A signal denoising method and device based on sparse reconstruction

Abstract

The application relates to the field of signal processing, and particularly relates to a signal denoising method and device based on sparse reconstruction. The method comprises the following steps: acquiring a response time-frequency signal of Lamb waves; dividing the response time-frequency signal of Lamb waves into M frames of signals with a length of N, and recording the signal frame sequence; constructing a dictionary atom based on a sinusoidal function modulated by a Hann window according to the signal frame length N; combining and extending the dictionary atom into an over-complete dictionary; performing sparse decomposition on the M frames of time-frequency signals by using the over-complete dictionary to obtain M sparse representation vectors; reconstructing M frames of reconstructed signals according to the M sparse representation vectors and the over-complete dictionary; and sequentially recombining the M frames of reconstructed signals according to the signal frame sequence to obtain a denoised Lamb wave time-frequency signal. The signal-to-noise ratio of the target signal can be effectively improved, the denoising effect is achieved, and the method has the advantages of accuracy and reasonable design.

Description

Technical Field

[0001] The present application relates to the field of signal processing, in particular to mechanical wave signal processing technology, and more specifically to a signal denoising method and device based on sparse reconstruction. Background Art

[0002] Signal representation involves decomposing a given signal onto a known set of functions (or vectors), then expressing the original signal in the transform domain. Traditional methods like wavelet transforms and DCT transforms represent signals using a linear combination of a set of orthogonal basis functions. Because these basis functions are unrelated to the statistical properties of the signal itself, a complete representation of the original signal requires a significant expansion of the computational coefficients. This significantly increases the complexity of data storage and computation for signal representation, and sparsity often cannot be guaranteed.

[0003] To reduce the cost of signal processing during signal representation and improve signal processing efficiency, scholars have attempted to develop signal denoising methods with higher accuracy, lower computational complexity, and better sparsity. During signal processing, the sparse and redundant representation (SRR) method decomposes the signal based on a given overcomplete dictionary. Based on the decomposition results, the original signal is accurately represented using a linear combination of as few dictionary atoms as possible. The process of obtaining a sparse representation of the signal is therefore signal sparse decomposition. The SRR method facilitates rapid signal processing and extracts the characteristic information contained in the signal. Therefore, the SRR method reduces signal processing costs and improves signal compression efficiency.

[0004] Sparse and redundant representation algorithms provide extremely flexible signal representation methods. Given an appropriate overcomplete atom dictionary, they can solve for the sparse representation of the signal within that dictionary, thereby better capturing the signal's useful information, extracting its characteristic parameters, and reducing the complexity of subsequent signal processing. Therefore, selecting and constructing an overcomplete atom dictionary that effectively reveals the signal's key characteristics is crucial. Summary of the Invention

[0005] The embodiments of the present application provide a signal denoising method and apparatus based on sparse reconstruction, which constructs an overcomplete dictionary and reduces the reconstruction error through the solution process of the overcomplete dictionary. Through the solution process, clear sample information is obtained, a constructed signal is obtained, and the noise in the signal is reduced.

[0006] In order to achieve the above objectives, the technical solutions adopted in the embodiments of the present application are as follows:

[0007] In a first aspect, a signal representation method based on sparse reconstruction includes the following methods:

[0008] Obtain the response time-frequency signal of Lamb wave;

[0009] Divide the Lamb wave response time-frequency signal into blocks, divide the Lamb wave response time-frequency signal into M frames of length N, and record the signal frame sequence;

[0010] According to the signal frame length N, a dictionary atom based on the sine function modulated by the Hann window is constructed;

[0011] Expanding the dictionary atoms into an overcomplete dictionary;

[0012] Using the overcomplete dictionary, sparsely decompose the M frames of time-frequency signals to obtain M sparse representation vectors;

[0013] Reconstructing M frames of reconstructed signals according to the M sparse representation vectors and the overcomplete dictionary;

[0014] According to the signal frame sequence, the M frames of reconstructed signals are recombined in sequence to obtain a Lamb wave time-frequency signal after noise reduction.

[0015] In a first possible implementation, in combination with the first aspect, specifically including constructing a dictionary atom based on a sine function modulated by a Hann window according to the signal frame length N, including constructing the dictionary atom through translation and modulation transformation of the sine function modulated by the Hann window.

