Dual-source magnetic field comprehensive detection analysis method and system based on buried metal pipeline

Abstract

This invention provides a method and system for integrated detection and analysis of dual-source magnetic fields for buried metal pipelines. The method includes the following steps: applying a composite magnetic field excitation to the target section of the pipeline, and synchronously acquiring a mixed magnetic response signal through a magnetic sensor array deployed along the pipeline axis. The mixed signal is separated into a slowly varying component representing quasi-static background stress and a sudden component representing pipeline stress events, and the sudden component is extracted. Dispersion cancellation inverse processing is performed on the extracted sudden component, and combined with the arrival time difference and energy attenuation of the original waveform characteristics on different sensors, the location information of pipeline stress events on the pipeline is accurately determined through joint optimization. The energy and location information of the pipeline stress events are input into the model, and the cumulative damage factor at specific locations on the pipeline is continuously calculated and updated. This invention achieves accurate identification and location of stress events in buried metal pipelines.

Description

Technical Field

[0001] This invention belongs to the field of magnetic field detection technology, specifically relating to a dual-source magnetic field comprehensive detection and analysis method and system based on buried metal pipelines. Background Technology

[0002] Buried metal pipelines, serving as the lifeblood of modern industry and urban infrastructure, bear the responsibility of transporting critical materials such as energy and water resources. However, during long-term service, these pipelines are inevitably affected by various factors, including soil corrosion, geological subsidence, external impacts, and internal fluid pressure fluctuations, leading to material fatigue, crack initiation and propagation, and even catastrophic leaks or ruptures. Traditional pipeline inspection technologies, such as visual inspection, ultrasonic testing, and radiographic testing, often require excavation or shutdown, which is time-consuming, labor-intensive, and costly, making it difficult to achieve continuous, real-time monitoring of large-scale, long-distance pipelines. Moreover, in practical applications, due to the complexity of the buried environment and the inherent material properties of the pipelines, existing inspection methods can usually only detect obvious existing defects, failing to capture the various stresses the pipelines experience under internal or external forces, severely impacting the safety and reliability of the pipelines. Summary of the Invention

[0003] This invention provides a method and system for comprehensive detection and analysis of dual-source magnetic fields based on buried metal pipelines to solve the above-mentioned technical problems.

[0004] In a first aspect, the present invention provides a method for comprehensive detection and analysis of dual-source magnetic fields based on buried metal pipelines, the method comprising the following steps:

[0005] A composite magnetic field consisting of a low-frequency strong magnetic field and a high-frequency weak magnetic field is applied to the target detection section of the buried metal pipeline for excitation, and a magnetic sensor array deployed along the axial direction of the buried metal pipeline is used to collect the mixed magnetic response signal in response to the composite magnetic field.

[0006] A blind source separation algorithm based on independent component analysis is used to separate the hybrid magnetic response signal into a slowly varying component representing quasi-static stress and a burst component representing transient stress, and the burst component is extracted.

[0007] Based on the preset stress wave-magnetic coupling dispersion model, the sudden component is processed by dispersion cancellation inverse operation to reconstruct the original waveform characteristics of the pipeline stress event at the initial position without propagation distortion.

[0008] By combining the arrival time difference and energy attenuation of the original stress wave corresponding to the original waveform characteristics on different sensors in the magnetic sensor array, the location information of pipeline stress events on buried metal pipelines is determined by joint optimization solution.

[0009] The cumulative damage level of the buried metal pipeline target detection section is calculated based on the energy and location information of the pipeline stress event located each time.

[0010] Optionally, the step of applying a composite magnetic field consisting of a low-frequency strong magnetic field and a high-frequency weak magnetic field to the target detection section of the buried metal pipeline for excitation, and acquiring the mixed magnetic response signal in response to the composite magnetic field through a magnetic sensor array deployed along the axial direction of the buried metal pipeline, includes the following steps:

[0011] The first excitation coil for generating a low-frequency strong magnetic field and the second excitation coil for generating a high-frequency weak magnetic field are coaxially mounted and deployed in the target detection section of the buried metal pipeline.

[0012] A circumferential background magnetization field is established inside the wall of the buried metal pipeline by applying a direct current through the first excitation coil.

[0013] Based on the circumferential background magnetization field, a high-frequency alternating current is applied through the second excitation coil to superimpose a high-frequency detection magnetic field inside the tube wall to form a composite magnetic field;

[0014] A magnetic sensor array, linearly deployed at equal intervals along the axial direction of a buried metal pipeline, is used to sense changes in magnetic flux density in response to a composite magnetic field.

[0015] The magnetic flux density changes sensed by the magnetic sensor array are converted into multiple hybrid magnetic response signals by a multi-channel synchronous data acquisition device.

[0016] Optionally, the step of using a blind source separation algorithm based on independent component analysis to separate the hybrid magnetic response signal into a slowly varying component representing quasi-static background stress and a sudden component representing pipeline stress events, and extracting the sudden component, includes the following steps:

[0017] The acquired multi-channel hybrid magnetic response signals are segmented and an observation signal matrix is ​​constructed.

[0018] The optimization objective is to maximize non-Gaussianity. The fast fixed-point algorithm is used to iteratively calculate the observed signal matrix to obtain the unmixing matrix.

[0019] The observed signal matrix is ​​linearly transformed using the unmixing matrix to output multiple independent signal components.

[0020] Calculate the statistical characteristic parameters of multiple signal components, identify the slowly varying components and burst components based on the statistical characteristic parameters, and extract the signal components identified as burst components.

[0021] Optionally, the step of calculating the statistical characteristic parameters of multiple signal components, identifying the slowly varying components and burst components based on the statistical characteristic parameters, and extracting the signal components identified as burst components includes the following steps:

[0022] Calculate the kurtosis and sparsity values ​​for each signal component separately;

[0023] For any signal component, if the kurtosis value is greater than a preset first kurtosis threshold and the sparsity value is greater than a preset first sparsity threshold, then the signal component is identified as a burst component and the burst component is extracted.

[0024] If a signal component has a kurtosis value less than a preset second kurtosis threshold and a sparsity value less than a preset second sparsity threshold, then the signal component is identified as a slowly varying component, wherein the first kurtosis threshold is greater than the second kurtosis threshold and the first sparsity threshold is greater than the second sparsity threshold.

[0025] Optionally, the step of performing dispersion cancellation inverse operation on the sudden components based on a preset stress wave-magnetic coupling dispersion model to reconstruct the original waveform characteristics of the pipeline stress event at its initial position without propagation distortion includes the following steps:

[0026] The initial spectrum of the burst component is obtained by transforming the burst component from the time domain to the frequency domain;

[0027] A stress wave-magnetic coupling dispersion model for buried metal pipelines is pre-established, and the nonlinear mapping relationship between stress wave frequency and phase velocity is obtained through the stress wave-magnetic coupling dispersion model.

