Partial discharge detection method and device based on adjustable optical fiber

Abstract

The invention relates to the technical field of optical fiber discharge detection, in particular to a partial discharge detection method and device based on an adjustable optical fiber, and the method comprises the steps: determining a fluctuation difference value through calculating the difference between the mark diameter of each test position and the mark diameter of an adjacent position, and combining a diameter fitting error in a neighborhood; dividing an identification interval according to the dispersion degree of the fluctuation difference value in the neighborhood and the difference of adjacent positions; comparing the diameter distribution characteristics of adjacent intervals and the dispersion degree in the intervals, and combining the intervals to form an identification section; and fitting the mark diameter in each mark section, and establishing an accurate relation model between the diameter of the optical fiber interference ring and the partial discharge position. According to the method, the problem that overfitting is easily generated in the overall fitting process of a traditional method is solved, and the sensitivity and reliability of partial discharge detection are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of optical fiber discharge detection, in particular to a partial discharge detection method and device based on an adjustable optical fiber. BACKGROUND

[0002] The partial discharge detection of the optical fiber is performed by detecting the ultrasonic signal generated by the partial discharge, the ultrasonic signal causes the mechanical vibration of the optical fiber, the mechanical vibration causes the change of the Rayleigh scattering signal in the optical fiber, and the data of the optical fiber interference ring detection changes, so that the position and intensity of the partial discharge are detected. Due to the advantages of high sensitivity and anti-electromagnetic interference of the optical fiber technology, the optical fiber interference ring detection technology has been rapidly applied to the actual partial discharge detection of the power transmission line such as the cable and the line.

[0003] The traditional optical fiber interference ring partial discharge detection method usually adopts a fixed diameter design, and a relationship model between the partial discharge position and the diameter of the interference ring is established by nonlinear fitting. However, due to the significant nonlinear characteristics of the attenuation effect of the ultrasonic wave in the air, the traditional method is prone to overfitting problem in the overall fitting process, which leads to poor generalization ability of the model and cannot accurately reflect the diameter change rule at different distances. This defect directly causes inaccurate adjustment of the diameter of the interference ring, and further reduces the sensitivity and reliability of the partial discharge detection. SUMMARY

[0004] In order to solve the above technical problems, the purpose of the present application is to provide a partial discharge detection method and device based on an adjustable optical fiber, and the technical solution is as follows: In a first aspect, the present application provides a partial discharge detection method based on an adjustable optical fiber, which comprises the following steps: Acquiring the ultrasonic signal collected after simulating the discharge of the optical fiber interference ring at each test position under different diameters; At each test position, comparing the center frequency of the ultrasonic signal under all diameters with a preset frequency, and screening out the marker diameter of each test position from all diameters; determining the diameter characteristic value at each test position based on the difference between the marker diameter of each test position and the previous test position; comparing the diameter characteristic value of each test position with that of the next test position, and combining the fitting error when fitting the diameter characteristic value of all test positions in the neighborhood of each test position to determine the fluctuation difference value of each test position; Based on the discrete degree of the fluctuation difference value of all test positions in the neighborhood, and comparing the fluctuation difference value of each test position with that of the adjacent test position, the identification interval is obtained by dividing all test positions; comparing the distribution of the marker diameter of all test positions between each identification interval and the next identification interval, and combining the discrete degree of the marker diameter of all test positions in each identification interval, the identification section is obtained by merging the identification interval; Fitting all the diameters of the light interference ring to all the diameters of the marks in each mark section, a model of the diameters of the light interference ring and the location of the partial discharge is established.

[0005] Preferably, the screening of the mark diameter of each test position from all the diameters comprises: For each test position, the diameter of the light interference ring with the center frequency of the ultrasonic signal greater than the preset frequency and closest to the preset frequency in all the diameters is taken as the identification diameter of each test position. On the basis of the identification diameter, the value is sequentially increased, and the identification diameter and the result of the increase based thereon are fitted to the corresponding center frequency, and the fitting value at the position corresponding to the preset center frequency in the fitting curve is taken as the mark diameter of each test position.

[0006] Preferably, the diameter characteristic value of each test position is the difference between the mark diameter of each test position and the previous test position divided by the corresponding mark diameter.

[0007] Preferably, the determination method of the fluctuation difference value of each test position is: The diameter characteristic value of each test position is compared with that of the next test position, so as to screen the homomorphic position from all the test positions in the neighborhood of each test position. The difference in quantity between all the test positions in the neighborhood of each test position and all the homomorphic positions is calculated and recorded as the quantity difference, and the product of the fitting error when fitting the diameter characteristic value of all the test positions in the neighborhood of each test position and the quantity difference is taken as the fluctuation difference value of each test position.