[0016] In a second possible implementation, specifically including the first aspect, the dictionary expression is:

[0017]

[0018] in: is the time-frequency parameter;

[0019] Where u is the translation factor, v is the frequency modulation factor, w is the phase shift factor; H(t) is the Heaviside function; and n0 is the number of Lamb peaks.

[0020] In a third possible implementation manner, in combination with the first aspect or the second possible implementation manner in the first aspect, expanding the dictionary atom combination into an overcomplete dictionary includes:

[0021] performing discrete processing on the translation factor, the frequency modulation factor, and the phase shift factor to obtain a Hann atom corresponding to each set of discrete time-frequency parameters;

[0022] The time-frequency parameters after the discrete processing are:

[0023]

[0024] In a fourth possible implementation manner, in combination with the first aspect or the third possible implementation manner in the first aspect, Δu=t s , Δv=f c / 10f s , Δw=π / 6, 0≤l<N, 0<m<M, 0<n<12;

[0025] Among them, t s is the sampling interval, f s is the sampling frequency, f s =1 / t s .

[0026] In a fifth possible implementation manner, in combination with the first aspect or the fourth possible implementation manner in the first aspect, expanding the dictionary atom combination into an overcomplete dictionary further includes:

[0027] Based on the value orientations and discretization steps of the translation factors, the frequency modulation factors, and the phase shift factors corresponding to different Hann atoms, index evaluation is performed in sequence.

[0028] In a sixth possible implementation manner, in combination with the implementation manner of the first aspect, the method of using the overcomplete dictionary to perform sparse decomposition on the M frames of time-frequency signals to obtain M sparse representation vectors includes the following method:

[0029] Perform sparse decomposition based on the sparse model to obtain a sparse representation vector;

[0030] The sparse model includes:

[0031]

[0032] Where: r m is the residual signal, D is the dictionary, x is the coefficient representation vector and is a non-zero value coefficient, x γ is the decomposition coefficient, d γ is an atom in the dictionary D.

[0033] In a seventh possible implementation manner, in combination with the first aspect or the sixth possible implementation manner of the first aspect, the method includes sequentially recombining the M-frame reconstructed signals according to the signal frame sequence to obtain a Lamb wave time-frequency signal after noise reduction, including the following method:

[0034] For each frame of data block, the atom that best matches the target signal is selected from the overcomplete atom dictionary, and the OMP decomposition is performed on the data block separately. When the iterative termination condition is met, the decomposition is terminated, and the signal is represented by a linear combination of the selected best-matching atoms to obtain a frame of denoised signal; each segment of the signal is sparsely represented and reconstructed to obtain M frames of denoised signals, and the M frames of denoised signals are recombined to obtain the denoised Lamb wave signal.

[0035] In an eighth possible implementation, in combination with the first aspect or the seventh possible implementation of the first aspect, further comprising selecting the atom that best matches the target signal b in the overcomplete dictionary using an orthogonal projection method. So that the following formula holds:

[0036]

[0037] Among them, R n b is the residual signal obtained after the nth iteration, is the atom matched in the nth iteration, and b is the target signal.

[0038] In a second aspect, a device is provided, comprising:

[0039] It includes an acquisition module for acquiring the response time-frequency signal of the Lamb wave;

[0040] The data segmentation module is used to divide the Lamb wave response time-frequency signal into blocks, into M frames of length N, and record the sequence of the M frames of signals;

[0041] The dictionary atom construction module constructs dictionary atoms based on the sine function modulated by the Hann window according to the signal frame length N;

[0042] Overcomplete dictionary building module, used to expand dictionary atoms into an overcomplete dictionary;

[0043] A calculation module is used to perform sparse decomposition on the response time-frequency signal of the M-frame Lamb wave using an overcomplete dictionary to obtain a sparse representation vector;

[0044] The signal reconstruction module is used to reconstruct M frames of signals based on the sparse representation vector and the overcomplete dictionary, and to sequentially reorganize the M frames of signals to obtain the response time-frequency signal of the reconstructed Lamb wave.

[0045] The technical solution provided in the embodiment of the present application is accurate and efficient, with simple steps, reasonable design, and easy implementation: the target signal is decomposed and reconstructed using the OMP algorithm, which can effectively improve the signal-to-noise ratio of the target signal and achieve a denoising effect. Experiments have verified that it has the advantages of accuracy and reasonable design; the block processing method can greatly improve the speed of sparse decomposition, solving the problem of an over-complete dictionary being too large and the large amount of sparse decomposition calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0047] The methods, systems, and / or programs in the accompanying drawings will be further described according to exemplary embodiments. These exemplary embodiments will be described in detail with reference to the drawings. These exemplary embodiments are non-limiting exemplary embodiments, wherein example numerals represent similar structures in the various views of the drawings.