[0028] Based on the nonlinear mapping relationship, frequency-related phase correction is performed on each frequency component in the initial spectrum to obtain the corrected spectrum;

[0029] The corrected spectrum is converted back to the time domain by inverse Fourier transform to obtain the compressed pulse signal. The waveform of the pulse signal is used as the original waveform feature of the pipeline stress event at the initial position without propagation distortion.

[0030] Optionally, pre-establishing a pipeline stress wave-magnetic coupling dispersion model specifically includes the following steps:

[0031] A three-dimensional finite element model of the buried metal pipeline was established based on its material properties and geometric parameters.

[0032] A broadband impact load is applied at a predetermined location in a three-dimensional finite element model to excite test stress waves of various modes.

[0033] The propagation speed of stress waves of different frequencies along the axial direction of buried metal pipelines was tested by transient dynamic simulation calculation.

[0034] By fitting propagation velocity data of different frequency components, a nonlinear mapping relationship between stress wave frequency and phase velocity is established, and the nonlinear mapping relationship is solidified into a stress wave-magnetic coupling dispersion model for buried metal pipelines.

[0035] Optionally, the step of determining the location information of the pipeline stress event on the buried metal pipeline by combining the arrival time difference and energy attenuation of the original stress wave corresponding to the original waveform characteristics on different sensors in the magnetic sensor array, through joint optimization solution, includes the following steps:

[0036] Extract the pulse peak arrival time and pulse peak amplitude of the original stress wave corresponding to the original waveform characteristics from each magnetic sensor channel in the magnetic sensor array;

[0037] Calculate the arrival time difference between any two magnetic sensors based on the pulse peak arrival time;

[0038] Calculate the energy attenuation of each magnetic sensor relative to the reference sensor based on the pulse peak amplitude;

[0039] A joint optimization objective function containing arrival time difference error term and energy attenuation error term is constructed. A nonlinear least squares algorithm is used to iteratively optimize the joint optimization objective function. The solution that minimizes the objective function is determined as the location of the pipeline stress event.

[0040] Optionally, the step of calculating the cumulative damage level of the buried metal pipeline target detection section based on the energy and location information of the pipeline stress event at each location includes the following steps:

[0041] The three-dimensional finite element model of the buried metal pipeline is discretized into multiple damage calculation units along the pipeline axis;

[0042] Calculate the equivalent stress amplitude of each damage calculation unit based on the energy and location information of the pipeline stress event;

[0043] Based on the material SN curve of buried metal pipelines, find the allowable number of cycles corresponding to the equivalent stress amplitude;

[0044] The reciprocal of the allowable number of cycles is taken as the damage increment caused by this pipeline stress event to the corresponding damage calculation unit;

[0045] The damage increment is added to the cumulative damage factor of the corresponding damage calculation unit to complete the cumulative damage calculation of the target detection section of the buried metal pipeline.

[0046] Secondly, the present invention also provides a dual-source magnetic field comprehensive detection and analysis system based on buried metal pipelines, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the dual-source magnetic field comprehensive detection and analysis method based on buried metal pipelines as described in the first aspect.

[0047] Thirdly, the present invention also provides a computer-readable storage medium storing instructions that, when executed by a processor, configure the processor to perform the method for integrated detection and analysis of dual-source magnetic fields for buried metal pipelines according to the first aspect.

[0048] The beneficial effects of this invention are:

[0049] This invention employs a blind source separation algorithm based on independent component analysis, which intelligently separates complex mixed magnetic response signals into slowly varying components representing quasi-static background stress and sudden components representing pipeline stress events. This enables accurate extraction and denoising of key transient events, significantly improving the signal-to-noise ratio and recognition accuracy of the target signal. Dispersion cancellation inverse processing is performed on the extracted sudden components to reconstruct the original waveform characteristics of the pipeline stress event at its initial position without propagation distortion, avoiding misjudgments caused by signal propagation distortion. This allows for high-precision determination of the specific location information of pipeline stress events on the pipeline, providing crucial information for precise repair and maintenance. Finally, the cumulative damage factor at specific locations on the pipeline is calculated and updated in real time based on the pipeline stress events, enabling dynamic and quantitative assessment and prediction of pipeline fatigue damage, effectively reducing potential safety risks and operating costs. Attached Figure Description

[0050] Figure 1 This is a flowchart illustrating a dual-source magnetic field integrated detection and analysis method based on buried metal pipelines in one embodiment of this application.

[0051] Figure 2 This is a schematic diagram of the process of applying magnetic field excitation and acquiring response signals in one embodiment of this application.

[0052] Figure 3 This is a schematic diagram of the process for reconstructing the original waveform features in one embodiment of this application. Detailed Implementation

[0053] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0054] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0055] Figure 1 This is a flowchart illustrating a dual-source magnetic field integrated detection and analysis method for buried metal pipelines in one embodiment. It should be understood that, although... Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps. For example Figure 1 As shown, the dual-source magnetic field integrated detection and analysis method for buried metal pipelines disclosed in this invention specifically includes the following steps:

[0056] S101. A composite magnetic field consisting of a low-frequency strong magnetic field and a high-frequency weak magnetic field is applied to the target detection section of the buried metal pipeline for excitation, and a magnetic sensor array deployed along the axial direction of the buried metal pipeline is used to collect the mixed magnetic response signal in response to the composite magnetic field.

[0057] The process involves inducing magnetic response signals within the pipeline through external magnetic field excitation, signals that can be captured by sensors and contain rich information about the pipeline's stress state. Specifically, a first excitation coil for generating a low-frequency, strong magnetic field and a second excitation coil for generating a high-frequency, weak magnetic field are coaxially mounted and deployed in the target detection section. First, a DC or ultra-low-frequency current is applied through the first excitation coil to establish a stable circumferential background magnetization field within the pipeline wall; this magnetic field characterizes the quasi-static stress state of the pipeline. Then, a high-frequency alternating current is applied through the second excitation coil to superimpose a high-frequency detection magnetic field within the pipeline wall, forming a composite magnetic field. This high-frequency magnetic field is more sensitive to transient stress events. Subsequently, a magnetic sensor array, linearly deployed at equal intervals along the axial direction of the buried metal pipeline, senses changes in magnetic flux density in response to the composite magnetic field. Finally, a multi-channel synchronous data acquisition device converts the simulated magnetic flux density changes sensed by the magnetic sensor array into multi-channel digitized hybrid magnetic response signals.

