An on-line detection method for the insulation layer thickness of a power cable

Abstract

The present invention relates to the field of thickness detection technology, and specifically to an online detection method for the thickness of the insulation layer of a power cable. The method decomposes a signal into decomposed signals of different scales through a multi-level decomposition technology, and obtains the scale contribution through the similarity of the decomposed signals under the scale; performs similarity analysis in both the frequency domain and the signal near the decomposed signal corresponding to the peak group through a matching algorithm, and obtains the consistency index by combining the scale contribution to screen out the echo peak group; obtains the corrected time difference through the time difference of the echo peak group and the consistency index; according to the relationship between the temperature and sound velocity of the probe coolant environment, adjusts the sound velocity based on the change of the current temperature, and obtains the insulation layer thickness by combining the corrected time difference. Based on the temperature-corrected sound velocity, the present invention corrects the echo time difference from the local similarity characteristics of the scale and the peak through multi-level decomposition, effectively improving the accuracy of insulation layer thickness detection.

Description

Technical Field

[0001] The present invention relates to the technical field of thickness detection, and in particular to an online detection method for the thickness of an insulation layer of a power cable. Background Art

[0002] Power cables are used to transmit and distribute electrical energy. They primarily consist of a conductor core, insulation, shielding, and protective layers. The insulation layer, as the first line of defense, protects the internal conductor and prevents current from entering the external environment, playing a crucial role in power cables. Appropriate insulation thickness effectively protects the cable's internal conductors, and uniform insulation thickness evenly distributes the electric field within the conductor. Therefore, to ensure the safety and longevity of power cables, insulation thickness testing is necessary during production. Current insulation testing methods rely on the propagation characteristics of ultrasonic waves within the insulation layer, measuring their reflection time to calculate the thickness.

[0003] The basic process of ultrasonic testing of insulation thickness is to use the different speeds of ultrasonic waves when propagating in different media. When ultrasonic waves propagate from one medium to another, reflection and refraction will occur at the interface. Since the cable conductor and the insulation layer are different media, when the ultrasonic waves encounter the insulation layer-air interface and the conductor-insulation layer interface, reflection and refraction will occur. The external probe receives two returned ultrasonic echo signals, and the thickness of the insulation layer is obtained based on them.

[0004] However, during the production process of power cables, since the cable conductors and insulation layers are newly produced, the insulation layers are at high temperatures and need to be cooled. When temperatures fluctuate, the propagation speed of ultrasonic waves at these temperatures varies, causing the ultrasonic signal to vary when it encounters the cable interface and propagates through the medium, resulting in a broadening of the detected echo signal. Existing methods for determining the time difference between echo reception typically analyze the peaks of the echo signal through methods such as packet detection. However, because the echo signal is broadened, and because the uneven twisting surface of the conductors within the cable causes diffuse reflection, the echo signal is further dispersed, resulting in multiple groups of peaks in the echo signal. This affects the determination of the correspondence between the two echo signals, leading to errors in the time difference between the two detected echo signals, and thus inaccurate ultrasonic thickness testing results. Summary of the Invention

[0005] In order to solve the technical problem in the prior art that echo signals are broadened and diffuse reflection occurs, which further disperses the echo signals, resulting in multiple groups of peaks in the echo signals and affecting the determination of the correspondence between the two echo signals, the purpose of the present invention is to provide an online detection method for the thickness of the insulation layer of a power cable. The technical solution adopted is as follows:

[0006] The present invention provides a method for online detection of the thickness of the insulation layer of a power cable, the method comprising:

[0007] Acquire two echo signals at the detection point through the probe, and obtain the peak value in the echo signal; group each two different peak values into a peak group, where the two peak values in the peak group are located in the two echo signals respectively; perform multi-level decomposition on the echo signal to obtain the decomposed signal at each scale;

[0008] The scale contribution of each scale is obtained based on the correlation between the two decomposed signals at each scale. The two decomposed signals at each scale are matched within the preset neighborhood of each peak group to obtain a moment matching group. The matchability of each peak group at each scale is obtained based on the energy distribution similarity and signal similarity of the matching group of the two decomposed signals at each moment.

