of the conductor.

[0083] Substitute the distance calculation function into the magnetic field calculation function, and decompose the magnetic field generated by the three-phase conductor at the field point P(r,φ) into the ρ and φ directions, and the results are as follows:

[0084]

[0085] The above formula can be used to obtain the distribution of the magnetic field in the space around the cable. Taking into account the industrial frequency AC current passing through the three-phase conductor, the change of the magnetic induction intensity at the field point P with time can also be obtained.

[0086] The three-core power cable has three-phase conductors that generate magnetic fields that are superimposed on each other, which not only makes the magnetic field distribution no longer uniform, but also reduces the modulus of magnetic induction intensity, thereby reducing the output voltage of the induction coil. The purpose of optimizing the layout of the energy extraction module is to effectively utilize the magnetic field in the space around the three-core cable to obtain a higher output voltage and achieve the power output required by the load. The embodiment of the present application determines the optimal layout of the energy extraction module by calculating the output voltage of the induction coil located in the space around the cable.

[0087] Take the square cross-section coil placed on the cable surface as an example, assuming that the energy extraction coil is designed as a square coil and has no iron core. The cross-sectional dimensions of the coil are 40 mm long and 30 mm wide. The diameter of the copper wire of the coil is d = 0.5 mm, the number of turns of a single-layer coil is n = 80, and there are 4 layers wound in total. The total number of turns of the energy extraction coil is N = 320. Figure 5 A schematic diagram of a layout of an energy-harvesting coil provided in an embodiment of the present application, the layout of the energy-harvesting coil is as follows Figure 5 shown.

[0088] The magnetic induction intensity B at the field point P around the known cable P .

[0089] B P (r,φ)=B Pρ +B Pφ ;

[0090] in,

[0091]

[0092] So we can get the magnetic induction intensity B perpendicular to the coil cross section Pz .

[0093] B Pz (r,φ)=B Pφ cosφ+B Pρ sinφ;

[0094] Assume the coil angular position is φ 0 , calculate the magnetic flux of the i-th coil (i is ±1, ±2···±n / 2 along the positive and negative directions of the z-axis respectively) It is assumed that the cross section of the coil is along the y direction B z unchanged, while B z will change along the x and z directions. In order to calculate the magnetic flux of any turn of the coil It is necessary to integrate the microelement. Take the area microelement dS = y × dx. Where y = 30mm, which is the length of the coil, so the magnetic flux microelement can be obtained for:

[0095]

[0096] in,

[0097]

[0098] Then the magnetic flux of the entire coil domain can be obtained:

[0099]

[0100] From this we can get the coil induced voltage,

[0101]

[0102] The present application embodiment proposes a simplified cable calculation model to calculate and analyze the magnetic field distribution characteristics around a three-core power cable and the output voltage of the induction coil. The magnetic field distribution around a three-core cable is uneven, and the tangential magnetic induction intensity B φ And radial magnetic induction intensity B φ There are extreme values ​​at three positions, and their phases also change continuously along the circumferential direction. The output voltage of the induction coil arranged around the cable also changes with its relative position. Therefore, a core layout method suitable for magnetic field energy collection of three-core cables is proposed. Based on a full understanding of the internal structure and operating conditions of the three-core cable, an energy harvesting module optimization design method is proposed. The key structural parameters of the energy harvesting module include the number of coil turns, wire diameter, core thickness and core length. Finite element simulation software is used to simulate and calculate different structural parameters, and the parameters are optimized by comparing the output power of the energy harvesting module under different parameters. In addition, different core materials are compared and analyzed, and ferrite is finally determined as the core material of the energy harvesting module.

[0103] In one embodiment of the present application, based on the circuit topology of the half-wave rectifier boost circuit, the output voltage of each induction coil is rectified by half-wave rectification to obtain a DC output voltage, and the output voltages of the three induction coils are superimposed in series to increase the output voltage of the energy acquisition device. The output voltage of the half-wave rectifier boost circuit is stabilized based on the energy management chip. Multiple induction coils are connected in parallel, and the parallel-connected coils are output according to the rectifier boost circuit to increase the output power of the energy acquisition module. An energy management module is constructed by a half-wave rectifier boost circuit, an energy management chip, and multiple parallel-connected induction coils.

