A cold shrink cable terminal installation defect detection method based on ultrasonic guided waves

Abstract

The present application relates to the technical field of cold shrink cable defect detection, and specifically discloses a cold shrink cable terminal installation defect detection method based on ultrasonic guided waves, five ultrasonic receiving array elements are arranged at equal intervals along the axial direction on the outer wall of the cold shrink cable terminal, and the middle array element is set as the center reference. The guided wave is excited at the basic test frequency, and the echo arrival time and peak amplitude of each array element are collected; the maximum amplitude array element and the first wave arrival array element are screened, the amplitude asymmetry, time delay and position deviation are calculated, and the eccentricity characteristic evaluation value is obtained by weighted summation. The value is preliminarily determined, if it is in the preset threshold interval, it is switched to high-frequency test frequency re-measurement and the high-frequency evaluation value is obtained; the ratio of the absolute value of the difference between the two values to the original value is calculated as the characteristic change rate, which is compared with the preset sensitivity threshold, and the eccentricity defect detection result is output. The defect recognition sensitivity and noise robustness are significantly improved, and a quantitative basis is provided for the cable terminal insulation state evaluation.

Description

Technical Field

[0001] This invention belongs to the field of cold shrink cable defect detection technology, and relates to a method for detecting defects in the terminal installation of cold shrink cables based on ultrasonic guided waves. Background Technology

[0002] As the core hub of the power distribution network, the assembly quality of the internal stress cones of 1kV to 110kV medium and high voltage cable terminals directly determines the uniformity of the local electric field distribution and the long-term operational safety of the insulation system. Meanwhile, ultrasonic guided wave technology, as an advanced non-destructive testing method, plays an increasingly crucial role in diagnosing hidden defects in complex pipelines and layered electrical equipment due to its long propagation distance and high sensitivity to the continuity of the medium.

[0003] It is worth noting that, compared to prefabricated components that typically employ a coaxial integrated vulcanization process for voltage levels above 110kV, cable terminals for voltage levels from 1kV to 110kV mostly rely on on-site assembly, inevitably resulting in minor installation deformations or structural eccentricities. The unique conical geometry of the stress cone, when faced with ultrasonic guided wave penetration, can alter the focal point and energy propagation path of the sound waves even with an extremely small axial offset of less than 0.5 mm inside the terminal.

[0004] However, traditional detection methods rely excessively on single acoustic amplitude or time thresholds, neglecting the multidimensional acoustic field distortion effects caused by stress cone structures. Field practice shows that even with intact internal insulation, inherent focusing shifts caused by simple installation deformation often lead to localized energy anomalies at the receiver. This reflects a significant flaw in assessments that focus solely on single-point acoustic values ​​while ignoring waveform spatial asymmetry. This makes existing condition warning strategies highly susceptible to misjudgment, unable to accurately distinguish the physical boundary between geometric distortion and actual eccentricity defects in complex field environments, resulting in blind and delayed overall operation and maintenance decisions. Summary of the Invention

[0005] In view of the problems existing in the prior art, the present invention provides a method for detecting installation defects in cold shrink cable terminals based on ultrasonic guided waves, in order to solve the above-mentioned technical problems.

[0006] To achieve the above and other objectives, the technical solution adopted by the present invention is as follows:

[0007] This invention provides a method for detecting installation defects in cold-shrink cable terminations based on ultrasonic guided waves, the method comprising:

[0008] Five ultrasonic receiving array elements are set at equal intervals along the axial direction on the outer wall of the cold shrink cable terminal, and the ultrasonic receiving array element in the middle position is set as the center array element.

[0009] By exciting ultrasonic guided waves at a set basic test frequency, the arrival time and peak amplitude of the echoes received by each ultrasonic receiving array element are obtained.

[0010] Based on the peak amplitude of the echo, the array element number corresponding to the largest amplitude is selected, and based on the echo arrival time, the array element number corresponding to the first wave arrival is selected.

[0011] Extract the echo peak amplitudes of the adjacent array elements on both sides of the largest amplitude array element, and calculate the ratio of the difference and sum of the echo peak amplitudes of the adjacent array elements on both sides to obtain the amplitude asymmetry.

[0012] The time delay is obtained by calculating the difference between the echo arrival time of the array element with the largest amplitude and the echo arrival time of the first wave to the corresponding array element.

[0013] Calculate the positional deviation between the largest amplitude array element and the central array element, and then sum the positional deviation, amplitude asymmetry, and time delay by weight to obtain the eccentricity feature evaluation value.

[0014] The eccentricity feature evaluation value is initially determined. When the eccentricity feature evaluation value is between the first preset threshold and the second preset threshold, the frequency of the ultrasonic guided wave is switched to the high-frequency test frequency, and the same calculation process is used to re-acquire the high-frequency eccentricity feature evaluation value.

[0015] Calculate the absolute value of the difference between the high-frequency eccentricity feature evaluation value and the eccentricity feature evaluation value, and set the ratio of the absolute value of the difference to the eccentricity feature evaluation value as the feature change rate;

[0016] The characteristic change rate is compared with a preset sensitivity threshold, and the corresponding cable terminal eccentricity defect detection result is output based on the comparison result.

[0017] As described above, the present invention provides a method for detecting installation defects in cold-shrink cable terminals based on ultrasonic guided waves, which has at least the following beneficial effects:

[0018] This invention constructs a multi-channel surround guided wave signal receiving architecture by arranging five ultrasonic receiving array elements at equal intervals along the axial direction of the outer wall of the cold-shrink cable terminal and using the central array element as a spatial reference. Based on this, ultrasonic guided waves are excited at a basic test frequency, and the echo arrival time and peak amplitude of each array element are acquired simultaneously. The array element with the largest amplitude and the array element corresponding to the first wave arrival are then selected. The echo peak amplitude of the array elements on both sides of the array element with the largest amplitude is extracted, and the amplitude asymmetry is constructed by calculating the ratio of the difference and sum of the amplitudes on both sides. At the same time, the difference between the echo arrival time of the array element with the largest amplitude and the echo arrival time of the array element with the first wave arrival is calculated as the time delay. Combined with the positional deviation between the array element with the largest amplitude and the central array element, the three are weighted and summed to obtain a comprehensive eccentricity characteristic evaluation value. The above three physical features characterize the eccentricity state from three independent dimensions: spatial energy distribution asymmetry, wavefront arrival time difference, and geometric position offset. The three features have good complementarity and redundancy. Weighted fusion can effectively suppress the risk of misjudgment such as single features being susceptible to coupled noise interference and insignificant response to small-angle eccentricity, and significantly improve the sensitivity and anti-interference ability of eccentricity defect detection.

[0019] This invention further incorporates an adaptive dual-frequency switching verification mechanism in the initial inspection and judgment stage. When the eccentricity feature evaluation value falls within the ambiguity range between a first preset threshold and a second preset threshold, a hard decision is not made directly. Instead, the excitation frequency is switched to a higher frequency, and the high-frequency eccentricity feature evaluation value is re-acquired following the same process. The ratio of the absolute value of the difference between the high-frequency evaluation value and the basic evaluation value to the basic evaluation value is then calculated as the feature change rate. Finally, this ratio is compared with a preset sensitivity threshold to output the final detection result. The core of this mechanism lies in the significant differences in the response sensitivity of guided waves at different frequencies to eccentric defects. If the feature change rate is significant within the ambiguity range, it indicates that the eccentric signal has frequency-sensitive characteristics and originates from actual structural eccentricity rather than coupling interference or system noise. Conversely, the judgment result is corrected to no defect. This invention replaces the single-threshold hard decision method with a dual-frequency differential criterion of initial screening at the base frequency and verification at a high frequency. While ensuring a rapid response to significant defects, it significantly reduces the probability of false alarms and false negatives in edge states. In particular, it has outstanding reliability in judging the degree of eccentricity in the critical range. At the same time, dual-frequency verification is triggered only for a small number of samples that fall into the fuzzy range during the initial screening, without the need to perform it on all samples. While ensuring detection accuracy, it effectively controls detection time and computational overhead, which is significantly better than traditional ultrasonic methods that rely on a single fixed threshold or single-frequency detection. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Fig. 1 This is a schematic diagram showing the connections between the steps of the method of the present invention.