[0009] For each peak group, the consistency index is obtained by combining the scale contribution and matchability of each scale; the echo peak groups are screened based on the consistency index; the current corrected time difference is obtained based on the differences between the corresponding peak times in all echo peak groups and the consistency index of the echo peak groups;

[0010] The probe detects the changing relationship between the temperature and sound velocity of the ambient coolant, and adjusts the calibrated sound velocity based on the temperature deviation of the current detection environment to obtain the current corrected sound velocity. The insulation layer thickness at the current detection point is obtained by combining the current corrected time difference and the corrected sound velocity.

[0011] Furthermore, the method for obtaining the scale contribution includes:

[0012] For any scale, the DTW distance between the two decomposed signals is calculated at the scale to perform negative correlation mapping and normalization to obtain the scale contribution of the scale.

[0013] Furthermore, the method for obtaining the time matching group includes:

[0014] For any peak group, each scale is used as the analysis scale in turn. Within the preset neighborhood range of the peak corresponding to the moment in the peak group, the two decomposed signals under the analysis scale are matched using the DTW algorithm, and each pair of matching moments is regarded as a moment matching group.

[0015] Furthermore, the method for obtaining the matchability includes:

[0016] For any peak group, within the preset neighborhood of the peak group, the two decomposed signals obtain frequency domain features at the corresponding time in the matching group at each moment, and the similarity is analyzed to obtain the frequency domain similarity index of the two decomposed signals matching group at each moment;

[0017] Perform negative correlation mapping on the difference in signal values of the two decomposed signals at the corresponding time in the matching group at each time, and obtain the signal similarity index of the two decomposed signals at each time matching group;

[0018] The product of the frequency domain similarity index and the signal similarity index of the matching group at each moment is used as the similarity index of the matching group at each moment;

[0019] Within the preset neighborhood of the peak group, the sum of the similarity indices of the matching groups of the two decomposition signals at all times is used as the numerator, and the sum of the frequency domain similarity indices of the matching groups of the two decomposition signals at all times is used as the denominator to obtain the matching degree of the peak group at the corresponding scale of the two decomposition signals.

[0020] Furthermore, the method for obtaining the frequency domain similarity index includes:

[0021] The frequency domain characteristics of the two decomposed signals in time series are obtained by fast Fourier transform;

[0022] For any moment matching group, the decomposed signal corresponding to each moment in the moment matching group is used as the correlation signal of each moment; the period consisting of each moment in the moment matching group and the two adjacent moments in time sequence is used as the characteristic period of each moment;

[0023] In the two feature time periods in the matching group at this moment, the correlation of the frequency domain features between the corresponding correlation signals is calculated as the frequency domain similarity index of the matching group of the two decomposed signals at this moment.

[0024] Furthermore, the method for obtaining the consistency index includes:

[0025] For any peak group, at the peak group, the product of the matchability and scale contribution at each scale is used as the matching possibility index of each scale;

[0026] The sum of the matching possible indices of all scales is used as the numerator, and the sum of the scale contributions of all scales is used as the denominator to obtain the consistency index of the peak group.

[0027] Furthermore, the method for obtaining the echo peak group includes:

[0028] The peak group whose consistency index is greater than the preset judgment threshold is regarded as the echo peak group.

[0029] Furthermore, the method for obtaining the corrected time difference includes:

[0030] Calculate the difference between the corresponding moments of the two peaks in each echo peak group as the time difference of each echo peak group;

[0031] The product of the time difference of each echo peak group and the consistency index is used as the weighted time difference index of each echo peak group;

[0032] The sum of the weighted time difference indices of all echo peak groups is used as the numerator, and the sum of the consistency indices of all echo peak groups is used as the denominator to obtain the current corrected time difference.

[0033] Furthermore, the method for obtaining the corrected sound speed includes:

[0034] Obtain the temperature coefficient of sound velocity through preliminary experiments;

[0035] The product of the difference between the current detection temperature and the calibration temperature and the sound velocity temperature coefficient is used as the sound velocity adjustment degree;

[0036] The product of the calibrated sound speed and the sound speed adjustment degree is used as the sound speed adjustment value; the sum of the calibrated sound speed and the sound speed adjustment value is used as the corrected sound speed.

[0037] Furthermore, the method for obtaining the thickness of the insulating layer includes:

[0038] The product of the current corrected time difference and the corrected sound velocity is divided by 2 to obtain the insulation layer thickness at the current detection point.

