Fusion processing method for fatigue crack Lamb wave signal of orthotropic steel bridge deck

Abstract

The invention relates to the technical field of fatigue crack nondestructive testing, in particular to an orthotropic steel bridge deck fatigue crack Lamb wave signal fusion processing method which comprises the steps that theoretical analysis and numerical simulation are conducted on an orthotropic steel bridge deck ultrasonic Lamb wave signal, and Lamb wave detection basic parameters are obtained; based on Lamb wave detection basic parameters and original signals, baseline correction is carried out on the Lamb wave signals, signal baseline drift caused by the load effect is eliminated, and to-be-separated signals without baseline interference are obtained; for a to-be-separated signal without baseline interference, performing multi-mode separation on the Lamb wave signal after baseline correction, extracting a low-order symmetric vibration mode signal and a low-order anti-symmetric vibration mode signal of the Lamb wave, and obtaining a single-mode original signal; and processing the separated single-mode signal, and correcting the frequency dispersion compensation process by combining with the thickness change caused by the fatigue crack damage of the orthotropic steel bridge deck to obtain a single-mode non-frequency dispersion signal. And a high-quality detection signal can still be obtained in a non-ideal environment.

Description

Technical Field

[0001] This invention relates to the technical field of nondestructive testing of fatigue cracks, and in particular to a method for fusing and processing Lamb wave signals of fatigue cracks in orthotropic steel bridge decks. Background Technology

[0002] Orthotropic steel bridge decks are subjected to repeated vehicle loads during construction and operation, which can easily lead to fatigue cracks and other forms of damage at the welded joints between the top plate and the U-ribs. Fatigue cracks often begin to appear inside the weld, making them difficult to detect early using conventional visual methods. The development of fatigue cracks significantly reduces the load-bearing capacity and durability of the structure, and in extreme cases, may lead to serious engineering accidents, threatening the safety performance of the bridge structure. Currently, commonly used non-destructive testing methods for fatigue cracks in steel bridge structures include X-ray imaging, magnetic particle inspection, penetrant testing, eddy current testing, acoustic emission testing, and guided wave testing. X-ray imaging can image internal defects, but its application to large structures is limited by equipment size. Magnetic particle inspection, penetrant testing, and eddy current testing are mainly for surface and near-surface cracks. Magnetic particle inspection and penetrant testing require contact with the tested surface, while eddy current testing is non-contact. Acoustic emission testing is a passive monitoring technology, and its accuracy is easily affected by environmental noise.

[0003] Ultrasonic guided waves primarily refer to guided waves propagating within waveguide structures such as plates or tubes. Unlike bulk waves, ultrasonic guided waves, excited at a single point on the structure, can propagate in any direction along the structure. Due to minimal attenuation along the propagation path, they can travel long distances, allowing for the determination of defects based on information along the entire propagation path. Furthermore, because guided waves propagate along the entire plate structure, they can detect defects at different locations within the plate structure. In addition, guided waves possess multimodal and dispersion characteristics; each type of guided wave has multiple modes, each with different wave structures and propagation velocities. Even within the same mode, the propagation velocity varies depending on the frequency product. Based on the characteristic phenomena exhibited after the guided wave interacts with defects, the multimodal and dispersion characteristics of guided waves can be used for various detection needs. These unique advantages make ultrasonic guided waves one of the most promising technologies in the fields of non-destructive testing and structural health monitoring.

[0004] Currently, in bridge structural health monitoring applications, the signals acquired by guided wave detection technology typically exhibit significant nonlinear and non-stationary characteristics, displaying complex propagation properties. These mainly manifest as baseline drift, multimodal propagation, and dispersion effects. These characteristics affect the accuracy of structural condition evaluation and damage analysis for steel bridges. Therefore, a method for analyzing and processing guided wave detection signals for fatigue crack damage in orthotropic steel bridge decks is urgently needed. Summary of the Invention

[0005] This invention provides a method for fatigue crack Lamb wave signal fusion processing of orthotropic steel bridge decks that can still obtain high-quality detection signals in undesirable environments and improve the accuracy of bridge health monitoring. This method can effectively solve the problems in the background art.

