Method for testing the insulation of a wire or cable conductor

Abstract

The application belongs to the technical field of wire and cable detection, and discloses a wire and cable conductor insulation performance inspection method, which aims to solve the problem that traditional electrical measurement method is difficult to identify early insulation micro-deterioration. The method comprises the following steps: applying a nanosecond voltage pulse excitation to the conductor to excite mechanical vibration waves in the insulation medium; collecting shear wave time domain signals by using an acoustic sensor array surrounding the cable and extracting the propagation speed; calculating the spatial distribution of the elastic modulus by using a reverse problem reconstruction algorithm; synchronously obtaining local discharge signal characteristics; aligning and fusing the time and space of the two, constructing a comprehensive evaluation model, and outputting the quantitative results of the micro-mechanical embrittlement degree and irreversible thermal aging grade. By introducing the shear wave elasticity imaging principle and combining multi-source information fusion analysis, the application realizes non-destructive, high-sensitivity early diagnosis of the insulation state of the in-service cable, improves the early warning capability, and is suitable for online detection and intelligent operation and maintenance in various laying scenes.

Description

Technical Field

[0001] This invention belongs to the field of wire and cable testing technology, and specifically relates to a method for testing the insulation performance of wire and cable conductors. Background Technology

[0002] With the continuous advancement of power infrastructure construction, wires and cables, as the core carriers of energy transmission, play an irreplaceable role in maintaining the long-term stable operation of the entire power grid system due to the integrity of their insulation layers. During long-term operation, the insulating medium is subjected to multiple coupled effects of electric field stress, thermal load, and mechanical stress, causing its physical and chemical properties to undergo slow and complex evolution. To ensure the safe operation of the power system, real-time monitoring and periodic assessment of cable insulation performance have become key research topics in the field of power maintenance.

[0003] The testing of wire and cable insulation performance mainly focuses on capturing and diagnosing the electrical parameters of the dielectric, aiming to predict the risk of insulation failure through specific testing methods. Conventional performance evaluation typically relies on applying high-voltage loads to the conductor and collecting key indicators such as partial discharge pulses or leakage current intensity in real time to determine the electrical response of the insulation layer under a strong electric field environment. This testing mode emphasizes the end-stage identification of insulation failure results, attempting to indirectly map the electrical strength and service life of the insulation material through abnormal fluctuations in electrical parameters.

[0004] Traditional electrical testing methods struggle to effectively assess the degree of internal microscopic mechanical embrittlement and early irreversible thermal aging in insulating materials before significant electrical breakdown or macroscopic air gap discharge occurs, resulting in a significant lag in early warning and prevention. Existing technologies rely excessively on the sensitivity of electrical characteristics, lacking the ability to perceive the evolution of microscopic mechanical parameters in the insulating medium, and are unable to achieve non-destructive quantitative assessment of the insulation layer before electrical damage occurs. Furthermore, traditional evaluation systems fail to deeply integrate acoustic-mechanical responses with electrical state characteristics, resulting in a lack of accurate feature extraction and correlation algorithms when facing multi-source heterogeneous signals, and an inability to construct an intuitive and high-precision physical state perception system. Summary of the Invention

[0005] The purpose of this invention is to provide a method for testing the insulation performance of conductors in wires and cables, thereby solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides a method for testing the insulation performance of conductors in wires and cables, comprising the following steps:

[0007] A voltage pulse with a preset amplitude and pulse width is applied to the conductor of the wire or cable under test, causing mechanical vibration waves to be generated inside the insulating medium due to the electrostriction effect.

[0008] The time-domain signal of the mechanical vibration wave propagating in the insulating medium is synchronously acquired using an acoustic sensor array arranged around the outer surface of the cable;

[0009] Based on the acquired time-domain signal, the propagation velocity information of the shear wave at different locations in the insulating medium is extracted; using an inverse problem reconstruction algorithm, the elastic modulus values ​​corresponding to each region inside the insulating medium are calculated based on the propagation velocity information, and a spatial distribution map is generated.

[0010] The partial discharge signal generated during the application of the voltage pulse excitation is acquired synchronously, and its amplitude, phase and repetition frequency characteristics are extracted.

[0011] The spatial distribution map and the partial discharge signal characteristics are spatiotemporally aligned and data fused to construct a comprehensive insulation status evaluation model.

[0012] Based on the comprehensive evaluation model, the micromechanical embrittlement degree and irreversible thermal aging level of the insulating medium are output.

[0013] Preferably, applying a voltage pulse excitation with a preset amplitude and pulse width to the conductor of the wire or cable under test includes: the voltage pulse excitation using a unipolar or bipolar square wave pulse;

[0014] The amplitude of the voltage pulse excitation is adaptively adjusted according to the rated voltage level of the wire and cable under test, and the width of the voltage pulse excitation is in the range of 50 nanoseconds to 500 nanoseconds.

[0015] The electric field change rate generated by the rising edge time of the voltage pulse excitation causes the molecules inside the insulating medium to undergo instantaneous polarization orientation and displacement, thereby inducing high-frequency mechanical strain.

[0016] Preferably, the time-domain signal of the mechanical vibration wave propagating in the insulating medium is synchronously acquired using an acoustic sensor array arranged around the outer surface of the cable, including: the acoustic sensor array includes a piezoelectric ceramic sensor or a fiber Bragg grating sensor.

[0017] The acoustic sensor array is integrated into a flexible ring clamp and attached to the surface of the cable's metal shielding layer or outer sheath.

[0018] The acoustic sensor array presents a two-dimensional sensing network structure in space. It is arranged at preset intervals along the cable axis, and multiple sensing units are evenly distributed in the circumferential direction of each axial section to capture the propagation characteristics of shear waves in the radial, axial and circumferential dimensions.