[0039] The present invention has the following beneficial effects:

[0040] The present invention first decomposes the signal into signals of different scales using a multi-level decomposition technique, which helps reduce the impact of noise caused by conductor surface roughness or coolant flow. Based on the decomposed signals, the scale contribution is calculated by comparing the similarity of the decomposed signals at each scale. The echo similarity characteristics reflect the reliability of the analysis at that scale. A matching algorithm then performs similarity analysis in both the frequency domain and the signal near the corresponding decomposed signals of the peak groups. The degree of peak matching within the peak groups is reflected by the similarity of local details. Furthermore, a consistency index is derived from the scale contribution to filter out echo peak groups, reducing multi-peak interference. By adjusting the time difference of the echo peak groups and the consistency index, a more accurate corrected time difference is obtained. Furthermore, based on the relationship between the coolant temperature and sound velocity detected in real time by the probe, the calibrated sound velocity is adjusted based on the current temperature, eliminating the impact of sound velocity drift on thickness calculation in dynamic temperature environments and improving adaptability to complex working conditions. The corrected time difference and corrected sound velocity are combined to determine the insulation thickness. Based on the temperature-corrected sound velocity, the present invention uses multi-level decomposition to correct the echo time difference based on the scale and local similarity characteristics of the peaks, effectively improving the accuracy of insulation thickness detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0042] Figure 1 A flow chart of an online detection method for the thickness of the insulation layer of a power cable provided by one embodiment of the present invention;

[0043] Figure 2 A schematic diagram of the structure of a probe and a cable according to an embodiment of the present invention;

[0044] Figure 3 A schematic diagram of two echo signals provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0045] To further illustrate the technical means and effectiveness of the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features, and effectiveness of an online detection method for the thickness of a power cable insulation layer proposed by the present invention. In the following description, different references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0046] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0047] The specific scheme of the on-line detection method for the thickness of the insulation layer of a power cable provided by the present invention is described in detail below with reference to the accompanying drawings.

[0048] During the production of power cables, the copper wires of the cables are first straightened and twisted, and then a screw of a predetermined shape is used in an extruder to melt the insulating plastic under high temperature conditions. The screw is then extruded forward so that the insulating plastic passes through a mold to form an insulating layer around the cable conductor.

[0049] During this process, the plastic is heated to a high temperature, then cooled by coolant and air before proceeding to the next step of installing shielding and protective layers. When ultrasonic testing is used to detect the thickness of the cable insulation layer, the ultrasonic probe is placed in the coolant. Since the insulation layer has just been produced and is at a high temperature, and the cable is still in the coolant, the cable conductor and insulation layer gradually dissipate heat, while the coolant temperature gradually rises. The speed of ultrasound waves changes in coolant and insulation layers at different temperatures, causing the time between two echo signals received by the ultrasonic probe to be affected by temperature during testing.

[0050] Therefore, when making corrections to ultrasonic testing, the temperature influence of the testing insulation layer must also be considered. The ultrasonic echo signals received in the current time period and the temperature must be combined to make corrections to improve the accuracy of thickness testing. Figure 1 , which shows a flow chart of a method for online detection of the thickness of the insulation layer of a power cable provided by one embodiment of the present invention, the method comprising the following steps:

[0051] S1: Acquire two echo signals at the detection point through the probe and obtain the peak value in the echo signal; group every two different peak values into a peak group, where the two peak values in the peak group are located in the two echo signals respectively; perform multi-level decomposition on the echo signal to obtain the decomposed signal at each scale.

[0052] First, on the production line, the power cable is stretched straight by two rollers and immersed in coolant. Then, ultrasonic probes and temperature probes are installed at the bottom of the production line to detect the reflection of ultrasonic waves at various locations on the power cable and the temperature of the cable. Figure 2 , which shows a schematic diagram of the structure of a probe and a cable portion provided by an embodiment of the present invention.

[0053] In the embodiment of the present invention, the coolant can be preset to water, and a set of probes is set every 10 cm, including an ultrasonic probe and a temperature measuring probe. The temperature measuring probe can be an infrared temperature sensor. The ultrasonic probe needs to be perpendicular to the cable surface, and the center frequency of the ultrasonic wave is preset to be , the sampling rate of the ultrasonic receiver is The frequency of ultrasonic pulse emission and the detection frequency of infrared temperature sensor are both set to 10 times / second. The specific acquisition settings can be adjusted by the implementer according to the specific implementation scenario, and there is no restriction here.