[0006] To achieve the above objectives, the present invention provides a method for fusing and processing Lamb wave signals of fatigue cracks in orthotropic steel bridge decks, comprising: Theoretical analysis and numerical simulation were performed on the ultrasonic Lamb wave signal of orthotropic steel bridge deck to obtain the basic parameters for Lamb wave detection. Based on the basic parameters of Lamb wave detection and the original signal, baseline correction is performed on the Lamb wave signal to eliminate the signal baseline drift caused by the load and obtain the signal to be separated without baseline interference. For signals to be separated without baseline interference, multi-mode separation is performed on the baseline-corrected Lamb wave signal to extract the low-order symmetric vibration mode and low-order antisymmetric vibration mode signals of the Lamb wave, and obtain the original signal of a single mode. The separated single-mode signals are processed, and the dispersion compensation process is combined with the thickness change caused by fatigue crack damage of orthotropic steel bridge deck to obtain a single-mode non-dispersion signal.

[0007] In one possible design, the basic parameters for Lamb wave detection include dispersion characteristics, modal characteristics, propagation characteristics, and defect response characteristics.

[0008] In one possible design, the dispersion characteristic parameters include the propagation velocity, dispersion curve, and cutoff frequency of each mode at different frequencies.

[0009] In one possible design, the modal characteristic parameters include wave structure, sensitivity to different types of defects, and propagation characteristics under different frequency thick layers.

[0010] In one possible design, the propagation characteristic parameters include the optimal detection mode and frequency, propagation speed, and attenuation coefficient.

[0011] In one possible design, the defect response parameters represent the response characteristics of different modes to different types of defects, used to identify and distinguish different types of defects.

[0012] In one possible design, based on the fundamental parameters of the Lamb wave detection and the original signal, baseline correction is performed on the Lamb wave signal to obtain a baseline-free signal to be separated, including: The original Lamb wave signal is filtered, and the filtered signal is normalized. Based on the propagation velocity and dispersion curve in the basic parameters of Lamb wave detection, the theoretical time window for Lamb wave propagation in the signal is determined; In the original signal, a time period with stable signal intensity and no defect reflection signal is selected as the baseline reference area; combined with the theoretical propagation time window, the signal in this area is statistically analyzed to extract baseline feature values; Baseline model values ​​are constructed using the extracted baseline feature values; Subtracting the constructed baseline model value from the original signal yields the corrected signal, which is the signal to be separated without baseline interference.

[0013] In one possible design, for the baseline-corrected Lamb wave signal to be separated without baseline interference, multimode separation is performed to extract the low-order symmetric vibration mode and low-order antisymmetric vibration mode signals of the Lamb wave, obtaining the original single-mode signal, including: Based on the dispersion curve in the fundamental parameters of Lamb wave detection, the propagation velocity range of low-order symmetric vibration mode and low-order antisymmetric vibration mode at different frequencies is determined. Short-time Fourier transform is performed on the baseline-corrected Lamb wave signal to convert the time-domain signal into a time-frequency domain signal, thus obtaining the time-frequency distribution map of the signal; Based on the propagation velocity range determined by the dispersion curve, the frequency time region containing low-order symmetric vibration modes and low-order antisymmetric vibration modes is initially screened in the time-frequency distribution diagram. Based on the initially selected frequency and time regions, separate filters are constructed for low-order symmetric vibration modes and low-order antisymmetric vibration modes. The baseline-corrected Lamb wave signal is passed through the constructed low-order symmetric vibration mode filter and low-order antisymmetric vibration mode filter to obtain the filtered signal. The filtered signal is reconstructed to obtain low-order symmetric vibration mode signals and low-order antisymmetric vibration mode signals; The extracted low-order symmetric vibration mode signals and low-order antisymmetric vibration mode signals were verified to obtain the original signals of a single mode.

[0014] In one possible design, the separated single-mode signals are processed, and a dispersion compensation process is combined with the thickness variation correction due to fatigue crack damage in the orthotropic steel bridge deck to obtain a single-mode non-dispersion signal, including: Using high-precision measuring equipment, the thickness of the fatigue-cracked area and the surrounding normal area of ​​the orthotropic steel bridge deck were measured; the thickness data at different locations were recorded, the thickness variation caused by crack damage was analyzed, and a database of the relationship between thickness variation and location was established. Based on the material parameters and geometric dimensions of orthotropic steel bridge decks, and combined with Lamb wave propagation theory, dispersion curve models for different thicknesses are established. Based on the measured thickness variation pattern, the original single-mode signal is divided into several small segments along the propagation path; For each small signal segment, the propagation characteristic parameters of the corresponding mode Lamb wave at that thickness are obtained from the dispersion curve model based on the thickness at its location. Dispersion compensation is calculated for each small segment of the signal using propagation characteristic parameters; By splicing the small signal segments that have undergone dispersion compensation together in their original order, a single-mode non-dispersion signal is obtained.