[0008] Preferably, the homomorphic position is the test position with the diameter characteristic value less than the preset threshold value among all the test positions in the neighborhood of each test position.

[0009] Preferably, the identification interval obtained by dividing all the test positions comprises: The discrimination degree of each test position is determined based on the dispersion degree of the fluctuation difference value of all the test positions in the neighborhood of each test position and the difference in fluctuation difference value between each test position and its adjacent test position. The discrimination degrees of all the test positions are arranged in ascending order according to the distance from the test position to the optical fiber test ring, and the peak detection algorithm is used to extract the wave peaks and wave troughs in the arrangement result, and all the test positions between each wave peak and its adjacent previous wave trough form an identification interval.

[0010] Preferably, the expression of the discrimination degree of each test position is: ; in the formula, The discrimination degree of test position i is represented. The dispersion degree of the fluctuation difference value of all the test positions in the neighborhood of test position i is represented. , respectively represent the difference between the fluctuation difference value of the test position i and the difference between the fluctuation difference value of the test position i and the fluctuation difference value of the test position i-1; represents a constant greater than 0.

[0011] Preferably, the merging of the identification intervals to obtain the identification section comprises: Based on the difference between the average distribution of the marker diameter of all test positions between each identification interval and the next identification interval, and the dispersion degree of the marker diameter of all test positions in each identification interval, a non-merging coefficient between each identification interval and the next identification interval is determined. If the non-merging coefficient between the mth identification interval and the next identification interval is less than 0, the mth identification interval and the next identification interval are merged into one identification section, otherwise, no merging is performed, and all identification sections are obtained by traversing all identification intervals.

[0012] Preferably, the expression of the non-merging coefficient between each identification interval and the next identification interval is: ; In the formula, represents the non-merging coefficient between the interval j and the next identification interval; represents the difference between the average value of the marker diameter of all test positions between the identification interval j and the next identification interval; represents the dispersion degree of the marker diameter of all test positions in the identification interval j. In the second aspect, the embodiments of the present application also provide a local discharge detection device based on an adjustable optical fiber, which comprises a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the steps of the local discharge detection method based on the adjustable optical fiber according to any one of the above embodiments when executing the computer program.

[0013] The present application has at least the following beneficial effects: The application realizes the non-linear degree quantitative evaluation of the change trend of the fiber interference ring diameter with the local discharge distance by combining the marker diameter screening, diameter characteristic value calculation and fluctuation difference value analysis, effectively reflects the complexity and propagation stability of ultrasonic wave attenuation, provides a reliable basis for the optimization selection of the fiber ring diameter in the local discharge detection, and improves the precision and adaptability of the local discharge detection; further, the application realizes the accurate segmentation of the test position by combining the dispersion degree of the fluctuation difference value and the adjacent difference calculation distinguish degree, and intelligently merges the marker interval by using the non-merging coefficient, and finally obtains the marker segment which can accurately reflect the non-linear gradual change characteristics of the ultrasonic wave attenuation, effectively improves the adaptability and accuracy of the fiber interference ring diameter selection and the local discharge detection; further, the application effectively overcomes the overall fitting overfitting problem caused by the non-linearity of ultrasonic wave attenuation by segment fitting of the marker diameter in the marker segment, improves the accuracy and generalization ability of the relationship model between the fiber interference ring diameter and the local discharge position, and significantly improves the accuracy of the local discharge detection, thereby providing more accurate diameter adjustment basis for the local discharge detection, and improving the sensitivity and reliability of the local discharge detection. BRIEF DESCRIPTION OF DRAWINGS

[0014] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0015] Figure 1 The step flow chart of the local discharge detection method based on the adjustable optical fiber provided by an embodiment of the present application is shown in the figure. Figure 2 The homomorphism position screening process flow chart provided by an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0016] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the specific implementation, structure, features and effects of the local discharge detection method and device based on the adjustable optical fiber according to the present application are described in detail as follows by combining the drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0017] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0018] The application provides a partial discharge detection method and device based on an adjustable optical fiber.

[0019] Please refer to Figure 1 which shows a step flow chart of the partial discharge detection method based on the adjustable optical fiber according to an embodiment of the application, and the method comprises the following steps. Step S1: Obtain the ultrasonic signals collected after simulating the discharge at each test position of the optical fiber interference ring with different diameters.