[0054] The probe is placed at the test point on the cable insulation layer and emits an ultrasonic wave. The ultrasonic wave propagates through the coolant and, upon encountering the insulation surface, is reflected for the first time and received by the probe, recording as the first echo signal. Simultaneously, part of the ultrasonic wave is refracted and enters the insulation layer, continuing its journey. Upon encountering the conductor surface, it is reflected again and, after further propagation, received by the probe, recording as the second echo signal. Because the insulation layer and twisted conductor bundles of newly manufactured cables have slightly rough surfaces, and the insulation layer and coolant are in motion, the two interfaces reflect the ultrasonic signal with diffuse reflection. This causes the resulting echo signal to appear broadened upon reception, resulting in multiple peaks. See [Note: The following sentences appear unrelated and should be omitted for clarity.] Figure 3 , which shows a schematic diagram of two echo signals provided by an embodiment of the present invention, where M1 represents the first echo signal and M2 represents the second echo signal.

[0055] In this embodiment of the present invention, the upper and lower envelopes of the two echo signals are first constructed using a Hilport transform. The peaks of the two echo signals, also known as local maxima, are then determined using these envelopes. The time difference between the peaks is used for thickness analysis, but due to the influence of multiple peaks, mismatched values can be selected, resulting in errors and requiring further correction.

[0056] Every two different peaks in the two echo signals form a peak group, and the two peaks in the peak group are respectively located in the two echo signals, that is, one peak in the peak group is located in the first echo signal, and the other peak is located in the second echo signal.

[0057] The echo signal is further decomposed at multiple levels to obtain decomposed signals at each scale. In the embodiment of the present invention, a wavelet with a waveform similar to that of the echo signal is used to perform multi-level decomposition on the two echo signals. The number of decomposition layers is preset to 6, and there are decomposed signals at 6 scales. It should be noted that the multi-level decomposition performed by wavelet decomposition is a technical means well known to those skilled in the art, such as using Wavelet, etc., will not be elaborated or limited here.

[0058] S2: Obtain the scale contribution of each scale based on the correlation between the two decomposed signals at each scale; match the two decomposed signals at the scale within the preset neighborhood of each peak group to obtain a moment matching group; obtain the matchability of each peak group at each scale based on the similarity of the energy distribution and signal similarity of the matching groups of the two decomposed signals at each moment.

[0059] First, consider the presence of noise signals in the echo signal. These signals are of little help in determining the accurate echo time, and may even have a negative effect. Conversely, valid, non-noise signals are more helpful in determining the echo time. Furthermore, when testing the insulation thickness, the ultrasonic signal emitted is emitted by the same ultrasonic transmitter. Reflection and refraction from the insulation surface and the cable conductor surface have little effect on the ultrasonic signal, so the waveform information of the valid echo signal is relatively consistent. Therefore, the higher the signal consistency between the two echo signals at scale, the greater the contribution of the signal analysis at scale to determining the echo time.

[0060] Preferably, in an embodiment of the present invention, the method for obtaining the scale contribution includes: for any scale, calculating the DTW distance between the two decomposed signals at the scale, performing negative correlation mapping and normalization processing, and obtaining the scale contribution of the scale. The smaller the DTW distance, the more similar the decomposed signals are, the less noise they contain, and the higher their contribution to determining the echo time.

[0061] It should be noted that the DTW algorithm, negative correlation mapping, and normalization algorithm corresponding to the DTW distance are technical means well known to those skilled in the art. The negative correlation mapping can adopt a negative exponential power or an inverse proportional form, etc., and the normalization option can be linear normalization or standard normalization, etc., which will not be elaborated here.

[0062] Secondly, in the decomposed signals at each scale, because the two echo signals are emitted simultaneously and directly reflected from the same location, the reflection of the same ultrasonic signal in the two received echo signals does not significantly alter the signal. In other words, the frequency and other signal energy distributions of the same ultrasonic signal are consistent in the two echo signals. Furthermore, based on the more localized similarities in the decomposed signals at each scale, we can identify the likelihood that peaks in a peak group correspond to the same echo.

