Pipeline acoustic emission signal noise reduction method

Abstract

The invention relates to the technical field of noise reduction processing, in particular to a pipeline acoustic emission signal noise reduction method. The invention provides a comprehensive noise reduction strategy, the parameters of the VMD algorithm are optimized through the ISSA algorithm, and then the signals are decomposed into IMF components for further processing. According to the method, the advantages of the intelligent optimization algorithm and the signal processing technology are combined, the noise in the pipeline acoustic emission signal can be effectively removed, the signal-to-noise ratio and definition of the signal are improved, and more accurate data support is provided for subsequent signal analysis, fault detection, positioning and the like.

Description

Technical Field

[0001] The invention relates to the technical field of noise reduction processing, and in particular to a pipeline acoustic emission signal noise reduction method. Background Art

[0002] There are all kinds of pipelines for transporting water, gas and oil buried in the city. Once a pipeline leaks, it will inevitably cause serious ecological damage and huge economic and property losses. The expansion of crack defects in oil and gas pipelines will lead to their fracture and damage. In this process, acoustic emission signals will be generated, which need to be detected by highly sensitive electronic instruments (such as acoustic emission detectors).

[0003] Normally, the signals monitored by the detector contain not only the target signal but also the factory environment noise, vibration noise from the operation of adjacent equipment, etc. These noises will seriously affect the accuracy of the diagnosis results. Therefore, it is necessary to pre-process the acoustic emission signal to denoise it in order to improve the accuracy of subsequent diagnosis results. Summary of the invention

[0004] In order to avoid and overcome the technical problems existing in the prior art, the present invention provides a pipeline acoustic emission signal noise reduction method. The present invention can effectively reduce the noise in the acoustic emission signal.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A pipeline acoustic emission signal noise reduction method, comprising the following noise reduction steps:

[0007] S1. Use the ISSA algorithm to obtain the penalty factor α and the number of modal components K that meet the value requirements in the VMD algorithm within the value range;

[0008] S2, inputting the penalty factor ɑ and the number of modal components K into the VMD algorithm, and decomposing the original pipeline acoustic emission signal into IMF components through the VMD algorithm;

[0009] S3, dividing the IMF component into an effective signal component and a noise signal component;

[0010] S4, performing wavelet threshold denoising processing on the noise signal component to obtain the denoised noise signal component;

[0011] S5. Combining and reconstructing the effective signal component and the denoised noise signal component, and obtaining the denoised original pipeline acoustic emission signal through combined reconstruction.

[0012] As a further solution of the present invention: the specific content of step S1 is as follows:

[0013] S11, setting the value range of the penalty factor ɑ and the value range of the modal component number K; combining the value ranges of the two to form a two-dimensional value space;

[0014] S12, initializing the parameters in the ISSA algorithm, setting the size of the sparrow population, and dividing the sparrow population into discoverers, followers, and guards;

[0015] S13, iteratively updating the positions of the discoverer, follower and sentinel in the two-dimensional value space using a position update formula;

[0016] S14. When the fitness value of the discoverer, follower or sentinel is less than the fitness threshold or reaches the maximum number of iterations, stop the iteration and output the penalty factor α and the number of modal components K at this time. The penalty factor α and the number of modal components K are the penalty factor α and the number of modal components K that meet the value requirements.

[0017] As a further solution of the present invention: the location update formula of the discoverer is as follows:

[0018]

[0019] In the formula, Indicates the position of the i-th sparrow in the j-th dimension at the t+1th iteration, i∈[1,I], I represents the total number of sparrows in the sparrow population, j=1,2; Indicates the position of the i-th sparrow in the j-th dimension at the t-th iteration; Indicates the position of the sparrow with the smallest fitness value among all sparrows in the tth iteration; r'1 indicates the initial optimization step length; r'1' indicates the later optimization step length; r2 indicates the moving distance parameter of the sparrow, r2∈[0,,2π]; r3 indicates the random parameter, r3∈[0,,2]; sin indicates the sine function; cos indicates the cosine function; Ra indicates the alarm value; TH indicates the safety threshold; if indicates the conditional symbol;

[0020] The calculation formula for the initial optimal step length is as follows:

[0021]

[0022] In the formula, β represents the correction coefficient; η represents the adjustment coefficient; iter max represents the maximum number of iterations; t represents the current number of iterations;

[0023] The calculation formula for the later step length is as follows:

[0024]

[0025] Wherein, e represents a natural constant.