[0020] The position of the optical fiber interference ring is fixed, and the partial discharge is simulated at each test position with different distances in sequence, the diameter of the optical fiber interference ring is changed for multiple times to measure the ultrasonic signals at each test position, and the detailed process is as follows. In this embodiment, the position of the optical fiber interference ring is fixed, the partial discharge is simulated at a distance L from the optical fiber interference ring, the diameter of the optical fiber interference ring is changed for multiple times to obtain the ultrasonic signals at the test position with different diameters, further, the partial discharge is simulated at a distance 2L from the optical fiber interference ring, and the ultrasonic signals at the test position with different diameters of the optical fiber interference ring are obtained at the test position, the collection time is set as t and the collection frequency is set as f, and the above process is repeated, and N times of tests are performed, that is, the ultrasonic signals at N test positions with different distances from the optical fiber interference ring are obtained, wherein the adjustment range of the diameter of the optical fiber interference ring is [5, 30], the unit of the diameter is cm, the ultrasonic signals at each test position are tested from the initial value 5 cm, and the diameter is increased by 1 cm on the basis of the previous diameter, and the value of N is 100 in this embodiment, in actual application, as other embodiments, the implementer can set the value according to the specific condition, and the implementer can set the test position, which is not specially limited in this embodiment.

[0021] It is to be supplemented that the values of the collection time and the collection frequency are artificially set, the value of the collection time t is 0.01 s in this embodiment, and the value of the collection frequency f is 100 kHz, in actual application, as other embodiments, the implementer can set the values according to the specific condition, and this embodiment is not specially limited.

[0022] Further, in order to remove the interference of the noise on the ultrasonic signals, the wavelet threshold denoising algorithm is used to perform the denoising processing on the ultrasonic signals in this embodiment, in actual application, the implementer can use other denoising algorithms according to the specific condition, and the selection of the denoising algorithm is not specially limited in this embodiment.

[0023] The wavelet threshold denoising algorithm is a known technology, and the specific process of denoising the ultrasonic signals by using the wavelet threshold denoising algorithm is not repeated.

[0024] Step S2: comparing the center frequency of the ultrasonic signal under all diameters with the preset frequency at each test position, screening the marker diameter of each test position from all diameters; determining the diameter characteristic value at each test position based on the difference of the marker diameter between each test position and the previous test position; comparing the diameter characteristic value of each test position with that of the next test position, and determining the fluctuation difference value of each test position in combination with the fitting error when fitting the diameter characteristic values of all test positions in the neighborhood of each test position.

[0025] Since the ultrasonic signal at the partial discharge is a high-frequency signal, the signal frequency is distributed in 6-40 kHz, and the center frequency is near 7 kHz. Therefore, the center frequency of the ultrasonic signal is obtained by using the Fourier transform algorithm, and further, the difference between the center frequency of the ultrasonic signal under all diameters of the fiber interference ring and the preset frequency is compared at each test position to screen the marker diameter of each test position from all diameters. Specifically, In this embodiment, for each test position, the diameter of the fiber interference ring, in which the center frequency of the ultrasonic signal is greater than the preset frequency and closest to the preset frequency, is taken as the marker diameter of each test position.

[0026] It should be noted that the value of the preset frequency is artificially set, and in this embodiment, the value of the preset frequency is 6 kHz.

[0027] The fiber interference ring is a device manufactured by Sagnac effect. The Sagnac effect is proportional to the area of the ring and the rotation angle of the ring, and inversely proportional to the wavelength, and the center frequency is also inversely proportional to the wavelength. Therefore, when the diameter of the fiber interference ring increases, a higher frequency can be detected. Therefore, as the diameter of the fiber interference ring increases, the detected center frequency will also increase. Due to the propagation of ultrasonic waves in air or other substances, the substances interfere with the transmission of ultrasonic waves to different degrees of attenuation, so that the diameter of the fiber interference ring and the detected center frequency are not a linear relationship, but a nonlinear relationship.

[0028] Therefore, the embodiment further increases the value on the basis of the marker diameter at each test position, and fits the marker diameter and the result increased on the basis thereof with the corresponding center frequency, wherein the center frequency is the independent variable in the fitting process, and the marker diameter and the result increased on the basis thereof are the dependent variable in the fitting process. The fitting value at the position corresponding to the preset center frequency in the fitting curve is taken as the marker diameter of each test position.

[0029] It should be noted that the value of the preset center frequency is artificially set, and in this embodiment, the value of the preset center frequency is 7 kHz.