[0063] To analyze similarities from a local perspective, the two decomposed signals at the peak group are matched for time correspondence, and similarities are analyzed from the perspective of time matching. Preferably, in an embodiment of the present invention, the method for obtaining a time matching group includes: for any peak group, taking each scale as the analysis scale in turn, matching the two decomposed signals at the analysis scale using the DTW algorithm within a preset neighborhood range of the peak corresponding time in the peak group, and treating each pair of matched time moments as a time matching group. In this embodiment, the preset neighborhood range is a range centered at the time moment with a radius of 10 time moments, which can be adjusted by the implementer.

[0064] For example, the echo signals where the peaks in the peak group are located are M1 and M2, and the corresponding times are t1 and t2. For the two decomposition signals of the analysis scale, DTW matching is performed on the decomposition signal part corresponding to M1 within the neighborhood range of time t1 and the decomposition signal part corresponding to M2 within the neighborhood range of time t2, and the time when the matching is completed is used as the matching group for each time.

[0065] Then, similarity is analyzed from each matching group at each moment. The more similar the energy distribution of the decomposed signal at the matching moment is to the signal, the more likely it is that the peak group under the moment analysis is a matching two-interface reflection. Preferably, in an embodiment of the present invention, the method for obtaining the matching degree of each peak group at each scale includes:

[0066] First, for any peak group, within a preset neighborhood of the peak group, the two decomposed signals are respectively subjected to frequency domain features at the corresponding time in each matching group. The similarity is analyzed to obtain a frequency domain similarity index for the two decomposed signals at each matching group, and similarity analysis is performed based on the energy distribution in the frequency domain space. In this embodiment of the present invention, the method for obtaining the frequency domain similarity index includes:

[0067] The frequency domain characteristics of the two decomposed signals in time series are obtained by fast Fourier transform. Frequency domain characteristics are quantitative indicators used to describe the spectral characteristics of the signal after the signal is converted from the time domain to the frequency domain by Fourier transform, mainly including the center of gravity frequency, mean square frequency, root mean square frequency, frequency variance and frequency standard deviation. In an embodiment of the present invention, the frequency variance in the frequency domain space can be used as a frequency domain feature. The larger the variance, the more dispersed the energy distribution, which can intuitively reflect the distribution concentration characteristics of the spectrum energy. It should be noted that the method of obtaining frequency domain characteristics by Fourier transform is a technical means well known to those skilled in the art, and will not be elaborated or limited here.

[0068] For any moment matching group, the decomposed signal corresponding to each moment in the moment matching group is used as the correlation signal for each moment, and the moments in the moment matching group are labeled accordingly with the corresponding decomposed signal. The period consisting of each moment in the moment matching group and the two adjacent moments in time sequence is used as the characteristic period of each moment, and the period consisting of each moment and the two moments before and after is used as the characteristic period to implement local feature correlation analysis.

[0069] The correlation of the frequency domain features between the corresponding correlated signals in the two characteristic time periods in the moment matching group is calculated as the frequency domain similarity index of the two decomposed signals in the moment matching group. The similarity of the energy distribution is reflected by the degree of correlation of the time series changes in the frequency domain features corresponding to the two decomposed signals in the corresponding time periods at the two moments in the moment matching group. It should be noted that the calculation of time series data correlation is a technical means well known to those skilled in the art, such as the sampled DTW algorithm or the Pearson correlation coefficient, and is not limited here.

[0070] Furthermore, the difference in signal values of the two decomposed signals at corresponding moments in the matching group at each moment is negatively correlated to obtain the signal similarity index of the two decomposed signals in the matching group at each moment, and the similarity between moments in signal amplitude is analyzed.

[0071] Finally, the frequency domain feature similarity index is used as the weight for weighted normalization, and the product of the frequency domain similarity index and the signal similarity index of each moment matching group is used as the similarity index of each moment matching group, and the similarity degree is analyzed by frequency domain similarity weighting.

[0072] Within the preset neighborhood of the peak group, the sum of the similarity indices of the two decomposed signals at all times is used as the numerator, and the sum of the frequency domain similarity indices of the two decomposed signals at all times is used as the denominator. This yields the matchability of the peak group at the corresponding scales of the two decomposed signals. Weighted normalization is performed using this ratio to ensure that regions with high frequency domain consistency contribute more to the final matchability score. A higher matchability indicates a higher likelihood that the peak group corresponds to a corresponding echo relationship at a single scale.