[0026] As a further solution of the present invention: the position update formula of the follower is as follows:

[0027]

[0028] In the formula, Indicates the position of the sparrow with the smallest fitness value among all sparrows in the t+1th iteration; G u represents random numbers that follow a normal distribution; Indicates the position of the sparrow with the largest fitness value among all sparrows in the tth iteration; C + represents a random matrix; V represents a matrix of all 1s.

[0029] As a further solution of the present invention: the position update formula of the alerter is as follows:

[0030]

[0031]

[0032]

[0033] In the formula, represents random numbers that follow a normal distribution; represents the fitness value of the i-th sparrow in the t-th iteration; F g Indicates the fitness value of the best sparrow in the current global situation; F w represents the fitness value of the worst sparrow in the current global situation; τ represents a random number that obeys a normal distribution; μ represents a random non-zero constant; Levy(θ) represents a flight calculation function; θ represents a learning factor; B1 and B2 both represent random constants; ε represents a random constant; Γ represents a gamma function.

[0034] As a further solution of the present invention: the specific steps of step S2 are as follows:

[0035] S21, obtaining an original pipeline acoustic emission signal, and using a VMD algorithm to decompose the original pipeline acoustic emission signal into K modal components;

[0036] S22, setting the number of iterations, initializing the modal components, center frequencies and Lagrangian factors;

[0037] S23, start iteration, and update the modal components, center frequency and Lagrangian factors; the update formula of the modal components is expressed as follows:

[0038]

[0039] In the formula, represents the value of the kth modal component with respect to the center frequency ω at the n+1th iteration; express Fourier transform of ; f(ω) represents the photovoltaic power output function about the center frequency ω, represents the Fourier transform of f(ω); represents the value of the rth modal component at the nth iteration with respect to the center frequency ω; express Fourier transform of n (ω) represents the value of the Lagrangian factor at the nth iteration with respect to the center frequency ω; Represents λ n Fourier transform of (ω); represents the value of the center frequency ω of the kth modal component at the nth iteration;

[0040] The update formula of the center frequency is as follows:

[0041]

[0042] In the formula, represents the value of the center frequency ω of the kth modal component at the n+1th iteration; dω represents the differential of the center frequency ω;

[0043] The update formula of the Lagrangian factor is expressed as follows:

[0044]

[0045] λ n+1 (ω) represents the value of the Lagrangian factor with respect to the center frequency ω at the n+1th iteration; Represents λ n+1 Fourier transform of (ω); Ω represents the limiting factor;

[0046] S24, establish a judgment criterion. If the judgment criterion is met, the iteration is terminated. Otherwise, the iteration is repeated according to the content of step S23 until the maximum number of iterations is reached or the judgment criterion is met. The judgment criterion is expressed as follows:

[0047]

[0048] In the formula, express The second-order norm of ; express The quadratic value of express The second-order norm of ; express The quadratic value of; E represents the judgment accuracy;

[0049] S25. Output each IMF component.

[0050] As a further solution of the present invention: the specific content of step S3 is as follows:

[0051] S31, calculating the permutation entropy of various IMF components using the permutation entropy calculation formula;

[0052] S32, marking the IMF components whose permutation entropy is greater than a set threshold as noise IMF components, and marking the remaining IMF components as valid signal IMF components.