[0030] It should be noted that there are many commonly used fitting algorithms, and in the present embodiment, the nonlinear least squares method is used to fit the diameter and center frequency. In actual application, as an alternative embodiment, the implementer can also use other fitting algorithms such as polynomial function fitting method according to the specific circumstances. The selection of the fitting algorithm is not particularly limited in the present embodiment.

[0031] Among them, the process of obtaining the center frequency of the signal by using the Fourier transform algorithm, and the process of fitting by using the nonlinear least squares method are all known technologies, and will not be repeated here.

[0032] It should be noted that, except for special instructions, the nonlinear least squares method is used in the present embodiment for all content related to fitting.

[0033] Further, in the process of propagation, the ultrasonic wave will attenuate with distance, and when the distance is relatively short, the attenuation has a certain linear relationship, but as the distance increases, the attenuation of the ultrasonic wave will have the characteristics of intensification, so that the attenuation of the ultrasonic wave has a nonlinear relationship. Therefore, at different distance stages, the degree of ultrasonic wave attenuation is different, so in the present embodiment, the difference between the characteristic diameter of each test position and the previous test position is used to determine the diameter characteristic value at each test position, to indirectly reflect the trend of the ultrasonic signal at the same test position. Specifically: In the present embodiment, the difference between the characteristic diameter of each test position and the previous test position is used as the diameter characteristic value at each test position, which is used to represent the rate of change of the diameter of the fiber interference ring with the distance of partial discharge. If the diameter characteristic value of the current test position is larger, it means that the diameter increases rapidly with the increase of distance, indicating that the ultrasonic wave attenuates very sharply, which may be due to the propagation medium, obstacles or distance response causing the signal to decay rapidly. On the contrary, if the diameter characteristic value of the current test position is smaller, it means that the diameter increases slowly with the increase of distance, indicating that the ultrasonic wave attenuates relatively gently, and the signal propagation is relatively stable, which may be due to the uniform propagation medium, fewer obstacles or better linear distance response, resulting in slower signal decay.

[0034] Further, in the present embodiment, the difference between the diameter characteristic value of each test position and the next test position is used to screen the homomorphism position from all test positions in the neighborhood of each test position. The process flow chart of the homomorphism position screening provided by the present embodiment is shown in Figure 2 The specific process of homomorphism position screening is as follows: The neighborhood W is divided with each test position as the center, and the radius of the neighborhood W is 2, that is, the test positions corresponding to the previous two tests and the test positions corresponding to the next two tests before and after each test position, and the test positions constitute the neighborhood of each test position. The radius of the neighborhood W is artificially set, and in actual application, as other embodiments, the implementer can also set it by himself according to the specific situation, and the present embodiment does not make special limitation.

[0035] Further, in the present embodiment, the test positions with the diameter characteristic value less than the preset threshold value among all the test positions in the neighborhood of each test position are regarded as the homomorphic positions of each test position, which are used to represent the consistency of the diameter change trend of the fiber interference ring at the adjacent test positions. The number of the homomorphic positions indicates that the more consistent the diameter change trend of the fiber interference ring at the adjacent test positions is.

[0036] It should be noted that the value of the preset threshold is artificially set, and the value of the preset threshold in the present embodiment is 0.05. In actual application, as other embodiments, the implementer can also set it by himself according to the specific situation, and the present embodiment does not make special limitation.

[0037] Further, the present embodiment determines the fluctuation difference value of each test position based on the fitting error when fitting the diameter characteristic values of all the test positions in the neighborhood of each test position, and the difference in number between all the test positions in the neighborhood and all the homomorphic positions, to measure the nonlinearity degree of the diameter change trend of the fiber interference ring. Specifically, As a specific embodiment, in the present embodiment, the difference in number between all the test positions in the neighborhood of each test position and all the homomorphic positions is calculated and recorded as the number difference. The product of the fitting error when fitting the diameter characteristic values of all the test positions in the neighborhood of each test position and the number difference is taken as the fluctuation difference value of each test position.

[0038] It should be noted that there are many methods for measuring the difference between data. In the present embodiment, the absolute value of the difference in number between all the test positions in the neighborhood of test position i and all the homomorphic positions is taken as the difference in number between all the test positions in the neighborhood of test position i and all the homomorphic positions. In actual application, as other embodiments, the implementer can also use the square or ratio of the difference value and other methods for measuring the difference between data. The present embodiment does not make special limitation on the selection of the method for measuring the difference between data.

[0039] It should be further noted that, except for special description, the present embodiment uses the method of taking the absolute value of the difference value for all the contents related to the difference between data.

