A multi-modal guided wave industrial pipeline diagnosis method

Abstract

The application relates to the technical field of nondestructive testing of industrial pipelines, and discloses a multi-modal guided wave industrial pipeline diagnosis method. The method generates annular detection sections along the pipeline and separates corresponding data clusters from signals; material characteristic records containing inverted acoustic impedance distribution are established for each section. By comparing the acoustic impedance of all sections, a continuous section set is identified to define a continuous main body, and different lining characteristic areas are divided according to step changes. Independent wall thickness reference models are constructed for each lining characteristic area, homogeneous sound velocity estimates are calculated based on the data of the continuous main body area, and then the theoretical expected interval of the wall thickness of each section is derived. The actual wall thickness parameters of each section are compared with the theoretical interval, and when deviation occurs, the interval center value is replaced to generate a corrected diagnosis basis. The method realizes the structure differentiation and parameter self-checking of the composite pipeline, and improves the accuracy and reliability of the diagnosis result.

Description

Technical Field

[0001] This invention relates to the field of non-destructive testing technology for industrial pipelines, specifically a diagnostic method for industrial pipelines based on multimode guided waves. Background Technology

[0002] In the field of guided wave nondestructive testing (NDT) for industrial pipelines, existing technologies typically analyze the entire pipeline as a homogeneous whole, or rely solely on single signal reflection characteristics for defect identification and location. For pipelines with composite structures or axial material variations, these methods cannot effectively distinguish whether the guided wave signal response originates from changes in the pipeline's main structure, alterations in the lining, or interface effects at the junction of the two. This leads to ambiguity in defect type identification, making it difficult to accurately delineate the structural integrity boundaries of the pipeline body and its lining, thus affecting the specificity of the diagnosis.

[0003] Existing wall thickness assessment techniques primarily rely on the direct inversion calculation of the relationship between guided wave propagation velocity and signal arrival time. This method implicitly assumes that the acoustic properties of the pipe material are constant and known within the detection range. However, in real industrial environments, material aging, stress state changes, and measurement noise can all cause local fluctuations in sound velocity and signal distortion. These factors directly lead to deviations in the wall thickness inversion results, and existing methods lack effective internal verification mechanisms to identify and correct such systematic or random errors, posing challenges to the reliability and robustness of the diagnostic results.

[0004] There is a need for a guided wave diagnostic method that can adaptively distinguish complex pipeline structures and effectively self-verify key inversion parameters in order to improve the accuracy and reliability of condition assessment for complex industrial pipelines. Summary of the Invention

[0005] The purpose of this invention is to provide a diagnostic method for industrial pipelines based on multimode guided waves, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides a diagnostic method for industrial pipelines based on multimode guided waves, the method comprising: Several preset ring detection sections are generated based on the pipeline extension path, and the original data clusters corresponding to each ring detection section are separated from the multi-mode guided wave detection signal. Material feature records are established for each annular detection section. The material feature records include the acoustic impedance distribution of the main pipe structure and the acoustic impedance distribution of the pipe lining structure generated by inversion using the original data cluster of the section. By comparing the acoustic impedance distribution of the main pipe structure in all annular inspection sections, a set of sections with continuous acoustic impedance distribution is identified to define continuous main sections. By comparing the acoustic impedance distribution of the pipe lining structure in all annular inspection sections, the boundaries of sections with step changes are identified to divide them into different lining feature areas. An independent wall thickness benchmark model is constructed for each lining feature area, and the homogeneous sound velocity estimate of the main pipe structure is calculated based on the material feature records of each annular detection section within the continuous main body section. By combining the homogeneous sound velocity estimate of the main pipe structure with the wall thickness benchmark model, the theoretical expected range of wall thickness for each annular detection section is derived. For each annular detection segment, the actual wall thickness parameters contained in its material feature record are compared and verified with the theoretical expected range of wall thickness. When the actual wall thickness parameters deviate, the central reference value of the theoretical expected range of wall thickness is used for replacement, generating a corrected diagnostic basis.

[0007] Preferably, the step of generating several preset annular detection segments based on the pipeline extension path and separating the original data clusters corresponding to each annular detection segment from the multimode guided wave detection signal includes: A series of equally spaced virtual inspection sections are set along the pipeline axis; With each virtual detection section as the center, a preset capture length is extended to both sides of the pipeline axis, thereby forming a ring detection section around the pipeline; Receive a multimode guided wave detection signal, wherein the multimode guided wave detection signal contains a mixed echo sequence generated by the interaction of guided waves of different modes; Based on the propagation speed of the guided wave in the pipe and the geometric center position of each annular detection section, the theoretical two-way propagation time window of each mode guided wave from emission to arrival at the annular detection section and return is calculated. Based on the theoretical two-way propagation time window, all waveform amplitude data within a specific time segment associated with each ring detection segment are extracted from the mixed echo sequence; All waveform amplitude data corresponding to each ring detection segment are categorized and stored according to the guided wave mode type to form the original data cluster of the segment.

[0008] Preferably, the step of establishing a material feature record for each annular detection segment includes, A time-frequency joint analysis was performed on the original data cluster of the section to separate the reflectance energy spectrum components corresponding to the main pipe material and the reflectance energy spectrum components corresponding to the pipe lining material; Backscattering analysis is performed on the reflection energy spectrum components of the pipeline body material to reconstruct the acoustic impedance profile of the pipeline body structure distributed circumferentially along the annular detection section, and it is recorded as the acoustic impedance distribution of the pipeline body structure. The reflectance spectrum components of the pipe lining material are processed by tomographic imaging to invert the thickness and density changes of the pipe lining layer in the annular detection section, and quantify them as the acoustic impedance distribution of the pipe lining structure. By integrating the acoustic impedance distribution of the main pipe structure with that of the pipe lining structure, a complete material characteristic record of the annular detection section is formed.

[0009] Preferably, the step of comparing the acoustic impedance distribution of the main pipe structure in all annular detection sections to identify the set of sections with continuous acoustic impedance distribution includes, Calculate the cross-correlation coefficient between the acoustic impedance distributions of the main pipe structure in two adjacent annular detection sections; Set a continuity threshold for the cross-correlation coefficient; Starting from the first loop inspection section at the beginning of the pipeline, each loop inspection section is scanned sequentially. If the cross-correlation coefficient between the current ring detection segment and the next ring detection segment is higher than the continuity determination threshold, then the next ring detection segment is included in the current continuous segment set. If the cross-correlation coefficient between the current ring detection segment and the next ring detection segment is lower than the continuity determination threshold, then the expansion of the current segment set is terminated, and the formed segment set is marked as an independent continuous main segment. Starting from the next ring detection segment, repeat the expansion and termination process until all ring detection segments are classified into a single continuous main segment.