[0104] Specifically, the energy acquisition module is used to collect magnetic field energy. The output voltages of the three induction coils of the energy acquisition module are all industrial frequency AC voltages, and the phases are 120 degrees apart. Since the power supply voltage required by the load is a 3.3 / 5V DC voltage, the energy management module is required to regulate the output voltage of the energy acquisition module and finally output a DC voltage that meets the load requirements.

[0105] Furthermore, the working principle of the energy harvesting module is to use Faraday's law of electromagnetic induction to collect the magnetic field energy in the space around the cable through the induction coil, and finally form a voltage output. Therefore, the circuit model of the energy harvesting module can be established based on this, which is convenient for designing the circuit topology of the energy management module and analyzing the impact of the electrical parameters of the energy harvesting module on the output performance of the entire device. Figure 6 A schematic diagram of a circuit model of an induction coil provided in an embodiment of the present application. The energy acquisition module has three induction coils. The circuit model of each induction coil is as follows: Figure 6 As shown, the circuit includes the induced voltage E of the coil and the coil resistance R s , coil inductance L s , compensation capacitor C s .

[0106] Without adding compensation capacitors, the rectifier filter circuit and load in the energy management module are regarded as pure resistive loads. At this time, the energy acquisition module is regarded as an induction coil with an iron core. The coil is wound with multiple turns and has its own internal resistance. At the same time, the inductance of the coil is also large. In order to increase the output power of the coil, a capacitor C can be connected in series at each coil output end. s , so that the series capacitor and the coil's own inductance L s Series resonance occurs at the power frequency, compensating for the power loss of the coil's self-inductance and increasing the output power.

[0107] Furthermore, the output voltage of the induction coil of the energy acquisition module is an industrial frequency sinusoidal voltage, and the entire energy acquisition module includes a total of three induction coils. In the design of the rectifier boost circuit, the layout of the induction coil of the energy acquisition module and the output voltage characteristics should be considered. The output voltage should be increased while rectifying, and the high-frequency noise interference and AC pulse influence in the output voltage waveform of the rectifier circuit should be eliminated through the filtering link, so as to obtain a DC voltage that meets the input requirements of the DC voltage stabilization circuit. The energy acquisition module includes three induction coils, and the output voltage of each induction coil is an industrial frequency sinusoidal waveform, and the phases are 120° apart. The embodiment of the present application adopts a half-wave rectifier boost circuit, which uses a circuit topology to simultaneously realize the rectification and boosting of AC power. Figure 7 A schematic diagram of a half-wave rectifier boost circuit provided in an embodiment of the present application, the circuit topology is as follows Figure 7 As shown. This circuit topology uses half-wave rectification to rectify the output voltage of each induction coil to obtain a DC output voltage, and then superimposes the output voltages of the three induction coils in series, thereby increasing the output voltage of the entire energy harvesting device. Compared with the three-phase rectifier circuit, this circuit topology only contains three diodes, which reduces the complexity of the circuit topology and the voltage loss of electrical components. a 、E b 、E c is the industrial frequency AC voltage output by the induction coil, R s and Ls are the resistance and inductance of the induction coil itself. Using the series resonance method mentioned above, by matching L s With C s The parameters of the parallel capacitor C cause series resonance, compensating the coil's own inductance, thereby increasing the power output of the energy harvesting device. a , C b , C c It is a filter capacitor to reduce the output voltage ripple.

[0108] Furthermore, under actual working conditions, the current carrying capacity of the cable will change with the load. When the load current increases or decreases, the output voltage of the rectifier boost circuit will also increase or decrease accordingly. The load of the energy extraction device requires a stable DC power supply voltage. In order to solve the above problems, an energy management chip is required. The energy management chip used in the embodiment of the present application is LTC-3331. When the voltage at the input end of the chip changes within a specified range, the output voltage can be stabilized at a set value, thereby meeting the load requirements.

[0109] Furthermore, without changing the parameters of the wire and the magnetic core structure, the embodiment of the present application reduces the equivalent internal resistance of the energy taking module by connecting multiple coils in parallel. Figure 8 A schematic diagram of a coil parallel circuit model provided in an embodiment of the present application, such as Figure 8 As shown. The induction coils on each iron core are divided into three groups, ABC, and each group of coils with the same name are connected in parallel, and finally the three groups of coils are output according to the rectifier boost circuit.

