Lamb wave signal matching pursuit denoising method and system based on particle swarm optimization

Abstract

The invention discloses a Lamb wave signal matching pursuit de-noising method and system based on particle swarm optimization, which are applied to the technical field of structural health monitoring signal processing, and the method comprises the steps: collecting a Lamb wave signal of a structure, determining an atomic model according to the Lamb wave signal, and calculating a particle fitness value according to the Lamb wave signal and the atomic model; constructing an objective function based on the particle fitness value, and iteratively updating particle swarm parameters according to the objective function to obtain optimized parameters; and substituting the optimization parameter into the atomic model, performing sparse decomposition on the Lamb wave signal to obtain a decomposition result, and performing de-noising processing on the decomposition result to obtain a de-noised Lamb wave signal. In this way, the particle swarm optimization matching pursuit algorithm is utilized, the optimal parameter set of the atomic model is generated, signal sparse decomposition is achieved, the denoising time is greatly shortened, and therefore the matching pursuit denoising efficiency is improved.

Description

Technical Field

[0001] The present disclosure relates to the technical field of structural health monitoring signal processing, and in particular to a Lamb wave signal matching pursuit denoising method and system based on particle swarm optimization. Background Art

[0002] In recent years, with the development of science and technology, the requirements for health monitoring of structures in service have become increasingly higher. To this end, relevant scholars have proposed structural health monitoring technology, which collects monitoring signals in real time by placing sensors on the surface of the structure or integrating them into the structure, and then combines relevant technologies to monitor the health status of the structure. Among them, Lamb wave signals are widely used as information carriers of structural health monitoring technology because of their sensitivity to damage. Structural health monitoring technology based on Lamb waves is one of the important development directions in the field of structural health monitoring in the future.

[0003] In the health monitoring of structures, due to the influence of the external environment in service, there will be noise signals in the collected Lamb wave signals, which will affect the subsequent signal processing. In order to reduce the influence of noise, the collected Lamb wave signals need to be denoised. At present, the commonly used Lamb wave signal denoising methods mainly include wavelet transform, singular value decomposition and variational mode decomposition. Although these methods have a very fast denoising time, the denoising effect is very effective, especially in the case of strong noise, the noise cannot be effectively removed.

[0004] Compared with the above methods, the matching pursuit denoising method utilizes the principle of signal sparse decomposition, decomposing the signal into an effective part and a noise part, and removing the noise part to obtain the denoised Lamb wave signal. However, before performing signal sparse decomposition, all atoms need to be generated, which takes a long time, resulting in a long running time when using the matching pursuit denoising method to denoise the Lamb wave signal, which is not conducive to the real-time denoising of the Lamb wave signal.

[0005] Therefore, there is an urgent need for a technical solution that can utilize the matching pursuit denoising method of signal sparse decomposition to achieve real-time signal denoising. Summary of the invention

[0006] The present invention provides a Lamb wave signal matching pursuit denoising method and system based on particle swarm optimization. By optimizing the matching pursuit denoising method based on the particle swarm algorithm, sparse decomposition of the signal is achieved, which at least solves the technical problems of the prior art that the signal denoising process takes a long time and is inefficient.

[0007] According to a first aspect of the present disclosure, a Lamb wave signal matching pursuit denoising method based on particle swarm optimization is provided, comprising the following steps:

[0008] Collecting Lamb wave signals of the structure, determining an atomic model according to the Lamb wave signals, and calculating a particle fitness value according to the Lamb wave signals and the atomic model;

[0009] Constructing an objective function based on the fitness value of the particles, iteratively updating the particle swarm parameters according to the objective function, and obtaining optimized parameters;

[0010] The optimized parameters are substituted into the atomic model, the Lamb wave signal is sparsely decomposed to obtain a decomposition result, and the decomposition result is denoised to obtain a denoised Lamb wave signal.

[0011] According to the above aspects and any possible implementation, an implementation is further provided, wherein the process of collecting the Lamb wave signal of the structure, determining the atomic model according to the Lamb wave signal, and calculating the particle fitness value according to the Lamb wave signal and the atomic model is as follows:

[0012] Acquire Lamb wave excitation signals of the structure, define amplitude and phase as a parameter group, calculate different atoms according to the excitation signals and adopt amplitude scaling and phase shifting methods to construct an atomic model;

[0013] The Lamb wave time domain signal is expanded on the atom, the noise and the effective part in the Lamb wave time domain signal are calculated, the inner product value of the atom model and the Lamb wave time domain signal is obtained, and the inner product value is used as the fitness value of the particle swarm algorithm.