[0015] In one possible design, when performing dispersion compensation on the original single-mode signal, a time-domain inverse filtering method is used, and the parameters of the compensation filter are determined based on the propagation characteristic parameters of the mode at the corresponding thickness in the dispersion curve model.

[0016] The technical solution of this invention can achieve the following technical effects: Through theoretical analysis and numerical simulation, the fundamental parameters for Lamb wave detection were obtained, ensuring an accurate understanding of the characteristics and defects of the structure under test. Baseline correction of the signal eliminates baseline drift caused by load. The corrected signal not only avoids interference from baseline drift in subsequent processing but also improves the accuracy of the detection results. By separating signals of different modes, the relationship between each mode and the crack can be analyzed in a targeted manner, improving the accuracy of crack identification. Due to variations in the thickness of the steel bridge deck and the presence of damage such as cracks, dispersion effects inevitably occur during the propagation of Lamb waves, causing changes in the signal propagation speed and waveform. By correcting this dispersion effect, the signal can be restored to its original state, so that the detection results are no longer affected by dispersion. This not only improves the signal quality but also compensates for signal distortion caused by structural inhomogeneity, improving the authenticity of the final signal. Through multi-step integrated processing, high-quality detection signals can still be obtained in undesirable environments. The multi-mode and dispersion characteristics of Lamb waves are fully utilized in fatigue crack detection, thereby improving the accuracy of bridge health monitoring and ensuring the scientific nature of structural safety assessment. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a method for fusing and processing Lamb wave signals of fatigue cracks in orthotropic steel bridge decks. Detailed Implementation

[0019] This application will now be described with reference to the accompanying drawings.

[0020] like Figure 1 As shown, the present invention provides a method for fusing and processing Lamb wave signals of fatigue cracks in orthotropic steel bridge decks, which specifically includes the following steps: S1. Theoretical analysis and numerical simulation of ultrasonic Lamb wave signals of orthotropic steel bridge decks are performed to obtain basic parameters for Lamb wave detection. S2. Based on the basic parameters of Lamb wave detection and the original signal, perform baseline correction on the Lamb wave signal to obtain the signal to be separated without baseline interference; S3. For the signal to be separated without baseline interference, perform multi-mode separation on the baseline-corrected Lamb wave signal, extract the low-order symmetric vibration mode and low-order antisymmetric vibration mode signals of the Lamb wave, and obtain the original signal of a single mode. S4. Process the separated single-mode signals, and combine the thickness change correction dispersion compensation process caused by fatigue crack damage of orthotropic steel bridge deck to obtain single-mode non-dispersion signals.

[0021] In this embodiment, the basic parameters for Lamb wave detection are obtained through theoretical analysis and numerical simulation, ensuring an accurate understanding of the characteristics and defects of the structure under test. Baseline correction of the signal eliminates baseline drift caused by load. The corrected signal not only avoids interference from baseline drift in subsequent processing but also improves the accuracy of the detection results. By separating signals of different modes, the relationship between each mode and the crack can be analyzed in a targeted manner, improving the accuracy of crack identification. Due to variations in the thickness of the steel bridge deck and the presence of damage such as cracks, dispersion effects inevitably occur during the propagation of Lamb waves, causing changes in the signal propagation speed and waveform. By correcting this dispersion effect, the signal can be restored to its original state, so that the detection results are no longer affected by dispersion. This not only improves the signal quality but also compensates for signal distortion caused by structural inhomogeneity, improving the authenticity of the final signal. Through multi-step integrated processing, high-quality detection signals can still be obtained in undesirable environments. The multi-mode and dispersion characteristics of Lamb waves are fully utilized in fatigue crack detection, thereby improving the accuracy of bridge health monitoring and ensuring the scientific nature of structural safety assessment.