[0040] The calculation method of the fitting error is a known technology, and the specific calculation process will not be repeated here.

[0041] According to the fluctuation difference value of each test position, it can be understood that the fluctuation difference value integrates the local fitting error and the number of homomorphic positions, and is used to measure the nonlinearity degree of the diameter change trend of the fiber interference ring; if the fitting error of the diameter characteristic value fitting of all test positions in the neighborhood of the current test position is larger, the difference between the diameter characteristic value and the fitting value is larger, which indicates that the nonlinearity of the diameter change rate of the fiber interference ring is stronger, and the diameter change is irregular, which means that the ultrasonic wave attenuation is more complex, and therefore the corresponding fluctuation difference value is larger; at the same time, if the difference in the number of all test positions in the neighborhood of the current test position and all homomorphic positions is larger, that is, the number difference is larger, the number of homomorphic positions is smaller, the diameter trend change of the fiber interference ring is larger, which indicates that the nonlinearity of the ultrasonic wave attenuation is more significant, and the corresponding fluctuation difference value is larger. On the contrary, if the fitting error in the neighborhood of the current test position is smaller, the difference between the diameter characteristic value and the fitting value is smaller, which indicates that the nonlinearity of the diameter change rate of the fiber interference ring is weaker, and the diameter change is more regular, which means that the ultrasonic wave attenuation is relatively stable, and therefore the corresponding fluctuation difference value is smaller; at the same time, if the difference in the number of all test positions in the neighborhood of the current test position and all homomorphic positions is smaller, that is, the number difference is smaller, the number of homomorphic positions is larger, the diameter trend change of the fiber interference ring is more consistent, which indicates that the nonlinearity of the ultrasonic wave attenuation is less significant, and the fluctuation difference value is also smaller accordingly.

[0042] So far, by combining the flag diameter screening, diameter characteristic value calculation and fluctuation difference value analysis, the nonlinearity degree of the diameter change trend of the fiber interference ring with the local discharge distance is quantitatively evaluated, the complexity and propagation stability of the ultrasonic wave attenuation are effectively reflected, a reliable basis is provided for the optimization selection of the fiber ring diameter in the partial discharge detection, and the precision and adaptability of the partial discharge detection are improved.

[0043] Step S3: based on the dispersion degree of the fluctuation difference values of all test positions in the neighborhood, and comparing the fluctuation difference values of each test position and its adjacent test position, the identification intervals are divided for all test positions; comparing the distribution of the flag diameters of all test positions between each identification interval and the next identification interval, and combining the dispersion degree of the flag diameters of all test positions in each identification interval, the identification intervals are merged to obtain the identification section.

[0044] Since there is still a certain subtle difference in the change of data in the same change stage, but the trend of ultrasonic wave attenuation is unchanged with the increase of distance, therefore, the nonlinearity of data in the same stage has a certain gradual change characteristic, that is, the data has the same trend change, but the change amount of different stages is different, that is, the gradient of the gradual change is different. With the increase of distance, the diameter of the fiber interference ring becomes larger, because a larger diameter can detect the center frequency of the partial discharge fault, and ensure that the partial discharge fault occurs in the line.

[0045] Based on the above analysis, the embodiment divides all test positions into identification intervals based on the dispersion degree of fluctuation difference values of all test positions in the neighborhood of each test position and compares the fluctuation difference values between each test position and its adjacent test positions; and the embodiment merges the identification intervals to obtain identification segments by comparing the distribution of all test position mark diameters between each identification interval and the next identification interval and combining the dispersion degree of all test position mark diameters in each identification interval, specifically as follows: In the embodiment, first, the discrimination degree at each test position is determined based on the dispersion degree of fluctuation difference values of all test positions in the neighborhood of each test position and the difference between the fluctuation difference values of each test position and its adjacent test positions, so as to represent whether the test position can be used as a demarcation point of the ultrasonic wave attenuation trend, i.e., whether the diameter change rate has a significant turning point, specifically as follows: As an implementation manner, in the embodiment, the discrimination degree at the test position i is The expression of the discrimination degree is as follows: ; in the expression, σi represents the dispersion degree of fluctuation difference values of all test positions in the neighborhood of the test position i; Δi+1 represents the difference between the fluctuation difference values of the test position i and the next test position, and Δi-1 represents the difference between the fluctuation difference values of the test position i and the previous test position; , is a preset constant greater than 0, which is used to prevent the denominator from being 0, The value of α is artificially set, and in the embodiment, the value of α is 0.01. On the premise of ensuring that the denominator is not 0 and not excessively affecting the calculation result, the implementer can also set it according to the specific circumstances, and the embodiment does not make special limitations.