[0014] According to the above aspects and any possible implementation, an implementation is further provided, wherein the atomic model is specifically:

[0015] g i (t) = A i u i (t+iΔt)

[0016] Among them, g i (t) is the i-th atom; A i is the amplitude expansion factor, Δt is the phase shift factor, u i is the excitation signal, i is the atomic number, and t is the time.

[0017] According to the above aspects and any possible implementation, an implementation is further provided, wherein the Lamb wave time domain signal is expanded on atoms, the noise and effective part in the Lamb wave time domain signal are calculated, and the inner product value of the atomic model and the Lamb wave time domain signal is obtained, and the inner product value is used as the fitness value of the particle swarm algorithm, specifically:

[0018]

[0019] Where k is the number of iterations, ranging from 0<k≤K; i represents the atomic number; is the inner product value of the ith atom and the signal in the kth iteration, and f(t) is the Lamb wave signal in the time domain.

[0020] According to the above aspects and any possible implementation, an implementation is further provided, wherein the objective function is constructed based on the particle fitness value, and the particle swarm parameters are iteratively updated according to the objective function to obtain the optimized parameters in the following process:

[0021] Initialize the particle swarm parameters and calculate the fitness value corresponding to each particle according to the objective function;

[0022] The particle swarm parameters are iteratively optimized according to the fitness value to obtain optimized parameters.

[0023] According to the above aspects and any possible implementation, an implementation is further provided, wherein the process of initializing the particle swarm parameters and calculating the fitness value corresponding to each particle according to the objective function is as follows:

[0024] Initialize the particle swarm size, number of evolutions, initial position and speed, and calculate the fitness value corresponding to each particle;

[0025] The individual historical optimal value is obtained according to the fitness value corresponding to each particle, and the group historical optimal value is obtained according to the objective function.

[0026] According to the above aspects and any possible implementation, an implementation is further provided, wherein the process of iteratively optimizing the particle swarm parameters according to the fitness value to obtain the optimized parameters is:

[0027] A self-learning factor, a social learning factor and an inertia weight factor are defined, and particle velocity and position parameters are calculated according to the individual historical optimal value and the group historical optimal value;

[0028] The calculation process of the particle speed and position parameters is continuously iterated, and when the maximum number of iterations is reached, the optimal particle speed and position parameters are obtained.

[0029] According to the above aspects and any possible implementation, an implementation is further provided, wherein the optimization parameters are substituted into the atomic model, the Lamb wave signal is sparsely decomposed to obtain a decomposition result, and the decomposition result is denoised to obtain a denoised Lamb wave signal, wherein the process is as follows: the optimal speed and position of the particle are substituted into the atomic model, the Lamb wave signal is sparsely decomposed to obtain a valid signal and a residual signal;

[0030] The residual signal is used as the original signal, and the sparse decomposition process is repeated. When the maximum number of sparse decompositions is reached, a final effective signal and a final residual signal are output;

[0031] The difference between the Lamb wave signal and the final residual signal is calculated to obtain a denoised Lamb wave signal.

[0032] According to a second aspect of the present disclosure, a Lamb wave signal matching pursuit denoising system based on particle swarm optimization is provided, which is used to implement the Lamb wave signal matching pursuit denoising method based on particle swarm optimization as described in the first aspect, including: a particle fitness value calculation module, an optimization parameter calculation module and a Lamb wave signal denoising module;

[0033] The particle fitness value calculation module is used to collect Lamb wave signals of the structure, determine the atomic model according to the Lamb wave signals, and calculate the particle fitness value according to the Lamb wave signals and the atomic model;

[0034] The optimization parameter calculation module is used to construct an objective function based on the particle fitness value, iteratively update the particle swarm parameters according to the objective function, and obtain the particle swarm optimization parameters;

[0035] The Lamb wave signal denoising module is used to substitute the particle swarm optimization parameters into the atomic model, perform sparse decomposition on the Lamb wave signal to obtain a decomposition result, and perform denoising on the decomposition result to obtain a denoised Lamb wave signal.