[0022] In some embodiments of the present invention, for step S1, theoretical analysis and numerical simulation are performed on the ultrasonic Lamb wave signal of orthotropic steel bridge deck to obtain the basic parameters for Lamb wave detection. Theoretical analysis of ultrasonic Lamb wave signals of orthotropic steel bridge decks is performed, including: This study delves into the geometry of orthotropic steel bridge decks, precisely measuring the thickness of the top plate, the dimensions of the U-ribs, and the welding parameters between the top plate and the U-ribs. Simultaneously, it clarifies the anisotropic properties of the material, obtaining mechanical parameters such as the elastic modulus and Poisson's ratio in different directions. Based on elastic dynamics theory, it establishes the wave equation for the propagation of ultrasonic Lamb waves in orthotropic steel bridge decks. Considering the boundary conditions of the structure, it solves the wave equation. The dispersion equation of the Lamb wave is obtained by solving the wave equation; the propagation velocity of each mode of the Lamb wave at different frequencies is obtained by solving the dispersion equation using numerical calculation methods; dispersion curves are plotted to visually show the relationship between the propagation velocity of each mode of the Lamb wave and the frequency; the characteristics of the dispersion curves are analyzed to determine the cutoff frequency of different modes and clarify the frequency range in which each mode exists. Based on the solutions to the wave equation, the wave structure of Lamb waves in different modes is analyzed, and the sensitivity of different modes to different types of defects is studied. At the same time, the propagation characteristics of different modes under different frequency thickness products are analyzed. Numerical simulation of ultrasonic Lamb wave signals for orthotropic steel bridge decks, including: Using finite element analysis software, an accurate finite element model is constructed based on the actual dimensions and geometry of the orthotropic steel bridge deck. The mesh is rationally divided, and the mesh is refined in key areas. The anisotropic properties of the material are set, and the mechanical parameters obtained from theoretical analysis are input into the model. The boundary conditions of the model are defined to simulate the working state of the actual structure. Set up an excitation source in the model to simulate the device that generates ultrasonic Lamb waves in actual testing; select an appropriate excitation method, such as piezoelectric ceramic excitation or electromagnetic excitation, and set the parameters of the excitation signal, such as frequency, amplitude, and excitation time; select an appropriate excitation frequency and mode for excitation based on theoretical analysis results; arrange receiving points reasonably in the model to receive the propagated Lamb wave signal; the location of the receiving points should cover areas where defects may exist in order to comprehensively detect the health status of the structure. Numerical simulation calculations were performed using a finite element model to simulate the propagation process of ultrasonic Lamb waves in orthotropic steel bridge decks; signal data at different receiving points at different times were recorded, including physical quantities such as displacement, velocity, and stress; the simulated signal data were processed and analyzed to extract key feature parameters; Based on the theoretical analysis results, the data obtained from the numerical simulation are analyzed in depth; the correctness of the theoretical analysis is verified, and complex factors that are difficult to consider in the theoretical analysis are supplemented; based on the simulation results, the optimal detection mode and frequency are determined, that is, the mode and frequency that are most sensitive to fatigue cracks and have the most stable propagation characteristics under specific working conditions; the propagation speed, attenuation coefficient and other parameters of Lamb wave under different modes and frequencies are obtained, as well as the response characteristics of different modes to different types of defects, and the basic parameters of Lamb wave detection are obtained. The basic parameters for Lamb wave detection include dispersion characteristics, modal characteristics, propagation characteristics, and defect response characteristics. The dispersion characteristic parameters include: The propagation velocities of each mode at different frequencies are obtained by solving the dispersion equation and using numerical calculation methods, including group velocity and phase velocity. The dispersion curve, plotted with frequency on the x-axis and propagation speed on the y-axis, visually demonstrates the relationship between the propagation speed of each mode of the Lamb wave and frequency. Cutoff frequency: The cutoff frequency of different modes is determined by analyzing the characteristics of the dispersion curve, thus clarifying the frequency range in which each mode exists; Modal characteristic parameters include: Wave structure: Based on the solution of the wave equation, analyze the wave structure characteristics such as displacement distribution and stress distribution of Lamb waves in different modes; The sensitivity of different modes to different types of defects was studied. For example, some modes produce obvious reflection and scattering phenomena at cracks, while other modes are not sensitive to cracks. Propagation characteristics under different frequency thickness products: Analysis of the propagation characteristics of different modes under different frequency thickness products, including changes in propagation speed, attenuation, etc. Propagation characteristic parameters include: The optimal detection modes and frequencies are determined based on simulation results, identifying the modes and frequencies that are most sensitive to fatigue cracks and have the most stable propagation characteristics under specific working conditions. Propagation speed: Obtain the propagation speed of the Lamb wave in different modes and frequencies for subsequent signal propagation time calculation and defect location; The attenuation coefficient is a parameter that reflects the degree of energy attenuation of a Lamb wave during propagation. The attenuation coefficient varies for different modes and frequencies, affecting the propagation distance and detection range of the signal. Defect response parameters are the response characteristics of different modes to different types of defects, used to identify and distinguish different types of defects.