[0048] Figure 1 is a flow chart of a method according to some embodiments of the present application;

[0049] Figure 2 is a flow chart of a method according to another embodiment of the present application;

[0050] Figure 3 is a structural diagram of a device according to an embodiment of the present application;

[0051] Figure 4 It is a structural diagram of a device according to another embodiment of the present application. DETAILED DESCRIPTION

[0052] In order to better understand the above technical solution, the technical solution of the present application is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical solution of the present application, rather than limitations on the technical solution of the present application. In the absence of conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other.

[0053] In the following detailed description, numerous specific details are set forth by way of example in order to provide a thorough understanding of the relevant teachings. However, it will be apparent to one skilled in the art that the present application can be practiced without these details. In other instances, well-known methods, procedures, systems, compositions, and / or circuits have been described at a relatively high level, without detail, to avoid unnecessarily obscuring aspects of the present application.

[0054] Flowcharts are used in this application to illustrate the execution processes performed by the system according to the embodiments of the present application. It should be clearly understood that the execution processes of the flowcharts may not be executed in sequence. Instead, these execution processes may be executed in reverse order or simultaneously. In addition, at least one additional execution process may be added to the flowchart. One or more execution processes may be deleted from the flowchart.

[0055] See also Figure 1 This embodiment provides a signal denoising method based on sparse reconstruction, which specifically includes the following steps:

[0056] 101. Obtain the Lamb wave response time-frequency signal.

[0057] In this embodiment, the response time-frequency signal of the Lamb wave is mainly acquired through piezoelectric ceramics, and the acquired ultrasonic time-frequency signals of all positions are stored.

[0058] 102. Divide the Lamb wave response time-frequency signal into blocks, into M frames of length N, and record the signal frame sequence.

[0059] Because the size of the overcomplete atom dictionary used in the sparse decomposition process depends on the length of the signal being processed, the longer the signal, the larger the number of atoms contained in the dictionary. This increases the computational complexity of the atom matching process in the decomposition algorithm, and the computational speed decreases. To address this issue, we draw on the idea of ​​block decomposition in image processing and decompose the collected signal data into blocks. A frame of block data is only a portion of the complete data sequence.

[0060] 103. According to the signal frame length N, a dictionary atom based on the sine function modulated by Hann window is constructed.

[0061] In this embodiment, the dictionary is implemented by obtaining a set of time-frequency parameters (u, v, w) by shifting and modulating a sine function modulated by a Hann window, thereby obtaining a dictionary. In this embodiment, the dictionary is expressed as follows:

[0062]

[0063] Among them, u is the translation factor, v is the frequency modulation factor, w is the phase shift factor, and t is time.

[0064] 104. Extend dictionary atom combinations to overcomplete dictionaries.

[0065] In this embodiment, the overcomplete dictionary includes M dictionaries, where M is a positive integer and the length of each dictionary is n, where n is a positive number and m<M. The specific expansion process includes:

[0066] The translation factor, frequency modulation factor and phase shift factor are discretized to obtain each set of discretized time-frequency parameters and a dictionary corresponding to the time-frequency parameters, which are then indexed in sequence to obtain an expanded over-complete dictionary.

[0067] In this embodiment, the time-frequency parameters after discretization are:

[0068]

[0069] The resulting overcomplete dictionary is:

[0070] D=[d1,d2,…,d γ ,…,d Γ ]∈R N×Γ (N<Γ), where D is an overcomplete dictionary, d γ is an atom in the dictionary D.

[0071] 105. Using an overcomplete dictionary, perform sparse decomposition on M frames of time-frequency signals to obtain M sparse representation vectors.

[0072] In this embodiment, the M dictionaries in the overcomplete dictionary are used to perform sparse decomposition on the response time-frequency signal of the Lamb wave to obtain a sparse representation vector, including the following method:

[0073] Perform sparse decomposition based on the sparse model to obtain a sparse representation vector;

[0074] The sparse model includes:

[0075]

[0076] The model of the target signal is:

[0077]

[0078] Where: r m is the residual signal, D is the dictionary, x is the coefficient representation vector and is a non-zero value coefficient, x γ is the decomposition coefficient, d γ It is an atom of the dictionary D, specifically a normalized N-dimensional column vector.