[0036] Compared with the prior art, the present invention has the following technical effects:

[0037] The present invention utilizes a particle swarm algorithm to optimize the optimal atom generation method in a matching pursuit algorithm, obtains the optimal parameter group of the atomic model through particle swarm optimization, generates the optimal atom based on the parameter group, realizes sparse decomposition of the signal, and then removes the noise signal; the present invention does not need to generate all atoms in the sparse decomposition, and only requires the optimal atom in each iteration, so the denoising time of the signal by the matching pursuit algorithm can be greatly reduced, the matching pursuit denoising efficiency of the signal can be effectively improved, and a new method is provided for the rapid denoising of Lamb wave signals.

[0038] It should be understood that the contents described in the summary of the invention are not intended to limit the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, among which:

[0040] Figure 1 A schematic diagram of a process of matching pursuit denoising of Lamb wave signals based on particle swarm optimization according to an embodiment of the present disclosure is shown;

[0041] Figure 2 A schematic diagram of the structure of a Lamb wave signal matching pursuit denoising system based on particle swarm optimization according to an embodiment of the present disclosure is shown;

[0042] Figure 3 A physical picture of a carbon fiber composite material plate in an experiment of a Lamb wave signal matching pursuit denoising method based on particle swarm optimization according to an embodiment of the present disclosure is shown;

[0043] Figure 4 A time domain diagram of an excitation signal of a Lamb wave signal matching pursuit denoising method based on particle swarm optimization according to an embodiment of the present disclosure is shown;

[0044] Figure 5 A schematic diagram of a rectangular coordinate system of a carbon fiber composite material plate constructed according to an embodiment of a Lamb wave signal matching pursuit denoising method based on particle swarm optimization according to an embodiment of the present disclosure is shown;

[0045] Figure 6 The figure shows the experimental result of denoising the Lamb wave damage monitoring signal under the noise level of 10 dB using a matching pursuit denoising method for Lamb wave signals based on particle swarm optimization according to an embodiment of the present disclosure and empirical mode decomposition (EMD), singular value decomposition (SVD), wavelet transform and basis pursuit denoising (MP) methods;

[0046] Figure 7 The graph shows the experimental results of Lamb wave damage monitoring signal denoising under 5 dB noise level using a Lamb wave signal matching pursuit denoising method based on particle swarm optimization according to an embodiment of the present disclosure and EMD, SVD, wavelet transform and MP methods;

[0047] Figure 8The graph shows the experimental results of Lamb wave damage monitoring signal denoising at a noise level of 0 dB using a Lamb wave signal matching pursuit denoising method based on particle swarm optimization according to an embodiment of the present disclosure and EMD, SVD, wavelet transform and MP methods. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the embodiments of the present disclosure clearer, the technical solution in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0049] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0050] Reference Figure 1 As shown, this embodiment provides a Lamb wave signal matching pursuit denoising method based on particle swarm optimization, comprising the following steps:

[0051] S101, collecting Lamb wave signals of the structure, determining the atomic model according to the Lamb wave signals, and calculating the particle fitness value according to the Lamb wave signals and the atomic model.

[0052] In this embodiment, the atomic model is constructed by collecting the frequency domain excitation signal, and the fitness value is calculated according to the atomic model. The specific process is:

[0053] In this embodiment, the frequency domain expression of the Lamb wave signal is:

[0054] F(ω)=|R(ω)|e jθ(ω) U(ω) (1)

[0055] Among them, U(ω) is the frequency domain excitation signal; F(ω) is the frequency domain Lamb wave signal; |R(ω)| is the amplitude change coefficient, and θ(ω) is the phase change coefficient.

[0056] From equation (1), we can see that the propagated signal only has amplitude and phase changes relative to the time domain excitation signal u(t). Therefore, the amplitude and phase can be defined as a parameter group, and different atoms can be obtained by amplitude scaling and phase shift. The i-th atomic model is represented as follows:

[0057] g i (t) = A i u i (t+iΔt) (2)

[0058] Among them, g i (t) is the i-th atom; A i is the amplitude expansion factor, Δt is the phase shift factor, u i is the excitation signal, i is the atomic number, and t is the time.

