Method for rapidly extracting effective signal in acoustic emission of circumferential crack of pipeline circumferential weld

Abstract

The invention relates to the technical field of acoustic emission signal processing, in particular to an effective signal rapid extraction method in pipeline circumferential weld circumferential crack acoustic emission. The method comprises the following steps: step 1, collecting a circumferential crack acoustic emission signal of the circumferential weld of the pipeline, setting a threshold value, and extracting a peak-to-peak signal; 2, processing the signal, and based on VMD, obtaining a stationary eigenmode function component containing a potential acoustic emission multi-mode component after optimizing the number of decomposition layers through center frequency selection; 3, initially determining a frequency value of a potential multi-mode component through spectral analysis, and taking the frequency value as a driving force parameter set; and 4, sending the signal to the improved Duffing oscillator system after parameter setting, solving an equation by using a four-order Runge-Kutta method, drawing a phase plane trajectory diagram, and determining whether the signal contains the above-mentioned power parameter set frequency according to whether the phase diagram transitions from critical chaos to a large-scale periodic state so as to eliminate false weak signals and extract effective signals.

Description

Technical Field

[0001] The invention relates to the technical field of acoustic emission signal processing, and in particular to a method for quickly extracting effective signals in acoustic emission of circumferential cracks of a pipeline girth weld. Background Art

[0002] During the installation and service of the pipeline, circumferential cracks are clearly identified as the main crack type, which frequently appears in the girth weld area. During the welding process, circumferential cracks are easy to form due to the hardened structure of the weld joint and the diffusion of hydrogen. During the service of the pipeline, the combined effects of cyclic loads, impacts and residual stresses further lead to the expansion of circumferential cracks.

[0003] At present, the acoustic emission signals generated by circumferential cracks in pipeline girth welds face the serious problem of low signal-to-noise ratio, which greatly hinders the accurate extraction and identification of effective signals. Traditional methods based on chaos theory to detect such weak signals often rely on inefficient frequency sweeping processes or the method of calculating frequencies with the help of prior knowledge to set the driving force parameters of the Duffing oscillator. However, in the complex environment of circumferential crack acoustic emission, the frequency of multimodal components is uncertain. This characteristic poses a significant challenge to the precise setting of driving force parameters, limiting the effectiveness and efficiency of traditional methods. Summary of the invention

[0004] In view of the technical problems in the above-mentioned prior art, the present invention proposes a method for quickly extracting effective signals from the acoustic emission of circumferential cracks in a pipeline girth weld, which can quickly extract effective signals from the acoustic emission of circumferential cracks in a pipeline girth weld.

[0005] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solutions:

[0006] A method for quickly extracting effective signals from acoustic emission of circumferential cracks in a pipeline girth weld comprises the following steps:

[0007] Step 1: Collect the acoustic emission signal of the circumferential crack of the pipeline girth weld, set the threshold, and extract the peak value signal;

[0008] Step 2: Process the extracted signal, based on VMD, and obtain the stabilized eigenmode function components containing potential acoustic emission multimodal components by optimizing the number of decomposition layers through center frequency selection;

[0009] Step 3: Preliminarily determine the frequency values ​​of potential multimodal components through spectral analysis and use them as the driving force parameter set;

[0010] Step 4: Send the signal to the improved Duffing oscillator system after parameter setting, use the fourth-order Runge-Kutta method to solve the equation, draw the phase plane trajectory diagram, and determine whether the signal contains the frequency of the aforementioned driving force parameter set based on whether the phase diagram transitions from critical chaos to a large-scale periodic state, so as to eliminate false weak signals and extract effective signals.

[0011] Further as a preferred technical solution of the present invention, the step 2 comprises the following steps:

[0012] Step 2.1: Initialize VMD algorithm parameters, including decomposition level K and modal function u k , center frequency ω k , number of cycles n, Lagrange multiplier λ and convergence accuracy ε;

[0013] Step 2.2: Use Hilbert transform to transform each modal function u k Convert to a single-sided spectrum, then modulate each mode to the baseband of the corresponding center frequency to obtain their respective analytical signals:

[0014]

[0015] Where t represents time, IMF k is the kth eigenmode component, {u k (t)}:={u 1 (t),u 2 (t),...,u k (t)},{ω k}:={ω 1 ,...,ω k} are the modes and their respective center frequencies, δ(t) is the Dirac pulse;

[0016] Step 2.3, introduce the quadratic penalty parameter and Lagrange multiplier as follows:

[0017]