[0010] Preferably, the step of comparing the acoustic impedance distribution of the pipe lining structure in all annular detection sections and identifying the boundary of sections where the acoustic impedance distribution exhibits a step change includes, Calculate the difference in the average acoustic impedance distribution of the pipe lining structure between adjacent annular inspection sections along the pipe extension direction; Simultaneously calculate the difference in the variance of the acoustic impedance distribution of the pipe lining structure between adjacent annular detection sections; Based on historical pipeline operating data, set a step threshold for the average difference and a stability threshold for the variance difference; Adjacent ring detection segment boundaries that simultaneously meet the following conditions are selected: the average difference is greater than the step threshold and the variance difference is less than the stability threshold. The boundaries of the selected adjacent ring-shaped detection sections are defined as the boundaries of sections where the acoustic impedance distribution exhibits a step change.

[0011] Preferably, the step of dividing the lining into different feature areas includes, The pipeline is divided into different lining feature zones based on the section boundaries; Along the pipeline extension path, the boundaries of all identified sections where the acoustic impedance distribution exhibits a step change are used as dividing points. All annular detection segments located between two adjacent dividing points are classified into the same lining feature area; If there is a ring-shaped inspection section between the beginning of the pipeline and the first dividing point, it shall be classified as an independent lining feature area; If there is a ring-shaped detection section between the last dividing point and the end of the pipe, it is classified as another independent lining feature area.

[0012] Preferably, the step of constructing an independent wall thickness reference model for each lining feature area includes, The construction of the wall thickness reference model depends on the acoustic impedance distribution of the pipe lining structure in the entire annular detection section covered by the lining feature area. Extract the acoustic impedance distribution of the pipe lining structure for all annular detection sections belonging to the same lining feature area; The acoustic impedance distribution of all pipe lining structures within the lining feature area is statistically aggregated to generate a lining acoustic impedance probability density function that characterizes the overall properties of the lining feature area. Based on the material acoustic properties database, a mapping relationship from lining acoustic impedance to equivalent thickness is established; The probability density function of the lining acoustic impedance is transformed through the mapping relationship to generate the wall thickness reference probability distribution of the lining feature area, which constitutes the wall thickness reference model.

[0013] The wall thickness reference model defines the statistically normal range of wall thickness values ​​within the lining feature area.

[0014] Preferably, the step of deriving the theoretical expected range of wall thickness for each annular detection section by combining the homogeneous sound velocity estimate of the main pipe structure with the wall thickness benchmark model includes: For each annular detection section, determine the continuous main section to which it belongs, and obtain the homogeneous sound velocity estimate of the pipeline main structure of the continuous main section. Determine the lining feature area to which the annular detection section belongs, and obtain the wall thickness reference model of the lining feature area; Based on the theory of elastic wave propagation in multilayer media, a calculation model is established. The input parameters of the calculation model include the homogeneous sound velocity estimate of the main pipe structure and the wall thickness value in the wall thickness benchmark model. The output is the theoretical guided wave dispersion curve. The actual guided wave propagation parameters derived from the material feature record are input into the calculation model to solve the inverse problem, thereby obtaining a series of wall thickness solution sets that can match the theoretical guided wave dispersion curve with the actual guided wave propagation parameters. From the wall thickness solution set, the wall thickness value with the highest probability of occurrence is selected as the central reference value; Based on the statistical characteristics of the wall thickness benchmark model, a symmetrical tolerance band is set around the central benchmark value. The upper and lower limits of the symmetrical tolerance band constitute the theoretical expected range of the wall thickness.

[0015] Preferably, the step of comparing and verifying the actual wall thickness parameters contained in the material feature record of each annular detection segment with the theoretical expected range of wall thickness includes: The actual wall thickness parameters characterizing the thickness are extracted from the acoustic impedance distribution of the pipe lining structure recorded by the material characteristics of the annular test section. The theoretical expected range of wall thickness derived from the annular detection section is read.

[0016] Determine whether the value of the actual wall thickness parameter is between the lower limit and the upper limit of the theoretical expected range of wall thickness; Record the comparison and verification results.

[0017] Preferably, the step of replacing the actual wall thickness parameter with the central reference value of the theoretical expected range of the wall thickness to generate a corrected diagnostic basis when the actual wall thickness parameter deviates includes: When the actual wall thickness parameter is within the theoretical expected range of the wall thickness, the material characteristic record of the annular detection section is confirmed as a valid diagnostic basis. When the actual wall thickness parameter deviates from the theoretical expected range of the wall thickness, the actual wall thickness parameter is replaced by the central reference value of the theoretical expected range of the wall thickness to generate a corrected diagnostic basis.

[0018] Compared with the prior art, the beneficial effects of the present invention are: By inverting and comparing acoustic impedance distributions, the set of continuous acoustic impedance segments in the main pipe structure and the boundary segments where the acoustic impedance of the lining structure changes abruptly are separated and identified from the multimodal guided wave signals. This technique can objectively divide complex pipe complexes into continuous main pipe segments with stable materials and lining regions with different characteristics based on the structural features revealed by the data itself. This process enables independent identification of the state of the main pipe and lining structures, allowing subsequent analysis to adopt differentiated strategies for structural parts with different physical properties, solving the problem of mixed signal responses, and enhancing the specificity of diagnosis.