[0023] In this embodiment, the propagation and dispersion characteristics of Lamb waves in orthotropic steel bridge decks were accurately obtained through theoretical analysis and numerical simulation. By solving the wave equation and performing dispersion analysis, the propagation velocities of each mode at different frequencies were obtained, and dispersion curves were plotted, providing a basis for further selection of appropriate detection frequencies and modes. By analyzing the dispersion curves and the propagation characteristics of different modes, the cutoff frequency and frequency range of each mode were clarified, thus enabling the selection of the optimal detection mode and frequency for actual detection. This not only improves the detection sensitivity but also ensures the stability of the signal propagation process. Numerical simulation of the propagation process of Lamb waves under different working conditions was used to demonstrate the sensitivity of Lamb waves to fatigue cracks. The simulation analysis effectively identifies different types of defects; the simulation results verify the correctness of the theoretical analysis and supplement complex factors that may not be considered in actual situations, thus improving the reliability of the detection results; this step, through precise modeling and reasonable selection of excitation sources and receiving points, can simulate the propagation characteristics of Lamb waves under actual working conditions; this not only ensures that the simulation results are close to reality, but also provides more comprehensive data support for detection under different working conditions; this step, through the combination of theoretical analysis and numerical simulation, accurately obtains the propagation characteristics and defect response features of Lamb waves in orthotropic steel bridge decks, providing reliable parameter support for the entire detection process, optimizing the detection scheme, and improving detection accuracy and reliability.

[0024] In some embodiments of the present invention, for step S2, based on the basic parameters of Lamb wave detection and the original signal, baseline correction is performed on the Lamb wave signal to obtain a signal to be separated without baseline interference. The original Lamb wave signal is filtered to remove high-frequency noise and interference; the filtered signal is then normalized to adjust the amplitude range to a uniform interval, which facilitates subsequent baseline correction calculations. Based on the propagation velocity and dispersion curve in the basic parameters of Lamb wave detection, the theoretical time window for Lamb wave propagation in the signal is determined. Under defect-free conditions, the propagation time of Lamb wave in the structure is relatively stable, and the theoretical propagation time of Lamb wave in orthotropic steel bridge deck at different frequencies and modes can be obtained by theoretical calculation. In the original signal, a time period with relatively stable signal intensity and no obvious defect reflection signal is selected as the baseline reference area; combined with the theoretical propagation time window, the signal in this area is statistically analyzed, such as calculating the average value and median value, and the baseline feature value is extracted. The extracted baseline feature values ​​are used to construct the baseline model values. Polynomial fitting, spline interpolation and other methods can be used to fit the baseline curve over the entire signal time range based on the signal data in the baseline reference area. When constructing the baseline model, the time-varying characteristics of the signal are considered and a dynamic adjustment method is adopted so that the baseline model can adapt to the baseline changes of the signal in different time periods. Subtracting the constructed baseline model value from the original signal yields the corrected signal, i.e., the signal to be separated without baseline interference; that is, for each sampling point in the signal, the baseline model value at the corresponding time point is subtracted from the original signal value to obtain the signal to be separated without baseline interference.