[0019] Preferably, based on the acquired time-domain signal, the propagation velocity information of the shear wave at different positions in the insulating medium is extracted, including: performing bandpass filtering preprocessing on the time-domain signal received by each channel to filter out low-frequency electromagnetic interference and high-frequency system noise;

[0020] The process of extracting the shear wave propagation velocity is achieved by performing cross-correlation analysis on the signal sequences received by two adjacent sensors along the wave propagation direction.

[0021] The cross-correlation analysis includes using a sliding window mechanism to extract the effective data segment containing the first wavefront oscillation, and calculating the time offset corresponding to when the cross-correlation function reaches its maximum value. The time offset is determined as the time delay of the shear wave propagating within the distance between the two sensors.

[0022] The wave velocity in a local area is calculated using the quotient of the physical distance between the sensors and the time delay.

[0023] In the process of the cross-correlation analysis, the signal sequence is first subjected to Hilbert transform to extract the instantaneous envelope of the signal, thereby reducing the impact of phase fluctuations on time delay estimation.

[0024] Calculate the integral value of the product of two envelope sequences on the time axis, and find the time offset that maximizes the integral value by continuously shifting the signal sequence;

[0025] Parabolic interpolation is used to fit discrete data near the cross-correlation peak point to obtain sub-sampling period time delay accuracy;

[0026] The sliding window continuously scans the acquired sequence signals to extract the wave velocity evolution patterns along the axial and circumferential directions within the insulating medium.

[0027] Preferably, an inverse problem reconstruction algorithm is used to calculate the elastic modulus values ​​corresponding to each region inside the insulating medium based on the propagation speed information, and generate a spatial distribution map, including: establishing a mapping relationship between shear wave velocity and material mechanical parameters;

[0028] The mapping relationship satisfies that the square of the propagation velocity of a shear wave in an isotropic medium is equal to the quotient of the shear modulus and the density.

[0029] The inverse problem reconstruction algorithm adopts an iterative optimization strategy, using the collected multipath wave velocity observations as constraints. By minimizing the sum of squared residuals between the measured wave velocity and the wave velocity predicted by the theoretical model, the elastic modulus value of each voxel in the three-dimensional spatial grid inside the insulating medium is solved.

[0030] The spatial distribution map is stored in the form of a three-dimensional matrix, with each coordinate point corresponding to a specific elastic modulus value;

[0031] During the execution of the inverse problem reconstruction algorithm, an iterative optimization solver based on the regularized Gauss-Newton algorithm is used; in each iteration, the theoretical propagation time on each observation path is calculated using the forward model of the wave equation based on the current initial value of the elastic modulus.

[0032] Calculate the residual vector between the measured time and the theoretical propagation time, and construct the Jacobian matrix that reflects the sensitivity of wave velocity to changes in local elastic modulus;

[0033] A Tikhonov regularization term is added to the objective function to constrain the spatial smoothness of the elastic modulus distribution; the elastic modulus value is updated by solving a system of linear equations until the L2 norm of the residual decreases below the preset convergence threshold.

[0034] Preferably, the partial discharge signal generated during the application of the voltage pulse excitation is acquired synchronously, including: acquiring the current signal using a high-frequency current transformer or capacitive coupling probe installed on the grounding lead of the cable terminal.

[0035] The sampling frequency of the partial discharge signal is not less than 100 MHz;

[0036] The extracted features include the apparent amplitude of the discharge charge, the phase distribution relative to the excitation pulse period, and the pulse repetition frequency within the preset observation time.

[0037] The partial discharge signal reflects the distribution of electrical weaknesses in the insulating medium under the action of an electric field, and complements the mechanical weaknesses reflected by the elastic modulus value.

[0038] Preferably, the spatial distribution map and the partial discharge signal characteristics are spatiotemporally aligned and fused to construct a comprehensive insulation state evaluation model, including:

[0039] The spatial coordinates of the acoustic signal location are correlated with the time of partial discharge signal generation based on the sound speed propagation delay;

[0040] The data fusion process adopts a multi-source information weighted association strategy, which is achieved by establishing a multi-dimensional feature vector that includes the elastic modulus deviation, partial discharge amplitude density, and discharge phase clustering index.

[0041] The weighted correlation strategy includes determining the overlap between the region of decreased elastic modulus and the region of active partial discharge in spatial topology. If there is overlap, the location is determined to be a high-risk aging point, and a higher confidence weight is assigned according to the correlation strength.

[0042] In the data fusion process, the weight matrix is ​​determined based on the principle of information entropy. By analyzing the mutual contribution of elastic modulus characteristics and partial discharge characteristics in identifying known defect samples, the weighting coefficients of the two are dynamically adjusted.

[0043] When the service life of the cable under test exceeds the preset aging threshold, the weight of the elastic modulus characteristic is increased.

[0044] When the cable under test is in the new installation stage, the focus is on the partial discharge characteristics to capture construction process defects;

[0045] The comprehensive evaluation model calculates a comprehensive score representing insulation health by introducing nonlinear fusion weights.

[0046] Preferably, the comprehensive evaluation model outputs the degree of micromechanical embrittlement and the irreversible thermal aging level of the insulating medium, including:

[0047] The degree of micromechanical embrittlement is characterized by calculating the relative ratio of the average local elastic modulus obtained from the current measurement to the reference elastic modulus value of the same type of cable in the new cable state.

[0048] The irreversible thermal aging levels are divided into normal operation level, early aging level, mid-term degradation level and critical level based on multiple preset discrete threshold ranges.

[0049] When the relative ratio drops to the first threshold, the system issues a warning of mechanical performance degradation.

[0050] When the relative ratio is lower than the second threshold and the partial discharge frequency exceeds the background noise threshold, it is determined to be a severe emergency level.