[0053] As a further solution of the present invention: the calculation formula of the permutation entropy is as follows:

[0054] H m =ln(P m log(P m ));

[0055] Where P m represents the probability of occurrence of the mth IMF component; ln represents the logarithmic function with a natural constant as the base; H m represents the permutation entropy of the mth IMF component.

[0056] As a further solution of the present invention: the specific content of step S4 is as follows:

[0057] S41, performing wavelet transform on the identified noise IMF component to obtain its wavelet coefficient, and recording the wavelet coefficient as the original wavelet coefficient;

[0058] S42, applying an improved wavelet threshold function to perform threshold processing on the original wavelet coefficients to obtain denoised wavelet coefficients, and recording the wavelet coefficients as improved wavelet coefficients;

[0059] S43, performing inverse wavelet transform on the improved wavelet coefficients to obtain an IMF component after the noise IMF component is denoised, and the IMF component is recorded as the denoised IMF component.

[0060] As a further solution of the present invention: the improved wavelet threshold function is specifically expressed as follows:

[0061]

[0062]

[0063] Where W z,c represents the original wavelet coefficient, z represents the scale parameter of wavelet decomposition, and c represents the position parameter of wavelet decomposition; Represents the original wavelet coefficient W z,cThe improved wavelet coefficients obtained after threshold processing; sgn represents the sign function; W u,v represents the original wavelet coefficient at scale u and position v; b and d are both adjustable parameters; ψ represents the denoising threshold; |W z,c | indicates W z,c The absolute value of ; σ represents the standard deviation of the noise IMF component; N represents the signal length of the noise IMF component.

[0064] Compared with the prior art, the present invention has the following beneficial effects:

[0065] 1. This application provides a comprehensive noise reduction strategy, which optimizes the parameters of the VMD algorithm through the ISSA algorithm, and then decomposes the signal into IMF components for further processing. This application combines the advantages of intelligent optimization algorithms and signal processing technology, can effectively remove noise from pipeline acoustic emission signals, improve the signal-to-noise ratio and clarity of the signal, and provide more accurate data support for subsequent signal analysis, fault detection and location.

[0066] 2. In the process of describing the use of the ISSA algorithm to optimize the penalty factor and the number of modal components of the VMD algorithm, by setting a reasonable value range and introducing three roles: discoverer, follower, and sentinel, efficient search and optimization of parameters are achieved, which can avoid the difficulty of parameter selection in traditional methods and improve the performance and stability of the VMD algorithm.

[0067] 3. The position update formula of the discoverer introduces sine and cosine functions, as well as dynamically adjusted early and late optimization steps, so that the discoverer can explore more effectively in the search space. This design helps to improve the search efficiency and global optimization ability of the algorithm, ensuring that a better penalty factor and number of modal components are found.

[0068] 4. The position update formula of the follower maintains the diversity of the population by imitating the optimal individual and introducing randomness, which helps to avoid premature convergence of the algorithm and improve the global search ability of the algorithm. At the same time, the introduction of random matrices and all-1 matrices further enhances the flexibility and adaptability of the algorithm.

[0069] 5. The position update formula of the sentinel helps maintain the diversity of the population by monitoring the fitness changes of the entire population and adjusting its position in time, which helps prevent the algorithm from falling into the local optimal solution and improves the global convergence of the algorithm. At the same time, the introduction of flight calculation function and learning factor further enhances the search ability and efficiency of the sentinel.

[0070] 6. The implementation process of the VMD algorithm, including the steps of iteratively updating the modal components, center frequency and Lagrangian factors, can accurately decompose the original pipeline acoustic emission signal into IMF components, providing clear and independent signal components for subsequent signal processing. At the same time, the introduction of the judgment criteria ensures the accuracy and effectiveness of the iteration.

[0071] 7. By calculating the permutation entropy of the IMF component, the effective signal component and the noise signal component can be effectively distinguished, and the noise IMF component can be accurately identified, providing a reliable basis for subsequent denoising. At the same time, the calculation formula of the permutation entropy is simple and easy to implement, with high calculation efficiency and accuracy.