[0022] Fig. 2 This is a schematic diagram of the initial judgment logic connection for the eccentricity feature evaluation value in this invention.

[0023] Fig. 3 This is a schematic diagram showing the variation of the high-frequency eccentricity feature evaluation value with the position of the array element provided by the present invention. Detailed Implementation

[0024] The present invention will be further described in detail below with reference to specific embodiments. The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the inventive concept or exceed the scope defined by the claims, all of which should fall within the protection scope of the present invention.

[0025] Please see Figs. 1-3 As shown, a method for detecting installation defects in cold-shrink cable terminals based on ultrasonic guided waves includes:

[0026] Five ultrasonic receiving array elements are set at equal intervals along the axial direction on the outer wall of the cold shrink cable terminal, and the ultrasonic receiving array element in the middle position is set as the center array element.

[0027] By exciting ultrasonic guided waves at a set basic test frequency, the arrival time and peak amplitude of the echoes received by each ultrasonic receiving array element are obtained.

[0028] Preferably, the acquisition of the echo arrival time and echo peak amplitude received by each ultrasonic receiving array element includes:

[0029] During the period when the ultrasonic guided wave is excited at the basic test frequency, the ultrasonic echo signals captured by each ultrasonic receiving array element are collected.

[0030] The first-order difference calculation of the ultrasonic echo signal is used to obtain the signal rate of change, and the position of the first zero crossing of the signal rate of change is extracted as the oscillation start time.

[0031] A signal interception window is defined with the start time of vibration as the starting point, and the data of the corresponding interval of the ultrasonic echo signal is intercepted as the main wave signal.

[0032] Extract the absolute values ​​of all local maxima in the main wave signal;

[0033] The maximum value among the absolute values ​​of all local maxima is taken as the echo peak amplitude of the array element, and the time when the maximum value is extracted is recorded as the echo arrival time.

[0034] In practice, five ultrasonic receiving array elements are evenly spaced along the axial direction on the outer wall of the cold-shrink cable terminal, with the element in the middle designated as the central array element. In actual arrangement, the axial spacing between adjacent ultrasonic receiving array elements is typically set according to the cable connector size, for example, between 20mm and 40mm. This symmetrical layout with equal spacing allows the central array element to serve as the reference origin.

[0035] After setup, ultrasonic guided waves are excited at the set fundamental test frequency to obtain the echo arrival time and peak amplitude of each ultrasonic receiving element. The fundamental test frequency is set based on the acoustic attenuation characteristics of the silicone or EPDM rubber material used in cold-shrink cable terminations, and an adjustable range of 50kHz to 150kHz is recommended. To ensure the guided waves can penetrate multiple insulation layers and possess suitable defect resolution, 60kHz is typically preferred as the fundamental test frequency.

[0036] In acquiring the arrival time and peak amplitude of the echoes received by each ultrasonic receiving element, the system first synchronously acquires the ultrasonic echo signals captured by each ultrasonic receiving element during the excitation of the ultrasonic guided wave at the basic test frequency, converting them into digital quantities containing time points and voltage amplitudes. To overcome low-frequency baseline drift caused by external electromagnetic interference during field testing and accurately locate the initial evolution of the waveform, first-order differential calculation is needed to obtain the signal rate of change from the ultrasonic echo signal. Specifically, assuming the discretized ultrasonic echo signal is S(n), where n is the sampling point number and the sampling interval is... The calculation process for the first-order difference rate of change D(n) is as follows:

[0037] ;

[0038] In this formula, S(n+1) and S(n-1) represent the voltage amplitudes at adjacent positions before and after the current sampling point, respectively, in millivolts; the calculated D(n) represents the slope of the signal change per unit time, in millivolts. .

[0039] To avoid misjudgments caused by ambient noise, a noise suppression threshold is preset before extracting the position of the first zero-crossing of the signal rate of change as the oscillation start time. The setting is based on the following: in a silent state without ultrasonic excitation, the system continuously collects the system background noise within a preset period, extracts the root mean square of the maximum peak-to-peak value as the noise floor reference amplitude, and divides this noise floor reference amplitude by the system's full-scale amplitude. The result is then set as the noise suppression threshold.

[0040] When the absolute value of D(n) first exceeds the noise suppression threshold, and a sampling point immediately follows where the product of D(n) and D(n+1) has opposite signs (i.e., the product is less than zero), the physical time corresponding to this zero-crossing point is extracted as the oscillation start time. This moment represents the initial point at which the main wave packet of the guided wave energy arrives at the receiving array element, causing the piezoelectric crystal to undergo substantial oscillation.

[0041] After determining the oscillation initiation time, a signal interception window is defined starting from the oscillation initiation time, and data from the corresponding interval of the ultrasonic echo signal is extracted as the main wave signal. The time length of the interception window... Adaptive matching is performed based on the period of the fundamental test frequency, typically taking 3 to 5 complete oscillation cycles of the fundamental test frequency. For example, at a fundamental test frequency of 60 kHz, a single cycle is approximately 16.67 microseconds, and the intercept window length can be set to 80 microseconds. This time interception operation can separate the main beam data from the subsequent structural clutter and multimodal dispersion waveforms generated by reflections from the pipe wall ends.

[0042] After truncation, the main wave signal data within the truncation window is traversed, and the absolute values ​​of all local maxima in the main wave signal are extracted. In specific data processing, this involves finding values ​​that satisfy... The system samples all points m under the given conditions and calculates the absolute voltage values ​​at these points. Finally, it compares the absolute values ​​of all local maxima and selects the maximum value as the echo peak amplitude of the ultrasonic receiving element. This value directly represents the intensity of sound energy concentration at the receiving point. Simultaneously, this maximum value will be obtained. The instantaneous sampling time is recorded as the echo arrival time. .

[0043] Based on the peak amplitude of the echo, the array element number corresponding to the largest amplitude is selected, and based on the echo arrival time, the array element number corresponding to the first wave arrival is selected.

[0044] Extract the echo peak amplitudes of the adjacent array elements on both sides of the largest amplitude array element, and calculate the ratio of the difference between the echo peak amplitudes of the adjacent array elements on both sides to the sum of the values ​​to obtain the amplitude asymmetry.

[0045] Preferably, the echo peak amplitudes of the adjacent array elements on both sides of the largest amplitude array element are extracted, and the amplitude asymmetry is calculated, including:

[0046] Identify the position of the maximum amplitude array element among the five ultrasound receiving array elements;

[0047] When the largest amplitude matrix element is at the leftmost position, it is determined that the data element of its left adjacent matrix element is missing, and the amplitude asymmetry is directly assigned to the preset first extreme value constant.

[0048] When the largest amplitude matrix element is at the rightmost position, it is determined that the adjacent matrix element to its right is missing, and the amplitude asymmetry is directly assigned to the preset second extreme value constant.

[0049] When the element with the largest amplitude is neither at the leftmost nor the rightmost end, extract the amplitude of the element adjacent to its left and the amplitude of the element adjacent to its right.

[0050] Subtract the amplitude of the adjacent array element on the left from the amplitude of the adjacent array element on the right to get the difference value. Add the amplitude of the adjacent array element on the left to the amplitude of the adjacent array element on the right to get the sum value. Divide the difference value by the sum value to get the final amplitude asymmetry.

[0051] In one specific embodiment, after acquiring the echo arrival time and echo peak amplitude of each ultrasonic receiving array element, the system reads these five sets of data according to the physical arrangement order of the array elements, for example, labeling them sequentially as numbers 1 to 5 along the cable axis. By traversing and comparing the echo peak amplitude data of the five array elements, the system finds the largest value and records the physical location of the array element with the largest value as the array element number with the largest amplitude. Similarly, the echo arrival times of the five array elements are arranged in ascending order from smallest to largest, and the data with the smallest time value is extracted, indicating that the ultrasonic energy was received first at that location. The array element corresponding to that time is recorded as the array element number corresponding to the first wave arrival.