[0110] S103, construct a harmonic detection device based on the power acquisition device and the power distribution network cable harmonic abnormality characteristic live detection device.

[0111] In one embodiment of the present application, a harmonic detection device is constructed by constructing an energy collection device and a distribution network cable harmonic abnormal characteristic live detection device. The changing magnetic field generated by the cable load current in the surrounding space is used to provide a stable energy supply for the load. The energy collection device is mainly divided into two modules, an energy collection module and an energy management module, which collect the magnetic field energy in the space around the cable to achieve the purpose of outputting a 5V DC voltage.

[0112] S104, performing harmonic data detection on the cable to be tested by using a harmonic detection device.

[0113] In one embodiment of the present application, a harmonic detection device is used to perform multiple harmonic signal measurements on different parts of the intermediate joint of the cable to be tested, and a non-contact electromagnetic coupling method is used to measure the cable harmonic signal characteristics. The extracted cable harmonic characteristics are analyzed through the change trend and relative proportional relationship of the harmonic characteristic fingerprint, so as to realize the insulation defect status evaluation of the cable.

[0114] S105. Process the harmonic data by using a preset detection algorithm to obtain harmonic characteristic data, so as to determine the defect type and status assessment information corresponding to the cable to be tested based on the harmonic characteristic data.

[0115] In one embodiment of the present application, sample data of the proportion of 2-11 harmonic components is used as input, and a covariance matrix is ​​constructed according to the standardized data set. The eigenvalues ​​of the covariance matrix are determined and the corresponding eigenvectors are determined. Based on the eigenvalues ​​and eigenvectors, the contribution rate and cumulative contribution rate corresponding to each principal component are determined. Harmonics are eliminated based on the contribution rate and cumulative contribution rate. Clustering is performed based on the remaining harmonic component data after elimination, and the influence relationship between different operating conditions and harmonic characteristics is obtained based on the clustering results; wherein the different operating conditions factors include at least one of metal shielding, hot and cold states, and load current.

[0116] Specifically, the embodiments of the present application determine the correlation between "insulation defects-operating conditions-harmonic characteristics" of distribution network cables by adopting data statistical analysis algorithms such as principal component analysis and cluster analysis, clarify the influence of operating conditions such as metal shielding, hot and cold states, and load current on harmonic characteristics, and accumulate experimental data for proposing a normalized algorithm for harmonic characteristic parameter indicators under different load conditions.

[0117] Furthermore, the harmonic proportion of the 2nd to 11th harmonic components was used as a quantitative variable (x 1 ,x 2 ,…,x 10 ) input, calculate the covariance matrix R based on the standardized data set:

[0118]

[0119] Calculate the eigenvalue λ of the matrix R 1 ≥λ 2 ≥…≥λ n ≥0 and the corresponding eigenvector u 1 ,u 2 ,…,u n , where u j =(u 1j ,u 2j ,…,u nj ),u nj Represents the nth component of the jth eigenvector; n new indicator variables are composed of eigenvectors:

[0120]

[0121] In the formula, y 1 is the first principal component, y 2 is the second principal component, …, y nis the nth principal component. Calculate each principal component y j Contribution rateb j (j=1,2,...,n) and y 1 ,y 2 , …, y n The cumulative contribution rate α of (p≤n) p .

[0122]

[0123] After calculation and analysis, principal component 1 contains the 2nd, 3rd, 5th, 7th, and 11th harmonics, which can be used for further quantitative characterization and analysis. The 4th, 6th, 8th, and 9th harmonics have the lowest contribution and the common factors have the lowest dependence on them, so the corresponding harmonics can be eliminated. Therefore, the 2nd, 3rd, 5th, 7th, and 11th harmonic components can be selected as feature quantities for further data analysis and research.

[0124] Randomly select k initial cluster centers C from the data set of harmonic components i (1≤i≤k), calculate the remaining data objects and the C in the cluster i The Euclidean distance is used to find the cluster center Ci closest to the target data object and assign the data object to the cluster center C. i Then calculate the average value of the data objects in each cluster as the new cluster center, and perform the next iteration until the cluster center no longer changes or the maximum number of iterations is reached.