[0072] 8. The calculation formula of permutation entropy can quantify the complexity of IMF components by calculating the probability of occurrence of IMF components and introducing logarithmic functions to calculate permutation entropy, which helps to identify noise IMF components. At the same time, the calculation formula is simple, intuitive, easy to understand and implement.

[0073] 9. The noise IMF component is denoised by wavelet transform and improved wavelet threshold function, which can effectively remove noise while retaining the detailed characteristics of the signal. The improved wavelet threshold function improves the denoising effect and signal reconstruction quality by introducing adjustable parameters and information such as the standard deviation of the noise IMF component.

[0074] 10. The improved wavelet threshold function realizes flexible processing of the original wavelet coefficients by introducing information such as sign function, adjustable parameters and standard deviation of noise IMF components. It can adaptively adjust the threshold according to the characteristics of noise IMF components, improve the denoising effect and signal reconstruction quality. At the same time, the formula is simple, easy to implement, and has high computational efficiency and practicality. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 It is a noise reduction flow chart of the present invention.

[0076] Figure 2 It is the minimum envelope entropy iteration curve of the SSA algorithm in the present invention.

[0077] Figure 3 It is the IMF component diagram in the present invention.

[0078] Figure 4 It is the IMF component frequency diagram in the present invention.

[0079] Figure 5 This is the original pipeline acoustic emission signal diagram after denoising in the present invention. DETAILED DESCRIPTION

[0080] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0081] See also Figure 1 In an embodiment of the present invention, a pipeline acoustic emission signal noise reduction method includes the following contents:

[0082] 1. Setting the key parameters of the VMD algorithm

[0083] The key parameters of the VMD algorithm include the penalty factor α and the number of modal components K that meet the value requirements. Using the ISSA algorithm and taking fitness as the objective function, the penalty factor α and the number of modal components K of the VMD algorithm are combined and optimized to obtain the best [K, α] combination. The specific steps are as follows:

[0084] S11. Set the value range of the penalty factor α to [10, 2500], and set the value range of the number of modal components K to [1, 10]. Combine the value ranges of the two to form a two-dimensional value space.

[0085] S12, initialize the parameters in the ISSA algorithm, set the size of the sparrow population, and divide the sparrow population into discoverers, followers, and guards; at the same time, set the maximum number of iterations iter max =40.

[0086] S13. Use the position update formula in the two-dimensional value space to iteratively update the positions of the discoverer, follower and sentinel.

[0087] The specific formula for updating the location of the discoverer is as shown in formula (1):

[0088]

[0089] The value range of r2 is [0,2π]. The value range of r3 is [0,2], which is a random number indicating the influence of the optimal position on the next position.

[0090] The calculation formula for the initial optimal step length is shown in formula (2):

[0091]

[0092] In this embodiment, β=1, iter max =40.

[0093] The calculation formula for the later optimal step length is shown in formula (3):

[0094]

[0095] The follower's position update formula is shown in formula (4):

[0096]

[0097] The position update formula of the sentinel is shown in formula (5):

[0098]

[0099]

[0100]

[0101] Among them, B1 and B2 are both random numbers, and their value ranges are both [0,1].

[0102] S14. When the fitness value of the discoverer, follower or sentinel is less than the fitness threshold or reaches the maximum number of iterations, stop the iteration and output the penalty factor α and the number of modal components K at this time. The penalty factor α and the number of modal components K are the penalty factor α and the number of modal components K that meet the value requirements.

[0103] like Figure 2 The minimum envelope entropy iteration curve of the SSA algorithm shown is given by Figure 2 It can be seen that when the number of iterations reaches 35, the fitness value reaches the minimum, which is 6.6898. At this time, the number of modal components K=6, and the penalty factor α=2250.744.