[0018] Among them, α represents the penalty factor, λ represents the Lagrange multiplier, and f(t) is the original signal;

[0019] Step 2.4: Update the n+1th iteration in the non-negative frequency domain interval and

[0020]

[0021] in, They are Fourier transform corresponding to f(t) and λ(t);

[0022] Update the Lagrange multipliers:

[0023]

[0024] Where τ is the update step size parameter of the Lagrange multiplier, which is usually 0;

[0025] Step 2.5: Ending the loop of step 2.4 requires satisfying specific iterative convergence conditions:

[0026]

[0027] Among them, ε is the convergence accuracy, and each mode u is calculated k , we need to Perform inverse Fourier transform to obtain all the stationary intrinsic mode function components IMF obtained by VMD decomposition k .

[0028] Further as a preferred technical solution of the present invention, the step 4 comprises the following steps:

[0029] Step 4.1: Improve the Duffing oscillator equation to:

[0030]

[0031] Among them, k is the damping ratio, r is the amplitude of the periodic driving force, ω 0 is the angular frequency, rcos(ω 0 t) is the periodic driving force, s(t) is the weak signal to be detected;

[0032] Set the value of k and use IMF k Calculate the angular frequency of the fault signal to be detected, denoted as ω 1 , which will improve the ω in the Duffing oscillator equation 0 Let ω 1 ;

[0033] Step 4.2: When no external signal is added, use the Runge-Kutta method to solve the differential equation and denote the critical value of the Duffing oscillator as r 1 , set r to r 1 , so that the chaotic oscillator is in a critical state, and the phase plane trajectory diagram is drawn;

[0034] Step 4.3, add a signal to the system in the critical state, use the fourth-order Runge-Kutta method to solve the improved Duffing oscillator equation after adding the signal, and observe the changes in the phase plane trajectory diagram. If the measured signal causes the phase diagram to transition from critical chaos to a large-scale periodic state, it can be determined that the signal contains the frequency of the aforementioned driving force parameter set, and the false weak signal can be eliminated, thereby extracting the effective signal.

[0035] Further as a preferred technical solution of the present invention, in step 4.1, the traditional Holmes-type Duffing equation is in the form of:

[0036]

[0037] Among them, k is the damping ratio, r is the amplitude of the periodic driving force, ω is the angular frequency, and rcos(ωt) is the periodic driving force;

[0038] When the signal is input into the Duffing oscillator system, its system characteristics are mainly determined by the nonlinear restoring force. In order to improve the system's ability to detect weak signals, the Duffing equation needs to be improved as follows:

[0039]

[0040] The equivalent first-order differential equation of this formula is:

[0041]

[0042] In order to detect a periodic signal of any frequency, By performing a time scale transformation and adding the detected weak signal s(t), the improved Duffing oscillator equation can be obtained.

[0043] The method for quickly extracting effective signals from acoustic emission of circumferential cracks in a pipeline girth weld according to the present invention has the following technical effects compared with the prior art by using the above technical solution:

[0044] The present invention proposes for the first time that by combining VMD and Duffing oscillator system, noise interference can be effectively removed, the accuracy of acoustic emission signal processing of circumferential cracks in pipeline girth welds can be significantly improved, and different types of noise environments can be adapted. Meanwhile, less computing resources and storage space are required, the computing efficiency is high, and the computing cost is low. The method can be applied to real-time signal processing systems and is more universal and applicable. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a flow chart of a method for quickly extracting effective signals from acoustic emission of circumferential cracks in a pipeline girth weld according to the present invention;

[0046] Figure 2The figure is a flow chart of the method implementation in the embodiment of the present invention. DETAILED DESCRIPTION

[0047] The present invention is further explained below in detail with reference to the accompanying drawings so that those skilled in the art can more deeply understand the present invention and be able to implement it. However, the following reference examples are only used to explain the present invention and are not intended to limit the present invention.

[0048] like Figure 1-2 As shown, this embodiment proposes a method for quickly extracting effective signals from acoustic emission of circumferential cracks in a pipeline girth weld, comprising the following steps:

[0049] First, it is necessary to prepare a pipeline specimen containing circumferential cracks in the girth weld, and use a fatigue testing machine and a constructed test platform to carry out a fatigue loading test. Use n sensors to collect the acoustic emission signal near the moment of circumferential crack propagation, set the threshold, and extract the peak value signal.

[0050] The signals are then divided into two groups: the first group is sensor 1 signal, which is used to solve the frequency of the potential multimodal components of acoustic emission and form a driving force parameter set; the second group is sensor 2-n signal, which is used to determine whether the signal contains the characteristic frequency in the aforementioned driving force parameter set, and to eliminate false weak signals, thereby extracting effective signals.