[0125] The embodiment of the present application carries out cluster analysis on the harmonic measurement data of the damp cable samples, and takes into account the load current, armored or excluding armor, cooling and heating status, voltage level and other working conditions. The analysis results show that: armored or excluding armor, cooling and heating status have basically no effect on the clustering results; the clustering results of different load currents have obvious differentiation, indicating that the influence of local defects on the magnetic flux is related to the current amplitude; the data differentiation under high voltage is not obvious, but the differentiation of different load currents is still relatively high, indicating that the defects introduced by moisture are greatly affected by the current amplitude.

[0126] In one embodiment of the present application, the harmonic data is processed by a self-organizing map algorithm to obtain first harmonic characteristic data. A cable failure probability model and an operation risk assessment model are constructed by LASSO regression and Weibull distribution, and the harmonic data are processed by the model to obtain second harmonic characteristic data. The first harmonic characteristic data and the second harmonic characteristic data are respectively compared with a preset cable defect harmonic characteristic information table to obtain the defect type corresponding to the cable to be tested and the status assessment information corresponding to the cable to be tested.

[0127] Specifically, the self-organizing map (SOM) algorithm is a neural network model based on unsupervised learning. The training process includes: input selection, calculation of the neuron with the smallest distance, updating the weight vector, iterative training, model convergence, mapping and classification.

[0128] After the SOM model is trained, the harmonic characteristic data of the cable to be tested is input into the model. The model will map the input data to a neuron in the grid based on previous training experience. At this point, the type of cable defect can be identified by analyzing the distribution and adjacent relationship of neurons. For example, some areas where neurons are clustered may represent aging defects, while other areas may represent moisture defects.

[0129] Specifically, the harmonic current characteristics of XLPE materials under different degradation states were given in the previous study, but there is still a lack of quantitative indicators to judge the degree of cable degradation. For this reason, the LASSO regression analysis method is introduced to provide degradation degree discrimination. LASSO regression adds a regularization term to prevent overfitting on the basis of linear regression fitting data by minimizing the residual sum of squares. The objective function of LASSO regression is:

[0130]

[0131] In order to minimize the objective function, the λ value is first determined by the cross-validation method. Then, the variables screened by the model are determined by the λ and regression coefficient graph, where the standardized coefficient b i Variables that are zero can be considered to be eliminated by the LASSO regression model. Finally, the formula of the LASSO regression model is obtained:

[0132]

[0133] Take the normal heat-aged cable as an example, Fig. 9 A LASSO regression cross-validation diagram provided in the embodiment of the present application, such as Fig. 9 As shown in the figure, the vertical axis is the mean square error of the model, and the horizontal axis is λ. The optimization goal is to select a model with as few feature variables as possible and as small an error as possible corresponding to λ. There are two numerical dotted lines in the figure, the left one is the line with the lowest error, and the right one is the line with fewer feature variables.

[0134] When λ = 0, the minimum mean square error of the model is obtained. Within a variance range of λ = 0.0045, the model has the least characteristic variables. When the λ value reaches a certain size, continuing to increase the number of independent variables will not significantly improve the model performance.

[0135] For cables with normal electrical and thermal aging, Fig.10 A LASSO regression fitting coefficient trajectory diagram provided in the embodiment of the present application, such as Fig.10As shown in the figure, each curve represents the change trajectory of each independent variable coefficient, the ordinate is the value of the coefficient, and the abscissa is λ. As the value of λ changes, the corresponding variable whose coefficient is compressed to 0 later is more important. Fig.10 It can be found that the 6th, 7th, 8th, 9th, and 10th harmonics become zero first, so they are deleted. As the λ value changes, the 3rd and 5th harmonics are finally compressed to 0, so the 3rd and 5th harmonics are the most important for the dependent variable. It can be seen from the figure that if used for aging assessment, the 3rd harmonic has a more important role than the 5th harmonic. In addition, for heat-aged and water-immersion heat-aged cables, the difference in the 5th harmonic between the two is small. Therefore, the generalization of the obtained evaluation model will be better.