[0046] It should be noted that there are many methods for measuring the dispersion degree of a group of data. In the embodiment, the standard deviation of the fluctuation difference values of all test positions in the neighborhood of the test position i is used as the dispersion degree of the fluctuation difference values of all test positions in the neighborhood of the test position i. In actual application, as other implementation manners, the implementer can also use other methods for measuring the dispersion degree of data, such as variance or coefficient of variation. The embodiment does not make special limitations on the selection of the method for measuring the dispersion degree of data.

[0047] It should be further noted that, except for special descriptions, the embodiment uses the calculation method of the standard deviation for all contents related to the measurement of the dispersion degree of data.

[0048] ​​According to the discriminant of each test position, it can be understood that the discriminant represents whether the test position can be used as a demarcation point of the ultrasonic attenuation trend, that is, whether the diameter change rate has a significant turning point. The greater the dispersion degree of the fluctuation difference values of all test positions in the neighborhood of the test position i, the more significant the diameter change trend difference between the test position i and the test positions before and after it, the more inconsistent the fluctuation, and the more suitable it is as a segmentation node. The greater the corresponding discriminant is; at the same time, the greater the difference between the fluctuation difference values of the test position i and the test position after it than the difference between the fluctuation difference values of the test position i and the test position before it, that is, the greater, the greater the difference between the diameter characteristic values of the current test position and the test position after it than the difference between the diameter characteristic values of the current test position and the test position before it, indicating that the diameter change trend before and after the current test position has a significant turning point, and it is suitable to be used as a segmentation node. The corresponding discriminant is greater; On the contrary, the smaller the dispersion degree of the fluctuation difference values of all test positions in the neighborhood of the test position i, the more consistent the diameter change trend between the test position i and the test positions before and after it, and the more stable the fluctuation. It is not suitable to be used as a segmentation node. The smaller the corresponding discriminant is; at the same time, the smaller the difference between the fluctuation difference values of the test position i and the test position after it than the difference between the fluctuation difference values of the test position i and the test position before it, that is, the smaller, the difference between the diameter characteristic values of the current test position and the test position after it is close to or smaller than the difference between the diameter characteristic values of the current test position and the test position before it, indicating that the diameter change trend before and after the current test position has no obvious turning point, and it is not suitable to be used as a segmentation node. The corresponding discriminant is smaller.

[0049] Further, based on the discriminant, the embodiment divides all test positions to obtain an identification interval, which is used to represent a plurality of continuous intervals obtained by dividing all test positions according to the node discriminant. The diameter change trend inside each identification interval is relatively consistent, and there is a significant difference between different identification intervals. The specific determination process of the identification interval is as follows: In the embodiment, the discriminants of all test positions are arranged in ascending order according to the distance from the test position to the optical fiber test ring. The automatic multi-scale peak detection algorithm is used to extract the peaks and troughs in the arrangement result. All test positions between each peak and the adjacent previous trough form an identification interval.

[0050] Among them, there are many commonly used peak detection algorithms. In the embodiment, the automatic multi-scale peak detection algorithm is used. In actual application, as other implementation manners, the implementer can also use other peak detection algorithms according to the specific situation, which is not specially limited in the embodiment.

[0051] Among them, the automatic multi-scale peak detection algorithm is a known technology, and the specific process of detecting the peaks and troughs will not be repeated.

[0052] Further, the embodiment determines a non-merging coefficient between each identification interval and the next identification interval based on the difference between the average distribution of the sign diameters of all test positions between the identification interval and the next identification interval and the dispersion degree of the sign diameters of all test positions in the identification interval, so as to obtain an identification section by merging the identification intervals, specifically: In the embodiment, the expression of the non-merging coefficient between the identification interval j and the next identification interval is: ; in the expression, Δj represents the difference between the average distribution of the sign diameters of all test positions between the identification interval j and the next identification interval; and σj represents the dispersion degree of the sign diameters of all test positions in the identification interval j.