[0104] 2. Decomposition of original pipeline acoustic emission signals

[0105] The penalty factor α and the number of modal components K are input into the VMD algorithm, and the original pipeline acoustic emission signal is decomposed into IMF components through the VMD algorithm. The specific steps are as follows:

[0106] S21, obtaining the original pipeline acoustic emission signal, and using the VMD algorithm to decompose the original pipeline acoustic emission signal into K modal components.

[0107] The process of obtaining the original pipeline acoustic emission signal is as follows:

[0108] Build a pipeline pressure equipment platform and arrange two sensors at both ends of the pipeline to collect signals. Use the pressure equipment to inject water into the pipeline to simulate the actual working conditions. Open the acoustic emission signal acquisition software, perform a lead breaking experiment between the two sensors, and export the collected original acoustic emission signals. Add Gaussian white noise to the acquired signal to simulate the noise signal of the actual working environment.

[0109] S22. Set the number of iterations and initialize the modal components, center frequencies and Lagrangian factors.

[0110] S23, start iteration, and update the modal components, center frequency and Lagrangian factors; the update formula of the modal components is shown in formula (8):

[0111]

[0112] The update formula of the center frequency is shown in formula (9):

[0113]

[0114] The update formula of the Lagrangian factor is shown in formula (10):

[0115]

[0116] S24, establish a judgment criterion. If the judgment criterion is met, the iteration is terminated. Otherwise, the iteration is repeated according to the content of step S23 until the maximum number of iterations is reached or the judgment criterion is met. The judgment criterion is expressed as shown in formula (11):

[0117]

[0118] S25, output each IMF component. Figure 3 As shown, IMF components 1 to 6, their specific frequency diagrams are as follows Figure 4 shown.

[0119] 3. Obtaining effective signal components and noise signal components

[0120] The IMF components are divided into effective signal components and noise signal components.

[0121] S31. Calculate the permutation entropy of various IMF components using the permutation entropy calculation formula; the permutation entropy calculation formula is specifically shown in formula (12):

[0122] H m =ln(P m lpg(P m )) (12)

[0123] S32, marking the IMF components whose permutation entropy is greater than a set threshold as noise IMF components, and marking the remaining IMF components as valid signal IMF components.

[0124] In this embodiment, the threshold is set to 1, and the permutation entropies of IMF1-IMF6 are calculated to be 1.1989, 1.2685, 1.5688, 1.7326, 1.7906, and 1.7180. Components with permutation entropies greater than 1 are considered high entropy values, that is, they are significantly affected by noise. It can be seen that IMF1-IMF6 are all noise IMF components.

[0125] 4. Denoising the Noise Signal

[0126] The noise signal component is subjected to wavelet threshold denoising to obtain the denoised noise signal component. The specific process is as follows:

[0127] S41, performing wavelet transform on the identified noise IMF component to obtain its wavelet coefficient, and recording the wavelet coefficient as the original wavelet coefficient.

[0128] S42, applying the improved wavelet threshold function to perform threshold processing on the original wavelet coefficients to obtain the denoised wavelet coefficients, and recording the wavelet coefficients as the improved wavelet coefficients. The improved wavelet threshold function is specifically shown in formula (13):

[0129]

[0130]

[0131] In this embodiment, the value range of b is [0, 1].

[0132] S43, performing inverse wavelet transform on the improved wavelet coefficients to obtain an IMF component after the noise IMF component is denoised, and the IMF component is recorded as the denoised IMF component.

[0133] The wavelet waveform in this embodiment is DB3, and the number of wavelet decomposition levels is 5.

[0134] 5. Obtain the original pipeline acoustic emission signal after denoising

[0135] The effective signal component and the noise signal component after denoising are combined and reconstructed, and the original pipeline acoustic emission signal after denoising is obtained through combined reconstruction. The signal-to-noise ratio of the Gaussian white noise signal in this embodiment is 3.24dB, the signal-to-noise ratio after denoising is 4.00dB, and the root mean square error is 0.0012. The original pipeline acoustic emission signal after denoising is as follows Figure 5 As shown, the results show that the method proposed in this application has a higher signal-to-noise ratio than before denoising, and the denoising effect is better.