[0059] The key to sparse decomposition based on the matching pursuit algorithm is to find the atom that is most similar to the signal, expand the signal on the atom, and obtain the noise and effective part of the signal. The inner product can represent the similarity of two functions. The inner product value of the signal and the atom is used as the fitness value of the particle swarm algorithm, which is expressed as follows:

[0060]

[0061] Where k is the number of iterations, ranging from 0<k≤K; i represents the atomic number; is the inner product value of the ith atom and the signal in the kth iteration, and f(t) is the Lamb wave signal in the time domain.

[0062] Specifically, in this embodiment, it is necessary to determine the number of sparse decompositions in the matching pursuit algorithm to obtain a specific result of the inner product value.

[0063] S102, constructing an objective function based on the particle fitness value, and iteratively updating the particle swarm parameters according to the objective function to obtain optimized parameters.

[0064] In this embodiment, the maximum value of the fitness value is used as the objective function, specifically:

[0065]

[0066] in, is the maximum fitness value in the kth iteration.

[0067] In this embodiment, before iteratively updating the fitness value of the particle swarm, the particle swarm size, evolution times, initial position and speed in the particle swarm algorithm are initialized, and the fitness value corresponding to each particle is calculated, which is regarded as the individual historical optimal value, and the maximum value is regarded as the group historical optimal value. Then the speed and position of the particle swarm are updated as follows:

[0068]

[0069] Where t is time; p i (t) is the individual historical optimal fitness value of the i-th particle in the current iteration, p g (t) is the optimal fitness value of the swarm in the current iteration of the i-th particle; v i (t) and x i (t) are the velocity and position of the ith particle in the current iteration, vi (t+1) and x i (t+1) are the speed and position of the next iteration respectively; r1 and r2 are random numbers that obey the uniform distribution on (0, 1); c1 and c2 are the self-learning factor and social learning factor respectively; ω is the inertia weight factor.

[0070] After the particle swarm parameters are updated, the fitness value of each particle is calculated, the historical optimal value of the swarm is updated, and the corresponding speed and position are the optimal speed and position of the particle, and then the next iteration is entered. After reaching the maximum number of iterations, the optimal parameter group corresponding to the optimal position can be obtained.

[0071] S103, substituting the optimized parameters into the atomic model, performing sparse decomposition on the Lamb wave signal to obtain a decomposition result, and performing denoising on the decomposition result to obtain a denoised Lamb wave signal.

[0072] After obtaining the optimal parameter set, substitute the parameters into the atomic model according to this step and perform sparse decomposition on the Lamb wave signal as follows:

[0073]

[0074] Among them, the right side of the equation and They are respectively the effective signal and residual signal obtained by the k-th sparse decomposition, and the residual signal is a noisy signal.

[0075] Then, the residual signal is regarded as the original signal, and the steps are repeated for the next sparse decomposition process until the maximum number of sparse decompositions is reached.

[0076] The final denoised signal in the above steps can be obtained as follows:

[0077]

[0078] Among them, f K (t) They are the final denoised Lamb wave signal and noise signal respectively.

[0079] like Figure 2 As shown, in this embodiment, a Lamb wave signal matching pursuit denoising system based on particle swarm optimization is also provided, including: a particle fitness value calculation module 1, an optimization parameter calculation module 2 and a Lamb wave signal denoising module 3;

[0080] The particle fitness value calculation module 1 is used to collect the Lamb wave signal of the structure, determine the atomic model according to the Lamb wave signal, and calculate the particle fitness value according to the Lamb wave signal and the atomic model;

[0081] The optimization parameter calculation module 2 is used to construct an objective function based on the particle fitness value, iteratively update the particle swarm parameters according to the objective function, and obtain the particle swarm optimization parameters;

[0082] The Lamb wave signal denoising module 3 is used to substitute the particle swarm optimization parameters into the atomic model, perform sparse decomposition on the Lamb wave signal to obtain a decomposition result, and perform denoising on the decomposition result to obtain a denoised Lamb wave signal.