[0052] After determining the array element corresponding to the maximum amplitude, in order to evaluate the deflection of the acoustic field energy distribution inside the tube wall, it is necessary to extract the energy attenuation data on both sides of the central stress point. Specifically, this involves extracting the echo peak amplitude of the adjacent array elements on both sides of the maximum amplitude array element, and then calculating the amplitude asymmetry. Since the array elements are arranged in a limited axial direction on the outer wall of the cold-shrink cable terminal, it is necessary to identify the specific location of the maximum amplitude array element among the five ultrasonic receiving array elements before calculation to handle the missing adjacent data due to boundary constraints.

[0053] When the largest amplitude array element is identified as being at the leftmost position (element number 1), there are no receiving array elements to its left physically, and the system determines that its adjacent array data to the left is missing. In this case, instead of performing conventional difference division calculations, the amplitude asymmetry is directly assigned to a preset first extreme value constant. It is recommended that this first extreme value constant be set to -1, as this dimensionless value represents an extreme rightward bias in the acoustic energy distribution. Conversely, when the largest amplitude array element is at the rightmost position (element number 5), the system determines that its adjacent array data to the right is missing, and the amplitude asymmetry is directly assigned to a preset second extreme value constant. Correspondingly, the second extreme value constant is usually set to 1, representing an extreme leftward bias in the energy distribution. This direct assignment of extreme values ​​avoids program calculation errors caused by array out-of-bounds errors and also conforms to the characteristic of unilateral acoustic energy concentration caused by severe eccentricity in physical phenomena.

[0054] When the element with the largest amplitude is neither at the leftmost nor the rightmost end (i.e., it is numbered 2, 3, or 4), it indicates that complete acoustic monitoring data exists on both sides. In this case, the system extracts the amplitude of its left-side adjacent element in the direction of decreasing the position coordinate by one, and extracts the amplitude of its right-side adjacent element in the direction of increasing the position coordinate by one, based on the current number of the element with the largest amplitude. After obtaining these two adjacent amplitude data, the system calculates the final amplitude asymmetry using a normalized difference formula. The calculation formula is as follows:

[0055] ;

[0056] In this formula, It represents the magnitude asymmetry and takes values ​​within the closed interval [-1, 1]. This represents the peak echo amplitude of the element adjacent to the left, in millivolts, and is the absolute voltage value directly read from the previous processing step. This represents the peak echo amplitude of the adjacent array element on the right, also in millivolts.

[0057] Due to differences in the material of cold-shrink cable terminals from different batches, or uneven application of coupling agent on the sensor surface, there can be significant fluctuations in ultrasonic noise floor and overall signal attenuation. Simply calculating the difference is easily affected by interference from overall signal gain amplification. However, dividing the difference by the sum can effectively eliminate absolute amplitude interference caused by overall acoustic energy fluctuations, and purely extract the relative geometric features reflecting the eccentric compression of the internal structure of the pipe wall, making the asymmetry data under different detection environments and different base voltages horizontally comparable.

[0058] The time delay is obtained by calculating the difference between the echo arrival time of the array element with the largest amplitude and the echo arrival time of the first wave to the corresponding array element.

[0059] Preferably, the calculation of time delay includes:

[0060] Collect the surface temperature of the cable duct wall at the location of each ultrasonic receiving array element;

[0061] The average value of the surface temperatures distributed along the axial direction is calculated to obtain the equivalent ambient temperature along the ultrasonic guided wave propagation path.

[0062] Based on the temperature difference between the equivalent ambient temperature and the standard reference temperature, and combined with the sound velocity temperature drift coefficient of the pipe wall material itself, the corresponding temperature delay compensation is calculated.

[0063] The initial time difference is obtained by subtracting the echo arrival time of the first wave to the corresponding array element from the echo arrival time of the array element with the largest amplitude.

[0064] The initial time difference is smoothed by filtering to eliminate high-frequency jitter interference, and the filtered value is superimposed with the temperature delay compensation to obtain the calibrated time delay.

[0065] In a specific embodiment, after determining the array element with the largest amplitude and the array element corresponding to the first wave arrival, it is necessary to calculate the difference in echo arrival time between the two to quantify the time delay caused by structural eccentricity during sound wave propagation inside the pipe wall. Since cold-shrink cable terminals are generally made of insulating silicone or EPDM rubber, the acoustic properties of these polymers are highly sensitive to temperature changes. Fluctuations in ambient temperature directly cause changes in the sound velocity within the material, leading to baseline drift in the extracted echo arrival time. Therefore, before calculating the time delay, the system first synchronously collects the surface temperature of the cable pipe wall at the location of each array element using thermal sensors attached to the outside of each ultrasonic receiving array element. After obtaining five surface temperature values ​​distributed along the axial direction, these values ​​are arithmetically averaged to calculate the equivalent ambient temperature along the current ultrasonic guided wave propagation path.

[0066] Subsequently, the propagation time deviation caused by temperature changes is calculated based on the equivalent ambient temperature. Specifically, the temperature difference between the equivalent ambient temperature and the standard reference temperature is calculated, and the corresponding temperature delay compensation is determined by combining this with the sound velocity temperature drift coefficient of the pipe wall material itself. ;

[0067] The calculated temperature delay compensation is expressed in microseconds. This is the equivalent ambient temperature, expressed in degrees Celsius. This is a preset standard reference temperature, which is usually preset to a certain value. ; The equivalent time-temperature drift coefficient of the pipe wall material, in units of... .

[0068] It should be noted that, It is an empirical linear fitting constant for the change of sound wave propagation time per unit distance with temperature, calculated in advance by conducting acoustic transmission calibration experiments on the same batch of cold-shrink cable materials in a constant temperature chamber.

[0069] After the temperature compensation parameters are ready, the system extracts the echo arrival time of the element with the largest amplitude and subtracts the echo arrival time of the first wave to the corresponding element to obtain the initial time difference for a single measurement. Considering the analog-to-digital conversion error of the hardware sampling circuit and the electromagnetic radiation from surrounding high-voltage electrical equipment in actual field testing, high-frequency jitter interference will inevitably be introduced at the signal acquisition edge, causing random fluctuations in the initial time difference of a single measurement. To eliminate this high-frequency jitter, the system performs a moving average filtering process on the initial time difference of multiple continuously acquired pulse transmission cycles. Specifically, a smooth data window of length N is set. To balance the denoising effect with the real-time response speed of the system, N is usually an integer between 8 and 16. The initial time difference of the current measurement cycle and the previous N-1 cycles is accumulated and divided by N to obtain the filtered smooth time difference. .

[0070] Finally, the filtered smooth time difference is balanced with the calculated temperature delay compensation to obtain the final calibrated time delay. The formula is:

[0071] ;

[0072] In the formula, This is the time delay after calibration.

[0073] Calculate the positional deviation between the largest amplitude array element and the central array element, and then sum the positional deviation, amplitude asymmetry, and time delay by weight to obtain the eccentricity characteristic evaluation value.

[0074] Preferably, the eccentricity feature evaluation value is calculated, including:

[0075] Obtain the elastic modulus of the insulating silicone and the nominal wall thickness of the cold-shrink cable terminal;

[0076] The position weight coefficients are obtained by inversely scaling the preset reference weight coefficients using the elastic modulus of insulating silicone.

[0077] The delay weighting coefficient is obtained by scaling the baseline weighting coefficient proportionally using the nominal wall thickness.

[0078] Extract the preset material attenuation constant as the asymmetric weighting coefficient;

[0079] Calculate the absolute value of the distance difference between the largest amplitude array element and the central array element as a positional deviation with positive and negative directional information; for example, a positive value represents an offset to one end, and a negative value represents an offset to the other end, in order to preserve the different effects of the eccentric direction on the stress distribution of the cable terminal.

[0080] The position deviation is multiplied by the position weight coefficient to obtain the position weighted component. The absolute value of the amplitude asymmetry is extracted and multiplied by the asymmetry weight coefficient to obtain the amplitude weighted component. The time delay is multiplied by the delay weight coefficient to obtain the delay weighted component.

[0081] The position-weighted component, amplitude-weighted component, and delay-weighted component are summed to obtain the eccentricity feature evaluation value.