[0025] In this embodiment, by filtering, normalizing, and baseline modeling the original signal, baseline drift caused by load, temperature changes, or instrument drift can be eliminated, making the zero reference point of the signal more stable and ensuring the accuracy of subsequent signal processing. After filtering out high-frequency noise and interference, the signal is cleaner, eliminating random errors caused by the environment and instruments, improving the signal-to-noise ratio, and providing reliable input for mode separation and dispersion compensation. The use of a dynamically adjusted baseline modeling method allows the baseline to be adaptively corrected over time, ensuring that the baseline correction of the signal remains accurate under different time periods or different detection conditions. By unifying the signal amplitude range and baseline level, signals under different experiments or different fatigue crack conditions are comparable, reducing the impact of human and environmental factors on the results. This step, through filtering, normalizing, baseline modeling, and correction, can eliminate baseline drift and interference in the Lamb wave signal, obtaining a stable baseline-free interference signal, providing a high-quality and reliable input signal for subsequent multi-mode separation and dispersion compensation, thereby significantly improving the accuracy of fatigue crack detection in steel bridge decks.

[0026] In some embodiments of the present invention, for step S3, for the signal to be separated without baseline interference, the baseline-corrected Lamb wave signal is subjected to multi-mode separation to extract the low-order symmetric vibration mode and low-order antisymmetric vibration mode signals of the Lamb wave to obtain the original signal of a single mode. Based on the dispersion curve in the fundamental parameters of Lamb wave detection, the propagation velocity range of low-order symmetric vibration mode and low-order antisymmetric vibration mode at different frequencies is determined. Short-time Fourier transform is performed on the baseline-corrected Lamb wave signal to convert the time-domain signal into a time-frequency domain signal, thus obtaining the time-frequency distribution map of the signal; Based on the propagation velocity range determined by the dispersion curve, frequency time regions that may contain low-order symmetric vibration modes and low-order antisymmetric vibration modes are preliminarily screened in the time-frequency distribution diagram. An adaptive filter design method is employed. Based on the initially selected frequency-time regions, separation filters are constructed for low-order symmetrical and low-order antisymmetric vibration modes. For low-order symmetrical vibration modes, a bandpass filter is constructed based on its propagation velocity and frequency characteristics. The passband range is determined according to the frequency range of the mode on the dispersion curve, while also being appropriately adjusted considering the signal's spectral characteristics. Similarly, for low-order antisymmetric vibration modes, a bandpass filter is constructed, with its passband range determined according to the frequency range of the mode on the dispersion curve. An adaptive algorithm is used in the filter design to automatically adjust the filter parameters based on the actual signal conditions, thereby improving the filter's performance and adaptability. The baseline-corrected Lamb wave signal is passed through the constructed low-order symmetric vibration mode filter and low-order antisymmetric vibration mode filter to obtain the filtered signal. The filtered signal is reconstructed to obtain low-order symmetric vibration mode signals and low-order antisymmetric vibration mode signals. During the reconstruction process, the inverse short-time Fourier transform can be used to convert the time-frequency domain signal back to the time domain signal. The extracted low-order symmetric vibration mode signals and low-order antisymmetric vibration mode signals are verified. The extracted signals can be compared with theoretical models or simulated signals with known defects to check whether they meet the expected modal characteristics. If the extracted signals have errors or do not meet the requirements, the filter parameters are adjusted and optimized, and the mode separation and signal extraction are repeated until a satisfactory original signal of a single mode is obtained.

[0027] In this embodiment, through multimode separation and filtering, single-mode signals of low-order symmetric and low-order antisymmetric vibration modes can be extracted from the baseline-corrected Lamb wave signal, thereby effectively removing multimode interference and improving signal quality and accuracy. By combining dispersion curves, bandpass filters for different modes are designed and their parameters are adaptively adjusted to suppress signal distortion caused by dispersion, ensuring that the extracted mode signals retain their original characteristics and avoiding the influence of dispersion on the signal. Adaptive filter design method is used to automatically adjust filter parameters according to the actual frequency characteristics of the signal, which can improve the adaptability of the filter under different conditions and improve separation accuracy. Through short-time Fourier transform and inverse transform, the time-domain signal can be accurately converted into a time-frequency domain signal, and fine processing is performed in the time-frequency domain. Finally, the signal is restored to the time domain through reconstruction, obtaining a more accurate and stable low-order mode signal. This step, by combining dispersion compensation, mode separation, and adaptive filtering techniques, achieves accurate separation and processing of the Lamb wave signal, which can improve signal quality.