[0051] First, for the data of group 1, based on VMD, the center frequency is selected to optimize the number of decomposition layers to obtain the stabilized eigenmode function components containing potential acoustic emission multimodal components. Then, the frequency values ​​of the potential multimodal components are initially determined by spectral analysis and used as the driving force parameter set. The specific steps are as follows:

[0052] Step (1), initialize the VMD algorithm parameters, including the decomposition level K, the modal function u k , center frequency ω k , the number of cycles n, the Lagrange multiplier λ and the convergence accuracy ε. The choice of the decomposition level K is crucial. The number of decomposition layers is determined by the center frequency. K = 2 is first taken for pre-decomposition. The center frequency of the decomposition component is calculated. If the center frequency distribution is normal, the number of decomposition layers is increased by 1 and the decomposition is continued. If the modes are particularly close, the modes will be aliased, which is contrary to the original intention of the decomposition. At this time, the K value of the previous moment is selected.

[0053] Step (2), use Hilbert transform to transform each mode function u k Convert to a single-sided spectrum, then modulate each mode to the baseband of the corresponding center frequency to obtain their respective analytical signals:

[0054]

[0055] Where t represents time, IMFk is the kth eigenmode component, {u k (t)}:={u 1 (t),u 2 (t),...,u k (t)},{ω k}:={ω 1 ,...,ω k} are the modes and their respective center frequencies, δ(t) is the Dirac pulse;

[0056] Step (3), introduce the quadratic penalty parameter and Lagrange multiplier as follows:

[0057]

[0058] Among them, α represents the penalty factor, λ represents the Lagrange multiplier, and f(t) is the original signal;

[0059] Step (4), update the n+1th iteration in the non-negative frequency domain interval and

[0060]

[0061] in, They are The Fourier transform corresponding to f(t) and λ(t).

[0062] Update the Lagrange multipliers:

[0063]

[0064] Where τ is the update step size parameter of the Lagrange multiplier, which is usually 0;

[0065] Step (5) and the loop ending step (4) need to meet specific iterative convergence conditions:

[0066]

[0067] Among them, ε is the convergence accuracy, and each mode u is calculated k , we need to By performing inverse Fourier transform, we can obtain all the stationary intrinsic mode function components IMF obtained by VMD decomposition. k ;

[0068] Step (6): Preliminarily determine the frequency values ​​of potential multimodal components through spectral analysis and use them as the driving force parameter set.

[0069] Then, for the data of the second group, they are respectively sent to the Duffing oscillator system after parameter setting, and the fourth-order Runge-Kutta method is used to solve the equations, and the phase plane trajectory diagram of the second group of signals is drawn. According to whether the phase diagram transitions from critical chaos to a large-scale periodic state, it is determined whether the signal contains the frequency of the aforementioned driving force parameter set to eliminate false weak signals. The specific steps are as follows:

[0070] Step a: Improve the Duffing oscillator equation to:

[0071]

[0072] Among them, k is the damping ratio, r is the amplitude of the periodic driving force, ω 0 is the angular frequency, rcos(ω 0 t) is the periodic driving force, and s(t) is the weak signal to be detected.

[0073] Set the value of k and use IMF k Calculate the angular frequency of the fault signal to be detected, denoted as ω 1 , which will improve the ω in the Duffing oscillator equation 0 Let ω 1 ;

[0074] The specific process of obtaining the improved Duffing oscillator equation is:

[0075] The traditional Holmes-type Duffing equation is in the form:

[0076]

[0077] Among them, k is the damping ratio, r is the amplitude of the periodic driving force, ω is the angular frequency, and rcos(ωt) is the periodic driving force.

[0078] When the signal is input into the Duffing oscillator system, its system characteristics are mainly determined by the nonlinear restoring force. In order to improve the system's ability to detect weak signals, the Duffing equation needs to be improved as follows:

[0079]

[0080] The equivalent first-order differential equation of this formula is:

[0081]

[0082] In order to detect a periodic signal of any frequency, t=ω 0 τ, perform time scale transformation, and add the detected weak signal s(t) to obtain the improved Duffing oscillator equation.