[0079] In this embodiment, b is an N-dimensional column vector, which is the target signal.

[0080] 106. Reconstruct M frames of reconstructed signals based on M sparse representation vectors and overcomplete dictionary.

[0081] 107. The denoised Lamb wave time-frequency signal is obtained by sequentially recombining the M-frame reconstructed signals according to the signal frame sequence.

[0082] In this embodiment, the specific method is:

[0083] At least one sample subset of the signal samples is obtained by loop iteration overcomplete dictionary processing, and the loop iteration is terminated when a predetermined judgment condition is met, and the at least one signal sample is statistically analyzed to form a new signal.

[0084] Specifically, the OMP algorithm is used to perform signal sparse dispersion loop iteration. First, the orthogonal projection method is used to select the atom c that best matches the target signal b in the determined overcomplete dictionary D. γ , so that it meets the following conditions:

[0085]

[0086] Among them, R n b is the residual signal obtained after the nth iteration, The atom matched at the nth iteration.

[0087] And decompose the target signal into components and residuals on the best atom, that is,

[0088] R n b= <R n b, c γ >c γ +R1;

[0089] Where R1 is the atom c γ R n b The remaining part after matching.

[0090] The decomposition process of the formula is repeated for the residual after the best match, where the formula is:

[0091] R t = <R t , c γn >c γn +R t+1 .

[0092] Use the orthogonalization method to match the atoms c at each step of the decomposition. γn Orthogonalization is performed, which can not only speed up the convergence speed but also avoid the residual error in the atomic c γn Unnecessary components are introduced when projecting upward.

[0093] The target signal b is where R m b is the residual signal obtained after the mth iteration, that is, the residual.

[0094] This embodiment also provides a signal reconstruction device 100, which is specifically any single device in a signal processing system, such as a computer, a laptop computer, or any other device in the signal processing system that can implement a signal denoising method based on sparse reconstruction, and is subject to any of the above-mentioned sparse reconstruction signal reconstruction methods provided by the embodiments of the present invention, see Figure 3 Shown, including:

[0095] The acquisition module 110 is used to obtain the response time-frequency signal of the Lamb wave;

[0096] The data segmentation module 120 is used to divide the Lamb wave response time-frequency signal into blocks, into M frames of signal with a length of N, and record the sequence of the M frames of signal;

[0097] The dictionary atom construction module 130 constructs a dictionary atom based on a sine function modulated by a Hann window according to a signal frame length N;

[0098] An overcomplete dictionary construction module 140 is used to expand dictionary atoms into an overcomplete dictionary;

[0099] The calculation module 150 is used to perform sparse decomposition on the response time-frequency signal of the Lamb wave of the M-frame signal using an overcomplete dictionary to obtain a sparse representation vector.

[0100] The signal reconstruction module 160 is used to reconstruct M frames of signals according to the sparse representation vector and the overcomplete dictionary, and to sequentially reassemble the M frames of signals to obtain a reconstructed Lamb wave response time-frequency signal.

[0101] See Figure 4 As shown, the signal reconstruction device also includes:

[0102] A setting unit 151 is used to set an iteration termination condition according to the establishment of the overcomplete dictionary;

[0103] The first calculation subunit 152 is used to select the atom that best matches the target signal from the overcomplete atom dictionary D for each frame data block, perform OMP decomposition on the data block separately, and update the signal residual;

[0104] A verification unit 153, configured to verify whether the iteration requirement is met;

[0105] The second calculation subunit 154 is used to represent the signal by a linear combination of the selected best-matching atoms to obtain a denoised signal of M frames;

[0106] The recombining unit 161 recombines the M frames of denoised signals to obtain denoised Lamb wave signals.

[0107] In one or more embodiments, a terminal device is disclosed, including a server. The server includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, a signal reconstruction method based on sparse reconstruction is implemented in accordance with an embodiment. For the sake of brevity, this description is omitted here.

[0108] It should be understood that in this embodiment, "module" may include a unit implemented in hardware, software, or firmware, and may be used interchangeably with other terms such as "logic," "logic block," "component," and "circuit." A module may be a single integrated component or the smallest unit or component of such a single integrated component that is adapted to perform one or more functions. For example, according to an embodiment, a module may be implemented in the form of an application-specific integrated circuit (ASIC).