[0073] S3: For each peak group, the consistency index is obtained by combining the scale contribution and matchability of each scale; the echo peak groups are selected based on the consistency index; the current corrected time difference is obtained based on the difference between the corresponding peak times in all echo peak groups and the consistency index of the echo peak groups.

[0074] If a single peak group is relatively matched across all scales, then the two peaks can be considered to be the result of reflections from the same ultrasonic signal across two interfaces. The consistency index is derived by also considering the contribution of the valid case and combining the matching degree of the single peak group across all scales.

[0075] In an embodiment of the present invention, a method for obtaining a consistency index includes: for any peak group, at the peak group, taking the product of the matchability at each scale and the scale contribution as the matching possibility index of each scale, and weighting it by the scale contribution. The scale with a higher contribution is more reliable in its matching analysis.

[0076] Then, the sum of the matching possible indices of all scales is used as the numerator, and the sum of the scale contributions of all scales is used as the denominator to obtain the consistency index of the peak group. The consistency index is obtained through weighted normalization. The larger the consistency index, the greater the possibility that the peaks of the peak group are the same ultrasonic signal. As an example, the expression of the consistency index is: , where Expressed as The consistency index of the peak group, Expressed as The peak group is in The matching degree under different scales, Expressed as The scale contribution of each scale, Expressed as the total number of scales. Expressed as The peak group is in The matching possibility index of each scale.

[0077] The peak group with the corresponding possible high value is screened by threshold judgment. In an embodiment of the present invention, the peak group whose consistency index is greater than the preset judgment threshold is used as the echo peak group. The preset judgment threshold is set to 0.5. When it is greater than the threshold, it indicates that the peak in the peak group is the same ultrasonic reflection, and is used as the echo peak group for subsequent time difference analysis.

[0078] However, there is a certain error in determining the time it takes for ultrasound to penetrate the insulation layer using a single echo signal point, and not every echo peak group is relatively accurate. That is, the more consistent the two echo signals in an echo peak group are, the more reliable the resulting time difference is. Therefore, the time difference is weighted by the consistency index of each echo peak group to obtain the corrected time difference between the two echo signals.

[0079] Preferably, in an embodiment of the present invention, the method for obtaining the corrected time difference includes:

[0080] Calculate the difference between the corresponding times of the two peaks in each echo peak group as the time difference for each echo peak group, and as the accepted time difference between the echoes. Multiply the time difference of each echo peak group by the consistency index to form the weighted time difference index for each echo peak group. Weighted by the consistency index, peak groups with higher consistency have a higher degree of credibility in the integrated time difference.

[0081] Then, the sum of the weighted time difference indices of all echo peak groups is used as the numerator, and the sum of the consistency indices of all echo peak groups is used as the denominator to obtain the current corrected time difference. Through weighted normalization correction, a more accurate echo time difference is obtained.

[0082] S4: Use the probe to detect the changing relationship between the temperature and sound velocity of the ambient coolant, adjust the calibrated sound velocity based on the temperature deviation of the current detection environment, and obtain the current corrected sound velocity; combine the current corrected time difference and the corrected sound velocity to obtain the insulation layer thickness at the current detection point.

[0083] Regarding the effect of temperature on the speed of sound waves, by pre-testing the speed-temperature relationship of the insulating plastic in a coolant environment, the speed of sound can be corrected in a timely manner by monitoring the real-time temperature deviation change. Preferably, in an embodiment of the present invention, the method for obtaining the corrected speed of sound includes:

[0084] The temperature coefficient of the speed of sound is obtained through preliminary experiments. The temperature coefficient of the speed of sound refers to the ratio of the speed of sound to the temperature when the sound wave propagates in the medium. In the embodiment of the present invention, the speed of sound-temperature relationship of the coolant and the insulating layer plastic is pre-tested in the laboratory. In the coolant environment, the resonance interferometry method is preset to measure the corresponding speed of sound at different temperatures, obtain a large number of speed of sound temperature data points, and then fit the speed of sound at various temperatures by the least squares method. Then, every Select a sound velocity data point to form a temperature-speed sequence. Then, use the difference method to calculate the difference sequence of the temperature-speed sequence to describe the effect of temperature changes in the medium on the sound velocity. Calculate the average value of the elements in the difference sequence to obtain the corresponding temperature coefficient of the coolant.