[0058] S102. A blind source separation algorithm based on independent component analysis is used to separate the hybrid magnetic response signal into a slowly varying component representing quasi-static stress and a burst component representing transient stress, and the burst component is extracted.

[0059] The process involves segmenting the acquired multi-channel mixed magnetic response signals and constructing an observation signal matrix. Maximizing non-Gaussianity is set as the optimization objective, and a fast fixed-point algorithm is used to iteratively calculate the observation signal matrix, yielding a demixing matrix. This demixing matrix is ​​then used to perform a linear transformation on the observation signal matrix, resulting in multiple independent signal components. To identify the physical meaning of these components, the kurtosis and sparsity values ​​are calculated for each independent signal component. First kurtosis thresholds and first sparsity thresholds are set to identify burst components. Signal components with kurtosis values ​​greater than both the first kurtosis threshold and the first sparsity threshold are identified as burst components. The signals identified as burst components are extracted from the multiple independent signal components, thereby extracting key information representing pipeline stress events from the complex mixed signal.

[0060] S103. Based on the preset stress wave-magnetic coupling dispersion model, the sudden component is processed by dispersion cancellation inverse operation to reconstruct the original waveform characteristics of the pipeline stress event at the initial position without propagation distortion.

[0061] The steps for pre-establishing a pipeline stress wave-magnetically coupled dispersion model include: establishing a three-dimensional finite element model of the buried metal pipeline based on its material properties and geometric parameters; applying a broadband impact load at a predetermined location on the model to excite multi-mode stress waves; calculating and extracting the propagation velocities of different frequency components of the stress wave along the pipeline axis through transient dynamic simulation; fitting these velocity data to establish a nonlinear mapping relationship between the stress wave frequency and phase velocity, and solidifying it into a pipeline stress wave-magnetically coupled dispersion model. In actual processing, the extracted burst components are first transformed from the time domain to the frequency domain to obtain their initial spectrum. Then, the pre-set dispersion model is loaded to obtain the nonlinear mapping relationship between the stress wave frequency and phase velocity. Based on the mapping relationship, frequency-related phase correction is performed on each frequency component in the initial spectrum to obtain the original waveform characteristics.

[0062] S104. Combining the arrival time difference and energy attenuation of the original stress wave corresponding to the original waveform characteristics on different sensors in the magnetic sensor array, the location information of the pipeline stress event on the buried metal pipeline is determined by joint optimization solution.

[0063] This process involves accurately extracting the pulse peak arrival time and amplitude from the raw waveform characteristics of each sensor channel. Based on the pulse peak arrival time, the arrival time difference between any two sensors is calculated. Simultaneously, the energy attenuation of each sensor relative to the reference sensor is calculated based on the pulse peak amplitude. A joint optimization objective function is constructed, incorporating both arrival time difference and energy attenuation error terms. A nonlinear least squares algorithm is used to iteratively optimize the joint objective function, and the solution that minimizes the objective function is determined as the precise location of the pipeline stress event. This step, by fusing time and energy information, effectively improves the positioning accuracy.

[0064] S105. The cumulative damage level of the buried metal pipeline target detection section is calculated based on the energy and location information of the pipeline stress event in each location.

[0065] This method discretizes the three-dimensional finite element model of buried metal pipelines along the axial direction into a series of damage calculation elements, each representing a specific region of the pipeline. Based on the energy and location of a stress event, the equivalent stress amplitude caused by the event to each damage calculation element is calculated. Using the SN curve (stress-cycle curve) of the buried metal pipeline material, the allowable number of cycles corresponding to the equivalent stress amplitude is determined. According to Miner's rule, the reciprocal of the allowable number of cycles is used as the damage increment caused by the event to the corresponding damage calculation element. Finally, this damage increment is accumulated into the cumulative damage factor of the corresponding damage calculation element. By quantifying the cumulative damage, a scientific basis is provided for pipeline remaining life assessment and maintenance decisions.

[0066] In one embodiment, applying a composite magnetic field consisting of a low-frequency strong magnetic field and a high-frequency weak magnetic field to the target detection section of the buried metal pipeline for excitation, and acquiring the mixed magnetic response signal in response to the composite magnetic field by a magnetic sensor array deployed along the axial direction of the buried metal pipeline, includes the following steps:

[0067] The first excitation coil for generating a low-frequency strong magnetic field and the second excitation coil for generating a high-frequency weak magnetic field are coaxially mounted and deployed in the target detection section of the buried metal pipeline.

[0068] A circumferential background magnetization field is established inside the wall of the buried metal pipeline by applying a direct current through the first excitation coil.

[0069] Based on the circumferential background magnetization field, a high-frequency alternating current is applied through the second excitation coil to superimpose a high-frequency detection magnetic field inside the tube wall to form a composite magnetic field;

[0070] A magnetic sensor array, linearly deployed at equal intervals along the axial direction of a buried metal pipeline, is used to sense changes in magnetic flux density in response to a composite magnetic field.

[0071] The magnetic flux density changes sensed by the magnetic sensor array are converted into multiple hybrid magnetic response signals by a multi-channel synchronous data acquisition device.

[0072] In this embodiment, a first excitation coil for generating a low-frequency strong magnetic field and a second excitation coil for generating a high-frequency weak magnetic field are coaxially mounted and deployed in the target detection section. The purpose is to simultaneously establish magnetic fields of different frequencies within the wall of the buried metal pipeline. Coaxial mounting of the two coils ensures that the generated magnetic fields are optimally coupled to the circumferential and axial directions of the pipeline, forming a uniform and effective excitation area. A direct current is applied through the first excitation coil, thereby utilizing the ferromagnetism of the metal pipeline and forming a stable magnetic field baseline within it through external excitation. The principle is that when a direct current or ultra-low frequency current passes through the first excitation coil, a magnetic field is generated around the coil. Since the pipeline is made of ferromagnetic material, this magnetic field penetrates the pipeline wall and induces magnetization in the circumferential direction, forming a circumferential background magnetization field. This background magnetization field is highly sensitive to quasi-static stresses in the pipeline material (such as residual stress and stress changes caused by long-term loads), because stress alters the material's permeability and magnetic domain structure through magnetostriction, thus affecting the distribution of the magnetization field. Establishing a background magnetization field can provide a reference for subsequent detection of high-frequency transient stress events, enabling the differentiation of magnetic responses from different sources in complex environments.