[0051] Preferably, the method further includes the step of establishing a historical detection database, which is used to store spatial distribution maps, partial discharge characteristic data and environmental parameters obtained from previous tests;

[0052] By using pattern recognition algorithms to perform time-series analysis on the detection data of the same cable at different time points, the evolution pattern of aging trend can be identified.

[0053] Calculate the aging rate, which is the rate of change of elastic modulus with operating time;

[0054] The remaining effective life of the cable is estimated based on the aging rate.

[0055] Preferably, the method further includes a step of real-time monitoring of the cable's ambient temperature and humidity before testing;

[0056] Real-time temperature field data is obtained using temperature measuring devices placed on the surface of the cable;

[0057] Environmental parameters are introduced as correction factors in the calculation of the elastic modulus value;

[0058] The calculated elastic modulus is normalized using a pre-calibrated temperature compensation curve to eliminate the interference of external environmental fluctuations on the sound wave propagation speed.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] 1. This invention solves the problem that traditional electrical testing methods rely solely on electrical parameters for insulation assessment. It introduces the shear wave elastic imaging principle from the medical field into the field of wire and cable insulation performance testing, enabling non-destructive perception of the microscopic mechanical state inside the insulating medium.

[0061] 2. By simultaneously acquiring mechanical vibration response and partial discharge signals and performing in-depth fusion analysis, the aging risk can be accurately identified in the early stage when the insulation layer has not yet undergone electrical breakdown and only the elastic modulus has decreased, thus improving the early warning capability and timeliness.

[0062] 3. The method does not require stripping the sheath or damaging the structure, making it suitable for periodic health assessments of in-service cables, extending equipment lifespan and reducing the probability of sudden failures.

[0063] 4. The constructed comprehensive evaluation model has good adaptability and scalability, and can perform customized diagnosis for cables of different materials, structures and operating environments, providing a brand-new technical path for condition-based maintenance of smart grids. Attached Figure Description

[0064] Figure 1 This is a schematic diagram of the overall technical solution architecture of the present invention;

[0065] Figure 2 This is a schematic diagram of the core principle framework for reconstructing the inverse problem of the elastic modulus of an insulating medium based on the shear wave propagation velocity in this invention.

[0066] Figure 3 This is a logical flowchart of the spatiotemporal alignment and data fusion of spatial distribution map and partial discharge signal characteristics in this invention;

[0067] Figure 4 This is a schematic diagram of the multi-level interaction relationship and data flow between the acoustic sensor array, the excitation device, and the comprehensive evaluation terminal in this invention;

[0068] Figure 5 This is a logical flowchart of the output of the micromechanical embrittlement degree and thermal aging level of the insulating medium based on the comprehensive evaluation model in this invention. Detailed Implementation

[0069] Example 1: Reference Figures 1 to 5 In a method for testing the insulation performance of wire and cable conductors according to the present invention, a deep understanding of the health status of the cable insulation medium is achieved by integrating acoustic detection technology and high-voltage electrical testing technology. The method is specifically implemented according to the following steps:

[0070] A voltage pulse with a preset amplitude and pulse width is applied to the conductor of the tested wire or cable to induce weak mechanical vibration waves within the insulating medium due to electrostriction. In practice, the voltage pulse excitation device is physically coupled to the cable's core conductor via a high-voltage coaxial connector. The voltage pulse excitation is set to a narrow pulse waveform that is either unipolar or bipolar.

[0071] To ensure that the excitation signal can induce a mechanical vibration signal with a sufficient signal-to-noise ratio without causing irreversible electrical breakdown of the insulation layer, the amplitude of the voltage pulse is adaptively adjusted according to the rated voltage level of the cable. For example, for a cross-linked polyethylene cable with a rated voltage of 10 kV, the pulse amplitude is controlled between 2 kV and 5 kV. The pulse width is controlled within the range of 50 nanoseconds to 500 nanoseconds. The rise time of the applied voltage pulse excitation is precisely controlled within the nanosecond range, typically 10 nanoseconds to 50 nanoseconds.

[0072] The rise time of the voltage pulse excitation can generate a high electric field change rate inside the insulating medium. According to the principle of electrostriction, this causes the cable insulation molecules to undergo instantaneous polarization orientation and displacement under the action of a strong field, thereby inducing high-frequency mechanical strain. This strain diffuses outward from inside the insulation layer in the form of sound waves, forming the original acoustic carrier required for subsequent detection.

[0073] A high-sensitivity acoustic sensor array, arranged around the outer surface of the cable, synchronously acquires the time-domain signal of the mechanical vibration wave propagating in the insulating medium. To achieve full circumferential and radial coverage of the cable without blind spots, the acoustic sensor array employs piezoelectric ceramic sensors or fiber Bragg grating sensors. These sensors are integrated within a flexible annular clamp of adjustable diameter, tightly fitted to the surface of the cable's metal shielding layer or outer sheath.

[0074] The sensor array is spatially distributed in a two-dimensional sensing network structure, with a group of sensors arranged at 5 to 10 cm intervals along the cable axis, and 8 to 16 sensing units evenly distributed circumferentially in each axial section. This configuration can capture the complete propagation characteristics of shear waves in the radial, axial, and circumferential dimensions. The acquisition system has a multi-channel synchronous triggering function. When the synchronous trigger pulse emitted by the voltage pulse excitation device reaches the acquisition module, all sensing channels simultaneously start high-frequency sampling. The sampling frequency is set between 20 MHz and 50 MHz to ensure that the captured time-domain signal contains rich high-frequency details of the shear wave and avoid signal distortion.