[0109] The above disclosures of the embodiments of this application are clear and complete for those skilled in the art. It should be understood that the deduction and analysis of unexplained technical terms by those skilled in the art based on the above disclosures are based on the contents described in this application, and therefore the above disclosures do not constitute a judgment on the creativity of the overall solution.

[0110] The basic concepts have been described above. It will be apparent to those skilled in the art that the detailed disclosure provided above is merely illustrative and does not limit the present application. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and amendments to the present application. Such modifications, improvements, and amendments are suggested in the present application and remain within the spirit and scope of the exemplary embodiments of the present application.

[0111] At the same time, this application uses specific terms to describe the embodiments of this application. For example, "one embodiment," "an embodiment," and / or "some embodiments" refer to a certain feature, structure, or characteristic related to at least one embodiment of this application. Therefore, it should be emphasized and noted that "one embodiment," "an embodiment," or "an alternative embodiment" mentioned twice or more in different parts of this specification does not necessarily refer to the same embodiment. In addition, certain features, structures, or characteristics in at least one embodiment of this application may be appropriately combined.

[0112] In addition, it will be understood by those skilled in the art that the various aspects of the present application can be illustrated and described by a number of patentable categories or situations, including any new and useful process, machine, product or material combination, or any new and useful improvement thereof. Accordingly, the various aspects of the present application can be performed entirely by hardware, can be performed entirely by software (including firmware, resident software, microcode, etc.), or can be performed by a combination of hardware and software. The above hardware or software can all be referred to as "units", "components" or "systems". In addition, the various aspects of the present application can be expressed as a computer product located in at least one computer-readable medium, and the product includes computer-readable program code.

[0113] A computer-readable signal medium may include a propagated data signal containing computer program code, such as in baseband or as part of a carrier wave. The propagated signal may have a variety of manifestations, including electromagnetic, optical, etc., or a suitable combination. A computer-readable signal medium may be any computer-readable medium other than a computer-readable storage medium that can be connected to an instruction execution system, device, or apparatus to communicate, propagate, or transmit a program for use. The program code on the computer-readable signal medium may be propagated via any suitable medium, including radio, cable, fiber optic cable, RF, or similar media, or any combination of the above media.

[0114] The computer program code required for the execution of various aspects of the present application can be written in any combination of one or more programming languages, including object-oriented programming, such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python, etc., or similar conventional programming languages, such as the "C" programming language, Visual Basic, Fortran 2003, Perl, COBOL 2002, PHP, ABAP, dynamic programming languages ​​such as Python, Ruby and Groovy or other programming languages. The programming code can be executed entirely on the user's computer, or as a separate software package on the user's computer, or partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the latter case, the remote computer can be connected to the user's computer through any network form, such as a local area network (LAN) or a wide area network (WAN), or connected to an external computer (e.g., via the Internet), or in a cloud computing environment, or as a service such as software as a service (SaaS).

[0115] Furthermore, the order of processing elements or steps, or the order in which they are presented, unless specifically stated in the claims, is not intended to be limiting, unless specifically stated in the claims. Although the above disclosure discusses some presently preferred embodiments of the application, the embodiments discussed are not intended to limit the application to those specifically disclosed. On the contrary, the application is intended to cover all modifications, and equivalents that fall within the scope of the present application; additionally, although the application is illustrated and described a number of embodiments, it is understood that such embodiments are merely exemplary of the application, and are not intended to limit the same. For example, although the system components described above can be implemented as hardware devices, they can also be implemented as software solutions, such as installing the described system on an existing server or mobile device.

[0116] It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting, unless specifically stated in the claims. It must be noted that, as used in the specification and the appended claims, the singular forms "a," "an" and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to "a component" or "the component" can include a plurality of such components, and so forth.