[0083] Example

[0084] This embodiment is a denoising experiment of Lamb wave damage monitoring signal under 10dB, 5dB and 0dB noise levels of the present invention. Figure 3 The figure shows the experimental carbon fiber composite material plate, which has a size of 500mm×500mm×2mm and a ply ratio of [45 / 0 / -45 / 90 / -45 / 0 / 45 / 90]s. Six circular thin-film piezoelectric sensors with a diameter of 10mm are arranged at the upper and lower ends of the plate for the excitation and reception of Lamb waves. The excitation signal is as follows: Figure 4 As shown, it is a five-peak sinusoidal modulation signal with a center frequency of 60kHz. The xy axis coordinate system is established with the lower left corner of the CFRP plate as the coordinate origin, the bottom boundary as the x axis, and the left boundary as the y axis, as shown in Figure 5 As shown, the range of both axes is 0-500mm, the blue circles on the upper and lower sides indicate the sensor position, and the red circle in the middle indicates the damage position. The damage response signal and the reference signal when there is no damage are collected respectively, and the damage monitoring signal is obtained by subtracting them. In addition to the present invention, this embodiment performs EMD, SVD, wavelet transform and MP algorithm denoising experiments as comparison. Figure 6 Figure 1 shows the experimental results of Lamb wave damage monitoring signal denoising under a noise level of 10 dB, where the signal diagram with serial number (a) represents the noisy signal, and the signal diagrams with serial numbers (b), (c), (d), (e), and (f) represent the denoised signal diagrams obtained after denoising the noisy signal using EMD, SVD, wavelet transform, MP, and particle swarm-MP algorithms, respectively. Figure 7 The graphs are the experimental results of Lamb wave damage monitoring signal denoising under 5 dB noise level, where the signal graph with serial number (a) represents the noisy signal, and the signal graphs with serial numbers (b), (c), (d), (e), and (f) represent the denoised signal graphs obtained after denoising the noisy signal using EMD, SVD, wavelet transform, MP, and particle swarm-MP algorithms, respectively. Figure 8 The figure shows the experimental results of Lamb wave damage monitoring signal denoising under 0 dB noise level, where the signal diagram of serial number (a) represents the noisy signal, and the signal diagrams of serial numbers (b), (c), (d), (e), and (f) represent the denoised signal diagrams obtained after denoising the noisy signal using EMD, SVD, wavelet transform, MP, and particle swarm-MP algorithms, respectively.

[0085] In order to compare the signal denoising effect of the embodiment, the signal to noise ratio (SNR), root mean square error (RMSE) and running time are used as evaluation indicators of the algorithm denoising performance. The larger the SNR and the smaller the RMSE, the better the denoising effect. The SNR and RMSE expressions are as follows:

[0086]

[0087] Where f(t) is the original signal without noise, is the signal after denoising, and N is the number of sampling points.

[0088] As shown in Table 1, this is the experimental average result obtained by conducting 20 denoising experiments. It can be seen that the EMD, SVD and wavelet transform algorithms run very fast and can be completed within milliseconds, but from the SNR and RMSE results, it can be seen that the denoising effects of these three denoising algorithms are poor and limited. In comparison, the MP method and the present invention have better denoising performance at different noise levels, especially at the 0dB level, they can still maintain a good denoising effect. Since all atoms need to be generated, the running time is longer. Compared with the MP method, the method of the present invention has a better denoising effect and greatly reduces the running time, effectively improving the matching pursuit denoising efficiency of Lamb wave signals.

[0089] Table 1 Average results of Lamb wave damage monitoring signal denoising experiments at different noise levels

[0090]

[0091]

[0092] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should be aware that the present disclosure is not limited by the order of the actions described, because according to the present disclosure, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present disclosure.

[0093] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0094] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A Lamb wave signal matching pursuit denoising method based on particle swarm optimization, characterized in that: The following steps are involved: Collecting Lamb wave signals of the structure, determining an atomic model according to the Lamb wave signals, and calculating a particle fitness value according to the Lamb wave signals and the atomic model; Constructing an objective function based on the fitness value of the particles, iteratively updating the particle swarm parameters according to the objective function, and obtaining optimized parameters; The optimized parameters are substituted into the atomic model, the Lamb wave signal is sparsely decomposed to obtain a decomposition result, and the decomposition result is denoised to obtain a denoised Lamb wave signal.

2. The Lamb wave signal matching pursuit denoising method based on particle swarm optimization according to claim 1 is characterized in that: The process of collecting the Lamb wave signal of the structure, determining the atomic model according to the Lamb wave signal, and calculating the particle fitness value according to the Lamb wave signal and the atomic model is as follows: Acquire Lamb wave excitation signals of the structure, define amplitude and phase as a parameter group, calculate different atoms according to the excitation signals and adopt amplitude scaling and phase shifting methods to construct an atomic model; The Lamb wave time domain signal is expanded on the atom, the noise and the effective part in the Lamb wave time domain signal are calculated, the inner product value of the atom model and the Lamb wave time domain signal is obtained, and the inner product value is used as the fitness value of the particle swarm algorithm.