[0075] Based on the acquired time-domain signals, the propagation velocity information of shear waves at different locations in the insulating medium is extracted. In the specific signal processing flow, the weak acoustic signals received by each channel are first preprocessed, including filtering out low-frequency electromagnetic interference and high-frequency system noise using bandpass filters with a passband range of 100 kHz to 2 MHz. Subsequently, the method for extracting the shear wave propagation velocity employs an advanced signal correlation analysis strategy. Specifically, signal sequences received by two adjacent sensors along the wave propagation direction are selected, and a sliding window mechanism is used to extract the effective data segment containing the first wavefront oscillation.

[0076] Cross-correlation analysis was performed on the two data segments to calculate the time offset corresponding to the maximum value of the cross-correlation function. This time offset represents the time delay of the shear wave propagating within the distance between the two sensors. Combined with the known physical distance between the sensors, the wave velocity in the local region was calculated using the quotient of distance and time delay. The process involves continuous scanning of the entire acquired long sequence signal using a sliding window, which reflects the evolution of wave velocity along the axial and circumferential directions within the insulation layer.

[0077] An inverse problem reconstruction algorithm is employed to calculate the elastic modulus values ​​corresponding to each region within the insulating medium based on the propagation velocity information, and to generate a spatial distribution map. The core physics lies in establishing a deterministic mapping relationship between shear wave velocity and material mechanical parameters. According to linear elasticity theory, the propagation velocity of transverse waves (i.e., shear waves) in an isotropic medium is related to the medium's shear modulus and density by the following condition: the square of the velocity equals the quotient of the shear modulus and the density.

[0078] In this invention, the physical relationship between shear wave velocity and elastic modulus is modeled as a nonlinear mapping function, further considering the conversion logic between Young's modulus, shear modulus, and Poisson's ratio. The inverse problem reconstruction process employs an iterative optimization strategy, using the collected multipath wave velocity observations as constraints. By minimizing the sum of squared residuals between the measured wave velocity and the theoretical model-predicted wave velocity, the elastic modulus value of each voxel in the three-dimensional spatial grid within the insulating medium is solved. The generated spatial distribution map is stored in the form of a three-dimensional matrix, with each coordinate point corresponding to a specific elastic modulus value, thus realizing a digital characterization of the distribution of mechanical properties within the insulating medium.

[0079] The partial discharge signal generated during the application of the voltage pulse excitation is simultaneously acquired, and its amplitude, phase, and repetition frequency characteristics are extracted. If air gaps, impurities, or electrical trees formed due to aging exist within the cable during the application of the high-voltage pulse excitation, local charge movement can be induced, resulting in partial discharge. The partial discharge signal is acquired using a high-frequency current transformer or a capacitively coupled probe, which is installed on the grounding lead at the cable terminal.

[0080] To capture nanosecond-level pulse components, the sampling frequency was set to at least 100 MHz. The acquired raw current signal was fed into a digital signal processing unit via a high-speed analog-to-digital converter for automatic pulse identification and extraction. The extracted features included: the apparent amplitude of the discharge charge, the phase distribution relative to the excitation pulse period, and the pulse repetition frequency within a preset observation time. These electrical parameters reflect the distribution of electrical weaknesses in the insulating medium under the influence of an electric field, complementing the observed mechanical weaknesses.

[0081] The spatial distribution map and the partial discharge signal characteristics are spatiotemporally aligned and fused to construct a comprehensive insulation status assessment model. Since the speed of sound propagation is much lower than that of electrical signals, strict temporal alignment is required before fusion. The spatial coordinates of the acoustic signal location are correlated with the time of partial discharge signal generation based on the sound propagation delay. A multi-source information weighted correlation strategy is adopted for data fusion. A multi-dimensional feature vector is established, including elastic modulus deviation, partial discharge amplitude density, and the clustering index of the discharge phase. If the elastic modulus of a certain region shows a significant decrease, and the region coincides with an active point of partial discharge in spatial topology, the correlation logic determines the location as a high-risk aging point, and assigns a higher confidence weight based on the correlation strength. The comprehensive assessment model calculates a comprehensive score representing insulation health by introducing nonlinear fusion weights.

[0082] The comprehensive evaluation model outputs quantitative results of the micromechanical embrittlement degree and irreversible thermal aging level of the insulating medium. The micromechanical embrittlement degree is characterized by the relative elastic modulus ratio, specifically defined as the ratio of the currently measured average local elastic modulus to the baseline elastic modulus value of the same type of cable in its new cable state. When the ratio drops to a preset first threshold (e.g., 0.7), the system prompts an alert for mechanical performance degradation. The irreversible thermal aging level is divided into four discrete levels based on multiple preset discrete threshold ranges: normal operation, early aging, intermediate degradation, and severe critical. For example, if the elastic modulus ratio is between 0.8 and 1.0 and the partial discharge level is below the background noise threshold, it is judged as the normal level; if the elastic modulus ratio is between 0.7 and 0.8 and the partial discharge level is below the background noise threshold, it is judged as the early aging level; if the elastic modulus ratio is below 0.7 and the partial discharge frequency does not exceed the background noise threshold, it is judged as the intermediate degradation level; if the elastic modulus ratio is below 0.5 and the partial discharge frequency increases, it is judged as the severe critical level. These quantitative results are directly output to the operation and maintenance decision support system to guide cable replacement or repair plans.

[0083] In an extended embodiment of the present invention, to further enhance the long-term effectiveness and trend prediction capability of the detection, the method further includes establishing a historical detection database. This historical detection database adopts a relational database architecture, storing spatial distribution maps, partial discharge characteristic data, and corresponding environmental parameters obtained from each inspection. By introducing pattern recognition algorithms, such as support vector machines or deep neural networks, time-series analysis is performed on the detection data of the same cable at different time points to automatically identify the evolution pattern of aging trends. The system can calculate the aging rate, i.e., the rate of change of the elastic modulus with operating time, and thereby estimate the remaining effective lifespan.