[0019] Based on the defined lining feature regions, corresponding wall thickness benchmark models are independently constructed. Combined with homogeneous sound velocity estimates calculated from continuous main body segment data, a theoretical expected wall thickness range is derived for each detection segment. The actual wall thickness parameters obtained from the measured inversion are compared with this theoretical range, and the central benchmark value of the range is used for correction when the data deviates. This method introduces an internal verification mechanism based on a physical model and statistical range, which can effectively identify and suppress abnormal inversion results caused by local material inhomogeneity, sound velocity fluctuations, or measurement noise, thus improving the anti-interference ability and overall reliability of the wall thickness assessment results. Attached Figure Description

[0020] Figure 1 This is a schematic diagram illustrating the working principle of the multi-mode guided wave-based industrial pipeline diagnostic method described in this invention. Figure 2 A flowchart for generating the original data clusters for the ring detection section and the separation section; Figure 3 A flowchart for establishing a record of material characteristics for a ring-shaped inspection section; Figure 4 Comparison chart showing the wall thickness verification of the circular inspection section of an industrial pipeline; Figure 5 A comparison chart showing the effects of correcting indicators for industrial pipeline diagnostic assessment. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Please see Figure 1This invention provides a diagnostic method for industrial pipelines based on multimode guided waves. The method includes: planning and generating a series of preset annular detection sections along the extension path of the industrial pipeline to be inspected; separating the original data clusters of each annular detection section from the acquired multimode guided wave detection signals according to the propagation time window; establishing an independent material characteristic record for each annular detection section, which includes the acoustic impedance distribution of the pipeline main structure and the acoustic impedance distribution of the pipeline lining structure obtained by inverting the original data clusters of the sections; comparing the overall acoustic impedance distribution of the pipeline main structure of all annular detection sections, and identifying continuous main sections composed of multiple sections by analyzing their continuity; comparing the acoustic impedance distribution of the pipeline lining structure of all annular detection sections, and delineating different lining feature areas by identifying step changes in acoustic impedance values; constructing an independent wall thickness benchmark model based on the internal acoustic impedance distribution data of the pipeline lining structure for each lining feature area; and calculating the homogeneous sound velocity estimate of the pipeline main material based on the material characteristic records of each area within the continuous main section. By combining the obtained homogeneous sound velocity estimate of the main pipeline structure with the wall thickness benchmark model of the lining feature area of ​​each annular inspection section, a theoretical expected wall thickness range is derived for each annular inspection section. For each annular inspection section, the actual wall thickness parameters parsed from its material feature record are compared and verified with the theoretical expected wall thickness range of that section. When the actual parameters deviate, the central benchmark value of the theoretical expected wall thickness range is used to replace them, thereby generating a more reliable diagnostic basis after consistency correction.

[0023] In one embodiment of the present invention, see [reference] Figure 2 Based on the pipeline's extension path, several pre-defined annular detection sections are generated, and the original data clusters corresponding to each annular detection section are separated from the multi-mode guided wave detection signals. A series of virtual detection sections are set at equal intervals along the pipeline's axial direction. A pre-defined capture length is symmetrically extended to both sides of the pipeline's axial direction from the center of each virtual detection section, thus forming an annular detection section surrounding the pipeline. The system receives multi-mode guided wave detection signals acquired by a sensor array, which contain a mixed echo sequence generated by the interaction of different guided wave modes. Based on the known propagation speed of each guided wave mode in the pipeline and the geometric center position of each annular detection section, the theoretical two-way propagation time window for each guided wave mode—from the transmitting probe, to the annular detection section, and back to the receiving probe—is calculated. According to the calculated theoretical two-way propagation time window, all waveform amplitude data within a specific time segment associated with each annular detection section are precisely extracted from the mixed echo sequence. All waveform amplitude data corresponding to each annular detection section are categorized and stored according to the guided wave mode type, forming the original data cluster for that section.

[0024] In specific implementation, several preset annular detection sections are generated based on the pipeline extension path, and the original data clusters corresponding to each annular detection section are separated from the multimodal guided wave detection signal. The implementation method is as follows: A series of virtual detection sections are set along the axial centerline of the pipeline to be detected at fixed intervals. The number of virtual detection sections depends on the total length of the pipeline and the preset detection resolution. In some embodiments, the interval length between virtual detection sections is set according to the dominant wavelength of the guided wave, and the interval length is an integer multiple of the dominant wavelength of the guided wave. Taking each virtual detection section as the axial center point, a preset capture length is extended upstream and downstream of the pipeline. The setting of the capture length takes into account the spatial resolution of the sensor array and the attenuation characteristics of the guided wave energy, thereby forming an annular detection section with an axial width on the outer surface of the pipeline. The axial width of the annular detection section is twice the capture length.

[0025] The system receives multimode guided wave detection signals excited and acquired by a piezoelectric sensor array. These signals are a mixed echo sequence formed by the superposition and interference of various guided wave modes, including longitudinal wave, torsional wave, and bending wave modes. The mixed echo sequence is recorded as a voltage-time series. Based on the group velocities of different guided wave modes in the pipe body material obtained from material handbooks or previous calibration experiments, and the axial distance from the geometric center of each annular detection section to the excitation probe, the theoretical two-way propagation time window for each guided wave mode from emission to arrival at the annular detection section and return to the receiving probe is calculated. In some embodiments, for the first... Waveguide mode and the first A ring-shaped detection section, theoretical two-way propagation time window Through formula Perform calculations, where Indicates the first The axial distance from the geometric center of each annular detection section to the excitation probe. Indicates the first The group velocity of each guided wave mode in the pipe. Based on the calculated theoretical two-way propagation time window, the timestamp falling within each window is precisely extracted from the complete mixed echo sequence. All waveform amplitude data within the interval, time deviation tolerance The setting depends on the system's time sampling accuracy and the duration of the guided wave pulse.

[0026] All waveform amplitude data corresponding to each annular detection segment are categorized and stored according to their respective guided wave modes. Mode classification is based on pre-established guided wave dispersion curves and signal characteristic templates for each mode. In practice, pattern recognition algorithms can be used to analyze the intercepted time segment signals, assigning them to P-wave, torsional wave, or flexural wave mode datasets based on their frequency components, polarization direction, and arrival time difference. The resulting raw data cluster is a structured dataset indexed by the annular detection segment, with each index storing subsets of waveform amplitude data corresponding to several guided wave modes, thus establishing an independent and complete raw signal file for each annular detection segment. Optionally, during storage, waveform amplitude data for different modes are normalized and time-aligned to eliminate the influence of excitation energy differences and initial trigger time deviations.

[0027] In one embodiment of the present invention, see [reference] Figure 3 Material characteristic records are established for each annular inspection section. These records include the acoustic impedance distributions of the main pipe structure and the pipe lining structure, generated by inversion from the original data cluster of the section. Time-frequency joint analysis is performed on the original data cluster of the section to separate the reflection energy spectrum components corresponding to the main pipe material and the pipe lining material. Backscattering analysis is performed on the reflection energy spectrum components of the main pipe material to reconstruct the circumferentially distributed acoustic impedance profile of the main pipe structure along the annular inspection section, which is recorded as the acoustic impedance distribution of the main pipe structure. Tomographic imaging is performed on the reflection energy spectrum components of the pipe lining material to invert the thickness and density variations of the pipe lining layer within the annular inspection section, quantifying them as the acoustic impedance distribution of the pipe lining structure. The acoustic impedance distributions of the main pipe structure and the pipe lining structure are integrated to form a complete material characteristic record for the annular inspection section.