[0085] The product of the difference between the current detection temperature and the calibration temperature and the sound velocity temperature coefficient is used as the sound velocity adjustment degree to reflect the rate of change of the sound velocity caused by the temperature difference. Therefore, the product of the calibration sound velocity and the sound velocity adjustment degree is used as the sound velocity adjustment value to represent the sound velocity adjustment amount. In the embodiment of the present invention, the calibration temperature is a preset standard temperature, which is set to The calibrated sound velocity is the sound velocity at the calibration temperature, and the specific value can be controlled by the implementer according to the specific implementation situation.

[0086] Finally, the sum of the calibrated sound speed and the sound speed adjustment value is used as the corrected sound speed to make more accurate corrections for ambient temperature fluctuations.

[0087] Based on the temperature information of the cable insulation layer, the echo signal is compensated and the insulation layer thickness can be obtained by combining the time difference. In this embodiment of the present invention, the product of the current corrected time difference and the corrected sound speed is divided by 2 to obtain the insulation layer thickness at the current detection point. As an example, the expression for the insulation layer thickness is: , where Expressed as the thickness of the insulation layer, Indicates the correction time difference; Expressed as the corrected speed of sound.

[0088] In the embodiment of the present invention, repeated measurement is a key step in ensuring data accuracy and reliability during ultrasonic testing of the insulation thickness of power cables. To obtain reliable measurement results, multiple sets of probes are set up at equal intervals throughout the production process. The same temperature compensation method used in the previous steps is used to calculate the insulation thickness at the cable test point. 、 Since the production speed of the power cable is constant and the probe setting interval is fixed, the time interval for each probe to detect the same position of the cable is determined, and then the average value of the thickness of the insulation layer detected multiple times is used as the final thickness of the insulation layer at the detection point.

[0089] In summary, the present invention first decomposes the signal into signals of different scales through a multi-level decomposition technique, which helps reduce the impact of noise caused by conductor surface roughness or coolant flow. Based on the decomposed signals, the scale contribution is obtained by analyzing the similarity of the decomposed signals at different scales. The reliability of the analysis at different scales is reflected by the echo similarity characteristics. A matching algorithm is then used to perform similarity analysis in both the frequency domain and the signal near the corresponding decomposed signals of the peak groups. The degree of peak matching of the peak groups is reflected by the similarity of local details. Furthermore, a consistency index is obtained by combining the scale contribution to filter out the echo peak groups, reducing multi-peak interference. By adjusting the time difference of the echo peak groups and the consistency index, a more accurate corrected time difference is obtained. Furthermore, based on the relationship between the coolant temperature and sound velocity detected in real time by the probe, the calibrated sound velocity is adjusted based on the current temperature change, eliminating the impact of sound velocity drift on thickness calculation in dynamic temperature environments and improving adaptability to complex working conditions. The corrected time difference and corrected sound velocity are combined to obtain the insulation layer thickness. Based on the temperature-corrected sound velocity, the present invention uses multi-level decomposition to correct the echo time difference based on the scale and local similarity characteristics of the peaks, effectively improving the accuracy of insulation layer thickness detection.

[0090] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0091] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. A method for online detection of the thickness of the insulation layer of a power cable, characterized in that: The method comprises: Acquire two echo signals at the detection point through the probe, and obtain the peak value in the echo signal; group each two different peak values into a peak group, where the two peak values in the peak group are located in the two echo signals respectively; perform multi-level decomposition on the echo signal to obtain the decomposed signal at each scale; The scale contribution of each scale is obtained based on the correlation between the two decomposed signals at each scale. The two decomposed signals at each scale are matched within the preset neighborhood of each peak group to obtain a moment matching group. The matchability of each peak group at each scale is obtained based on the energy distribution similarity and signal similarity of the matching group of the two decomposed signals at each moment. For each peak group, the consistency index is obtained by combining the scale contribution and matchability of each scale; the echo peak groups are screened based on the consistency index; the current corrected time difference is obtained based on the differences between the corresponding peak times in all echo peak groups and the consistency index of the echo peak groups; The probe detects the changing relationship between the ambient coolant temperature and the sound velocity, and adjusts the calibrated sound velocity based on the temperature deviation of the current detection environment to obtain the current corrected sound velocity. The insulation layer thickness at the current detection point is obtained by combining the current corrected time difference and the corrected sound velocity. The method for obtaining the consistency index includes: For any peak group, at the peak group, the product of the matchability and scale contribution at each scale is used as the matching possibility index of each scale; The sum of the matching possible indices of all scales is used as the numerator, and the sum of the scale contributions of all scales is used as the denominator to obtain the consistency index of the peak group.