[0083] Step b: When no external signal is added, the differential equation is solved using the Runge-Kutta method, and the critical value of the Duffing oscillator is recorded as r 1 , set r to r 1 , so that the chaotic oscillator is in a critical state, and the phase plane trajectory diagram is drawn;

[0084] Step c, add the second group of signals to the system in the critical state, solve the improved Duffing oscillator equation after adding the signals by the fourth-order Runge-Kutta method, and observe the changes in the phase plane trajectory diagram. If the measured signal causes the phase diagram to transition from the critical chaos to the large-scale periodic state, it can be determined that the signal contains the frequency of the aforementioned driving force parameter set, and the false weak signals can be eliminated, thereby extracting the effective signal.

[0085] The specific implementation scheme described above further describes in detail the purpose, technical scheme and beneficial effects of the present invention. It should be understood that the above is only a specific implementation scheme of the present invention and is not intended to limit the scope of the present invention. Any equivalent changes and modifications made by any technician in the field without departing from the concept and principle of the present invention should fall within the scope of protection of the present invention.

Claims

1. A method for quickly extracting effective signals from acoustic emission of circumferential cracks in pipeline girth welds, characterized in that: The following steps are involved: Step 1: Collect the acoustic emission signal of the circumferential crack of the pipeline girth weld, set the threshold, and extract the peak value signal; Step 2: Process the extracted signal, based on VMD, and obtain the stabilized eigenmode function components containing potential acoustic emission multimodal components by optimizing the number of decomposition layers through center frequency selection; Step 3: Preliminarily determine the frequency values ​​of potential multimodal components through spectral analysis and use them as the driving force parameter set; Step 4: Send the signal to the improved Duffing oscillator system after parameter setting, use the fourth-order Runge-Kutta method to solve the equation, draw the phase plane trajectory diagram, and determine whether the signal contains the frequency of the aforementioned driving force parameter set based on whether the phase diagram transitions from critical chaos to a large-scale periodic state, so as to eliminate false weak signals and extract effective signals.

2. The method for quickly extracting effective signals from acoustic emission of circumferential cracks in a pipeline girth weld according to claim 1 is characterized in that: The step 2 comprises the following steps: Step 2.1: Initialize VMD algorithm parameters, including decomposition level K and modal function u k , center frequency ω k , number of cycles n, Lagrange multiplier λ and convergence accuracy ε; Step 2.2: Use Hilbert transform to transform each modal function u k Convert to a single-sided spectrum, then modulate each mode to the baseband of the corresponding center frequency to obtain their respective analytical signals: Where t represents time, IMF k is the kth eigenmode component, {u k (t)}:={u1(t),u2(t),...,u k (t)},{ω k }:={ω1,...,ω k } are the modes and their respective center frequencies, δ(t) is the Dirac pulse; Step 2.3, introduce the quadratic penalty parameter and Lagrange multiplier as follows: Among them, α represents the penalty factor, λ represents the Lagrange multiplier, and f(t) is the original signal; Step 2.4: Update the n+1th iteration in the non-negative frequency domain interval and in, They are Fourier transform corresponding to f(t) and λ(t); Update the Lagrange multipliers: Where τ is the update step size parameter of the Lagrange multiplier, which is usually 0; Step 2.5: Ending the loop of step 2.4 requires satisfying specific iterative convergence conditions: Among them, ε is the convergence accuracy, and each mode u is calculated k , we need to Perform inverse Fourier transform to obtain all the stationary intrinsic mode function components IMF obtained by VMD decomposition k .

3. The method for quickly extracting effective signals from acoustic emission of circumferential cracks in a pipeline girth weld according to claim 1 is characterized in that: The step 4 comprises the following steps: Step 4.1: Improve the Duffing oscillator equation to: Where, k is the damping ratio, r is the amplitude of the periodic driving force, ω0 is the angular frequency, rcos(ω0t) is the periodic driving force, and s(t) is the weak signal to be detected; Set the value of k and use IMF k Calculate the angular frequency of the fault signal to be detected, record it as ω1, and set ω0 in the improved Duffing oscillator equation to ω1; Step 4.2, when no external signal is added, the differential equation is solved by the Runge-Kutta method, the critical value of the Duffing oscillator is recorded as r1, r is set to r1, so that the chaotic oscillator is in a critical state, and the phase plane trajectory diagram is drawn; Step 4.3, add a signal to the system in the critical state, use the fourth-order Runge-Kutta method to solve the improved Duffing oscillator equation after adding the signal, and observe the changes in the phase plane trajectory diagram. If the measured signal causes the phase diagram to transition from critical chaos to a large-scale periodic state, it can be determined that the signal contains the frequency of the aforementioned driving force parameter set, and the false weak signal can be eliminated, thereby extracting the effective signal.

Images