[0028] The acoustic impedance distributions of the main pipe structure in all annular inspection sections are compared to identify sets of sections with continuous acoustic impedance distributions. The cross-correlation coefficient between the acoustic impedance distributions of the main pipe structure in two adjacent annular inspection sections is calculated. A threshold for the cross-correlation coefficient used to determine continuity is set. Starting from the first annular inspection section at the beginning of the pipe, each annular inspection section is scanned sequentially. If the cross-correlation coefficient between the current annular inspection section and the next annular inspection section is higher than the continuity threshold, the next annular inspection section is included in the current set of continuous sections. If the cross-correlation coefficient between the current annular inspection section and the next annular inspection section is lower than the continuity threshold, the expansion of the current set of sections is terminated, and the formed set of sections is marked as an independent continuous main segment. Starting from the next annular inspection section, the above expansion and termination process is repeated until all annular inspection sections are classified into a single continuous main segment.

[0029] In specific implementation, a material characteristic record is established for each annular detection section. The implementation method is as follows: Time-frequency joint analysis is performed on the original data cluster of the section. Short-time Fourier transform or wavelet transform methods are used to convert the time-domain waveform signal into a time-frequency domain energy distribution matrix. From the energy distribution matrix, based on the differences in the spectral characteristics of the reflection signals from different material interfaces, the reflection energy spectrum components corresponding to the main pipe material and the reflection energy spectrum components corresponding to the pipe lining material are separated. Backscattering analysis is performed on the separated reflection energy spectrum components of the main pipe material. Using an inversion algorithm based on the Born approximate wave equation, the acoustic impedance profile of the main pipe structure, continuously distributed circumferentially along the annular detection section, is reconstructed. The acoustic impedance profile is represented by a sequence of acoustic impedance values ​​per unit length, which is recorded as the acoustic impedance distribution of the main pipe structure. It can be understood that the acoustic impedance distribution of the main pipe structure reflects the circumferential variation of the product of the main material density and the longitudinal wave velocity in the annular detection section.

[0030] The reflected energy spectrum components of the separated pipe lining material are processed using tomographic imaging. Utilizing the signal amplitude and phase information acquired from multiple sensor channels, the two-dimensional acoustic characteristic distribution of the pipe lining layer within the annular inspection section is inverted using algebraic reconstruction or synchronous iterative reconstruction techniques. The thickness variation sequence and density variation sequence of the pipe lining layer are quantified from the inversion results. The product of corresponding points in the thickness variation sequence and density variation sequence is recorded as the acoustic impedance distribution of the pipe lining structure. In some embodiments, the acoustic impedance distribution of the pipe lining structure is stored as a two-dimensional array, where the row index corresponds to the circumferential angle and the column index corresponds to the axial micro-segments. Integrating the acoustic impedance distributions of the main pipe structure and the pipe lining structure forms a complete material feature record for the annular inspection section. This material feature record exists in the form of an encapsulated data structure, which includes the acoustic impedance distribution arrays of the main pipe structure, the pipe lining structure, and their corresponding spatial coordinate information.

[0031] In practice, the acoustic impedance distributions of the main pipe structure in all annular detection sections are compared to identify a set of sections with continuous acoustic impedance distributions. The implementation method is as follows: The cross-correlation coefficient between the acoustic impedance distributions of the main pipe structure in two adjacent annular detection sections is calculated. The cross-correlation coefficient measures the consistency of the two acoustic impedance distributions in their circumferential shape. A threshold for the cross-correlation coefficient used to determine continuity is set, with a value between 0.7 and 0.9. Starting from the first annular detection section at the beginning of the pipe, each annular detection section is scanned sequentially according to the axial extension direction of the pipe. If the cross-correlation coefficient between the current annular detection section and the next annular detection section is higher than the continuity threshold, the next annular detection section is included in the current set of continuous sections. The set of continuous sections is a dynamically maintained list. If the cross-correlation coefficient between the current annular detection section and the next annular detection section is lower than the continuity threshold, the expansion of the current set of sections is terminated, and the formed set of sections is marked as an independent continuous main body section. A unique identifier is assigned to the continuous main body section. Starting with the next unclassified annular inspection segment, repeat the expansion and termination process described above until all annular inspection segments are classified into a single continuous main section. It is understood that a continuous main section may contain one or more annular inspection segments, depending on the axial homogeneity of the pipe body material. Optional, cross-correlation coefficients... The calculation formula is:

[0032] in: Indicates the first The and the first Cross-correlation coefficients of acoustic impedance distribution of the main structure of the pipeline in each annular testing section. and They represent the first The and the first The annular detection section is in the circumferential direction. Acoustic impedance values ​​at each sampling point and These represent the average values ​​of the corresponding acoustic impedance distributions. This represents the total number of circumferential sampling points.

[0033] In one embodiment of the present invention, the acoustic impedance distributions of the pipe lining structure in all annular inspection sections are compared to identify the boundary sections where the acoustic impedance distribution exhibits a step change. The difference in the average values ​​of the acoustic impedance distributions of the pipe lining structure between adjacent annular inspection sections is calculated along the pipe extension direction. Simultaneously, the difference in the variance of the acoustic impedance distributions of the pipe lining structure between adjacent annular inspection sections is calculated. Based on historical pipeline operating data or prior knowledge, a step threshold for the difference in average values ​​and a stability threshold for the difference in variance are set. The boundary sections of adjacent annular inspection sections that simultaneously meet the following conditions are selected: their average value difference is greater than the step threshold, and their variance difference is less than the stability threshold. These selected boundary sections of adjacent annular inspection sections are determined as the boundary sections where the acoustic impedance distribution exhibits a step change.

[0034] The pipeline is divided into different lining feature zones based on the identified section boundaries. Along the pipeline's extension path, the boundaries of all identified sections exhibiting a step-like change in acoustic impedance are used as dividing points. All loop-shaped detection sections located between two adjacent dividing points are classified into the same lining feature zone. If a loop-shaped detection section exists between the pipeline's starting point and the first dividing point, it is classified into a separate lining feature zone. If a loop-shaped detection section exists between the last dividing point and the pipeline's end, it is classified into another separate lining feature zone.