[0082] In a specific embodiment, to eliminate physical interference from cable terminals of different specifications and materials on feature evaluation, the system first reads the elastic modulus of the insulating silicone and the nominal wall thickness of the cold-shrink cable terminal through the configuration interface of the testing equipment. The elastic modulus of the insulating silicone characterizes the wall material's ability to resist elastic deformation, measured in megapascals (MPa). This data is typically provided by the cable manufacturer in its factory inspection report, with a typical value between 2.0 MPa and 4.5 MPa. The nominal wall thickness refers to the standard radial thickness of the insulation layer of the cold-shrink terminal on the design drawings, measured in millimeters.

[0083] After obtaining the aforementioned basic material parameters, they need to be converted into weighting coefficients for the final summation calculation. First, the position weighting coefficients are obtained by inversely scaling the preset baseline weighting coefficients using the elastic modulus of insulating silicone. The physical logic here is that materials with higher elastic modulus are harder, and under the same eccentric stress, they exhibit smaller physical deformation and acoustic energy shift. Therefore, for high-hardness materials, the sensitivity weight of their positional deviation needs to be amplified to allow for comparison with soft materials on the same evaluation scale. In the specific calculation, a baseline spatial constant related to spatial deformation is set. Its dimension is preset to MPa / mm, and this constant is a priori empirical value obtained through finite element analysis of standard coaxial cable joints. Position weighting coefficient. The calculation formula is: Where E is the elastic modulus of insulating silicone, the position weighting coefficient is obtained through this division variation. The dimensions are transformed .

[0084] It should be added that a reference space constant is introduced here. The term "stiffness attenuation" characterizes the resistance of a specific type of cable's insulating silicone material to radial spatial compressive deformation. In real-world cables, it corresponds to the elastic modulus-spatial displacement gradient parameter at the interface within the stress cone. MPa / mm is used as the unit because it quantifies the elastic mechanical relationship: a 1 mm shift in the ultrasonic propagation path reflects the magnitude of abnormal compressive stress generated within the material.

[0085] Reference space constant The method for determining the value is as follows: Take no less than 30 samples of cold-shrink cable terminals from the same batch that have passed the factory partial discharge and withstand voltage tests. Under standard ambient temperature, excite ultrasonic guided waves at the same basic test frequency as the subsequent field test and collect echo data of each array element. Calculate the basic position deviation of each qualified sample under the condition of no substantial eccentric defects. and the corresponding basic eccentricity characteristic evaluation value . The initial value formula is: The mean function represents the arithmetic mean. Based on this, a known radial offset is applied to the cable terminal of the same model through finite element simulation. The correspondence between the spatial offset of the ultrasonic guided wave propagation path and the eccentricity characteristic evaluation value is recorded in the simulation. The calibration results are then corrected to finally determine the final value. The specific values. For new cable specifications for which a sample library has not yet been established, equivalent acoustic tests can be conducted using standard test blocks made of the same insulating silicone material. The sound field change data under known offsets in the standard test blocks can be used to replace the sample statistical data for calculation. After calibration using the above method, for 1kV-110kV cold-shrink cable terminals, The empirical value range is 0.035MPa / mm to 0.085MPa / mm.

[0086] Next, the reference weighting coefficient is proportionally scaled using the nominal wall thickness to obtain the delay weighting coefficient. Since the propagation path of ultrasonic guided waves within the tube wall increases with wall thickness, a thicker wall naturally amplifies the time delay caused by internal structural asymmetry. To ensure the final evaluation index has cross-specification universality, proportional scaling logic is used to calibrate the weight of the time delay. A reference time constant is set. Its dimensions are Multiply the nominal wall thickness d by the reference time constant, i.e. The delay weighting coefficient is calculated. Its dimensions are correspondingly transformed into .

[0087] It should be added that, This is based on prior acoustic statistics of a large number of standard qualified cable terminal samples. Specifically, the system was pre-calibrated in a laboratory environment using ultrasonic transmission on qualified cable terminals of the same material that had passed factory partial discharge and withstand voltage tests. Due to the unavoidable slight micro-crosslinking unevenness within the silicone material, even qualified products will exhibit a very small fundamental time shift when sound waves propagate along different paths. Technicians extracted the maximum allowable normal time delay fluctuation per unit wall thickness for these qualified samples; the unit of the maximum normal time delay fluctuation is... The reciprocal of this value is defined as the reference time constant. .therefore, The dimensions are .

[0088] It is particularly important to note that, on the one hand, due to the differences in the elastic modulus of silicone rubber with different formulations, the larger the elastic modulus of a material, the more difficult it is to achieve the same spatial displacement due to stress deformation. Therefore, the system uses a built-in mechanism to assign a smaller position weight coefficient to materials with larger elastic moduli, avoiding misjudging normal minor installation tolerances of high-hardness materials as defects. On the other hand, the propagation background time delay of ultrasonic guided waves within the tube wall increases with wall thickness, and thicker tube walls naturally amplify the time delay caused by internal structural asymmetry. This invention directly binds the wall thickness to the time reference constant, so that the larger the wall thickness, the greater the time delay weight. This ability to dynamically and adaptively adjust the weight based on the inherent physical parameters of the cable effectively solves the defect of existing technologies that rely on fixed empirical thresholds, leading to failure upon model change.

[0089] Meanwhile, the system directly extracts the preset material attenuation constant as the asymmetric weighting coefficient. This is a dimensionless constant. The specific value of this preset value is assigned by looking up a table based on the inherent acoustic absorption rate of the silicone material at the ultrasonic working frequency. This reference data table was obtained through pulse echo attenuation calibration experiments on liquid silicone rubber with a specific formulation used in cable terminals. During the experiment, multiple sets of ultrasonic pulses with different center frequencies were emitted to homogeneous silicone test blocks with different standard step thicknesses using a broadband ultrasonic transducer, and the energy loss rate before and after the sound wave penetrated the test block was recorded. The system calculated the inherent acoustic absorption rate at different working frequencies, with the unit being dB / m, and proportionalized its attenuation relative to the fundamental reference frequency. The specific proportionalization formula is as follows: .in, These are asymmetric weighting coefficients. The intrinsic acoustic absorption rate of silicone is measured at the current ultrasonic operating frequency. Acoustic absorption rate at the base reference frequency The basic weighting constants for experimental calibration can be found by looking up a table. This ultimately solidifies into an asymmetric weighting coefficient mapping table, where the basic reference frequency is typically the lowest frequency band supported by the detection equipment.

[0090] The specific data mapping format and typical values ​​are shown in the table below:

[0091] Table 1: Mapping table of asymmetric weighting coefficients;

[0092] After the weighting parameters are established, the system calculates the absolute value of the physical distance difference between the largest array element and the central array element, which is used as the positional deviation. If the fixed axial spacing between adjacent array elements is known... The largest element in the matrix is ​​numbered as follows: The central array element is numbered as In a five equidistant array element, the central element is always numbered 3. The calculation process for the positional deviation is as follows: .

[0093] Finally, the system multiplies and linearly accumulates the acquired features and their corresponding weight coefficients, combining them to obtain the eccentricity feature evaluation value. Its calculation is expressed as:

[0094] ;

[0095] By adopting a weighted cumulative linear model, the compensatory effect of other dimensional features can be preserved, resulting in a stronger fault tolerance and data robustness in the final output of the eccentric feature evaluation value.

[0096] An initial assessment of the eccentricity feature evaluation value is performed. When the eccentricity feature evaluation value is between the first preset threshold and the second preset threshold, the frequency of the ultrasonic guided wave is switched to the high-frequency test frequency, and the same calculation process is used to re-acquire the value to obtain the high-frequency eccentricity feature evaluation value.

[0097] Preferably, the preliminary determination of the eccentricity feature evaluation value includes:

[0098] Compare the eccentricity feature evaluation value with the first preset threshold and the second preset threshold. The second preset threshold is greater than the first preset threshold.

[0099] When the eccentricity feature evaluation value is less than the first preset threshold, it is determined that the waveform propagation delay is within the normal tolerance range, the coaxiality of the cable terminal is deemed qualified, and the test round ends.

[0100] When the eccentricity feature evaluation value is greater than or equal to the second preset threshold, it is determined that the waveform has abnormal offset or distortion, a serious eccentricity defect is identified, and the result is output.

[0101] When the eccentricity feature evaluation value is greater than or equal to the first preset threshold and less than the second preset threshold, it is determined that there is a slight eccentricity defect, triggering the subsequent process of switching the test frequency to a high-frequency test frequency.