[0136] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A pipeline acoustic emission signal noise reduction method, characterized in that: The following noise reduction steps are included: S1. Use the ISSA algorithm to obtain the penalty factor α and the number of modal components K that meet the value requirements in the VMD algorithm within the value range; S2, inputting the penalty factor α and the number of modal components K into the VMD algorithm, and decomposing the original pipeline acoustic emission signal into IMF components through the VMD algorithm; S3, dividing the IMF component into an effective signal component and a noise signal component; S4, performing wavelet threshold denoising processing on the noise signal component to obtain the denoised noise signal component; S5. Combining and reconstructing the effective signal component and the denoised noise signal component, and obtaining the denoised original pipeline acoustic emission signal through combined reconstruction.

2. A pipeline acoustic emission signal noise reduction method according to claim 1, characterized in that: The specific content of step S1 is as follows: S11, setting the value range of the penalty factor α and the value range of the modal component number K; combining the value ranges of the two to form a two-dimensional value space; S12, initializing the parameters in the ISSA algorithm, setting the size of the sparrow population, and dividing the sparrow population into discoverers, followers, and guards; S13, iteratively updating the positions of the discoverer, follower and sentinel in the two-dimensional value space using a position update formula; S14. When the fitness value of the discoverer, follower or sentinel is less than the fitness threshold or reaches the maximum number of iterations, stop the iteration and output the penalty factor α and the number of modal components K at this time. The penalty factor α and the number of modal components K are the penalty factor α and the number of modal components K that meet the value requirements.

3. A pipeline acoustic emission signal noise reduction method according to claim 2, characterized in that: The discoverer's location update formula is as follows: In the formula, Indicates the position of the i-th sparrow in the j-th dimension at the t+1th iteration, i∈[1,I], I represents the total number of sparrows in the sparrow population, j=1,2; Indicates the position of the i-th sparrow in the j-th dimension at the t-th iteration; Indicates the position of the sparrow with the smallest fitness value among all the sparrows in the tth iteration; r'1 indicates the initial optimization step length; r″1 indicates the later optimization step length; r2 indicates the moving distance parameter of the sparrow, r2∈[0,,2π]; r3 indicates the random parameter, r3∈[0,,2]; sin indicates the sine function; cos indicates the cosine function; Ra indicates the alarm value; TH indicates the safety threshold; if indicates the conditional symbol; The calculation formula for the initial optimal step length is as follows: In the formula, β represents the correction coefficient; η represents the adjustment coefficient; iter max represents the maximum number of iterations; t represents the current number of iterations; The calculation formula for the later step length is as follows: Wherein, e represents a natural constant.

4. A pipeline acoustic emission signal noise reduction method according to claim 3, characterized in that: The follower's position update formula is as follows: In the formula, Indicates the position of the sparrow with the smallest fitness value among all sparrows in the t+1th iteration; G u represents random numbers that follow a normal distribution; Indicates the position of the sparrow with the largest fitness value among all sparrows in the tth iteration; C + represents a random matrix; V represents a matrix of all 1s.

5. A pipeline acoustic emission signal noise reduction method according to claim 4, characterized in that: The position update formula of the sentinel is as follows: In the formula, represents random numbers that follow a normal distribution; Indicates the fitness value of the i-th sparrow in the t-th iteration; F g Indicates the fitness value of the best sparrow in the current global situation; F w represents the fitness value of the worst sparrow in the current global situation; τ represents a random number that obeys a normal distribution; μ represents a random non-zero constant; Levy(θ) represents a flight calculation function; θ represents a learning factor; B1 and B2 both represent random constants; ε represents a random constant; Γ represents a gamma function.