[0073] Next, in the established circumferential background magnetization field When present, applying a high-frequency alternating current through the second excitation coil will generate a high-frequency detection magnetic field. This high-frequency detection magnetic field superimposes with the background magnetic field to form a composite magnetic field. High-frequency magnetic fields exhibit higher sensitivity to minute and rapid changes in permeability in pipe materials caused by transient stress waves (such as stress waves induced by crack initiation, propagation, or external impact). By superimposing a high-frequency weak magnetic field, rapidly changing pipe stress events indicating potential pipe damage can be effectively detected without interfering with the background magnetization field's quasi-static stress characterization. When stress changes occur inside the pipe, whether quasi-static stress or pipe stress events, they cause local changes in the permeability of the pipe material through magnetostriction, leading to corresponding changes in the magnetic flux density in the composite magnetic field. Multiple magnetic sensors linearly deployed at equal intervals along the pipe axis can synchronously sense and measure these local magnetic flux density changes. This array deployment not only provides magnetic response information at different locations along the pipe but also accurately determines the location of transient stress events using the distance between sensors and the signal arrival time difference, greatly improving the spatial resolution and localization capability of the detection.

[0074] Since the changes in magnetic flux density sensed by magnetic sensors are typically output as analog voltage signals, these analog signals must be converted into digital signals by an analog-to-digital converter. Therefore, a multi-channel synchronous data acquisition device can simultaneously acquire data from all sensors in the array, and the sampling of all channels is precisely synchronized by a common clock. This ensures that at any given sampling moment, the outputs of all sensors are recorded simultaneously, forming a digitized hybrid magnetic response signal.

[0075] In one implementation, a blind source separation algorithm based on independent component analysis is used to separate the hybrid magnetic response signal into a slowly varying component representing quasi-static background stress and a sudden component representing pipeline stress events. Extracting the sudden component includes the following steps:

[0076] The acquired multi-channel hybrid magnetic response signals are segmented and an observation signal matrix is ​​constructed.

[0077] The optimization objective is to maximize non-Gaussianity. The fast fixed-point algorithm is used to iteratively calculate the observed signal matrix to obtain the unmixing matrix.

[0078] The observed signal matrix is ​​linearly transformed using the unmixing matrix to output multiple independent signal components.

[0079] Calculate the statistical characteristic parameters of multiple signal components, identify the slowly varying components and burst components based on the statistical characteristic parameters, and extract the signal components identified as burst components.

[0080] In this implementation, the original continuous multi-channel hybrid magnetic response signal data volume can be extremely large, and direct processing would increase the computational burden. Data segmentation divides the long sequence signal into several more manageable short-time-window data segments, which also helps to satisfy the assumption typically made by independent component analysis algorithms that the signal is stationary or quasi-stationary within the analysis window. Constructing the observed signal matrix... This involves organizing the segmented multi-channel signals into a standard data structure, where each row can represent the sampling sequence of a sensor channel over a period of time, or each column can represent the sampling values ​​of all sensors at a certain moment. Since independent source signals are typically non-Gaussian distributed, while the linear mixture of multiple independent source signals tends to be Gaussian, maximizing the non-Gaussianity of the separated signal can effectively identify these independent source signals. The Fast Fixed-Point Algorithm in Independent Component Analysis (ICA) iteratively optimizes to find a projection direction that maximizes the non-Gaussianity of the projected signal. After multiple iterations, the algorithm eventually converges and solves for an unmixing matrix. This matrix contains all the information needed to convert the observed mixed signal back into independent source signals.

[0081] Next, a linear transformation is performed on the observed signal matrix using the demixing matrix, outputting multiple independent signal components. This step applies the demixing rules obtained in the previous step to the original mixed signal, thereby achieving signal separation. The demixing matrix is ​​a linear transformation matrix that can decouple the mixed signals in the observed signal matrix. This is achieved through simple matrix multiplication, i.e. The observed signal matrix is ​​linearly transformed into an independent component matrix. In the formula, This is an independent component matrix, where each row represents an estimated independent signal component. These output signal components are statistically independent, representing different physical processes or information sources; for example, one component might represent background stress, and another a transient event. The statistical characteristic parameters of these independent signal components can then be calculated, and gradually varying components and sudden burst components can be identified based on these parameters. Different types of physical events produce signals with different statistical properties. Pipeline stress events typically manifest as short-duration, high-amplitude pulse signals with sharp waveforms, resulting in high kurtosis and sparsity values. In contrast, signal changes caused by quasi-static background stress are usually smoother, more closely resembling a Gaussian distribution, with lower kurtosis and sparsity values. Therefore, the kurtosis and sparsity values ​​are calculated for each independent signal component. Kurtosis value The calculation formula is:

[0082] ,

[0083] in As independent signal components, Its mean, Its standard deviation.

[0084] sparsity value The calculation formula is: Based on preset kurtosis and sparsity thresholds, signal components exhibiting both high kurtosis and high sparsity that meet the burst characteristics are identified as burst components, while signal components meeting the gradual variation characteristics are identified as gradually varying components. Since pipeline stress events are a key factor causing pipeline fatigue damage, it is necessary to accurately filter all signals identified as burst components from the set of independent components. These extracted burst components have been freed from background noise and interference from gradually varying stress signals, resulting in a higher signal-to-noise ratio and purer physical information.

[0085] In one embodiment, the statistical characteristic parameters of multiple signal components are calculated respectively, and the slowly varying components and burst components are identified based on the statistical characteristic parameters. The extraction of the signal components identified as burst components includes the following steps:

[0086] Calculate the kurtosis and sparsity values ​​for each signal component separately;

[0087] For any signal component, if the kurtosis value is greater than a preset first kurtosis threshold and the sparsity value is greater than a preset first sparsity threshold, then the signal component is identified as a burst component and the burst component is extracted.

[0088] If a signal component has a kurtosis value less than a preset second kurtosis threshold and a sparsity value less than a preset second sparsity threshold, then the signal component is identified as a slowly varying component, wherein the first kurtosis threshold is greater than the second kurtosis threshold and the first sparsity threshold is greater than the second sparsity threshold.

[0089] In this embodiment, different types of signals exhibit significant differences in statistical distribution. Signals representing pipeline stress events typically have sharp waveforms and sparse energy distributions, demonstrating high non-Gaussianity. Signals representing quasi-static background stress, on the other hand, may be smoother, with a statistical distribution closer to a Gaussian distribution. A first kurtosis threshold and a first sparsity threshold are set to identify abrupt components. These thresholds can be set based on experience, experimental data, or domain knowledge; they represent the appropriate threshold range determined by analyzing known abrupt event signals. A second kurtosis threshold and a second sparsity threshold are set to identify slowly varying components. Slowly varying components typically represent quasi-static background stress or environmental noise in the pipeline. These signals are characterized by smooth changes, lacking sharp impulse characteristics, and their statistical distribution may be closer to a Gaussian distribution or have lower sparsity. Therefore, the second set of thresholds is lower than the first set of thresholds used to identify abrupt components to reflect the smoothness and non-impulse characteristics of the slowly varying component signals.