3. The particle swarm optimization-based Lamb wave signal matching pursuit denoising method according to claim 1, characterized in that: The atomic model is specifically: g i (t)=A i u i (t+iΔt) Among them, g i (t) is the i-th atom; A i is the amplitude expansion factor, Δt is the phase shift factor, u i is the excitation signal, i is the atomic number, and t is the time.

4. The particle swarm optimization-based Lamb wave signal matching pursuit denoising method according to claim 3, characterized in that: The Lamb wave time domain signal is expanded on the atom, the noise and the effective part in the Lamb wave time domain signal are calculated, and the inner product value of the atomic model and the Lamb wave time domain signal is obtained. The inner product value is used as the fitness value of the particle swarm algorithm. Specifically, Where k is the number of iterations, ranging from 0<k≤K; i represents the atomic number; is the inner product value of the ith atom and the signal in the kth iteration, and f(t) is the Lamb wave signal in the time domain.

5. The particle swarm optimization-based Lamb wave signal matching pursuit denoising method according to claim 1, characterized in that: The process of constructing an objective function based on the particle fitness value and iteratively updating the particle swarm parameters according to the objective function to obtain the optimized parameters is as follows: Initialize the particle swarm parameters and calculate the fitness value corresponding to each particle according to the objective function; The particle swarm parameters are iteratively optimized according to the fitness value to obtain optimized parameters.

6. The particle swarm optimization-based matching pursuit denoising method for Lamb wave signals according to claim 5, characterized in that: The process of initializing the particle swarm parameters and calculating the fitness value corresponding to each particle according to the objective function is as follows: Initialize the particle swarm size, number of evolutions, initial position and speed, and calculate the fitness value corresponding to each particle; The individual historical optimal value is obtained according to the fitness value corresponding to each particle, and the group historical optimal value is obtained according to the objective function.

7. The particle swarm optimization-based matching pursuit denoising method for Lamb wave signals according to claim 6, characterized in that: The process of iteratively optimizing the particle swarm parameters according to the fitness value to obtain the optimized parameters is: A self-learning factor, a social learning factor and an inertia weight factor are defined, and particle velocity and position parameters are calculated according to the individual historical optimal value and the group historical optimal value; The calculation process of the particle speed and position parameters is continuously iterated, and when the maximum number of iterations is reached, the optimal particle speed and position parameters are obtained.

8. The particle swarm optimization-based Lamb wave signal matching pursuit denoising method according to claim 7, characterized in that: The process of substituting the optimization parameters into the atomic model, performing sparse decomposition on the Lamb wave signal to obtain a decomposition result, and performing denoising on the decomposition result to obtain a denoised Lamb wave signal is as follows: Substituting the optimal speed and position of the particle into the atomic model, sparsely decomposing the Lamb wave signal to obtain an effective signal and a residual signal; The residual signal is used as the original signal, and the sparse decomposition process is repeated. When the maximum number of sparse decompositions is reached, a final effective signal and a final residual signal are output; The difference between the Lamb wave signal and the final residual signal is calculated to obtain a denoised Lamb wave signal.

9. A Lamb wave signal matching pursuit denoising system based on particle swarm optimization, used to implement the Lamb wave signal matching pursuit denoising method based on particle swarm optimization as claimed in any one of claims 1 to 8, characterized in that: include: A particle fitness value calculation module (1), an optimization parameter calculation module (2) and a Lamb wave signal denoising module (3); The particle fitness value calculation module (1) is used to collect Lamb wave signals of the structure, determine the atomic model according to the Lamb wave signals, and calculate the particle fitness value according to the Lamb wave signals and the atomic model; The optimization parameter calculation module (2) is used to construct an objective function based on the particle fitness value, and iteratively update the particle swarm parameters according to the objective function to obtain the particle swarm optimization parameters; The Lamb wave signal denoising module (3) is used to substitute the particle swarm optimization parameters into the atomic model, perform sparse decomposition on the Lamb wave signal to obtain a decomposition result, and perform denoising on the decomposition result to obtain a denoised Lamb wave signal.

Images