[0102] Preferably, obtaining high-frequency eccentricity feature evaluation values ​​includes:

[0103] After switching the test frequency to a high-frequency test frequency, the high-frequency ultrasonic echo signals generated by each ultrasonic receiving array element are acquired; wherein, the basic test frequency is set in the low-frequency range of 20kHz to 60kHz, and the high-frequency test frequency is set in the range of 100kHz to 250kHz, to ensure that the difference in ultrasonic length between the two reaches at least 2 times.

[0104] Construct a decaying envelope reference line with the same oscillation period as the high-frequency test frequency;

[0105] The high-frequency ultrasonic echo signals of each ultrasonic receiving array element are sequentially cross-correlated with the attenuation envelope reference line to generate a cross-correlation sequence.

[0106] Extract the time corresponding to the maximum peak coordinate in the cross-correlation sequence, take it as the arrival time of the high-frequency echo, and extract the original waveform amplitude corresponding to that time as the peak amplitude of the high-frequency echo;

[0107] Based on the arrival time and peak amplitude of the high-frequency echo, the same amplitude asymmetry calculation process and weighted summation process are used to calculate the high-frequency eccentricity characteristic evaluation value at the high-frequency test frequency.

[0108] Preferably, after outputting the severe eccentricity defect result, the three-dimensional boundary of the severe eccentricity defect is calculated and defined, including:

[0109] The echo peak amplitudes of the five ultrasonic receiving array elements are summarized, and ultrasonic receiving array elements whose echo peak amplitudes are lower than half of the preset standard reference amplitude are selected and grouped into the defective array element group.

[0110] The actual physical distance between the two ultrasonic receiving array elements at the farthest ends in the array element group corresponding to the defect is measured, and half of this actual physical distance is set as the tangential half-axis length of the defect.

[0111] The ratio of the extracted time delay to the echo arrival time of the first wave to the corresponding array element is used as a coefficient to characterize the degree of eccentricity of the compression unevenness.

[0112] By combining the eccentricity coefficient and the length of the defect tangential semi-axis, a three-dimensional orthogonal spatial expansion is carried out along the pipe wall normal direction to construct an eccentric influence envelope ellipsoid containing the insulation defect region. All three-dimensional coordinate points on the surface of this eccentric influence envelope ellipsoid are extracted as the three-dimensional defect boundary for the final assessment of the defect severity.

[0113] In one specific embodiment, after obtaining the eccentricity feature evaluation value, the system compares it with a preset threshold to make a preliminary classification judgment on the internal coaxiality status of the cable terminal. This process specifically includes: calculating the eccentricity feature evaluation value... With the first preset threshold Second preset threshold Perform a numerical comparison. Here... and This empirical dividing point is determined by statistically analyzing the distribution range of evaluation values ​​after testing a large number of qualified cable terminals and samples known to have varying degrees of eccentricity defects. Furthermore, it meets the following criteria: .

[0114] Specifically, the values ​​of the first and second preset thresholds follow the normal process tolerance distribution limits in statistics. The system pre-acquires the eccentricity characteristic evaluation values ​​of a large number of qualified samples and calculates their statistical mean μ and standard deviation σ. Based on the 3σ criterion, the first preset threshold is set to μ+3σ as the critical dividing point between benign appearance deformation and minor defects; the second preset threshold is set to μ+6σ as the critical dividing point between minor defects and severe eccentricity defects.

[0115] When the eccentricity feature evaluation value Less than the first preset threshold When the system determines that the waveform evolution difference, characterized by positional deviation, amplitude asymmetry, and time delay, is within the range allowed by normal manufacturing tolerances and installation deformation, it determines that the coaxiality of the cable terminal is qualified and ends the current round of inspection. When the eccentricity characteristic evaluation value... Greater than or equal to the second preset threshold When the system determines that the propagation characteristics of the ultrasonic guided wave show a significant abnormal shift or energy distortion, this phenomenon strongly points to serious eccentric defects such as uneven compression or poor interface bonding inside the cable terminal. In this case, the system will directly output the conclusion that a serious eccentric defect exists. And when the eccentricity characteristic evaluation value... The value is between the first preset threshold. With the second preset threshold If the result is within a certain range, it indicates that the test result is in a fuzzy range and there is a suspicion of a slight eccentricity defect, but it is not enough to make a direct judgment. At this time, the system will automatically trigger the subsequent high-frequency re-examination process, that is, by switching the frequency of the ultrasonic guided wave to the high-frequency test frequency, the same calculation process is used to re-acquire the result and obtain the high-frequency eccentricity feature evaluation value to verify the authenticity of the suspected defect.

[0116] In the high-frequency re-inspection process, the shorter wavelength of high-frequency ultrasonic guided waves provides a stronger ability to distinguish minute defects. However, their signals are also more susceptible to interference from noise and material dispersion effects, rendering the simple peak extraction method inapplicable. Therefore, after switching the test frequency to the high-frequency test frequency, the system acquires the high-frequency ultrasonic echo signals captured by each ultrasonic receiving array element. To accurately pinpoint the arrival time of the high-frequency echo in a noisy environment, the system first constructs an ideal attenuation envelope reference line in memory with the same center frequency and oscillation period as the current high-frequency test frequency. This reference line is a standard, noise-free Hanning window modulated sine wave R(t), whose mathematical form is: ,in For high-frequency testing, To determine the window length, the system then uses the high-frequency ultrasonic echo signals acquired by the five ultrasonic receiving array elements. Cross-correlation calculations are performed sequentially with the attenuation envelope reference line to generate cross-correlation sequences. The discretization process for cross-correlation calculation is as follows:

[0117] ;

[0118] The essence of this calculation is to find the original signal. The position that most closely resembles the template signal R(n) after shifting k sampling points on the time axis. When the cross-correlation sequence C(k) reaches its maximum value, the physical time represented by its corresponding horizontal coordinate k is the arrival time of the most concentrated main packet of signal energy. The system extracts the time corresponding to this maximum peak coordinate and uses it as the arrival time of the high-frequency echo of this array element. At the same time, at that moment The corresponding original high-frequency waveform The instantaneous amplitude is extracted as the peak amplitude of the high-frequency echo. Finally, the newly acquired high-frequency echo arrival times and peak amplitudes are substituted into the same amplitude asymmetry calculation process, temperature compensation process, and multi-feature weighted summation process used in the fundamental frequency test to calculate the high-frequency eccentricity characteristic evaluation value at that high-frequency test frequency. .

[0119] Preferably, the cross-correlation operation adopts a normalized cross-correlation form to eliminate the influence of echo amplitude differences on peak location results. The length of the analysis window is determined based on the duration of the main echo signal, the sampling frequency, and the background noise distribution, preferably covering the complete main wave energy segment and avoiding obvious pre-noise areas and tail clutter areas. If the window length is too short, it may lead to the loss of main peak information; if the window length is too long, it may introduce irrelevant noise and reduce peak location accuracy. Therefore, the window length should be sufficient to completely contain the main energy segment of the target echo.

[0120] In a preferred embodiment, the cross-correlation calculation can be expressed as follows: The correlation coefficient is calculated point-by-point across the sampled sequence within a preset time delay range, where the time delay range is pre-set based on the theoretical maximum time difference of the echo propagation path; when multiple local peaks appear in the cross-correlation sequence, the peak with the largest amplitude and located within the effective range of the main wave is taken as the target peak, and the corresponding time is taken as the arrival time of the high-frequency echo. To further improve noise immunity, amplitude normalization or bandpass pre-filtering can be performed on the original echo signal before cross-correlation.

[0121] If the system directly determines that a severe eccentricity defect exists during the initial inspection and judgment phase, it will initiate the calculation and definition process for the three-dimensional boundary of the severe eccentricity defect after outputting the results. This process first summarizes the echo peak amplitudes of the five ultrasonic receiving array elements. And filter out echo peak amplitudes that are lower than the preset standard reference amplitude. Half of the ultrasound receiving array elements, of which, The maximum echo peak amplitude of five array elements in the same test can be taken. All array elements that meet this condition are grouped into defect-corresponding array element groups. This means that at these locations, the acoustic energy is drastically attenuated due to severe defects such as internal air gaps or large-area debonding. Then, the actual physical distance between the two ultrasonic receiving array elements at the farthest ends in the defect-corresponding array element group is measured. And half of this distance is set as the semi-axial length of the defect-affected area in the tangential direction. To assess the impact of defects in the depth direction, the system extracts the previously calculated time delay. Echo arrival time of the array element corresponding to the first wave arrival The ratio of the dimensionless ratio As a coefficient representing the degree of eccentricity in the internal compression of the pipe wall.