[0090] For each independent signal component, the calculated kurtosis value is compared with a first kurtosis threshold, and its sparsity value is also compared with a first sparsity threshold. Only when both the kurtosis and sparsity values ​​of a signal component are greater than the first kurtosis threshold are the signal component explicitly identified as a sudden component representing a pipeline stress event. Signal components with kurtosis and sparsity values ​​less than a second kurtosis threshold are identified as slowly varying components. Similarly, for each independent signal component, the calculated kurtosis value is compared with a second kurtosis threshold, and its sparsity value is also compared with a second sparsity threshold. Only when both the kurtosis and sparsity values ​​of a signal component are less than the second kurtosis threshold are the signal component explicitly identified as a slowly varying component representing quasi-static background stress.

[0091] In one embodiment, the process of reconstructing the original waveform characteristics of the pipeline stress event at its initial position without propagation distortion by performing dispersion cancellation inverse operation on the burst component based on a preset stress wave-magnetic coupling dispersion model includes the following steps:

[0092] The initial spectrum of the burst component is obtained by transforming the burst component from the time domain to the frequency domain;

[0093] A stress wave-magnetic coupling dispersion model for buried metal pipelines is pre-established, and the nonlinear mapping relationship between stress wave frequency and phase velocity is obtained through the stress wave-magnetic coupling dispersion model.

[0094] Based on the nonlinear mapping relationship, frequency-related phase correction is performed on each frequency component in the initial spectrum to obtain the corrected spectrum;

[0095] The corrected spectrum is converted back to the time domain by inverse Fourier transform to obtain the compressed pulse signal. The waveform of the pulse signal is used as the original waveform feature of the pipeline stress event at the initial position without propagation distortion.

[0096] In this embodiment, the burst component is transformed from the time domain to the frequency domain. The time-domain signal contains the superposition information of all frequency components, but the amplitude and phase relationships of each frequency component are not directly apparent. Through Fourier transform, the time-domain signal can be decomposed into a series of sine and cosine waves of different frequencies, thereby obtaining the signal's spectrum, which contains the amplitude and phase information of each frequency component. Specifically, the time-domain burst component extracted from independent component analysis... The initial spectrum is obtained by applying the Fast Fourier Transform. Spectrum It is a complex function, where the magnitude represents the intensity of the frequency component and the phase represents the phase of the frequency component relative to the time origin. Next, a pipe stress wave-magnetically coupled dispersion model is loaded to obtain the nonlinear mapping relationship between the stress wave frequency and the phase velocity. Since stress waves of different frequencies typically propagate at different speeds in solid media, to accurately eliminate the dispersion effect in signal propagation, it is necessary to precisely understand the propagation speed of stress waves in the pipe material at different frequencies. The pipe stress wave-magnetically coupled dispersion model is pre-established through finite element simulation and data fitting, storing the stress wave frequency f and its corresponding phase velocity. The model establishes a nonlinear mapping relationship between these parameters. By loading this model, the precise phase velocity of a stress wave propagating in a pipe at any given frequency can be determined.

[0097] If different frequency components propagate at different speeds, their phase relative to the original waveform will shift upon reaching the sensor, resulting in waveform broadening and distortion. To restore the original waveform, a reverse phase shift needs to be applied to each frequency component so that it can be realigned upon reaching the sensor. In practice, for the initial spectrum... For each frequency component f in the spectrum, its corresponding phase velocity is obtained using the applied dispersion model. Assuming the propagation distance from the event's starting point to the sensor is L, then the cumulative phase shift of this frequency component is... To correct the offset, the initial spectrum is phase-adjusted, and the formula for calculating the corrected spectrum is as follows: , where j is the imaginary unit.

[0098] After phase correction of each frequency component in the frequency domain, the relative phase relationships of these components have been restored to their state at the starting point. The inverse Fourier transform can then be used to resynthesize these phase-corrected frequency components into a time-domain signal. Since the phases of all frequency components are correctly aligned, the waveform, which was originally broadened or distorted due to dispersion, will be refocused, forming a sharper and more concentrated pulse signal. Through the aforementioned series of signal processing steps, especially the dispersion cancellation inverse operation, the resulting compressed pulse signal has restored, to the greatest extent possible, the original form of the stress event at the starting point before propagation distortion. This waveform is no longer affected by the dispersion effect of the pipe material and can more accurately reflect the intrinsic characteristics of the event, such as intensity, duration, and energy distribution. Therefore, the shape and parameters of this compressed pulse signal can be directly used as the original waveform characteristics of the pipe stress event.

[0099] In one embodiment, the pre-establishment of the pipeline stress wave-magnetic coupling dispersion model specifically includes the following steps:

[0100] A three-dimensional finite element model of the buried metal pipeline was established based on its material properties and geometric parameters.

[0101] A broadband impact load is applied at a predetermined location in a three-dimensional finite element model to excite test stress waves of various modes.

[0102] The propagation speed of stress waves of different frequencies along the axial direction of buried metal pipelines was tested by transient dynamic simulation calculation.

[0103] By fitting propagation velocity data of different frequency components, a nonlinear mapping relationship between stress wave frequency and phase velocity is established, and the nonlinear mapping relationship is solidified into a stress wave-magnetic coupling dispersion model for buried metal pipelines.

[0104] In this embodiment, the finite element method can discretize complex continuous structures into a finite number of simple elements. By performing mechanical analysis on these elements and then integrating the results, the response of the entire structure can be approximately solved. Specifically, detailed material properties of the buried metal pipeline, such as elastic modulus, Poisson's ratio, and density, need to be input. These parameters determine the propagation characteristics of stress waves in the material. Simultaneously, the geometric dimensions of the pipeline, such as length, outer diameter, and wall thickness, are input to precisely define the physical boundaries of the model. A three-dimensional mesh model reflecting these physical properties is constructed using finite element software, and appropriate element types and mesh densities are set to ensure the model accurately captures the propagation details of stress waves. Then, a broadband impact load is applied at a predetermined location on the three-dimensional finite element model to excite multi-mode test stress waves, thereby generating stress waves covering a wide frequency range in the virtual pipeline for comprehensive analysis of their dispersion characteristics. To study the propagation speed of stress waves at different frequencies, an excitation source containing these frequencies is needed. Impact loads are typically characterized by concentrated energy and short duration. They exhibit broadband characteristics in the frequency domain and can excite various stress wave modes present in the pipeline, such as longitudinal waves, torsional waves, and bending waves.