[0028] In some embodiments of the present invention, for step S4, the separated single-mode signal is processed, and at the same time, the dispersion compensation process for thickness change caused by fatigue crack damage of orthotropic steel bridge deck is combined to obtain a single-mode non-dispersion signal. Using high-precision measuring equipment, the thickness of the fatigue-cracked area and the surrounding normal area of ​​the orthotropic steel bridge deck were measured; the thickness data at different locations were recorded, the thickness variation caused by crack damage was analyzed, and a database of the relationship between thickness variation and location was established. Based on the material parameters and geometric dimensions of orthotropic steel bridge decks, and combined with Lamb wave propagation theory, dispersion curve models for different thicknesses are established. This model can describe the propagation velocity and group velocity of a single-mode Lamb wave under different frequency-thickness products. Through numerical simulation software, simulation calculations are performed for different thicknesses to improve the dispersion curve model and ensure its accuracy. Based on the thickness change information obtained from the measurement, the separated single-mode signal is divided into several small segments along the propagation path; the length of each small segment is determined according to the thickness change gradient and signal characteristics to ensure that the thickness change within each small segment is relatively uniform. For each small signal segment, the propagation characteristic parameters of the corresponding mode Lamb wave at that thickness are obtained from the dispersion curve model based on the thickness at its location. Dispersion compensation is calculated for each small segment of the signal using propagation characteristic parameters. Specific calculation methods can include time-domain inverse filtering or frequency-domain phase correction to eliminate the dispersion effect caused by thickness variations. By splicing the small signal segments that have undergone dispersion compensation together in their original order, a single-mode non-dispersion signal is obtained.

[0029] In this embodiment, by combining the thickness changes caused by fatigue crack damage, the dispersion effect can be accurately corrected, ensuring that the final obtained single-mode signal is no longer affected by dispersion and guaranteeing signal accuracy. High-precision thickness measurement equipment accurately acquires the thickness data of the fatigue crack damage area and the surrounding normal area of ​​the orthotropic steel bridge deck, thereby analyzing the relationship between crack damage and thickness changes. This helps to further understand crack development and the health status of the bridge structure, providing a reliable basis for damage assessment. By establishing a dispersion curve model and combining it with numerical simulation, the dispersion effect can be accurately compensated for under different thickness conditions, thus eliminating the impact of thickness changes on signal propagation and ensuring signal accuracy. When processing the signal in segments... The length of each segment is determined reasonably based on the thickness variation gradient, making the dispersion compensation of each segment more accurate. Independent compensation of each segment is performed using methods such as time-domain inverse filtering or frequency-domain phase correction, which can minimize the dispersion error caused by thickness variations and thus improve the overall signal quality. By sequentially splicing the dispersion-compensated segments, a single-mode signal without dispersion effects is obtained, which can accurately reflect the state of the steel bridge deck and improve the accuracy of crack identification. This step, through precise thickness measurement, reasonable segmentation, and efficient dispersion compensation, improves the quality of the Lamb wave signal, providing more accurate and stable signal support for fatigue crack damage detection in steel bridges, thereby improving the accuracy of detection and monitoring.

[0030] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely exemplary illustrations of the application as defined herein, and are to be considered as covering any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Thus, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A method for fatigue crack Lamb wave signal fusion processing of orthotropic steel bridge deck, characterized in that, The application relates to a Lamb wave signal processing method for orthotropic steel bridge decks. Theoretical analysis and numerical simulation are conducted on an ultrasonic Lamb wave signal of an orthotropic steel bridge deck to obtain basic parameters of the Lamb wave detection and an original signal; Baseline correction is performed on the Lamb wave signal based on the basic parameters of the Lamb wave detection and the original signal to eliminate signal baseline drift caused by load action and obtain a baseline-free signal to be separated; Multi-modal separation is performed on the baseline-corrected Lamb wave signal for the baseline-free signal to be separated to extract low-order symmetric vibration mode and low-order anti-symmetric vibration mode signals of the Lamb wave and obtain single-mode original signals; The separated single-mode signals are processed, and a thickness change correction and dispersion compensation process caused by fatigue crack damage of the orthotropic steel bridge deck is combined to obtain single-mode non-dispersive signals.

2. The method of claim 1, wherein the method is characterized by: The basic parameters of the Lamb wave detection include dispersion characteristic parameters, mode characteristic parameters, propagation characteristic parameters and defect response characteristic parameters.