[0136] The results of LASSO regression show that based on the variable intercept term and the standardized coefficients of the 2nd to 11th harmonics, the variable intercept term, the 3rd harmonic, the 4th harmonic, the 5th harmonic, and the 11th harmonic are retained, and other harmonic variables are deleted. The standardized formula of the LASSO regression model for normal thermal aging cable is: y = 29.062 + 1876.404 × 3rd harmonic + 23477.972 × 4th harmonic + 5615.799 × 5th harmonic - 22796.236 × 11th harmonic. The root mean square error (RMSE) between the regression prediction value and the true value is calculated as an indicator to measure the accuracy of LASSO regression analysis.

[0137]

[0138] y i is the true value of the aging time, y m is the predicted value, and n is the number of test points. The results show that the LASSO regression analysis of normal thermally aged cables has a high accuracy, with a root mean square error of only 17.1, which accounts for only 0.14 of the actual aging time range. The predicted aging time of damp cables is longer and there is dispersion. The predicted aging time of damp cables is longer, which is consistent with the actual state. Moisture will accelerate the insulation aging rate. The main reason for the increase in dispersion is that the harmonic current test itself has a certain dispersion, and the effective insulation information and interference are intertwined, resulting in certain fluctuations in the evaluation results.

[0139] The embodiment of the present application uses the LASSO regression model to quantitatively analyze the degradation trend of typical defects of distribution network cables, and also covers four situations: semi-conductive suspension defects, metal tip defects, metal suspension defects on the insulation surface, and residual defects of insulation tape on the surface of the crimping tube. By introducing a regularization term, LASSO regression screens and reduces the dimensionality of high-dimensional feature data while ensuring the prediction accuracy of the model, thereby achieving accurate modeling and trend evaluation of the defect degradation process.

[0140] In one embodiment of the present application, the failure risk probability model based on the health state in the embodiments of the present application is an exponential function form failure probability model. This model believes that when the device state deteriorates, its failure rate increases exponentially, as shown in the following formula.

[0141] λ(CI) = K × e -C×CI

[0142] In the formula, K is the proportionality parameter; C is the curvature parameter, λ(CI) is the device failure occurrence probability, and CI is the failure state score value.

[0143] In one embodiment of the present application, a method for finding the device life distribution from data statistical analysis is proposed according to the Weibull analysis method, which lays a foundation for further analyzing device failures, predicting the development of failures, studying its failure mechanism, and formulating maintenance strategies.

[0144] Specifically, the curve of the failure rate of most electrical equipment changing with time is a classic bathtub curve, which is roughly divided into three stages: the early failure period, the accidental failure period, and the wear-out failure period. Fig.11 FIG. 11 is a schematic diagram of the bathtub curve of the failure rate of a power equipment provided by an embodiment of the present application. The two-parameter Weibull distribution has strong adaptability to these three failure periods. Its first parameter is the shape parameter k, which determines which distribution in the Weibull distribution has the best correlation. The second parameter is the scale parameter η, which defines most of the distribution states. The two parameters can be estimated from the data. The Weibull failure rate depends on the value of k, which shows the different failure possibilities of new or old components, thus depicting the bathtub curves of different components. The device failure rate λ(t) based on the Weibull distribution can be expressed as

[0145]

[0146] When k < 1, the failure rate shows a downward trend; when k = 1, the failure rate is a constant; when k > 1, the failure rate shows an upward trend. By analyzing the historical statistical data of the same type of electrical equipment, the failure rate curve based on the Weibull distribution can be segmented and fitted to obtain the parameters k and η of each stage.

[0147] The three stages of the bathtub curve: the early failure period, the accidental failure period, and the wear-out failure period.

[0148] Weibull distribution probability density function:

[0149] From the perspective of probability theory and statistics, the Weibull distribution is a continuous probability distribution, and the probability density of its two-parameter model is:

[0150] Where k and λ are the distribution state and scale parameter respectively.

[0151] In the Weibull distribution, the shape parameter k is the most important parameter. Its value determines the shape of the density function curve. When k = 1, the Weibull distribution is the exponential distribution. When 0 < k < 1, the graph is a curve asymptotic to the f(x) axis and the x-axis. When k > 1, the function graph has a maximum value. As k > 1 increases, the maximum value becomes larger and the graph becomes narrower. Fig.12 FIG. is a schematic diagram of the probability density curve of the Weibull distribution provided by an embodiment of the present application, as Fig.12 (a) shown; for the scale parameter λ, if the value of k remains unchanged, as the scale parameter η increases, the graph becomes flatter, as Fig.12 (b) shown.