[0122] Finally, combining the eccentricity coefficient and the tangential semi-axis length of the defect, the specific process of expanding the three-dimensional orthogonal space is as follows: taking the physical center of the corresponding array element of the defect as the origin of the spatial coordinates. Let 'a' be the semi-axis length along the circumferential tangential direction, i.e., the semi-axis length along the defect tangential direction, and 'b' be the semi-axis length along the axial direction, where 'b' is determined by Gaussian smoothing interpolation based on the spacing between adjacent array elements. The eccentricity coefficient is then normalized and mapped to represent the radial semi-axis length 'c'. The standard ellipsoidal parametric equations are then used... A three-dimensional spatial envelope calculation is performed along the normal direction of the cable conduit wall, and all three-dimensional coordinate points on the boundary surface of the equation are extracted as the three-dimensional defect boundary for the final assessment of the severity of the defect.

[0123] Calculate the absolute value of the difference between the high-frequency eccentricity feature evaluation value and the eccentricity feature evaluation value, and set the ratio of the absolute value of the difference to the eccentricity feature evaluation value as the feature change rate.

[0124] In one specific embodiment, after obtaining the high-frequency eccentricity characteristic evaluation value at the high-frequency test frequency, it is jointly compared with the eccentricity characteristic evaluation value obtained at the conventional test frequency. The physical basis for this cross-comparison is that the response mechanisms of ultrasonic waves at different frequencies to internal defects in insulating media are significantly different. Conventional low-frequency ultrasonic waves have longer wavelengths and stronger diffraction capabilities, mainly reflecting macroscopic gradual changes in pipe wall thickness or the overall geometric eccentricity state; while high-frequency ultrasonic waves have shorter wavelengths and are more sensitive to microscopic stress cracks, micro-gaps, or interface peeling within the medium.

[0125] If the cable termination only exhibits simple geometric deformation due to minor external contact or visible wiring, without altering the internal stress distribution, the evaluation results at both frequencies should remain at a similar level. However, cold-shrink cables, due to their strong radial shrinkage force, will inevitably experience excessive stress concentration on one side of the interface if substantial installation eccentricity occurs. This leads to increased local density, while the other side is weakened, potentially resulting in microscopic interface delamination or tiny air gaps. High-frequency ultrasonic waves, with their extremely short wavelengths, will experience strong scattering and local attenuation upon encountering these tiny air gaps and stress unevenness caused by eccentricity, leading to significant abrupt changes in high-frequency characteristic parameters.

[0126] It is important to note that cold-shrink cables differ from ordinary rigid conduits. They primarily rely on the enormous radial recoil force released by the expansion of highly elastic silicone rubber material to hold the cable in place. Therefore, once substantial installation eccentricity occurs internally, it inevitably leads to a severe imbalance in the internal stress field of the material: the stress on the eccentric compression side is extremely concentrated, while the recoil force on the eccentric back side is insufficient, making it highly susceptible to micro-peeling, surface gaps, or micro-cracks at the interface between the insulation layer and the stress cone.

[0127] Based on the above physical mechanism, the system extracts the calculated high-frequency eccentricity feature evaluation value and the eccentricity feature evaluation value, and first calculates the absolute value of the difference between the two. This absolute value of the difference reflects the absolute change in the feature parameters caused by frequency domain switching. Furthermore, this absolute value of the difference is compared with the eccentricity feature evaluation value at the conventional test frequency to characterize the proportion of additional defect response excited by the high-frequency signal. Considering that in the actual computing environment, the eccentricity feature evaluation value may be extremely small under certain extreme defect-free conditions, to prevent the mathematical singularity caused by the denominator approaching zero and the abnormal amplification of hardware noise, a very small regularization constant is introduced into the denominator for variation processing. The specific calculation formula is constructed as follows:

[0128] ;

[0129] In this calculation formula, The rate of change of representative characteristics is used to comprehensively quantify differences in frequency sensitivity. This is the evaluation value for high-frequency eccentricity characteristics; The evaluation value of the eccentricity feature extracted under the normal testing frequency; This is a preset regularization constant. Regarding the regularization constant... The system's preset logic is primarily determined by referencing the inherent quantization noise floor of the analog-to-digital conversion circuit within the ultrasonic testing equipment and the fluctuation deviation of ambient white noise. Its purpose is to act as a smoothing term when the denominator is extremely small, thus shielding division divergences caused by weak, disordered signals. In practical applications, it is recommended to set this parameter within the range of 0.01 to 0.05, which ensures the stability of numerical calculations without affecting the true rate of change.

[0130] The characteristic change rate is compared with a preset sensitivity threshold, and the corresponding cable terminal eccentricity defect detection result is output based on the comparison result.

[0131] Preferably, the comparison and determination of the feature change rate with a preset sensitivity threshold includes:

[0132] Determine whether the feature change rate is greater than the sensitivity threshold;

[0133] When the characteristic change rate is less than or equal to the sensitivity threshold, the current waveform offset phenomenon is determined to be geometric lens distortion caused by installation appearance deformation error, eliminating the risk of internal substantial installation defects, and outputting a test result of qualified coaxiality.

[0134] It should be added that the geometric lens distortion in this application is defined as the simple external physical deformation of the external stress cone of the cable terminal, including but not limited to overall bending, slight surface unevenness, etc., which produces an acoustic focusing or diverging effect similar to an optical lens on the ultrasonic guided wave, resulting in a regular spatial shift in the amplitude of the echo at the receiving end, but without the destruction of the internal continuity of the medium.

[0135] When the characteristic rate of change is greater than the sensitivity threshold, it is determined that the waveform has produced abnormal energy attenuation and phase velocity distortion at different frequencies. This is defined as a substantial eccentric defect caused by uneven stress at the internal interface of the cable terminal, and the detection result of a slight eccentric defect is output.

[0136] Preferably, after outputting the detection result indicating a slight eccentricity defect, the radial penetration depth of the slight eccentricity defect is further calculated, including:

[0137] The high-frequency eccentricity feature evaluation value and the eccentricity feature evaluation value are normalized and integrated to obtain the dual-frequency distortion feature quantity.

[0138] Based on the time delay, the equivalent refraction angle of ultrasonic guided waves propagating at the interface of a non-uniform medium inside the pipe wall is calculated using the law of refraction.

[0139] The initial radial penetration depth is calculated by multiplying the dual-frequency distortion characteristic quantity with the wavelength value at the basic test frequency and dividing the result by the sine of the deflection angle under equivalent refraction.

[0140] Based on the spatial distance between the largest amplitude array element and the central array element, the preliminary radial intrusion depth sequence is segmented and smoothed for noise reduction. The stable value after smoothing and noise reduction is used as the final radial intrusion depth data output.

[0141] In one specific embodiment, after obtaining the eccentricity feature evaluation values ​​at the basic test frequency and the high-frequency test frequency, the system analyzes the differences and correlations between the two to determine whether the currently detected waveform offset phenomenon originates from benign installation appearance deformation or is caused by a substantial internal eccentricity defect, and then performs a quantitative evaluation of the latter. The calculated feature change rate is then used to... Compared with the preset sensitivity threshold Compare the sensitivity threshold. The process involves conducting dual-frequency tests on a large number of qualified cable terminal samples that exhibit only minor surface deformations, such as slight dents or protrusions, and then statistically analyzing their characteristic change rates. The threshold is determined by taking the upper bound of the 95% confidence interval of the statistical distribution. Typically, the empirical value of this threshold ranges from 0.15 to 0.25. Less than or equal to When the frequency of the test is high, it indicates that the evaluation value changes little, and the characteristic evaluation value is not sensitive to frequency. This phenomenon is consistent with the acoustic characteristics of ultrasound propagation in a homogeneous medium due to changes in the geometric path. Therefore, the system can rule out the risk of substantial internal installation defects and output a test result indicating that the coaxiality is qualified. Conversely, when the frequency of the test is low, the evaluation value changes little, and the characteristic evaluation value is not sensitive to frequency. Greater than When the frequency of ultrasonic guided wave propagation is high, it indicates that the high-frequency ultrasonic guided wave has generated disproportionate additional energy attenuation or phase velocity distortion related to its wavelength during propagation. This dispersion effect is a typical acoustic sign of uneven stress or discontinuous medium interface. Therefore, the system determines that there is a substantial eccentric defect in the internal interface of the cable terminal and outputs a detection result indicating a slight eccentric defect.