[0035] In specific implementation, the acoustic impedance distribution of the pipe lining structure in all annular inspection sections is compared, and the boundary of sections with step changes in acoustic impedance distribution is identified. The implementation method is as follows: Calculate the difference in the average value of the acoustic impedance distribution of the pipe lining structure between adjacent annular inspection sections along the pipe extension direction. The average value of the acoustic impedance distribution is a scalar obtained by arithmetically averaging all acoustic impedance sample values ​​in the distribution. Simultaneously, calculate the difference in the variance of the acoustic impedance distribution of the pipe lining structure between adjacent annular inspection sections. Variance is used to describe the degree of dispersion of the internal values ​​of the acoustic impedance distribution of the pipe lining structure. Based on the lining material change records or maintenance log data recorded in the pipeline historical operating condition database, set a step threshold for the difference in average value and a stability threshold for the difference in variance. The step threshold is used to determine whether the change in average value is significant, and the stability threshold is used to determine whether the distribution pattern remains relatively stable. Adjacent annular inspection section boundaries that simultaneously meet the following conditions are selected: the difference in the average value of the acoustic impedance distribution of the pipe lining structure between adjacent annular inspection sections is greater than the step threshold, and the difference in the variance of the acoustic impedance distribution of the pipe lining structure between adjacent annular inspection sections is less than the stability threshold. The boundaries of adjacent ring-shaped detection sections selected from the screening are determined as the boundaries of sections where the acoustic impedance distribution exhibits a step change. The axial position information of the step change is recorded at the section boundaries.

[0036] In some embodiments, the first The and the first The average difference in acoustic impedance distribution of the pipe lining structure between the annular testing sections It can be done through the formula:

[0037] in: This represents the difference between the average values. and They represent the first The and the first The average acoustic impedance distribution of the pipe lining structure in each annular detection section. It can be understood that the screening logic aims to identify boundaries where the acoustic impedance level undergoes a significant jump while the uniformity of the distribution remains relatively unchanged. These boundaries often correspond to locations where changes occur in the type, thickness, or manufacturing process of the lining material.

[0038] In specific implementation, the pipeline is divided into different lining feature zones based on the identified segment boundaries. The implementation method is as follows: Along the pipeline extension path, the boundaries of all identified segments where the acoustic impedance distribution exhibits a step change are used as dividing points. These dividing points divide the pipeline axially into multiple continuous intervals. All annular detection segments located between two adjacent dividing points are classified into the same lining feature zone. The acoustic impedance distribution of the pipeline lining structure within the same lining feature zone is considered statistically homogeneous. If there are one or more annular detection segments between the pipeline start end and the first dividing point, these annular detection segments are classified into an independent lining feature zone. If there are one or more annular detection segments between the last dividing point and the pipeline end, these annular detection segments are classified into another independent lining feature zone. In some embodiments, the lining feature zone is assigned a unique number, and each number is associated with a list that stores the index identifiers of all annular detection segments belonging to that lining feature zone. Lining feature zones containing only a single annular detection segment are specially marked for later analysis.

[0039] In one embodiment of the present invention, an independent wall thickness reference model is constructed for each lining feature area. The construction of this model depends on the acoustic impedance distribution of the pipe lining structure in all annular detection sections covered by the lining feature area. The acoustic impedance distribution of the pipe lining structure in all annular detection sections belonging to the same lining feature area is extracted. Statistical aggregation is performed on the acoustic impedance distribution of all pipe lining structures within the lining feature area to generate a lining acoustic impedance probability density function characterizing the overall characteristics of the lining feature area. Based on a pre-established material acoustic property database, a mapping relationship from lining acoustic impedance to equivalent thickness is established. The specific process for establishing this mapping relationship is as follows: First, the material acoustic property database needs to be constructed in advance through laboratory calibration experiments. Standard test blocks containing various common lining materials (such as rubber, epoxy resin, ceramic) and different thickness specifications are prepared, and the acoustic impedance value of each standard test block is measured using the ultrasonic pulse-echo method. At the same time, its corresponding true thickness is obtained through precision mechanical measurement. Multiple sets of measurements were obtained. Data pairs are entered into the database, forming a basic data pair set. Based on this, for each type of lining material in the database, its acoustic impedance is... The actual thickness is the independent variable. Using the least squares method as the dependent variable, a regression model of acoustic impedance-thickness for this type of material is established. For example, for rubber linings, the fitted regression model can be expressed as follows:

[0040] in For equivalent thickness, Acoustic impedance, These are the regression coefficients determined using the least squares method. For acoustic impedance values ​​not directly covered in the database, interpolation is performed using this regression model, thus expanding discrete data pairs into continuous functional mappings. Furthermore, to ensure the accuracy of the mapping, the database is periodically calibrated by randomly sampling on-site pipe section lining samples, and newly measured values ​​are updated accordingly. The data is compared with the predicted values ​​of the regression model. If the deviation exceeds the preset threshold, the fitting process is re-executed to update the regression coefficients, so that the mapping relationship can always reflect the actual acoustic characteristics of the lining material under the current working conditions.

[0041] The probability density function of the acoustic impedance of the lining is transformed through a mapping relationship to generate the reference probability distribution of the wall thickness in the lining feature area. The reference probability distribution of the wall thickness is a function that describes the probability of the wall thickness value occurring. This distribution constitutes the reference model of the wall thickness, which defines the statistical normal range of the wall thickness values ​​in the lining feature area.

[0042] The probability density function of the lining acoustic impedance is transformed through the mapping relationship to generate the wall thickness reference probability distribution of the lining feature area. This distribution constitutes the wall thickness reference model, which defines the statistical normal range of the wall thickness values ​​in the lining feature area.

[0043] By combining the homogeneous sound velocity estimate of the main pipe structure with the wall thickness benchmark model, the theoretical expected range of wall thickness for each annular inspection section is derived. For each annular inspection section, its corresponding continuous main pipe section is determined, and the homogeneous sound velocity estimate of the main pipe structure for that continuous main pipe section is obtained. The lining feature region to which the annular inspection section belongs is determined, and the wall thickness benchmark model for that lining feature region is obtained. Based on the theory of elastic wave propagation in multilayer media, a calculation model is established. The input parameters of this calculation model include the homogeneous sound velocity estimate of the main pipe structure and the wall thickness value in the wall thickness benchmark model, and the output is the theoretical guided wave dispersion curve. The actual guided wave propagation parameters derived from the material feature record are input into the calculation model, and the inverse problem is solved to obtain a series of wall thickness solution sets that match the theoretical guided wave dispersion curve with the actual guided wave propagation parameters. From the wall thickness solution set, the wall thickness value with the highest probability of occurrence is selected as the central benchmark value. According to the statistical characteristics of the wall thickness benchmark model, a symmetrical tolerance band around the central benchmark value is set, and the upper and lower limits of this symmetrical tolerance band constitute the theoretical expected range of wall thickness.