[0152] The cumulative distribution function of the Weibull distribution:

[0153] The cumulative distribution function represents the probability of a failure occurring when the operating time of the device is less than or equal to x:

[0154]

[0155] As Fig.12 (a) shown, when the scale parameter λ is a fixed value, before reaching the intersection point in the figure, the larger the value of the shape parameter k, the smaller the possibility of early failure; as Fig.12 (b) shown, when the shape parameter k is a fixed value, as the scale parameter λ increases, the possibility of early failure of the device becomes smaller, and the possibility of the device safely passing through the early failure period becomes larger.

[0156] The reliability function of the device is defined by the Weibull cumulative distribution function. The probability that the device completes the specified function under the specified conditions and within the specified time is called the reliability of the device, denoted as R(x) or S(x).

[0157] R(x) = P(T > t) = 1 - P(T ≤ t) = 1 - F(x)

[0158] Of course, the reliability of the product is related to the time factor and thus related to the life of the product. However, the life of the product is also relative to certain usage conditions and refers to the "duration" during which the product has the specified function under certain usage conditions. The two are interdependent.

[0159] Failure rate function:

[0160] For a component that has worked until time t, the probability of failure occurring within the unit time after time t is called the failure rate of the component at time t, that is, the device failure rate, denoted as h(t):

[0161]

[0162] Obtained from the differential property and the conditional probability solution method:

[0163]

[0164] When k > 1, h(t) is an increasing function of t; when 0 < k < 1, h(t) is a decreasing function of t; when k = 1, h(t) = λ -1 is the failure rate function of the exponential distribution and is a constant λ -1 . Fig.14 is a schematic diagram of a failure rate function curve provided by an embodiment of the present application. From Fig.14 several failure rate curves with λ = 1 and m = 0.5, 1, 1.5, 5 can be obtained.

[0165] Furthermore, the embodiment of the present application uses the Levenberg - Marquardt algorithm to solve the fault probability function. The Levenberg - Marquardt algorithm is one of the optimization algorithms, mainly to find the parameter vector that makes the function value minimum.

[0166] According to the characteristics of harmonic components and the requirements of signal discrimination, the embodiment of the present application proposes the design requirements of non - contact harmonic component coupling technology. A live detection device for harmonic anomalies in distribution network cables integrating non - contact harmonic component detection sensors, signal modulation, interference suppression, and data analysis and processing units is constructed. To extend the power supply duration of the detection device, an online energy harvesting component is developed. By analyzing the influence of the layout of the energy harvesting coil, the number of coil turns, the core size, and the core material on the output voltage and power, an optimized design scheme and an energy management circuit suitable for magnetic field energy harvesting of three - phase cables are proposed, realizing voltage amplification, rectification filtering, and storage and distribution of the induced voltage. Through the study of the propagation characteristics of harmonic currents, the influence of key factors (such as the distance between the sensor and the measured cable, the relative distance from the defect position, the measurement angle, etc.) in the application process of non - contact sensors on the detection accuracy is analyzed, and a detection method suitable for harmonic signal acquisition of on - site distribution network cables and a harmonic anomaly detection application scheme are proposed.

[0167] Fig.15 is a schematic diagram of the structure of a distribution network cable defect detection device based on the characteristics of harmonic components provided by an embodiment of the present application, as Fig.15As shown, a distribution network cable defect detection device with harmonic component characteristics includes: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can: construct a distribution network cable harmonic abnormal characteristic live detection device based on a non-contact harmonic current sensor, a signal processing and analysis module, a power supply module and a display module; construct an energy acquisition module according to the magnetic field distribution characteristics around the cable and the output voltage of the induction coil, and construct an energy management module according to the output voltage and output power of the energy acquisition module, so as to construct an energy acquisition device based on the energy acquisition module and the energy management module; construct a harmonic detection device based on the energy acquisition device and the distribution network cable harmonic abnormal characteristic live detection device; perform harmonic data detection on the cable to be tested through the harmonic detection device; process the harmonic data through a preset detection algorithm to obtain harmonic characteristic data, so as to determine the defect type and status evaluation information corresponding to the cable to be tested based on the harmonic characteristic data.