3. The method of claim 2, wherein the method further comprises: The dispersion characteristic parameters include the propagation speed of each mode at different frequencies, dispersion curves and cutoff frequencies.

4. The method of claim 2, wherein the method is characterized by: The mode characteristic parameters include wave structures, sensitivity to different types of defects and propagation characteristics at different frequency-thickness products.

5. The method of claim 2, wherein the method further comprises: The propagation characteristic parameters include optimal detection modes and frequencies, propagation speeds and attenuation coefficients.

6. The method of claim 2, wherein the method is characterized by: The defect response parameters represent the response characteristics of different modes to different types of defects and are used for identifying and distinguishing different types of defects.

7. The method of claim 3, wherein the method further comprises: Baseline correction is performed on the Lamb wave signal based on the basic parameters of the Lamb wave detection and the original signal to obtain a baseline-free signal to be separated, including: Filtering is performed on the original Lamb wave signal, and normalization is performed on the filtered signal; The theoretical time window of Lamb wave propagation in the signal is determined according to the propagation speed and the dispersion curve in the basic parameters of the Lamb wave detection; In the original signal, a time period with stable signal intensity and without defect reflection signals is selected as a baseline reference region; statistical analysis is performed on the signals in the region in combination with the theoretical propagation time window to extract baseline characteristic values; A baseline model value is constructed by using the extracted baseline characteristic values; The original signal is subtracted from the constructed baseline model value to obtain the corrected signal, i.e. the baseline-free signal to be separated.

8. The method of claim 3, wherein the method is characterized by: Multi-modal separation is performed on the baseline-corrected Lamb wave signal for the baseline-free signal to be separated to extract low-order symmetric vibration mode and low-order anti-symmetric vibration mode signals of the Lamb wave and obtain single-mode original signals, including: The propagation speed range of the low-order symmetric vibration mode and the low-order anti-symmetric vibration mode at different frequencies is determined based on the dispersion curve in the basic parameters of the Lamb wave detection; Short-time Fourier transform is performed on the baseline-corrected Lamb wave signal to convert the time-domain signal into a time-frequency domain signal and obtain a time-frequency distribution diagram of the signal; The frequency-time region containing the low-order symmetric vibration mode and the low-order anti-symmetric vibration mode is preliminarily screened out in the time-frequency distribution diagram according to the propagation speed range determined by the dispersion curve; A separation filter for the low-order symmetric vibration mode and the low-order anti-symmetric vibration mode is constructed according to the preliminarily screened-out frequency-time region; The baseline-corrected Lamb wave signal is respectively filtered by the constructed low-order symmetric vibration mode filter and low-order anti-symmetric vibration mode filter to obtain filtered signals; The filtered signals are reconstructed to obtain low-order symmetric vibration mode signals and low-order anti-symmetric vibration mode signals; The extracted low-order symmetric vibration mode signals and low-order anti-symmetric vibration mode signals are verified to obtain single-mode original signals.

9. The method of claim 2, wherein the method is characterized by: The separated single-mode signals are processed, and a thickness change correction dispersion compensation process caused by fatigue crack damage of the orthotropic steel bridge deck is combined to obtain single-mode non-dispersive signals, including: A high-precision measuring device is used to measure the thickness of the orthotropic steel bridge deck in the fatigue crack damage area and the surrounding normal area; the thickness data at different positions are recorded, the thickness change rule caused by the crack damage is analyzed, and a database of the relationship between the thickness change and the position is established; According to the material parameters and geometric dimensions of the orthotropic steel bridge deck, a dispersion curve model under different thicknesses is established in combination with the Lamb wave propagation theory; According to the measured thickness change rule, the single-mode original signal is divided into several small segments along the propagation path; For each small segment of signal, the thickness at the position is obtained from the dispersion curve model to obtain the propagation characteristic parameters of the corresponding mode Lamb wave at the thickness; The propagation characteristic parameters are used to calculate the dispersion compensation of each small segment of signal. The dispersion-compensated small signals are spliced in the original order to obtain single-mode non-dispersive signals.

10. The method of claim 9, wherein the method is characterized by: When the single-mode original signal is subjected to dispersion compensation, a time-domain inverse filtering method is adopted, and the compensation filter parameters are determined according to the propagation characteristic parameters of the mode in the dispersion curve model at the corresponding thickness.

Images