[0142] After confirming the existence of a mild eccentricity defect, the system will calculate the equivalent radial penetration depth of the defect to further assess its severity. First, the high-frequency eccentricity characteristic evaluation value will be... Compared with the basic eccentricity characteristic evaluation value Normalization and integration were performed to obtain the dual-frequency distortion characteristic. The integration calculation here uses a weighted average method, and the formula is as follows:

[0143] ;

[0144] in, This is a characteristic quantity of dual-frequency distortion; This is a preset frequency weighting coefficient, a dimensionless coefficient ranging from 0 to 1, used to adjust the contribution ratio of the evaluation values ​​at the two frequencies. Considering that high-frequency signals are more sensitive to minute defects but may have a lower signal-to-noise ratio, The default logic is to balance sensitivity and stability, with a typical value set to 0.5, which is the arithmetic mean of the two evaluation values.

[0145] Next, based on the previously calculated time delay... Using a physical model of sound wave refraction, the equivalent refraction angle produced when an ultrasonic guided wave propagates in a non-uniform medium inside a pipe wall is estimated. This angle can be approximately calculated using the wavefront tilt caused by the time delay as the sound wave propagates along the axis of the sensor array. The formula is as follows: In this formula, This is the average propagation speed of ultrasonic guided waves in qualified silicone material, in m / s. This value is retrieved from the equipment material library according to the cable type. The time delay after calibration, in seconds; This represents the axial distance between the first wave arriving at the array element and the array element with the largest amplitude, expressed in meters (m). The physical meaning of this formula is that the greater the time delay, the more the wavefront tilts when propagating the same axial distance, and the greater the equivalent refraction deflection angle.

[0146] Then, by combining the dual-frequency distortion characteristic quantities, the wavelength of the basic test frequency, and the equivalent refraction down-angle, the preliminary radial intrusion depth is calculated. Its physical meaning lies in: a dimensionless dual-frequency distortion characteristic quantity. With wavelength The product of these two values ​​represents the equivalent sound path disturbance length corresponding to the distorted signal, expressed as the physical disturbance scale caused by the defect in the wave propagation direction; divided by the sine of the equivalent refraction angle... This is based on the principles of ultrasonic geometric acoustics, which projects the characteristic scale of the obliquely propagating disturbance onto the radial vertical depth of the material, thus obtaining the actual physical radial penetration depth. The calculation formula is:

[0147] ;

[0148] In the above formula, Basic test frequency The ultrasonic guided wave wavelength is determined by the formula. The calculation yields a result in meters (m).

[0149] Finally, since the initial radial penetration depth calculated at a single point may be affected by local noise, the system will collect a series of measurements formed by multiple measurements taken by the sensor moving axially along the cable terminal. Numerical values, based on the spatial distance between the largest array element and the central array element. The depth sequence is segmented and smoothed for noise reduction. Specifically, the system only performs segmented smoothing and denoising on this sequence. Within the measurement segment where non-zero values ​​indicate detected eccentricity, the Savitzky-Golay filtering algorithm is applied. The sequence is smoothed. This algorithm effectively filters out random noise while preserving the abrupt changes in defect boundaries by performing local polynomial fitting on the data within a sliding window. The stable values ​​output after smoothing and denoising are adopted as the final radial intrusion depth data and output.

[0150] Preferably, when smoothing and denoising the radial intrusion depth sequence obtained from the initial calculation, the Savitzky-Golay filtering algorithm is used to perform local polynomial fitting smoothing on the sequence. Specifically, a sliding window is constructed with each sampling point in the sequence to be processed as the center. A preset number of adjacent sampling points are selected within each sliding window. The data within the window is fitted with a least-squares polynomial of a preset order, and the function value of the fitted curve at the center sampling point is taken as the smoothing result of that sampling point.

[0151] Wherein, the length of the sliding window is denoted as N, and the order of the polynomial is denoted as p, satisfying N>p+1. Preferably, N is an odd number so that the central sampling point is located in the middle of the window, facilitating symmetrical smoothing; p is preferably of order 2 or 3 to preserve the local variation characteristics of the defect response while suppressing random noise. The selection of the window length N is based on the following criteria: the window should cover the local variation interval corresponding to the spacing between adjacent array elements in the radial intrusion depth sequence, while avoiding an excessively large window that weakens the abrupt change characteristics of the defect boundary. Further, N can be determined based on the sampling point density, the spacing between adjacent array elements, and the width of the echo main peak.

[0152] In a preferred embodiment, N can be set to 5, 7, or 9, and p can be set to 2. For sampling points at the boundary that are less than the full window length, compensation can be performed by mirror extension, repeating boundary points, or shortening the window to ensure that the entire sequence can be smoothly calculated.

[0153] It should be noted that the interval and threshold sizes are set for ease of comparison. The size of the threshold depends on the amount of sample data and the base number set by those skilled in the art for each set of sample data, as long as it does not affect the proportional relationship between the parameter and the quantized value. Furthermore, the above formulas are all dimensionless calculations, and the formulas are derived from software simulations using a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0154] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0155] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

[0156] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for detecting installation defects in cold-shrink cable terminals based on ultrasonic guided waves, characterized in that, include: Five ultrasonic receiving array elements are set at equal intervals along the axial direction on the outer wall of the cold shrink cable terminal, and the ultrasonic receiving array element in the middle position is set as the center array element. By exciting ultrasonic guided waves at a set basic test frequency, the arrival time and peak amplitude of the echoes received by each ultrasonic receiving array element are obtained. Based on the peak amplitude of the echo, the array element number corresponding to the largest amplitude is selected, and based on the echo arrival time, the array element number corresponding to the first wave arrival is selected. Extract the echo peak amplitudes of the adjacent array elements on both sides of the largest amplitude array element, and calculate the ratio of the difference and sum of the echo peak amplitudes of the adjacent array elements on both sides to obtain the amplitude asymmetry. The time delay is obtained by calculating the difference between the echo arrival time of the array element with the largest amplitude and the echo arrival time of the first wave to the corresponding array element. Calculate the positional deviation between the largest amplitude array element and the central array element, and then sum the positional deviation, amplitude asymmetry, and time delay by weight to obtain the eccentricity feature evaluation value. The eccentricity feature evaluation value is initially determined. When the eccentricity feature evaluation value is between the first preset threshold and the second preset threshold, the frequency of the ultrasonic guided wave is switched to the high-frequency test frequency, and the same calculation process is used to re-acquire the high-frequency eccentricity feature evaluation value. Calculate the absolute value of the difference between the high-frequency eccentricity feature evaluation value and the eccentricity feature evaluation value, and set the ratio of the absolute value of the difference to the eccentricity feature evaluation value as the feature change rate; The characteristic change rate is compared with a preset sensitivity threshold, and the corresponding cable terminal eccentricity defect detection result is output based on the comparison result.

2. The method for detecting installation defects in cold-shrink cable terminals based on ultrasonic guided waves according to claim 1, characterized in that, Obtain the arrival time and peak amplitude of the echoes received by each ultrasonic receiving array element, including: During the period when the ultrasonic guided wave is excited at the basic test frequency, the ultrasonic echo signals captured by each ultrasonic receiving array element are collected. The first-order difference calculation of the ultrasonic echo signal is used to obtain the signal rate of change, and the position of the first zero crossing of the signal rate of change is extracted as the oscillation start time. A signal interception window is defined with the start time of vibration as the starting point, and the data of the corresponding interval of the ultrasonic echo signal is intercepted as the main wave signal. Extract the absolute values ​​of all local maxima in the main wave signal; The maximum value among the absolute values ​​of all local maxima is taken as the echo peak amplitude of the array element, and the time when the maximum value is extracted is recorded as the echo arrival time.