[0084] In another detailed embodiment of the invention, considering the sensitivity of environmental factors to the speed of sound wave propagation, the method further includes real-time monitoring of the cable's ambient temperature and humidity before testing. A thermistor or infrared thermometer is placed on the cable surface to acquire real-time temperature field data. Since the elastic modulus of polymer materials exhibits a significant temperature dependence, increased temperature typically leads to material softening, i.e., a decrease in the elastic modulus value. Therefore, this embodiment incorporates environmental parameters as correction factors into the elastic modulus calculation model. The originally calculated elastic modulus is normalized using a pre-calibrated temperature compensation curve to eliminate interference from external environmental fluctuations, ensuring the comparability of in-service testing results under different seasons and climatic conditions.

[0085] In the hardware-integrated implementation, the acoustic sensor array and voltage excitation device are highly integrated into a portable testing terminal. This portable testing terminal features an electromagnetic interference-resistant shielded casing and incorporates a high-capacity lithium battery pack and an embedded high-performance computing platform. It supports rapid on-site deployment; technicians do not need to disassemble the cable or disconnect the power to strip the sheath. They can simply attach the flexible sensor probe to the area to be tested to achieve non-destructive online or semi-online testing. This implementation expands the application scenarios of the method, making it suitable not only for routine cable tunnel maintenance but also for rapid on-site fault location and condition assessment of complex overhead lines, buried underground cables, and submarine cables.

[0086] To assist non-experts in decision-making, the method also includes a visualization interface module. This module runs on a terminal display or remote monitoring computer, using a graphics rendering engine to present the spatial distribution map in three-dimensional space as a pseudo-color heatmap. Areas with high elastic modulus are displayed in blue, while areas with low elastic modulus, i.e., areas prone to embrittlement, are displayed in red. Simultaneously, hotspot markers for partial discharge are overlaid on the interface, with flashing icons representing the intensity of the discharge. This intuitive visual feedback enables a clear display of the cable insulation status and precise anomaly location.

[0087] The comprehensive evaluation model incorporates an expert rule base. This base contains templates representing typical aging characteristics for cables of different materials and structures. Each template records the expected evolution path of the material's elastic modulus and partial discharge characteristics over a specific number of years of operation. When the real-time acquired data deviates from the expected evolution path to a predetermined statistical deviation range, the expert system automatically triggers anomaly analysis to determine whether the aging is due to overall thermal degradation caused by long-term overload or localized damage caused by localized mechanical stress or moisture.

[0088] The specific implementation details of cross-correlation analysis are as follows: First, a Hilbert transform is performed on the acquisition sequences of sensor A and sensor B to extract the instantaneous envelope of the signal, thereby reducing the impact of phase fluctuations on time delay estimation. The integral value of the product of the two envelope sequences on the time axis is calculated. By continuously shifting the signal sequences, the time offset that maximizes the integral value is found. This time offset is considered as the group delay of the signal propagating from sensor A to sensor B. To improve resolution, parabolic interpolation is used to fit the discrete data near the cross-correlation peak point to obtain sub-sampling period time delay accuracy.

[0089] The iterative optimization solver used is based on the regularized Gauss-Newton algorithm. In each iteration, the theoretical propagation time on each observation path is first calculated using the forward model of the wave equation based on the current initial value of the elastic modulus. Then, the residual vector between the measured time and the theoretical propagation time is calculated, and a Jacobian matrix is ​​constructed, which reflects the sensitivity of wave speed to changes in the local elastic modulus. To prevent instability in the inverse problem solution process, a Tikhonov regularization term is added to the objective function to constrain the spatial smoothness of the elastic modulus distribution. The elastic modulus value is updated by solving a system of linear equations until the L2 norm of the residual decreases below a preset convergence threshold.

[0090] The weighting matrix is ​​determined based on the principle of information entropy. By analyzing the mutual contributions of elastic modulus characteristics and partial discharge characteristics in identifying known defect samples, the weighting coefficients of the two are dynamically adjusted. For example, when inspecting old cables that have been in operation for over 20 years, considering that overall mechanical embrittlement may occur earlier than partial discharge, the system automatically increases the weight of the elastic modulus characteristic. However, in the acceptance testing of newly installed cables, the system focuses on partial discharge characteristics to capture air gaps or impurities caused by construction defects.

[0091] Example 2: Based on Example 1, this example makes adaptive improvements to the sensor array and signal transmission method to meet the special monitoring needs of long-distance submarine cables.

[0092] In the above method, the acoustic sensor array uses a distributed fiber optic sensing system instead of traditional piezoelectric ceramic sensors. It utilizes the communication optical fiber integrated into the cable structure or specially laid sensing optical fiber as the sensing medium. By emitting highly coherent light pulses into the optical fiber and detecting the phase modulation of Rayleigh scattered light in the fiber, continuous sensing of mechanical vibration waves along the entire cable is achieved. This implementation increases the sensor distribution density to the meter or even sub-meter level, enabling seamless monitoring of cable insulation conditions over distances of several kilometers or even tens of kilometers.

[0093] In the above method, the applied voltage pulse excitation is coupled into the conductor through the terminal box at the end of the cable. Due to the large inductance and capacitance parameters of long-distance cables, to prevent severe dispersion and attenuation of the pulse waveform during long-distance transmission, the voltage pulse excitation device employs pulse shaping network technology. By adjusting the parameters of a multi-stage LC circuit, pre-distortion compensation is performed on the output pulse. This ensures that the pulse maintains sufficient rise edge steepness when it reaches the distant detection point, thus exciting a detectable mechanical response.