[0168] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0169] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the embodiments of the present application may have various changes and variations. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A distribution network cable defect detection method based on harmonic component characteristics, characterized in that: The method comprises: Based on non-contact harmonic current sensor, signal processing and analysis module, power module and display module, a live detection device for harmonic abnormality characteristics of distribution network cables is constructed; An energy acquisition module is constructed according to the magnetic field distribution characteristics around the cable and the output voltage of the induction coil, and an energy management module is constructed according to the output voltage and output power of the energy acquisition module, so as to construct an energy acquisition device based on the energy acquisition module and the energy management module; Based on the energy taking device and the distribution network cable harmonic abnormality characteristic live detection device, a harmonic detection device is constructed; Performing harmonic data detection on the cable to be tested by the harmonic detection device; The harmonic data is processed by a preset detection algorithm to obtain harmonic characteristic data, so as to determine the defect type and status assessment information corresponding to the cable to be tested based on the harmonic characteristic data.

2. A distribution network cable defect detection method based on harmonic component characteristics according to claim 1, characterized in that: The energy acquisition module is constructed according to the magnetic field distribution characteristics around the cable and the output voltage of the induction coil, specifically including: The output voltage of the induction coil is determined by building a three-dimensional simulation model of the three-core cable; Based on the output voltage and the coil angle position and the current phase corresponding to the three-core cable, the layout of the energy acquisition module is determined; wherein the energy acquisition module is used to collect the magnetic field energy of the three-core cable; Inputting preset key structural parameters into the three-dimensional model of the constructed three-core cable, and determining the output power corresponding to the energy extraction module under different preset key parameters; wherein the preset key structural parameters include at least one of the number of coil turns, wire diameter, magnetic thickness and magnetic core length; An optimal parameter is determined based on the output power, and optimization is performed based on the optimal parameter to obtain the energy acquisition module.

3. A distribution network cable defect detection method based on harmonic component characteristics according to claim 2, characterized in that: Before determining the output voltage of the induction coil by constructing the three-dimensional simulation model of the three-core cable, the method further includes: The three-phase conductor corresponding to the three-core cable is used as an infinitely long straight conductor to pass a three-phase power frequency current, the magnetic field generated by the three-phase conductor in the surrounding space is obtained, and a magnetic field calculation function is established; According to the coordinates of the field point in the polar coordinate system and the distance from the cable center to the conductor center, determine the distance from the field point to the conductor center and establish a distance calculation function; The distance calculation function is substituted into the magnetic field calculation function, and the magnetic field generated by the three-phase conductor at the field point is directional-decomposed to obtain the spatial magnetic field distribution function around the cable, so as to simplify the cable model.

4. A distribution network cable defect detection method based on harmonic component characteristics according to claim 1, characterized in that: The spatial magnetic field distribution function around the cable is: Among them, B A is the magnetic field of the surrounding space corresponding to A; B B is the magnetic field in the surrounding space corresponding to B; B C is the magnetic field in the surrounding space corresponding to C; u0 is the magnetic permeability of vacuum; I A is the A phase power frequency current; I B is the power frequency current of phase B; I C is the C-phase power frequency current; r and φ are the coordinates of the field point in the polar coordinate system; a is the distance from the cable center to the conductor center.

5. A distribution network cable defect detection method based on harmonic component characteristics according to claim 1, characterized in that: The energy management module is constructed according to the output voltage and output power of the energy acquisition module, specifically comprising: Based on the circuit topology of the half-wave rectifier boost circuit, the output voltage of each induction coil is rectified by half-wave rectification to obtain a DC output voltage, and the output voltages of the three induction coils are superimposed in series to increase the output voltage of the energy extraction device; Based on the energy management chip, the output voltage of the half-wave rectifier boost circuit is stabilized; Connecting a plurality of the induction coils in parallel, and outputting the parallel-connected coils according to a rectifier boost circuit to increase the output power of the energy extraction module; The energy management module is constructed by using the half-wave rectification and boosting circuit, the energy management chip and the plurality of induction coils connected in parallel.