3. The method for detecting installation defects in cold-shrink cable terminals based on ultrasonic guided waves according to claim 1, characterized in that, Extract the echo peak amplitudes of the adjacent array elements on both sides of the largest amplitude array element, and calculate the amplitude asymmetry, including: Identify the position of the maximum amplitude array element among the five ultrasound receiving array elements; When the largest amplitude matrix element is at the leftmost position, it is determined that the data element of its left adjacent matrix element is missing, and the amplitude asymmetry is directly assigned to the preset first extreme value constant. When the largest amplitude matrix element is at the rightmost position, it is determined that the adjacent matrix element to its right is missing, and the amplitude asymmetry is directly assigned to the preset second extreme value constant. When the element with the largest amplitude is neither at the leftmost nor the rightmost end, extract the amplitude of the element adjacent to its left and the amplitude of the element adjacent to its right. The difference is obtained by subtracting the amplitude of the adjacent array element on the left from the amplitude of the adjacent array element on the right. The sum is obtained by adding the amplitude of the adjacent array element on the left to the amplitude of the adjacent array element on the right. The difference is divided by the sum to obtain the final amplitude asymmetry.

4. The method for detecting installation defects in cold-shrink cable terminals based on ultrasonic guided waves according to claim 1, characterized in that, The eccentricity feature evaluation value is calculated, including: Obtain the elastic modulus of the insulating silicone and the nominal wall thickness of the cold-shrink cable terminal; The position weight coefficients are obtained by inversely scaling the preset reference weight coefficients using the elastic modulus of insulating silicone. The delay weighting coefficient is obtained by scaling the baseline weighting coefficient proportionally using the nominal wall thickness. Extract the preset material attenuation constant as the asymmetric weighting coefficient; Calculate the absolute value of the distance difference between the largest array element and the central array element, as the positional deviation with positive and negative directional information; The position deviation is multiplied by the position weight coefficient to obtain the position weighted component. The absolute value of the amplitude asymmetry is extracted and multiplied by the asymmetry weight coefficient to obtain the amplitude weighted component. The time delay is multiplied by the delay weight coefficient to obtain the delay weighted component. The position-weighted component, amplitude-weighted component, and delay-weighted component are summed to obtain the eccentricity feature evaluation value.

5. The method for detecting installation defects in cold-shrink cable terminals based on ultrasonic guided waves according to claim 4, characterized in that, The initial assessment of the eccentricity feature evaluation value includes: Compare the eccentricity feature evaluation value with the first preset threshold and the second preset threshold. The second preset threshold is greater than the first preset threshold. When the eccentricity feature evaluation value is less than the first preset threshold, it is determined that the waveform propagation delay is within the normal tolerance range, the coaxiality of the cable terminal is deemed qualified, and the test round ends. When the eccentricity feature evaluation value is greater than or equal to the second preset threshold, it is determined that the waveform has abnormal offset or distortion, a serious eccentricity defect is identified, and the result is output. When the eccentricity feature evaluation value is greater than or equal to the first preset threshold and less than the second preset threshold, it is determined that there is a slight eccentricity defect, triggering the subsequent process of switching the test frequency to a high-frequency test frequency.

6. The method for detecting installation defects in cold-shrink cable terminals based on ultrasonic guided waves according to claim 1, characterized in that, The characteristic change rate is compared with a preset sensitivity threshold for determination, including: Determine whether the feature change rate is greater than the sensitivity threshold; When the characteristic change rate is less than or equal to the sensitivity threshold, the current waveform offset phenomenon is determined to be geometric lens distortion caused by installation appearance deformation error, eliminating the risk of internal substantial installation defects, and outputting a test result of qualified coaxiality. When the characteristic rate of change is greater than the sensitivity threshold, it is determined that the waveform has produced abnormal energy attenuation and phase velocity distortion at different frequencies. This is defined as a substantial eccentric defect caused by uneven stress at the internal interface of the cable terminal, and the detection result of a slight eccentric defect is output.

7. The method for detecting installation defects in cold-shrink cable terminals based on ultrasonic guided waves according to claim 1, characterized in that, Calculate the time delay, including: Collect the surface temperature of the cable duct wall at the location of each ultrasonic receiving array element; The average value of the surface temperatures distributed along the axial direction is calculated to obtain the equivalent ambient temperature along the ultrasonic guided wave propagation path. Based on the temperature difference between the equivalent ambient temperature and the standard reference temperature, and combined with the sound velocity temperature drift coefficient of the pipe wall material itself, the corresponding temperature delay compensation is calculated. The initial time difference is obtained by subtracting the echo arrival time of the first wave to the corresponding array element from the echo arrival time of the array element with the largest amplitude. The initial time difference is smoothed by filtering to eliminate high-frequency jitter interference, and the filtered value is superimposed with the temperature delay compensation to obtain the calibrated time delay.

8. The method for detecting installation defects in cold-shrink cable terminals based on ultrasonic guided waves according to claim 5, characterized in that, Obtain high-frequency eccentricity feature evaluation values, including: After switching the test frequency to a high-frequency test frequency, the high-frequency ultrasonic echo signals generated by each ultrasonic receiving array element were acquired. Construct a decaying envelope reference line with the same oscillation period as the high-frequency test frequency; The high-frequency ultrasonic echo signals of each ultrasonic receiving array element are sequentially cross-correlated with the attenuation envelope reference line to generate a cross-correlation sequence. Extract the time corresponding to the maximum peak coordinate in the cross-correlation sequence, take it as the arrival time of the high-frequency echo, and extract the original waveform amplitude corresponding to that time as the peak amplitude of the high-frequency echo; Based on the arrival time and peak amplitude of the high-frequency echo, the same amplitude asymmetry calculation process and weighted summation process are used to calculate the high-frequency eccentricity characteristic evaluation value at the high-frequency test frequency.

9. The method for detecting installation defects in cold-shrink cable terminals based on ultrasonic guided waves according to claim 6, characterized in that, After outputting the detection results for a slight eccentricity defect, the radial penetration depth of the slight eccentricity defect is further calculated, including: The high-frequency eccentricity feature evaluation value and the eccentricity feature evaluation value are normalized and integrated to obtain the dual-frequency distortion feature quantity. Based on the time delay, the equivalent refraction angle of ultrasonic guided waves propagating at the interface of a non-uniform medium inside the pipe wall is calculated using the law of refraction. The initial radial penetration depth is calculated by multiplying the dual-frequency distortion characteristic quantity with the wavelength value at the basic test frequency and dividing the result by the sine of the deflection angle under equivalent refraction. Based on the spatial distance between the largest amplitude array element and the central array element, the preliminary radial intrusion depth sequence is segmented and smoothed for noise reduction. The stable value after smoothing and noise reduction is used as the final radial intrusion depth data output.

10. The method for detecting installation defects in cold-shrink cable terminals based on ultrasonic guided waves according to claim 5, characterized in that, After outputting the results of the severe eccentricity defect, the three-dimensional boundary of the severe eccentricity defect is calculated and defined, including: The echo peak amplitudes of the five ultrasonic receiving array elements are summarized, and ultrasonic receiving array elements whose echo peak amplitudes are lower than half of the preset standard reference amplitude are selected and grouped into the defective array element group. The actual physical distance between the two ultrasonic receiving array elements at the farthest ends in the array element group corresponding to the defect is measured, and half of this actual physical distance is set as the tangential half-axis length of the defect. The ratio of the extracted time delay to the echo arrival time of the first wave to the corresponding array element is used as a coefficient to characterize the degree of eccentricity of the compression unevenness. By combining the eccentricity coefficient and the length of the defect tangential semi-axis, a three-dimensional orthogonal spatial expansion is carried out along the pipe wall normal direction to construct an eccentric influence envelope ellipsoid containing the insulation defect region. All three-dimensional coordinate points on the surface of this eccentric influence envelope ellipsoid are extracted as the three-dimensional defect boundary for the final assessment of the defect severity.

Images