[0105] Next, transient dynamic simulation is used to calculate and extract the propagation velocities of different frequency components of the stress wave as it propagates along the pipe axis. Transient dynamic simulation can simulate the structural response to dynamic loads in the time domain, recording the changes in displacement, velocity, or stress at any point inside the pipe over time. By setting virtual monitoring points at different locations along the pipe axis, time-series signals of the stress wave passing through these points can be recorded. Subsequently, these signals are processed, for example, by using Fourier transform to convert the time-domain signals to the frequency domain. Then, the propagation velocity of the frequency component is determined by calculating the phase difference or arrival time difference of the same frequency component between different monitoring points. Although transient dynamic simulation provides a large number of discrete data points, a continuous function that can describe the entire frequency range is needed for dispersion correction of arbitrary frequencies in practical applications. Therefore, curve fitting techniques are used to establish a nonlinear mapping relationship between the stress wave frequency and its phase velocity. Specifically, multiple propagation velocity data points can be used as input, and a suitable nonlinear function model (such as a polynomial function or exponential function) can be selected and fitted using the least squares method or other optimization algorithms to obtain an optimal fitting function. This function can accurately describe the dispersion characteristics of the stress wave propagating in the pipe, i.e., the difference in propagation velocities of different frequency components. Finally, the fitted function expression, function parameters, or lookup table can be saved as a separate software module or database file, which is the stress wave-magnetic coupling dispersion model. Here, the magnetic coupling is reflected in the fact that this dispersion model serves the processing of magnetic response signals induced by stress waves in magnetic field detection methods.

[0106] In one implementation, the stress wave-magnetic coupling dispersion model can be dynamically updated before performing dispersion inverse operations. Specific steps include:

[0107] Background vibration signals below a preset amplitude are continuously extracted from the hybrid magnetic response signal as calibration signals;

[0108] Cross-correlation interference calculations are performed on calibration signals acquired by two different magnetic sensors to obtain the empirical Green's function of the calibration wave propagating along the pipe axis;

[0109] The actual phase velocity of the calibration wave at different frequencies is obtained analytically from the empirical Green's function;

[0110] The nonlinear mapping relationship in the stress wave-magnetic coupling dispersion model is corrected based on the actual phase velocity to obtain the updated stress wave-magnetic coupling dispersion model.

[0111] In this embodiment, buried pipelines are subjected to low-amplitude background vibrations caused by various factors such as fluid flow, geological disturbances, and environmental noise during daily operation. These vibrations generate weak stress waves in the pipeline material, which in turn induce low-amplitude magnetic response signals that can be captured by magnetic sensors. By filtering the continuously acquired mixed magnetic response signals, these low-amplitude background vibration signals, which are unrelated to transient high-frequency stress events, can be effectively removed. The calibration signal is extracted from the complex signal. By performing cross-correlation calculations on the background noise signals collected by two different sensors, the Green's function (i.e., impulse response) of the propagation medium between the two sensors can be approximated. This is equivalent to virtually applying an impulse at the location of one sensor and observing the response at the other sensor. In specific implementation, for any two sensors a and b in a magnetic sensor array deployed along the pipe axis, the calibration signal is obtained... and Calculate their cross-correlation function:

[0112] ,

[0113] in It's a time lag. The empirical Green's function contains the time and phase information of the calibration wave propagating between the two sensors. By performing frequency domain analysis on the Green's function, the propagation characteristics of different frequency components can be separated. Specifically, the empirical Green's function can be Fourier transformed to obtain its frequency domain representation. From the frequency domain representation, the phase difference of the calibration wave between sensors a and b at different frequencies can be extracted. The distance between sensors a and b is known to be... The actual phase velocity of the calibration wave at frequency F can be calculated using the following formula: .

[0114] By comparing the actual phase velocity extracted from the background vibration with the model prediction, the parameters of the nonlinear mapping relationship in the dispersion model can be adjusted or refitted. Specifically, by using optimization algorithms (such as the least squares method) to correct the parameters describing the relationship between stress wave frequency and phase velocity in the pipeline stress wave-magnetic coupling dispersion model, an updated nonlinear mapping relationship can be obtained and solidified into the updated dispersion model.

[0115] In one implementation, the location information of the pipeline stress event on the buried metal pipeline is determined by combining the arrival time difference and energy attenuation of the original stress wave corresponding to the original waveform characteristics on different sensors in the magnetic sensor array, through joint optimization solution, including the following steps:

[0116] Extract the pulse peak arrival time and pulse peak amplitude of the original stress wave corresponding to the original waveform characteristics from each magnetic sensor channel in the magnetic sensor array;

[0117] Calculate the arrival time difference between any two magnetic sensors based on the pulse peak arrival time;

[0118] Calculate the energy attenuation of each magnetic sensor relative to the reference sensor based on the pulse peak amplitude;

[0119] A joint optimization objective function containing arrival time difference error term and energy attenuation error term is constructed. A nonlinear least squares algorithm is used to iteratively optimize the joint optimization objective function. The solution that minimizes the objective function is determined as the location of the pipeline stress event.

[0120] In this embodiment, after dispersion cancellation, the waveform of the transient high-frequency stress event is reconstructed into a sharper, more concentrated pulse signal. This clear pulse signal allows for high-precision identification of its peak position and peak intensity. Specifically, for the original waveform characteristics received by each sensor channel, signal processing algorithms (such as finding the maximum signal value or curve fitting) can be used to determine the time point of the pulse peak and the corresponding signal amplitude. These precisely extracted time and amplitude values ​​directly reflect the time and intensity of the stress wave reaching each sensor. The time required for the stress wave to propagate from the origin point to sensors at different locations is different, and this time difference is closely related to the distance from the event origin point to each sensor and the propagation speed of the stress wave. By measuring these time differences, the origin location of the event can be geometrically located. For any two sensors k and l in the magnetic sensor array, the pulse peak arrival time extracted in the previous step is used... and Calculate the time difference of arrival between them: On the other hand, as stress waves propagate through the pipe, their energy gradually attenuates due to material damping, geometric diffusion, and other factors, resulting in a decrease in signal amplitude. The degree of attenuation is related to the propagation distance; the greater the distance, the greater the attenuation. Therefore, this amplitude attenuation pattern can serve as another important clue for event localization. A sensor with the largest signal amplitude or a sensor located at the center of the array should be selected as a reference sensor. Its pulse peak amplitude is Then, for each sensor in the array... Utilizing its pulse peak amplitude Calculate its amplitude ratio relative to the reference sensor: This amplitude ratio reflects the energy attenuation of the stress wave relative to the reference sensor as it propagates from its origin to the sensor.