[0044] In practical implementation, an independent wall thickness benchmark model is constructed for each lining feature area. The implementation method is as follows: The acoustic impedance distributions of the pipe lining structure in all annular detection sections belonging to the same lining feature area are extracted. These acoustic impedance distributions constitute the original dataset for that lining feature area. Statistical aggregation is performed on the acoustic impedance distributions of all pipe lining structures within the lining feature area. Specifically, the acoustic impedance values ​​in all distributions are merged into a single overall sample set. A kernel density estimation method is used to generate a lining acoustic impedance probability density function characterizing the overall properties of the lining feature area. This function depicts the probability distribution of acoustic impedance values. Based on a pre-established material acoustic property database, which stores acoustic parameters of different lining materials under specific working conditions, a mapping relationship from lining acoustic impedance to equivalent thickness is established. This mapping relationship is implemented in the form of a function or lookup table. The probability density function of the lining acoustic impedance is transformed through a mapping relationship to generate the baseline probability distribution of the wall thickness in the lining feature area. This baseline probability distribution is a function describing the probability of a given wall thickness value occurring; this distribution constitutes the baseline wall thickness model, which defines the statistically normal range of wall thickness values ​​within the lining feature area. It can be understood that the peak region of the baseline wall thickness probability distribution corresponds to the most likely wall thickness state for that lining feature area.

[0045] In some embodiments, from liner acoustic impedance to equivalent thickness The mapping relationship can be represented as ,in: Indicates the equivalent thickness. Indicates acoustic impedance. This represents a mapping function determined by the acoustic properties of the material. See Table 1, which shows a mapping table of acoustic impedance-equivalent thickness for a lining material.

[0046] Table 1: Acoustic Impedance-Equivalent Thickness Mapping Table for Lining Materials

[0047] In practical implementation, the theoretical expected range of wall thickness for each annular inspection section is derived by combining the homogeneous sound velocity estimate of the pipeline main structure and the wall thickness benchmark model. The implementation method is as follows: For each annular inspection section, the continuous main section to which the annular inspection section belongs is determined, and the homogeneous sound velocity estimate of the pipeline main structure of the continuous main section is obtained from the system memory. The homogeneous sound velocity estimate of the pipeline main structure is a representative sound velocity value derived by statistically analyzing the acoustic impedance distribution of the pipeline main structure of all annular inspection sections within the continuous main section. The lining characteristic area to which the annular inspection section belongs is determined, and the wall thickness benchmark model of the lining characteristic area is obtained from the system memory. Based on the theory of elastic wave propagation in multilayer media, a calculation model is established. The calculation model can be a numerical model or a semi-analytical model. The input parameters of the calculation model include the homogeneous sound velocity estimate of the pipeline main structure and the wall thickness value sampled from the wall thickness benchmark model. The output of the calculation model is the theoretical guided wave dispersion curve. The actual guided wave propagation parameters derived from the material characteristic record are input into the calculation model. These parameters include the measured wavenumber or phase velocity of guided waves in a specific mode. An iterative optimization algorithm is used to solve the inverse problem, yielding a series of wall thickness solution sets that match the theoretical guided wave dispersion curve with the actual guided wave propagation parameters. From the wall thickness solution set, the frequency of each wall thickness value is statistically analyzed, and the wall thickness value with the highest probability of occurrence is selected as the central reference value. Based on the statistical characteristics of the wall thickness reference model, a symmetrical tolerance band is defined around the central reference value. The upper limit of the symmetrical tolerance band is the central reference value plus a tolerance amount, and the lower limit is the central reference value minus a tolerance amount. The upper and lower limits of the symmetrical tolerance band constitute the theoretical expected range of wall thickness. The tolerance amount can be determined by multiplying the standard deviation of the wall thickness reference probability distribution by a coefficient. The theoretical expected range of wall thickness provides a wall thickness reference range based on the principle of overall characteristic consistency.

[0048] See Figure 4 This is a comparison chart of wall thickness verification for circular inspection sections of an industrial pipeline. It displays the "actual wall thickness," "center reference value," and "expected range" for 10 circular inspection sections, representing the core visualization results of the "expected wall thickness range verification" stage. In some sections (such as section 5), the actual wall thickness significantly exceeds both the center reference value and the expected range, indicating a risk of wall thickness deviation. By comparing the actual wall thickness with the expected range, sections with abnormal wall thickness can be quickly identified, providing a basis for pipeline defect location. For sections deviating from the expected range, the center reference value can be used for data correction, improving the reliability of subsequent diagnostic results. The distribution of abnormal wall thickness can reflect the degree of wear and corrosion in the pipeline, assisting in the formulation of maintenance strategies.

[0049] In one embodiment of the present invention, for each annular detection segment, the actual wall thickness parameter contained in its material feature record is compared and verified with the theoretical expected range of wall thickness. The actual wall thickness parameter, characterizing the thickness, is extracted from the acoustic impedance distribution of the pipe lining structure in the material feature record of the annular detection segment. The theoretical expected range of wall thickness derived for that annular detection segment is read. It is determined whether the value of the actual wall thickness parameter is between the lower and upper limits of the theoretical expected range. The comparison and verification results are recorded. When the actual wall thickness parameter deviates, the center reference value of the theoretical expected range is used for replacement, generating a corrected diagnostic basis. When the actual wall thickness parameter is within the theoretical expected range, the material feature record of that annular detection segment is confirmed as a valid diagnostic basis. When the actual wall thickness parameter deviates from the theoretical expected range, the center reference value of the theoretical expected range is used to replace the actual wall thickness parameter, generating a corrected diagnostic basis.