6. A method for detecting defects in distribution network cables based on harmonic component characteristics according to claim 1, characterized in that: Before the harmonic data is processed by a preset detection algorithm to obtain harmonic characteristic data, the method further includes: The sample data of the proportion of harmonic components from 2nd to 11th order are taken as input, and the covariance matrix is ​​constructed based on the standardized data set; Determining eigenvalues ​​and corresponding eigenvectors of the covariance matrix; Based on the eigenvalues ​​and the eigenvectors, determining the contribution rate and the cumulative contribution rate corresponding to each principal component; Eliminating harmonics based on the contribution rate and the cumulative contribution rate; Clustering is performed based on the remaining harmonic component data after elimination, and the influence relationship between different operating conditions and harmonic characteristics is obtained based on the clustering results; wherein the different operating conditions factors include at least one of metal shielding, hot and cold states, and load current.

7. A distribution network cable defect detection method based on harmonic component characteristics according to claim 1, characterized in that: The processing of the harmonic data by a preset detection algorithm to obtain harmonic characteristic data, and determining the defect type and status assessment information corresponding to the cable to be tested based on the harmonic characteristic data, specifically includes: Processing the harmonic data by a self-organizing map algorithm to obtain first harmonic characteristic data; A cable failure probability model and an operation risk assessment model are constructed by LASSO regression and Weibull distribution, and the harmonic data are processed by the model to obtain second harmonic characteristic data; The first harmonic characteristic data and the second harmonic characteristic data are respectively compared with a preset cable defect harmonic characteristic information table to obtain the defect type corresponding to the cable under test and the status evaluation information corresponding to the cable under test.

8. A method for detecting defects in distribution network cables based on harmonic component characteristics according to claim 7, characterized in that: Before the harmonic data is processed by the self-organizing map algorithm to obtain the first harmonic characteristic data, the method further includes: Acquiring sample harmonic data, and preprocessing the sample harmonic data; The preprocessed sample harmonic data is used as an input sample to construct and train a self-organizing map model to obtain a trained self-organizing map model; The processing of the harmonic data by principal component analysis and self-organizing mapping algorithm to obtain first harmonic characteristic data specifically includes: Inputting the harmonic data into the trained self-organizing map model, and mapping the harmonic data onto neurons in a grid through the trained self-organizing map model; Based on the distribution and adjacent relationship of the neurons, first harmonic characteristic data is obtained, so as to determine the defect type of the cable to be tested based on the first harmonic characteristic data.

9. A method for detecting defects in distribution network cables based on harmonic component characteristics according to claim 7, characterized in that: The cable failure probability model and operation risk assessment model are constructed by LASSO regression and Weibull distribution, specifically including: LASSO regression is used to fit the data by minimizing the residual sum of squares, and regularization terms are added to prevent overfitting, so as to construct the cable failure probability model to quantitatively analyze the deterioration trend of distribution network cable defects; wherein the distribution network cable defects include at least one of semi-conductive suspension defects, metal tip defects, insulation surface metal suspension defects, and residual defects of insulation tape on the surface of the crimping tube; Construct a failure risk probability model based on health status through a preset exponential function; Based on the Weibull distribution and fitting the bathtub curve, model parameters are determined, and a failure risk probability model based on the operating years is constructed based on the determined model parameters; The operation risk assessment model is obtained based on the health status-based failure risk probability model and the operation years-based failure risk probability model.

10. A distribution network cable defect detection device based on harmonic component characteristics, characterized in that: The device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to: Based on non-contact harmonic current sensor, signal processing and analysis module, power module and display module, a live detection device for harmonic abnormality characteristics of distribution network cables is constructed; An energy acquisition module is constructed according to the magnetic field distribution characteristics around the cable and the output voltage of the induction coil, and an energy management module is constructed according to the output voltage and output power of the energy acquisition module, so as to construct an energy acquisition device based on the energy acquisition module and the energy management module; Based on the energy taking device and the distribution network cable harmonic abnormality characteristic live detection device, a harmonic detection device is constructed; Performing harmonic data detection on the cable to be tested by the harmonic detection device; The harmonic data is processed by a preset detection algorithm to obtain harmonic characteristic data, so as to determine the defect type and status assessment information corresponding to the cable to be tested based on the harmonic characteristic data.

Images