[0121] The true location of the event should minimize the difference between the predicted signal arrival time and amplitude attenuation based on that location and the actual observed values. Therefore, we will integrate all the observed spatiotemporal information into a mathematical framework to find the optimal event location through optimization methods. Specifically, we can construct an objective function that quantifies the sum of these differences. This function consists of two parts: one part is the difference between the observed arrival time difference and the predicted event location. The objective function is composed of two parts: the sum of squared errors between the predicted time differences of arrival and the sum of squared errors between the observed amplitude ratio and the predicted amplitude ratio. The objective function takes the form:

[0122]

[0123] in, It is the predicted time difference of arrival. It is the predicted magnitude ratio.

[0124] The constructed joint optimization objective function is usually nonlinear, and its minimum cannot be directly solved using simple algebraic methods. Nonlinear least squares algorithms can iteratively adjust the estimated location of the event's effect, continuously approximating the global minimum of the objective function. The algorithm starts from an initial guessed location, and in each iteration, adjusts the estimate based on the objective function... The gradient information is used to calculate an update direction and step size, thereby updating... The value of the objective function is determined. This process continues until the value of the objective function converges to a minimum, or... The change is less than the preset threshold. Ultimately, this minimizes the objective function. The value is the precise location at which the pipeline stress event is determined.

[0125] In one embodiment, calculating the cumulative damage level of the buried metal pipeline target detection section based on the energy and location information of each pipeline stress event includes the following steps:

[0126] The three-dimensional finite element model of the buried metal pipeline is discretized into multiple damage calculation units along the pipeline axis;

[0127] Calculate the equivalent stress amplitude of each damage calculation unit based on the energy and location information of the pipeline stress event;

[0128] Based on the material SN curve of buried metal pipelines, find the allowable number of cycles corresponding to the equivalent stress amplitude;

[0129] The reciprocal of the allowable number of cycles is taken as the damage increment caused by this pipeline stress event to the corresponding damage calculation unit;

[0130] The damage increment is added to the cumulative damage factor of the corresponding damage calculation unit to complete the cumulative damage calculation of the target detection section of the buried metal pipeline.

[0131] In this embodiment, the continuous pipeline structure is transformed into a discrete analysis object for localized damage assessment. The principle is that pipeline fatigue damage is typically localized; to accurately track and predict damage occurrence and development, the entire pipeline needs to be divided into smaller, manageable regions. Specifically, the three-dimensional finite element model of the pipeline can be divided along its axis, forming a series of adjacent damage calculation units with the same or different lengths. Each unit represents a specific region of the pipeline, and its internal material properties and geometric features are considered homogeneous. Pipeline stress events (such as internal and external pressure, temperature changes, and stress changes caused by physical forces) generate stress waves. These stress waves propagate through the pipeline, causing instantaneous stress responses in the pipeline material. Since the energy of stress waves attenuates with increasing propagation distance, the stress amplitude caused by the event varies at different damage calculation units. For each located pipeline stress event, its energy is... The location of action is For each damage calculation unit on the pipeline. Its central position is The impact of this event on the stress propagation and attenuation model is calculated using a pre-defined stress propagation and attenuation model. The resulting equivalent stress amplitude The stress propagation and attenuation model can be expressed as:

[0132]

[0133] in It is the energy-to-stress conversion coefficient. It is the stress attenuation index. It is the energy of pipeline stress events. It is the center location of the damage calculation unit. It is the location where the pipeline stress event occurred.

[0134] Then, based on the SN curve of the buried metal pipeline material, the allowable number of cycles corresponding to the equivalent stress amplitude is found. The SN curve (stress-cycle curve) is a key characteristic curve describing the fatigue performance of materials, reflecting the number of cycles the material can withstand under different stress amplitudes. For metallic materials, the higher the stress amplitude, the shorter the fatigue life (allowable number of cycles). The equivalent stress amplitude of each damage calculation unit is then calculated. By querying or calculating the SN curve, the allowable number of cycles corresponding to the stress amplitude can be found. The allowable number of cycles represents the number of cycles required for the material to undergo fatigue failure at a given stress amplitude. Based on Miner's rule, it is assumed that the damage caused by each loading cycle is independent and can accumulate linearly. A single transient stress event can be considered as one loading cycle. If the material can withstand a certain stress amplitude for a total of [number of cycles] cycles, then [the allowable number of cycles] is the number of cycles required for fatigue failure. It will only fail after several cycles, so the damage ratio caused by a single cycle is... Therefore, for each damage calculation unit, the incremental damage caused by this event to that unit is calculated based on the allowable number of cycles obtained from its equivalent stress amplitude. .

[0135] Pipeline fatigue damage is a continuous cumulative process; each pipeline stress event causes some damage. This is addressed by accumulating the damage increment from each event into the cumulative damage factor of the corresponding damage calculation unit. In this system, the degree of damage at various locations along the pipeline can be tracked in real time. The update formula for the cumulative damage factor is as follows: ,when If a certain critical value is reached, the unit is considered to have undergone fatigue failure.

[0136] The present invention also discloses a dual-source magnetic field integrated detection and analysis system based on buried metal pipelines, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the dual-source magnetic field integrated detection and analysis method based on buried metal pipelines as described in any of the above embodiments.

[0137] The processor can be a central processing unit (CPU). Of course, depending on the actual use, it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), off-the-shelf programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc., and this application does not limit it.

[0138] The memory can be an internal storage unit of a computer device, such as a hard disk or RAM, or an external storage device, such as a plug-in hard disk, smart memory card (SMC), secure digital card (SD), or flash memory card (FC) provided on the computer device. Furthermore, the memory can be a combination of internal storage units and external storage devices of a computer device. The memory is used to store computer programs and other programs and data required by the computer device. The memory can also be used to temporarily store data that has been output or will be output. This application does not limit this.

[0139] The present invention also discloses a computer-readable storage medium storing instructions that, when executed by a processor, configure the processor to perform the dual-source magnetic field integrated detection and analysis method based on buried metal pipelines as described in any of the above embodiments.

[0140] The computer program can be stored in a machine-readable medium. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or certain middleware. The machine-readable medium includes any entity or device capable of carrying computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the machine-readable medium includes, but is not limited to, the above-mentioned components.

[0141] The dual-source magnetic field integrated detection and analysis method based on buried metal pipelines described in the above embodiments is stored in the computer-readable storage medium and loaded and executed on the processor to facilitate the storage and application of the above method.

[0142] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of protection of this application is limited to these examples; within the framework of this application, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of one or more embodiments of this application as described above, which are not provided in detail for the sake of brevity.

[0143] One or more embodiments in this application are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of this application. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of one or more embodiments in this application should be included within the protection scope of this application.