[0094] To address the high hydrostatic pressure characteristics of submarine cables in deep-water environments, the inverse problem reconstruction algorithm further introduces a pressure correction term. Hydrostatic pressure alters the prestress state of the insulation material, affecting the shear wave velocity. In this embodiment, water depth data is acquired and converted into pressure values. These pressure values ​​are then used as background stress parameters input into the linear elasticity model to correct the relationship between wave velocity and elastic modulus, ensuring the accuracy and objectivity of the elastic modulus calculated under different water depths.

[0095] In the above method, the partial discharge signal acquisition employs an embedded distributed measurement strategy. Monitoring nodes with edge computing capabilities are deployed at the cable mid-joints to process the partial discharge signal in real time and extract feature values. Only the compressed feature data is uploaded to the central evaluation terminal via a fiber optic network. This distributed processing mechanism solves the problem of decreased partial discharge sensitivity due to long-distance signal attenuation, while also reducing the bandwidth pressure on the backbone communication network.

[0096] In the above method, spatial positioning indicators are added to the output quantification results. Combined with the spatial resolution capability of distributed fiber optic sensing, the system can accurately provide the kilometer marker location of anomalies while outputting the insulation condition level, with a positioning accuracy of ±1 meter. This has crucial engineering value for the precise location and repair of submarine cables.

[0097] The comprehensive evaluation model in this embodiment adds a multi-phase state correlation analysis function. For three-core submarine cables, the system simultaneously compares the elastic modulus distribution of the three-phase cores. If performance degradation occurs in all three phases at the same location, it is usually determined to be aging caused by environmental factors; if only a single phase shows an abnormality, it is determined to be due to manufacturing defects or localized external force damage within the phase insulation material. This lateral comparison logic improves the accuracy of fault diagnosis.

[0098] When processing massive amounts of data acquired from distributed sensing, the sliding window mechanism is optimized into a multi-scale parallel processing mode. Utilizing a GPU parallel computing architecture, cross-correlation calculations are performed on data from thousands of sensing channels simultaneously. Each computing core is responsible for a specific length of cable segment. Through multi-threaded concurrent execution, the waveform extraction process, which originally took several minutes, is reduced to the second level, achieving near real-time scanning of the insulation status of the entire cable.

[0099] Example 3: This example addresses the frequent occurrence of water treeing aging in cross-linked polyethylene cables in power distribution networks. The method of the present invention has been refined, with a particular emphasis on enhancing the ability to identify microscopic defects.

[0100] In the above method, the applied voltage pulse excitation is set as a variable frequency sequence pulse. By changing the repetition frequency of the pulses (from 10 Hz to 1000 Hz), the mechanical resonant frequency of the damp region in the insulating medium is sought. At the resonant frequency, the amplitude of the mechanical vibration induced by the damp region is enhanced, enabling the early water tree aging zone to be distinguished from normal dry insulation with higher contrast.

[0101] In the above method, a spectral feature analysis step is added when extracting the propagation velocity. This is because water tree regions not only alter wave velocity but also cause strong scattering and frequency-dependent attenuation of sound wave energy. The system extracts feature parameters reflecting the microscopic size of the defect by calculating the wave velocity differences of different frequency components. These feature parameters are added as additional input vectors during data fusion, enhancing the model's sensitivity to microscopic mechanical embrittlement.

[0102] In the aforementioned method, the inverse problem reconstruction algorithm employs a high-precision strategy based on full waveform inversion (FWI). Unlike traditional methods that only utilize travel time information, FWI simulates the complete propagation trajectory of sound waves in a non-uniform medium and compares it with the measured time-domain waveform. By iteratively correcting the elastic modulus, damping coefficient, and density distribution of each grid point, an ultra-high resolution distribution map containing microstructural information is generated, achieving a spatial resolution at the millimeter level, sufficient to clearly display the outline of water tree branches.

[0103] In the methods described above, the extraction of partial discharge signals pays particular attention to the polarity characteristics of weak discharge pulses. Water trees often exhibit weak partial discharges in the later stages of aging, and these electrical signals are masked by background electromagnetic noise. This embodiment utilizes wavelet packet decomposition technology to decompose the signal into different frequency bands and employs statistical methods to extract the skewness and steepness of the discharge pulses, used to distinguish between corona discharge in cable accessories and surface discharge within the insulation.

[0104] In the above method, the output quantification result also includes an irreversible thermal aging risk index. The irreversible thermal aging risk index is obtained by matching the currently measured elastic modulus distribution with a preset "accelerated thermal aging test database". The system automatically retrieves the experimental sample with the closest characteristics to the current measurement, obtains the elongation at break and molecular weight distribution parameters of the experimental sample under experimental conditions, and infers the remaining mechanical life of the cable insulation from a chemical perspective.

[0105] To improve the efficiency of on-site inspection, the visualization interface module in this embodiment supports augmented reality (AR) display. Inspectors wear smart glasses to directly observe the cables in operation on-site. The system uses spatial coordinate mapping technology to overlay the calculated elastic modulus heat map onto the visual image of the actual cable in real time. This "transparent" display method allows inspectors to intuitively see the internal physical state of the cable, similar to using medical ultrasound, facilitating the on-site marking and verification of potential hazards.

[0106] This implementation also incorporates cable operating load current data acquired by current transformers. By analyzing the correlation between long-term load levels and the decrease in elastic modulus, the system can identify areas of "heat accumulation effect" caused by long-term overload. If the elastic modulus of a region continues to decrease, and the historical operating load corresponding to that region has been consistently high, the system will assign the region the highest aging risk level and provide recommendations for reducing load operation.

[0107] Example 4: This example describes an automatic inspection application scheme for densely laid cables inside large substations, which combines robotics technology to realize the large-scale application of this method.

[0108] In the above method, an acoustic sensor array is integrated into the end effector of an automated inspection robot. The robot travels along the cable tray track, using machine vision to locate the centerline of each cable. The robotic arm moves a semi-circular sensor array to engage with the cable surface, automatically releasing and moving to the next inspection point after data acquisition.