2. The method for online detection of the thickness of the insulation layer of a power cable according to claim 1, characterized in that: The method for obtaining the scale contribution includes: For any scale, the DTW distance between the two decomposed signals is calculated at the scale to perform negative correlation mapping and normalization to obtain the scale contribution of the scale.

3. The method for online detection of the thickness of the insulation layer of a power cable according to claim 1, characterized in that: The method for obtaining the time matching group includes: For any peak group, each scale is used as the analysis scale in turn. Within the preset neighborhood range of the peak corresponding to the moment in the peak group, the two decomposed signals under the analysis scale are matched using the DTW algorithm, and each pair of matching moments is regarded as a moment matching group.

4. The method for online detection of the thickness of the insulation layer of a power cable according to claim 1, characterized in that: The method for obtaining the matchability includes: For any peak group, within the preset neighborhood of the peak group, the two decomposed signals obtain frequency domain features at the corresponding time in the matching group at each moment, and the similarity is analyzed to obtain the frequency domain similarity index of the two decomposed signals matching group at each moment; Perform negative correlation mapping on the difference in signal values of the two decomposed signals at the corresponding time in the matching group at each time, and obtain the signal similarity index of the two decomposed signals at each time matching group; The product of the frequency domain similarity index and the signal similarity index of the matching group at each moment is used as the similarity index of the matching group at each moment; Within the preset neighborhood of the peak group, the sum of the similarity indices of the matching groups of the two decomposition signals at all times is used as the numerator, and the sum of the frequency domain similarity indices of the matching groups of the two decomposition signals at all times is used as the denominator to obtain the matching degree of the peak group at the corresponding scale of the two decomposition signals.

5. The method for online detection of the thickness of the insulation layer of a power cable according to claim 4, characterized in that: The method for obtaining the frequency domain similarity index includes: The frequency domain characteristics of the two decomposed signals in time series are obtained by fast Fourier transform; For any moment matching group, the decomposed signal corresponding to each moment in the moment matching group is used as the correlation signal of each moment; the period consisting of each moment in the moment matching group and the two adjacent moments in time sequence is used as the characteristic period of each moment; In the two feature time periods in the matching group at this moment, the correlation of the frequency domain features between the corresponding correlation signals is calculated as the frequency domain similarity index of the matching group of the two decomposed signals at this moment.

6. The method for online detection of the thickness of the insulation layer of a power cable according to claim 1, characterized in that: The method for obtaining the echo peak group includes: The peak group whose consistency index is greater than the preset judgment threshold is regarded as the echo peak group.

7. The method for online detection of the thickness of the insulation layer of a power cable according to claim 1, characterized in that: The method for obtaining the corrected time difference includes: Calculate the difference between the corresponding moments of the two peaks in each echo peak group as the time difference of each echo peak group; The product of the time difference of each echo peak group and the consistency index is used as the weighted time difference index of each echo peak group; The sum of the weighted time difference indices of all echo peak groups is used as the numerator, and the sum of the consistency indices of all echo peak groups is used as the denominator to obtain the current corrected time difference.

8. The method for online detection of the thickness of the insulation layer of a power cable according to claim 1, characterized in that: The method for obtaining the corrected sound speed includes: Obtain the temperature coefficient of sound velocity through preliminary experiments; The product of the difference between the current detection temperature and the calibration temperature and the sound velocity temperature coefficient is used as the sound velocity adjustment degree; The product of the calibrated sound speed and the sound speed adjustment degree is used as the sound speed adjustment value; the sum of the calibrated sound speed and the sound speed adjustment value is used as the corrected sound speed.

9. The method for online detection of the thickness of the insulation layer of a power cable according to claim 1, characterized in that: The method for obtaining the thickness of the insulating layer includes: The product of the current corrected time difference and the corrected sound velocity is divided by 2 to obtain the insulation layer thickness at the current detection point.

Images