[0050] In specific implementation, the actual wall thickness parameters contained in the material characteristic record of each annular detection section are compared and verified with the theoretical expected range of wall thickness. The implementation method is as follows: The actual wall thickness parameters, representing the thickness, are extracted from the acoustic impedance distribution of the pipe lining structure in the material characteristic record of the annular detection section. The acoustic impedance distribution of the pipe lining structure is a two-dimensional array containing acoustic impedance values. The parsing process first calculates the average acoustic impedance value of all values ​​in the acoustic impedance distribution of the pipe lining structure, and then converts the average acoustic impedance value into the actual wall thickness parameters by querying a pre-established acoustic impedance-thickness mapping table. The theoretical expected range of wall thickness derived for the annular detection section is read. The theoretical expected range of wall thickness contains a lower limit value and an upper limit value, which are directly obtained from the interval data structure stored in the system memory. It is determined whether the value of the actual wall thickness parameter is between the lower limit value and the upper limit value of the theoretical expected range of wall thickness. This process is performed by performing numerical comparison operations through conditional judgment statements in the computer program. Record the comparison and verification results. The comparison and verification results are saved in the form of record entries. Each record entry includes the ring detection section number, the actual wall thickness parameter value, the lower limit value of the theoretical expected wall thickness range, the upper limit value of the theoretical expected wall thickness range, and a status flag indicating whether it is within the range.

[0051] In some embodiments, the actual wall thickness parameter The analytical formula is:

[0052] in: This represents the actual wall thickness parameter. This represents the average acoustic impedance value, indicating the acoustic impedance distribution of the pipe lining structure. This represents the mapping function from acoustic impedance to thickness. It can be understood that the average acoustic impedance value... The result is obtained by summing all elements in the acoustic impedance distribution array of the pipe lining structure and then dividing by the total number of elements.

[0053] In specific implementation, when the actual wall thickness parameter deviates, the central benchmark value of the theoretical expected wall thickness range is used to replace and generate a corrected diagnostic basis. The implementation method is as follows: When the actual wall thickness parameter is within the theoretical expected wall thickness range, the material characteristic record of the annular detection section is confirmed as a valid diagnostic basis. The valid diagnostic basis is marked as verified by the system and directly used to generate a pipeline health status report. When the actual wall thickness parameter deviates from the theoretical expected wall thickness range, the actual wall thickness parameter is replaced by the central benchmark value of the theoretical expected wall thickness range. The central benchmark value is a value obtained by calculating the arithmetic mean of the lower limit and upper limit values ​​of the theoretical expected wall thickness range. A corrected diagnostic basis is generated. The corrected diagnostic basis is a new data object. The data object inherits most of the fields of the original material characteristic record but updates the value of the actual wall thickness parameter field to the central benchmark value. At the same time, a field is added to record the original actual wall thickness parameter value and metadata of the replacement operation. In some embodiments, the corrected diagnostic basis is stored in the form of an independent file and associated with the original diagnostic data. Optionally, the system creates an audit trail record for each annular detection section where parameter replacement occurs. The audit trail record records in detail the values ​​before and after replacement, the range boundaries, and the operation timestamp.

[0054] See Figure 5 This is a comparison chart of the correction effects of industrial pipeline diagnostic evaluation indicators. It covers five core evaluation items: wall thickness, acoustic impedance, lining, defect detection, and diagnostic consistency. The "corrected score" for all indicators is significantly higher than the "uncorrected score" and closer to the "ideal value." The "acoustic impedance continuity" indicator shows the smallest difference between its corrected score (≈0.95) and the ideal value, indicating the most significant correction effect. This visually demonstrates that correction strategies such as "wall thickness benchmark replacement" can improve the rationality of diagnostic indicators. By adjusting the correction magnitude of each indicator, weak points can be specifically strengthened. The fact that the corrected score approaches the ideal value indicates improved accuracy and consistency of the diagnostic results, reducing the risk of misjudgment.

[0055] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0056] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A diagnostic method for industrial pipelines based on multimode guided waves, characterized in that, include, Several preset ring detection sections are generated based on the pipeline extension path, and the original data clusters corresponding to each ring detection section are separated from the multi-mode guided wave detection signal. Material feature records are established for each annular detection section. The material feature records include the acoustic impedance distribution of the main pipe structure and the acoustic impedance distribution of the pipe lining structure generated by inversion using the original data cluster of the section. By comparing the acoustic impedance distribution of the main pipe structure in all annular inspection sections, a set of sections with continuous acoustic impedance distribution is identified to define continuous main sections. By comparing the acoustic impedance distribution of the pipe lining structure in all annular inspection sections, the boundaries of sections with step changes are identified to divide them into different lining feature areas. An independent wall thickness benchmark model is constructed for each lining feature area, and the homogeneous sound velocity estimate of the main pipe structure is calculated based on the material feature records of each annular detection section within the continuous main body section. By combining the homogeneous sound velocity estimate of the main pipe structure with the wall thickness benchmark model, the theoretical expected range of wall thickness for each annular detection section is derived. For each annular detection segment, the actual wall thickness parameters contained in its material feature record are compared and verified with the theoretical expected range of wall thickness. When the actual wall thickness parameters deviate, the central reference value of the theoretical expected range of wall thickness is used for replacement, generating a corrected diagnostic basis.

2. The industrial pipeline diagnostic method based on multimode guided waves according to claim 1, characterized in that, The steps of generating several preset ring-shaped detection sections based on the pipeline extension path and separating the original data clusters corresponding to each ring-shaped detection section from the multi-mode guided wave detection signal are as follows: include, A series of equally spaced virtual inspection sections are set along the pipeline axis; With each virtual detection section as the center, a preset capture length is extended to both sides of the pipeline axis, thereby forming a ring detection section around the pipeline; Receive a multimode guided wave detection signal, wherein the multimode guided wave detection signal contains a mixed echo sequence generated by the interaction of guided waves of different modes; Based on the propagation speed of the guided wave in the pipe and the geometric center position of each annular detection section, the theoretical two-way propagation time window of each mode guided wave from emission to arrival at the annular detection section and return is calculated. Based on the theoretical two-way propagation time window, all waveform amplitude data within a specific time segment associated with each ring detection segment are extracted from the mixed echo sequence; All waveform amplitude data corresponding to each ring detection segment are categorized and stored according to the guided wave mode type to form the original data cluster of the segment.

3. The industrial pipeline diagnostic method based on multimode guided waves according to claim 1, characterized in that, The step of establishing a material feature record for each annular detection segment includes, A time-frequency joint analysis was performed on the original data cluster of the section to separate the reflectance energy spectrum components corresponding to the main pipe material and the reflectance energy spectrum components corresponding to the pipe lining material; Backscattering analysis is performed on the reflection energy spectrum components of the pipeline body material to reconstruct the acoustic impedance profile of the pipeline body structure distributed circumferentially along the annular detection section, and it is recorded as the acoustic impedance distribution of the pipeline body structure. The reflectance spectrum components of the pipe lining material are processed by tomographic imaging to invert the thickness and density changes of the pipe lining layer in the annular detection section, and quantify them as the acoustic impedance distribution of the pipe lining structure. By integrating the acoustic impedance distribution of the main pipe structure with that of the pipe lining structure, a complete material characteristic record of the annular detection section is formed.