[0109] In the above method, voltage pulse excitation is applied remotely through a pre-installed high-voltage injection port within the station. The excitation signal transmitter maintains clock synchronization with the inspection robot via the station control layer communication network, employing a high-precision Precision Time Protocol (PTP) to ensure that the synchronization error between the robot's acquisition system and the excitation end is less than 1 microsecond. This distributed collaborative working mode enables batch and automated performance testing of hundreds of cables throughout the station.

[0110] Due to the complex electromagnetic environment within the substation, an ultra-high frequency (UHF) antenna array was used to collect partial discharge signals. The antenna array was fixed to the back of the inspection robot and received electromagnetic wave signals radiated from the cable fault point via spatial wave reception. Using time difference of arrival (TDOA) multi-antenna positioning technology, the partial discharge source was located in three-dimensional space, and its coordinates were matched and fused with the coordinates of the mechanical defects located by the acoustic sensors in three-dimensional space.

[0111] In the above method, the output quantification results are synchronized to the substation intelligent operation and maintenance platform in real time. The platform uses a built-in digital twin model to reproduce the physical state of the cable in virtual space. When the comprehensive evaluation model outputs abnormal results, the digital twin model will automatically perform multi-condition simulations to predict the transient withstand capability of the cable under aging conditions when encountering lightning strikes or short-circuit faults, providing the dispatching department with a worst-case safety margin warning.

[0112] Furthermore, this embodiment optimizes the signal extraction strategy for scenarios involving multi-layer cable stacking. Since adjacent cables can cause acoustic wave coupling and interference, this embodiment utilizes a blind source separation algorithm to extract the characteristic waveform of the target cable from the complex aliased sound field signal. This blind source separation algorithm maximizes the independence of the signal components, filtering out vibration interference from adjacent cables as background noise, thus ensuring detection accuracy in densely laid environments.

[0113] The study incorporates historical overvoltage records for cables. By interfacing with fault recorders within the substation, the frequency of lightning or switching overvoltages experienced by the cable over the past year is obtained. These transient electrical stress accumulation effects are used as correction factors to increase the weight of partial discharge characteristics in the evaluation model. This method, based on the fusion of data throughout the entire operational lifecycle, enhances the scientific rigor of determining the irreversible thermal aging level of the insulation medium.

[0114] The system in this embodiment also features a self-calibration function. Before each inspection begins, the robotic arm first performs a test on a standard cable segment with known mechanical characteristics. By comparing the measured wave velocity with the standard value, it automatically compensates for systematic errors such as sensor coupling pressure and ambient temperature drift. This online calibration mechanism ensures that the detection results maintain a high degree of consistency and accuracy even after long-term operation.

[0115] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0116] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for testing the insulation performance of conductors in electric wires and cables, characterized in that, Includes the following steps: A voltage pulse with a preset amplitude and pulse width is applied to the conductor of the wire or cable under test, causing mechanical vibration waves to be generated inside the insulating medium due to the electrostriction effect. The time-domain signal of the mechanical vibration wave propagating in the insulating medium is synchronously acquired using an acoustic sensor array arranged around the outer surface of the cable; Based on the collected time-domain signals, the propagation velocity information of shear waves at different locations in the insulating medium is extracted; An inverse problem reconstruction algorithm is used to calculate the elastic modulus values ​​of each region inside the insulating medium based on the propagation speed information, and generate a spatial distribution map, including: establishing a mapping relationship between shear wave velocity and material mechanical parameters; The mapping relationship satisfies that the square of the propagation velocity of a shear wave in an isotropic medium is equal to the quotient of the shear modulus and the density. The inverse problem reconstruction algorithm adopts an iterative optimization strategy, using the collected multipath wave velocity observations as constraints. By minimizing the sum of squared residuals between the measured wave velocity and the wave velocity predicted by the theoretical model, the elastic modulus value of each voxel in the three-dimensional spatial grid inside the insulating medium is solved. The spatial distribution map is stored in the form of a three-dimensional matrix, with each coordinate point corresponding to a specific elastic modulus value; During the execution of the inverse problem reconstruction algorithm, an iterative optimization solver based on the regularized Gauss-Newton algorithm is used; in each iteration, the theoretical propagation time on each observation path is calculated using the forward model of the wave equation based on the current initial value of the elastic modulus. Calculate the residual vector between the measured time and the theoretical propagation time, and construct the Jacobian matrix that reflects the sensitivity of wave velocity to changes in local elastic modulus; A Tikhonov regularization term is added to the objective function to constrain the spatial smoothness of the elastic modulus distribution; the elastic modulus value is updated by solving a system of linear equations until the L2 norm of the residual decreases below the preset convergence threshold. The partial discharge signal generated during the application of the voltage pulse excitation is acquired synchronously, and its amplitude, phase and repetition frequency characteristics are extracted. By spatiotemporally aligning and fusing the spatial distribution map with the partial discharge signal characteristics, a comprehensive insulation state assessment model is constructed, including: The spatial coordinates of the acoustic signal location are correlated with the time of partial discharge signal generation based on the sound speed propagation delay; The data fusion process adopts a multi-source information weighted association strategy, which is achieved by establishing a multi-dimensional feature vector that includes the elastic modulus deviation, partial discharge amplitude density, and discharge phase clustering index. The weighted correlation strategy includes determining the overlap between the region of decreased elastic modulus and the region of active partial discharge in spatial topology. If there is overlap, the location is determined to be a high-risk aging point, and a higher confidence weight is assigned according to the correlation strength. In the data fusion process, the weight matrix is ​​determined based on the principle of information entropy. By analyzing the mutual contribution of elastic modulus characteristics and partial discharge characteristics in identifying known defect samples, the weighting coefficients of the two are dynamically adjusted. When the service life of the cable under test exceeds the preset aging threshold, the weight of the elastic modulus characteristic is increased. When the cable under test is in the new installation stage, the focus is on the partial discharge characteristics to capture construction process defects; The comprehensive evaluation model calculates a comprehensive score representing insulation health by introducing nonlinear fusion weights. Based on the comprehensive evaluation model, the micromechanical embrittlement degree and irreversible thermal aging level of the insulating medium are output; The acoustic sensor array presents a two-dimensional sensing network structure in space. It is arranged at preset intervals along the cable axis, and multiple sensing units are evenly distributed in the circumferential direction of each axial section to capture the propagation characteristics of shear waves in the radial, axial and circumferential dimensions.