Claims

1. A dual-source magnetic field integrated detection analysis method based on buried metal pipeline, characterized in that, Includes the following steps: A composite magnetic field consisting of a low-frequency strong magnetic field and a high-frequency weak magnetic field is applied to the target detection section of the buried metal pipeline for excitation, and a magnetic sensor array deployed along the axial direction of the buried metal pipeline is used to collect the mixed magnetic response signal in response to the composite magnetic field. A blind source separation algorithm based on independent component analysis is used to separate the hybrid magnetic response signal into a slowly varying component representing quasi-static stress and a burst component representing transient stress, and the burst component is extracted. The initial spectrum of the burst component is obtained by transforming the burst component from the time domain to the frequency domain; A three-dimensional finite element model of the buried metal pipeline was established based on its material properties and geometric parameters. A broadband impact load is applied at a predetermined location in a three-dimensional finite element model to excite test stress waves of various modes. The propagation speed of stress waves of different frequencies along the axial direction of buried metal pipelines was tested by transient dynamic simulation calculation. By fitting propagation velocity data of different frequency components, a nonlinear mapping relationship between stress wave frequency and phase velocity is established, and the nonlinear mapping relationship is solidified into a stress wave-magnetic coupling dispersion model for buried metal pipelines. The nonlinear mapping relationship between stress wave frequency and phase velocity was obtained by using a stress wave-magnetic coupling dispersion model. Based on the nonlinear mapping relationship, frequency-related phase correction is performed on each frequency component in the initial spectrum to obtain the corrected spectrum; The corrected spectrum is converted back to the time domain by inverse Fourier transform to obtain the compressed pulse signal. The waveform of the pulse signal is used as the original waveform feature of the pipeline stress event at the initial position without propagation distortion. By combining the arrival time difference and energy attenuation of the original stress wave corresponding to the original waveform characteristics on different sensors in the magnetic sensor array, the location information of pipeline stress events on buried metal pipelines is determined by joint optimization solution. The three-dimensional finite element model of the buried metal pipeline is discretized into multiple damage calculation units along the pipeline axis; Calculate the equivalent stress amplitude of each damage calculation unit based on the energy and location information of the pipeline stress event; Based on the material SN curve of buried metal pipelines, find the allowable number of cycles corresponding to the equivalent stress amplitude; The reciprocal of the allowable number of cycles is taken as the damage increment caused by this pipeline stress event to the corresponding damage calculation unit; The damage increment is added to the cumulative damage factor of the corresponding damage calculation unit to complete the cumulative damage calculation of the target detection section of the buried metal pipeline.

2. The buried metal pipeline-based dual-source magnetic field comprehensive detection analysis method according to claim 1, characterized in that, The process of applying a composite magnetic field consisting of a low-frequency strong magnetic field and a high-frequency weak magnetic field to the target detection section of the buried metal pipeline for excitation, and acquiring the mixed magnetic response signal in response to the composite magnetic field through a magnetic sensor array deployed along the axial direction of the buried metal pipeline, includes the following steps: The first excitation coil for generating a low-frequency strong magnetic field and the second excitation coil for generating a high-frequency weak magnetic field are coaxially mounted and deployed in the target detection section of the buried metal pipeline. A circumferential background magnetization field is established inside the wall of the buried metal pipeline by applying a direct current through the first excitation coil. Based on the circumferential background magnetization field, a high-frequency alternating current is applied through the second excitation coil to superimpose a high-frequency detection magnetic field inside the tube wall to form a composite magnetic field; A magnetic sensor array, linearly deployed at equal intervals along the axial direction of a buried metal pipeline, is used to sense changes in magnetic flux density in response to a composite magnetic field. The magnetic flux density changes sensed by the magnetic sensor array are converted into multiple hybrid magnetic response signals by a multi-channel synchronous data acquisition device.

3. The method for comprehensive detection and analysis of dual-source magnetic fields based on buried metal pipelines according to claim 2, characterized in that, The method of using a blind source separation algorithm based on independent component analysis to separate the mixed magnetic response signal into a slowly varying component representing quasi-static background stress and a sudden component representing pipeline stress events, and extracting the sudden component includes the following steps: The acquired multi-channel hybrid magnetic response signals are segmented and an observation signal matrix is ​​constructed. The optimization objective is to maximize non-Gaussianity. The fast fixed-point algorithm is used to iteratively calculate the observed signal matrix to obtain the unmixing matrix. The observed signal matrix is ​​linearly transformed using the unmixing matrix to output multiple independent signal components. Calculate the statistical characteristic parameters of multiple signal components, identify the slowly varying components and burst components based on the statistical characteristic parameters, and extract the signal components identified as burst components.

4. The method for comprehensive detection and analysis of dual-source magnetic fields based on buried metal pipelines according to claim 3, characterized in that, The step of calculating the statistical characteristic parameters of multiple signal components, identifying slowly varying components and burst components based on the statistical characteristic parameters, and extracting the signal components identified as burst components includes the following steps: Calculate the kurtosis and sparsity values ​​for each signal component separately; For any signal component, if the kurtosis value is greater than a preset first kurtosis threshold and the sparsity value is greater than a preset first sparsity threshold, then the signal component is identified as a burst component and the burst component is extracted. If a signal component has a kurtosis value less than a preset second kurtosis threshold and a sparsity value less than a preset second sparsity threshold, then the signal component is identified as a slowly varying component, wherein the first kurtosis threshold is greater than the second kurtosis threshold and the first sparsity threshold is greater than the second sparsity threshold.

5. The buried metal pipeline-based dual-source magnetic field comprehensive detection analysis method according to claim 1, characterized in that, The process of determining the location information of pipeline stress events on buried metal pipelines by combining the arrival time difference and energy attenuation of the original stress wave corresponding to the original waveform characteristics on different sensors in the magnetic sensor array, and through joint optimization, includes the following steps: Extract the pulse peak arrival time and pulse peak amplitude of the original stress wave corresponding to the original waveform characteristics from each magnetic sensor channel in the magnetic sensor array; Calculate the arrival time difference between any two magnetic sensors based on the pulse peak arrival time; Calculate the energy attenuation of each magnetic sensor relative to the reference sensor based on the pulse peak amplitude; A joint optimization objective function containing arrival time difference error term and energy attenuation error term is constructed. A nonlinear least squares algorithm is used to iteratively optimize the joint optimization objective function. The solution that minimizes the objective function is determined as the location of the pipeline stress event.

6. A buried metal pipeline-based dual-source magnetic field comprehensive detection analysis system, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the dual-source magnetic field integrated detection and analysis method based on buried metal pipelines as described in any one of claims 1 to 5.

7. A computer-readable storage medium having stored thereon instructions, the computer-readable storage medium comprising: When executed by a processor, this instruction causes the processor to be configured to perform the dual-source magnetic field integrated detection and analysis method based on buried metal pipelines according to any one of claims 1 to 5.

Images