4. The industrial pipeline diagnostic method based on multimode guided waves according to claim 1, characterized in that, The step of comparing the acoustic impedance distribution of the main pipe structure in all annular detection sections to identify the set of sections with continuous acoustic impedance distribution is as follows. include, Calculate the cross-correlation coefficient between the acoustic impedance distributions of the main pipe structure in two adjacent annular detection sections; Set a continuity threshold for the cross-correlation coefficient; Starting from the first loop inspection section at the beginning of the pipeline, each loop inspection section is scanned sequentially. If the cross-correlation coefficient between the current ring detection segment and the next ring detection segment is higher than the continuity determination threshold, then the next ring detection segment is included in the current continuous segment set. If the cross-correlation coefficient between the current ring detection segment and the next ring detection segment is lower than the continuity determination threshold, then the expansion of the current segment set is terminated, and the formed segment set is marked as an independent continuous main segment. Starting from the next ring detection segment, repeat the expansion and termination process until all ring detection segments are classified into a single continuous main segment.

5. The industrial pipeline diagnostic method based on multimode guided waves according to claim 1, characterized in that, The step of comparing the acoustic impedance distribution of the pipe lining structure in all annular detection sections and identifying the boundary of sections where the acoustic impedance distribution exhibits a step change includes, Calculate the difference in the average acoustic impedance distribution of the pipe lining structure between adjacent annular inspection sections along the pipe extension direction; Simultaneously calculate the difference in the variance of the acoustic impedance distribution of the pipe lining structure between adjacent annular detection sections; Based on historical pipeline operating data, set a step threshold for the average difference and a stability threshold for the variance difference; Adjacent ring detection segment boundaries that simultaneously meet the following conditions are selected: the average difference is greater than the step threshold and the variance difference is less than the stability threshold. The boundaries of the selected adjacent ring-shaped detection sections are defined as the boundaries of sections where the acoustic impedance distribution exhibits a step change.

6. The industrial pipeline diagnostic method based on multimode guided waves according to claim 1, characterized in that, The step of dividing the lining into different feature areas includes, The pipeline is divided into different lining feature zones based on the section boundaries; Along the pipeline extension path, the boundaries of all identified sections where the acoustic impedance distribution exhibits a step change are used as dividing points. All annular detection segments located between two adjacent dividing points are classified into the same lining feature area; If there is a ring-shaped inspection section between the beginning of the pipeline and the first dividing point, it shall be classified as an independent lining feature area; If there is a ring-shaped detection section between the last dividing point and the end of the pipe, it is classified as another independent lining feature area.

7. The industrial pipeline diagnostic method based on multimode guided waves according to claim 1, characterized in that, The step of constructing an independent wall thickness reference model for each lining feature area includes, The construction of the wall thickness reference model depends on the acoustic impedance distribution of the pipe lining structure in the entire annular detection section covered by the lining feature area. Extract the acoustic impedance distribution of the pipe lining structure for all annular detection sections belonging to the same lining feature area; The acoustic impedance distribution of all pipe lining structures within the lining feature area is statistically aggregated to generate a lining acoustic impedance probability density function that characterizes the overall properties of the lining feature area. Based on the material acoustic properties database, a mapping relationship from lining acoustic impedance to equivalent thickness is established; The probability density function of the lining acoustic impedance is transformed through the mapping relationship to generate the wall thickness reference probability distribution of the lining feature area. This distribution constitutes the wall thickness reference model. The wall thickness reference model defines the statistically normal range of wall thickness values ​​within the lining feature area.

8. The industrial pipeline diagnostic method based on multimode guided waves according to claim 1, characterized in that, The step of combining the homogeneous sound velocity estimate of the main pipe structure with the wall thickness benchmark model to derive the theoretical expected range of wall thickness for each annular detection section. include, For each annular detection section, determine the continuous main section to which it belongs, and obtain the homogeneous sound velocity estimate of the pipeline main structure of the continuous main section. Determine the lining feature area to which the annular detection section belongs, and obtain the wall thickness reference model of the lining feature area; Based on the theory of elastic wave propagation in multilayer media, a calculation model is established. The input parameters of the calculation model include the homogeneous sound velocity estimate of the main pipe structure and the wall thickness value in the wall thickness benchmark model. The output is the theoretical guided wave dispersion curve. The actual guided wave propagation parameters derived from the material feature record are input into the calculation model to solve the inverse problem, thereby obtaining a series of wall thickness solution sets that can match the theoretical guided wave dispersion curve with the actual guided wave propagation parameters. From the wall thickness solution set, the wall thickness value with the highest probability of occurrence is selected as the central reference value; Based on the statistical characteristics of the wall thickness benchmark model, a symmetrical tolerance band is set around the central benchmark value. The upper and lower limits of the symmetrical tolerance band constitute the theoretical expected range of the wall thickness.

9. The industrial pipeline diagnostic method based on multimode guided waves according to claim 1, characterized in that, For each annular detection segment, the actual wall thickness parameters contained in its material feature record are compared and verified with the theoretical expected range of wall thickness, including: The actual wall thickness parameters characterizing the thickness are extracted from the acoustic impedance distribution of the pipe lining structure recorded by the material characteristics of the annular test section. The reading is the theoretical expected range of wall thickness derived from the annular detection section; Determine whether the value of the actual wall thickness parameter is between the lower limit and the upper limit of the theoretical expected range of wall thickness; Record the comparison and verification results.

10. The industrial pipeline diagnostic method based on multimode guided wave as described in claim 1, characterized in that, The step of replacing the actual wall thickness parameter with the central reference value of the theoretical expected range of the wall thickness to generate a corrected diagnostic basis when the actual wall thickness parameter deviates includes... When the actual wall thickness parameter is within the theoretical expected range of the wall thickness, the material characteristic record of the annular detection section is confirmed as a valid diagnostic basis. When the actual wall thickness parameter deviates from the theoretical expected range of the wall thickness, the actual wall thickness parameter is replaced by the central reference value of the theoretical expected range of the wall thickness to generate a corrected diagnostic basis.

Images