2. The method for testing the insulation performance of conductors in wires and cables according to claim 1, characterized in that, Applying a voltage pulse excitation with a preset amplitude and pulse width to the conductor of the wire or cable under test includes: the voltage pulse excitation adopts a unipolar or bipolar square wave pulse; The amplitude of the voltage pulse excitation is adaptively adjusted according to the rated voltage level of the wire and cable under test, and the width of the voltage pulse excitation is in the range of 50 nanoseconds to 500 nanoseconds. The electric field change rate generated by the rising edge time of the voltage pulse excitation causes the molecules inside the insulating medium to undergo instantaneous polarization orientation and displacement, thereby inducing high-frequency mechanical strain.

3. The method for testing the insulation performance of conductors in wires and cables according to claim 2, characterized in that, The time-domain signal of the mechanical vibration wave propagating in the insulating medium is synchronously acquired using an acoustic sensor array arranged around the outer surface of the cable, including: the acoustic sensor array includes a piezoelectric ceramic sensor or a fiber Bragg grating sensor. The acoustic sensor array is integrated within a flexible ring clamp and attached to the surface of the cable's metal shielding layer or outer sheath.

4. The method for testing the insulation performance of conductors in wires and cables according to claim 3, characterized in that, Based on the acquired time-domain signals, the propagation velocity information of shear waves at different positions in the insulating medium is extracted, including: bandpass filtering preprocessing of the time-domain signals received by each channel to filter out low-frequency electromagnetic interference and high-frequency system noise; The process of extracting the shear wave propagation velocity is achieved by performing cross-correlation analysis on the signal sequences received by two adjacent sensors along the wave propagation direction. The cross-correlation analysis includes using a sliding window mechanism to extract the effective data segment containing the first wavefront oscillation, and calculating the time offset corresponding to when the cross-correlation function reaches its maximum value. The time offset is determined as the time delay of the shear wave propagating within the distance between the two sensors. The wave velocity in a local area is calculated using the quotient of the physical distance between the sensors and the time delay. In the process of the cross-correlation analysis, the signal sequence is first subjected to Hilbert transform to extract the instantaneous envelope of the signal, thereby reducing the impact of phase fluctuations on time delay estimation. Calculate the integral value of the product of two envelope sequences on the time axis, and find the time offset that maximizes the integral value by continuously shifting the signal sequence; Parabolic interpolation is used to fit discrete data near the cross-correlation peak point to obtain sub-sampling period time delay accuracy; The sliding window continuously scans the acquired sequence signals to extract the wave velocity evolution patterns along the axial and circumferential directions within the insulating medium.

5. The method for testing the insulation performance of conductors in wires and cables according to claim 4, characterized in that, Synchronous acquisition of partial discharge signals generated during the application of the voltage pulse excitation includes: acquiring current signals using a high-frequency current transformer or capacitive coupling probe installed on the grounding lead of the cable terminal; The sampling frequency of the partial discharge signal is not less than 100 MHz; The extracted features include the apparent amplitude of the discharge charge, the phase distribution relative to the excitation pulse period, and the pulse repetition frequency within the preset observation time. The partial discharge signal reflects the distribution of electrical weaknesses in the insulating medium under the action of an electric field, and complements the mechanical weaknesses reflected by the elastic modulus value.

6. The method for testing the insulation performance of conductors in wires and cables according to claim 5, characterized in that, Based on the comprehensive evaluation model, the micromechanical embrittlement degree and irreversible thermal aging level of the insulating medium are output, including: The degree of micromechanical embrittlement is characterized by calculating the relative ratio of the average local elastic modulus obtained from the current measurement to the reference elastic modulus value of the same type of cable in the new cable state. The irreversible thermal aging levels are divided into normal operation level, early aging level, mid-term degradation level and critical level based on multiple preset discrete threshold ranges. When the relative ratio drops to the first threshold, the system issues a warning of mechanical performance degradation. When the relative ratio is lower than the second threshold and the partial discharge frequency exceeds the background noise threshold, it is determined to be a severe emergency level.

7. The method for testing the insulation performance of conductors in wires and cables according to claim 6, characterized in that, The method also includes the step of establishing a historical detection database, which is used to store spatial distribution maps, partial discharge characteristic data and environmental parameters obtained from previous tests; By using pattern recognition algorithms to perform time-series analysis on the detection data of the same cable at different time points, the evolution pattern of aging trend can be identified. Calculate the aging rate, which is the rate of change of elastic modulus with operating time; The remaining effective life of the cable is estimated based on the aging rate.

8. The method for testing the insulation performance of conductors in wires and cables according to claim 7, characterized in that, The method also includes a step of real-time monitoring of the cable's ambient temperature and humidity before testing; Real-time temperature field data is obtained using temperature measuring devices placed on the surface of the cable; Environmental parameters are introduced as correction factors in the calculation of the elastic modulus value; The calculated elastic modulus is normalized using a pre-calibrated temperature compensation curve to eliminate the interference of external environmental fluctuations on the sound wave propagation speed.

Images