6. A pipeline acoustic emission signal noise reduction method according to any one of claims 1 to 5, characterized in that: The specific steps of step S2 are as follows: S21, obtaining an original pipeline acoustic emission signal, and using a VMD algorithm to decompose the original pipeline acoustic emission signal into K modal components; S22, setting the number of iterations, initializing the modal components, center frequencies and Lagrangian factors; S23, start iteration, and update the modal components, center frequency and Lagrangian factors; the update formula of the modal components is expressed as follows: In the formula, represents the value of the kth modal component with respect to the center frequency ω at the n+1th iteration; express Fourier transform of ; f(ω) represents the photovoltaic power output function about the center frequency ω, represents the Fourier transform of f(ω); represents the value of the rth modal component at the nth iteration with respect to the center frequency ω; express Fourier transform of n (ω) represents the value of the Lagrangian factor at the nth iteration with respect to the center frequency ω; Represents λ n Fourier transform of (ω); represents the value of the center frequency ω of the kth modal component at the nth iteration; The update formula of the center frequency is as follows: In the formula, represents the value of the center frequency ω of the kth modal component at the n+1th iteration; dω represents the differential of the center frequency ω; The update formula of the Lagrangian factor is expressed as follows: λ n+1 (ω) represents the value of the Lagrangian factor with respect to the center frequency ω at the n+1th iteration; Represents λ n+1 Fourier transform of (ω); Ω represents the limiting factor; S24, establish a judgment criterion. If the judgment criterion is met, the iteration is terminated. Otherwise, the iteration is repeated according to the content of step S23 until the maximum number of iterations is reached or the judgment criterion is met. The judgment criterion is expressed as follows: In the formula, express The second-order norm of ; express The quadratic value of express The second-order norm of ; express The quadratic value of; E represents the judgment accuracy; S25. Output each IMF component.

7. A pipeline acoustic emission signal noise reduction method according to claim 6, characterized in that: The specific content of step S3 is as follows: S31, calculating the permutation entropy of various IMF components using the permutation entropy calculation formula; S32, marking the IMF components whose permutation entropy is greater than a set threshold as noise IMF components, and marking the remaining IMF components as valid signal IMF components.

8. A pipeline acoustic emission signal noise reduction method according to claim 7, characterized in that: The calculation formula of permutation entropy is as follows: A m =ln(P m log(P m )); Where P m represents the probability of occurrence of the mth IMF component; ln represents the logarithmic function with a natural constant as the base; H m represents the permutation entropy of the mth IMF component.

9. A pipeline acoustic emission signal noise reduction method according to claim 8, characterized in that: The specific content of step S4 is as follows: S41, performing wavelet transform on the identified noise IMF component to obtain its wavelet coefficient, and recording the wavelet coefficient as the original wavelet coefficient; S42, applying an improved wavelet threshold function to perform threshold processing on the original wavelet coefficients to obtain denoised wavelet coefficients, and recording the wavelet coefficients as improved wavelet coefficients; S43, performing inverse wavelet transform on the improved wavelet coefficients to obtain an IMF component after the noise IMF component is denoised, and the IMF component is recorded as the denoised IMF component.

10. A pipeline acoustic emission signal noise reduction method according to claim 9, characterized in that: The improved wavelet threshold function is specifically expressed as follows: Where W z,c represents the original wavelet coefficient, z represents the scale parameter of wavelet decomposition, and c represents the position parameter of wavelet decomposition; Represents the original wavelet coefficient W z,c The improved wavelet coefficients obtained after threshold processing; sgn represents the sign function; W u,v represents the original wavelet coefficient at scale u and position v; b and d are both adjustable parameters; ψ represents the denoising threshold; |W z,c | indicates W z,c The absolute value of ; σ represents the standard deviation of the noise IMF component; N represents the signal length of the noise IMF component.

Images