[0053] According to the non-merging coefficient between each identification interval and the next identification interval, it can be understood that the non-merging coefficient reflects the similarity of the diameter trends between the adjacent two identification intervals; if the difference between the average distribution of the sign diameters of all test positions between the identification interval j and the next identification interval is larger, it indicates that the diameter trends of the identification interval j and the next identification interval are significantly different, which is not suitable for merging, and therefore, the corresponding non-merging coefficient is larger; at the same time, if the dispersion degree of the sign diameters of all test positions in the identification interval j is smaller, it indicates that the internal data fluctuation of the identification interval is small, the trend is stable, and it is likely to be affected by noise, and tends to remain independent, and is not merged with the adjacent interval, and therefore, the corresponding non-merging coefficient is larger. On the contrary, if the difference between the average distribution of the sign diameters of all test positions between the identification interval j and the next identification interval is smaller, it indicates that the diameter trends of the identification interval j and the next identification interval are similar, which is suitable for merging, and therefore, the corresponding non-merging coefficient is smaller; at the same time, if the dispersion degree of the sign diameters of all test positions in the identification interval j is larger, it indicates that the internal data fluctuation of the identification interval is large, the trend is unstable, and it is likely to be significantly affected by noise, and tends to be merged with the adjacent interval to reduce the noise interference, and therefore, the corresponding non-merging coefficient is smaller.

[0054] Further, the embodiment merges the identification intervals based on the non-merging coefficient to obtain an identification section, which is used to determine the relationship between the diameter and the partial discharge position, specifically: In the embodiment, if the non-merging coefficient between the mth identification interval and the next identification interval is smaller than 0, the mth identification interval and the next identification interval are merged into one identification section, otherwise, no merging is performed, and all identification sections are obtained by traversing all identification intervals.

[0055] ​​​The final diameter segment obtained after non-merging coefficient evaluation and merging processing of the identification segment, each of which represents the highly consistent diameter change trend inside the segment, avoids overfitting problem caused by overall fitting through segment fitting, and improves the generalization ability of the model, so as to more accurately adjust the fiber ring diameter in partial discharge detection.

[0056] So far, the embodiment realizes accurate segmentation of the test position by combining the discrete degree of fluctuation difference value with the adjacent difference calculation discriminability, and intelligently merges the identification interval by using the non-merging coefficient, and finally obtains the identification segment that can accurately reflect the nonlinear gradual change characteristics of ultrasonic wave attenuation, effectively improving the adaptability and accuracy of fiber interference ring diameter selection and partial discharge detection.

[0057] Step S4: fitting all the marker diameters in each identification segment, and establishing a relationship model between the fiber interference ring diameter and the partial discharge position.

[0058] Based on the division of the identification segment completed in step S3, segment fitting is performed to avoid overfitting problem caused by overall fitting, improve the generalization ability of the model, and more accurately adjust the fiber ring diameter in partial discharge detection. The specific fitting process is as follows: The marker diameters of all test positions in each identification segment are taken as the input of the polynomial function fitting method. Since the attenuation of ultrasonic waves conforms to the exponential decay model, a larger interference ring diameter is required during detection. Therefore, in the present embodiment, the independent variable of the polynomial fitting is the exponential function value with the natural constant as the base number and the interference ring diameter as the exponent, and the dependent variable is the distance between the fiber interference ring and the test position. The diameter-partial discharge position fitting curve of each identification segment is output, which is used to determine the relationship between the fiber interference ring diameter and the partial discharge position, so as to determine the position of the partial discharge based on the fitting curve model.

[0059] Among them, the polynomial function fitting algorithm is a known technology, and the specific process of data fitting using it will not be repeated here.

[0060] So far, the embodiment effectively overcomes the overfitting problem caused by the nonlinear ultrasonic wave attenuation through segment fitting of the marker diameters in the identification segment, improves the accuracy and generalization ability of the relationship model between the fiber interference ring diameter and the partial discharge position, and significantly improves the accuracy of the partial discharge detection, thereby providing more accurate diameter adjustment basis for the partial discharge detection and improving the sensitivity and reliability of the partial discharge detection.

[0061] Based on the same inventive concept as the above method, the embodiments of the present application also provide a partial discharge detection device based on an adjustable optical fiber, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor executes the computer program to implement the steps of any one of the above methods of the partial discharge detection method based on the adjustable optical fiber.

[0062] It should be noted that the above-mentioned embodiments of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments. And the above describes a specific embodiment of the present application. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are also possible or may be advantageous.

[0063] Each of the embodiments in the specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the difference from other embodiments.