Claims

1. A signal denoising method based on sparse reconstruction, characterized in that: The method comprises: Step 101, obtaining a Lamb wave response time-frequency signal; Step 102, dividing the Lamb wave response time-frequency signal into blocks, dividing the Lamb wave response time-frequency signal into M frames of length N, and recording the signal frame sequence; Step 103: construct a dictionary atom based on the sine function modulated by the Hann window according to the signal frame length N; wherein the expression of the dictionary is: ； in: = is the time-frequency parameter; in, is the translation factor, is the frequency modulation factor, is the phase shift factor; is the Heaviside function; is the Lamb peak number; Step 104: Expand the dictionary atom combination into an overcomplete dictionary. Expanding the dictionary atom combination into an overcomplete dictionary includes: performing discrete processing on the translation factor, the frequency modulation factor, and the phase shift factor to obtain a Hann atom corresponding to each set of discrete time-frequency parameters; The time-frequency parameters after the discrete processing are: ； in, , , , , , ; in, is the sampling interval, is the sampling frequency, ; Step 105: Using the overcomplete dictionary, sparsely decompose the M frames of time-frequency signals to obtain M sparse representation vectors; wherein using the overcomplete dictionary, sparsely decompose the M frames of time-frequency signals to obtain M sparse representation vectors, includes the following method: Perform sparse decomposition based on the sparse model to obtain a sparse representation vector; The sparse model includes: ； in: is the residual signal, For the dictionary, is a coefficient representation vector and is a non-zero coefficient. is the decomposition coefficient, For dictionary an atom of Step 106, reconstructing M frames of reconstructed signals according to the M sparse representation vectors and the overcomplete dictionary; Step 107: According to the signal frame sequence, the M frames of reconstructed signals are sequentially recombined to obtain a Lamb wave time-frequency signal after noise reduction.

2. The signal denoising method based on sparse reconstruction according to claim 1, characterized in that: The step 103 constructs dictionary atoms based on the Hann window modulated sine function according to the signal frame length N, including constructing the dictionary atoms through translation and modulation transformation of the Hann window modulated sine function.

3. The signal denoising method based on sparse reconstruction according to claim 1, characterized in that: In step 104, expanding the dictionary atoms into an overcomplete dictionary further includes: Based on the value orientations and discretization steps of the translation factors, the frequency modulation factors, and the phase shift factors corresponding to different Hann atoms, index evaluation is performed in sequence.

4. The signal denoising method based on sparse reconstruction according to claim 1, characterized in that: In step 107, according to the signal frame sequence, the M frames of reconstructed signals are sequentially recombined to obtain a Lamb wave time-frequency signal after noise reduction, including the following method: For each frame of data block, the atom that best matches the target signal is selected from the overcomplete atom dictionary, and the OMP decomposition is performed on the data block separately. When the iterative termination condition is met, the decomposition is terminated, and the signal is represented by a linear combination of the selected best-matching atoms to obtain a frame of denoised signal; each segment of the signal is sparsely represented and reconstructed to obtain M frames of denoised signals, and the M frames of denoised signals are recombined to obtain the denoised Lamb wave signal.

5. The signal denoising method based on sparse reconstruction according to claim 4, characterized in that: It also includes the use of orthogonal projection method to select the target signal in the overcomplete dictionary. The best matching atom , so that the following equation holds: ； in, For the The residual signal obtained after iterations is For the The atoms matched by iterations, is the target signal.

6. A signal noise reduction device, characterized in that: include: Acquisition module, used to obtain the response time-frequency signal of Lamb wave; The data segmentation module is used to divide the Lamb wave response time-frequency signal into blocks, into M frames of length N, and record the sequence of the M frames of signals; The dictionary atom construction module constructs a dictionary atom based on the sine function modulated by the Hann window according to the signal frame length N; wherein the expression of the dictionary is: ； in: = is the time-frequency parameter; in, is the translation factor, is the frequency modulation factor, is the phase shift factor; is the Heaviside function; is the Lamb peak number; An overcomplete dictionary construction module is used to expand the dictionary atom combination into an overcomplete dictionary; wherein, expanding the dictionary atom combination into an overcomplete dictionary includes: performing discrete processing on the translation factor, the frequency modulation factor, and the phase shift factor to obtain a Hann atom corresponding to each set of discrete time-frequency parameters; The time-frequency parameters after the discrete processing are: ； in, , , , , , ; in, is the sampling interval, is the sampling frequency, ; A computing module is configured to perform sparse decomposition on the M frames of time-frequency signals using the overcomplete dictionary to obtain M sparse representation vectors; wherein the performing sparse decomposition on the M frames of time-frequency signals using the overcomplete dictionary to obtain M sparse representation vectors comprises the following method: Perform sparse decomposition based on the sparse model to obtain a sparse representation vector; The sparse model includes: ； in: is the residual signal, For the dictionary, is a coefficient representation vector and is a non-zero coefficient. is the decomposition coefficient, For dictionary an atom of The signal reconstruction module is used to reconstruct M frames of signals based on the sparse representation vector and the overcomplete dictionary, and to sequentially reorganize the M frames of signals to obtain the response time-frequency signal of the reconstructed Lamb wave.

Images