[0064] The above only describes the preferred embodiments of the present application, and does not limit the present application. Any modification, equivalent replacement, improvement, etc. made within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A partial discharge detection method based on a tunable optical fiber, characterized by, The method comprises the following steps: Obtaining the ultrasonic signals collected after simulating discharge at each test position of the optical fiber interference ring with different diameters; At each test position, comparing the center frequencies of the ultrasonic signals with different diameters with a preset frequency, and screening out the marker diameter of each test position from all diameters; determining the diameter characteristic value of each test position based on the difference between the marker diameter of each test position and the marker diameter of the previous test position; comparing the diameter characteristic values of each test position and the next test position, and combining the fitting error when fitting the diameter characteristic values of all test positions in the neighborhood of each test position to determine the fluctuation difference value of each test position; Based on the discrete degree of the fluctuation difference values of all test positions in the neighborhood, and comparing the fluctuation difference values of each test position and its adjacent test position, the identification interval is obtained by dividing all test positions; comparing the distribution of the marker diameters of all test positions between each identification interval and the next identification interval, and combining the discrete degree of the marker diameters of all test positions in each identification interval, the identification section is obtained by merging the identification intervals. Fitting all marker diameters in each identification section to establish a relationship model between the diameter of the optical fiber interference ring and the partial discharge position.

2. The tunable fiber-based partial discharge detection method of claim 1, wherein, The marker diameter of each test position is screened out from all diameters, which comprises: For each test position, the diameter of the ultrasonic signal center frequency greater than the preset frequency and closest to the preset frequency in all diameters of the light interference ring is taken as the identification diameter of each test position; On the basis of the identification diameter, the value is increased in turn, and the identification diameter and the result increased on the basis thereof are fitted with the corresponding center frequency, and the fitting value at the position corresponding to the preset center frequency in the fitting curve is taken as the marker diameter of each test position.

3. The tunable fiber-based partial discharge detection method of claim 1, wherein, The diameter characteristic value of each test position is the difference between the marker diameter of each test position and the marker diameter of the previous test position.

4. The tunable fiber-based partial discharge detection method of claim 1, wherein, The determination method of the fluctuation difference value of each test position is: Comparing the diameter characteristic values of each test position and the next test position to screen out the same state position from all test positions in the neighborhood of each test position; Calculating the difference in quantity between all test positions in the neighborhood of each test position and all same state positions, denoted as quantity difference, and taking the product of the fitting error when fitting the diameter characteristic values of all test positions in the neighborhood of each test position and the quantity difference as the fluctuation difference value of each test position.

5. The tunable fiber-based partial discharge detection method of claim 4, wherein, The same state position is the test position with a diameter characteristic value less than a preset threshold among all test positions in the neighborhood of each test position.

6. The tunable fiber-based partial discharge detection method of claim 1, wherein, The identification interval is obtained by dividing all test positions, which comprises: Determining the discrimination degree of each test position based on the discrete degree of the fluctuation difference values of all test positions in the neighborhood of each test position, and the difference between the fluctuation difference values of each test position and its adjacent test position; Arranging the discrimination degrees of all test positions in ascending order according to the distance from the test position to the optical fiber test ring, extracting the peaks and troughs in the arrangement result by using the peak detection algorithm, and grouping all test positions between each peak and its adjacent previous trough to form an identification interval.

7. The tunable fiber-based partial discharge detection method of claim 6, wherein, The expression of the distinguishability at each test position is: ; wherein, represents the distinguishability at test position i; represents the dispersion of fluctuation difference values of all test positions within the neighborhood of test position i; , respectively represent the difference of fluctuation difference values between test position i and its next test position, and between test position i and its previous test position; represents a preset constant greater than 0.

8. The tunable fiber-based partial discharge detection method of claim 1, wherein, The identification section is obtained by merging the identification intervals, which comprises: determining a non-merging coefficient between each identification interval and the next identification interval based on the difference between the average distribution of the sign diameters of all test positions between each identification interval and the next identification interval and the dispersion degree of the sign diameters of all test positions in each identification interval; if the non-merging coefficient between the mth identification interval and the next identification interval is less than 0, merging the mth identification interval and the next identification interval into one identification section, otherwise, not merging, and traversing all identification intervals to obtain all identification sections.

9. The tunable fiber-based partial discharge detection method of claim 8, wherein, The expression of the non-merging coefficient between the identification interval and the next identification interval is: ; wherein represents the non-merging coefficient between the interval j and the next identification interval; represents the difference of the mean of the flag diameters of all test positions between the identification interval j and the next identification interval; represents the dispersion of the flag diameters of all test positions within the identification interval j.

10. A partial discharge detection apparatus based on a tunable optical fiber, comprising a memory, a processor and a computer program stored in the memory and running on the processor, characterized in that, The processor implements the steps of the optical fiber based partial discharge detection method according to any one of claims 1-9 when executing the computer program.

Images