Acoustic emission signal adaptive spectral line enhancement method based on Hurst index

Abstract

The invention discloses an acoustic emission signal adaptive spectral line enhancement method based on a Hurst index, and belongs to the field of nondestructive testing. A Hurst index is introduced into an adaptive spectral line intensifier, iterative updating of an adaptive weight is realized by utilizing Hurst index minimization, an optimal parameter combination of a filter is optimized by utilizing a multi-parameter fuzzy evaluation model and a genetic algorithm, and the adaptive spectral line intensifier after parameter optimization is applied to typical strong noise acoustic emission signal processing. According to the method, the Hurst index is introduced into the self-adaptive spectral line intensifier, the convergence speed of the filter can be increased, meanwhile, multi-parameter fuzzy evaluation and a genetic algorithm are used for optimizing filter parameter combination, and the filtering effect of the self-adaptive spectral line intensifier is improved.

Description

Technical Field

[0001] The invention relates to an improved adaptive line spectrum enhancement method based on Hurst exponent. The method is suitable for extracting acoustic emission signals under strong noise conditions and belongs to the field of non-destructive testing. Background Art

[0002] Metal plates will be deformed or broken under the action of external forces, and these processes are often accompanied by the generation of acoustic emission signals. Various signal analysis methods have been used to try to extract acoustic emission signals, but when a large amount of noise is mixed in the acoustic emission signals, the extraction of signals will be greatly affected.

[0003] Domestic and foreign scholars have tried to apply a variety of signal analysis methods to acoustic emission signal extraction, such as wavelet transform, empirical mode decomposition, etc. Traditional non-stationary signal processing methods can effectively suppress non-structural noise, but the suppression effect on structural noise (structural vibration, factory noise, etc.) is not ideal. Sparse representation can achieve signal extraction in complex environments through sparse decomposition and sparse reconstruction, but the extraction effect is often poor in the case of strong system noise. Therefore, signal extraction methods based on specific features stand out. Dan et al. [1] proposed a signal extraction method based on Tsallis synchronous squeezing wavelet entropy, which uses Tsallis entropy to more accurately quantify the local changes of acoustic emission wavelet coefficients, improves the signal-to-noise ratio of the signal, and makes it easier to effectively detect and locate cracks. Xin et al. [2] The discrete wavelet transform is applied to the simulated contaminated rail crack signal, and then the rail crack acoustic emission signal is extracted by using the changing relationship between the wavelet coefficient and the Shannon entropy to realize the positioning of the rail crack acoustic emission signal.

[0004] Aiming at the extraction of acoustic emission signals in a strong noise environment, the present invention proposes an adaptive spectral line enhancement method for acoustic emission signals based on the Hurst exponent. The present invention introduces the Hurst exponent into the cost function of the adaptive filter, and uses the Hurst exponent and Lagrange multiplier minimization to realize the iterative update of the adaptive weights, so as to achieve the purpose of rapid adaptive convergence and signal enhancement. Summary of the invention

[0005] The structure of the adaptive spectral line enhancement method of acoustic emission signal based on Hurst exponent proposed by the present invention is as follows: Figure 1 As shown, the basic principle is:

[0006] 1) Carry out fractal feature analysis of acoustic emission noise signals. Calculate the mean and standard deviation of the noise signal at different scale increments to determine whether the noise signal conforms to fractal Brownian motion, that is, the increment Y of the noise signal m is a zero-mean Gaussian signal, and Ym The Gaussian statistical characteristics of are basically consistent with the normal distribution.

[0007] If the noise signal conforms to Brownian motion, calculate its Hurst exponent.

[0008]

[0009] Among them, α1 is the time shift 1 of the acoustic emission signal, and α2 is the time shift 2 of the acoustic emission signal.

[0010] 2) Propose an adaptive spectral line enhancement method based on Hurst exponent, such as Figure 1 As shown in the figure, the Hurst exponent and Lagrange multiplier are introduced into the cost function of the adaptive filter. By delaying the input data samples of the adaptive line spectrum enhancer, the broadband noise components in the data samples before and after the delay are made uncorrelated, while the line spectrum components remain correlated. The adaptive iteration of the weights is performed by minimizing the mean square error, Hurst exponent and Lagrange multiplier, so as to achieve the purpose of fast adaptive convergence and signal enhancement.

[0011] Constructed cost function Q w for:

[0012] Q w =E{e 2 (n)}+λH (2)

[0013] Among them, λ is the Lagrange multiplier and H is the Hurst index.

[0014] Using stochastic gradient descent method to w According to the solution, we get:

[0015]

[0016] Among them, i1 is the acoustic emission signal offset 1, and i2 is the acoustic emission signal offset 2.

[0017] The final recursive equation of the cost function is:

[0018]

[0019] Wherein, μ is the iteration step size, and i1 and i2 are the same as those in formula (3).

[0020] 3) The relative error of the time-reversed acoustic emission signal focusing distance and the Hurst index are used as characteristic parameters for evaluating the noise reduction effect of the adaptive spectral line enhancement technology.

[0021] The membership function μ of the small trapezoidal distribution of the relative error of the clustering distance of the time-reversal acoustic emission signal T δ (xδ,j ) is expressed as:

[0022]

[0023] Among them, x δ,j is the relative error index value of the convergence moment of the jth group of orthogonal tests (1≤j≤25).

[0024] Membership function μ of the small normal distribution with Hurst index H T δ (x δ,j ) is expressed as:

[0025]

[0026] Among them, x H,j is the Hurst index value of the jth group of orthogonal experiments (1≤j≤25).

[0027] 4) Establishment of comprehensive fuzzy evaluation model of relative error of clustering distance and Hurst index.

[0028] In order to comprehensively consider the influence of filter order L, Lagrange multiplier λ and two time offsets on the extraction effect of acoustic emission signals, an orthogonal experimental design was first adopted, and then the fuzzy theory was used to weight and quantify the two indicators to establish a comprehensive fuzzy signal extraction evaluation parameter model containing four characteristic indicators. Figure 2 A flow chart of the method is given.

[0029] According to the membership function value results of the evaluation index of the orthogonal test, the membership function values ​​of each evaluation index are averaged and weighted, and the comprehensive fuzzy evaluation model containing two evaluation indexes is obtained as follows:

[0030] f(x)=ω1·μ T δ (x δ,j )+ω2·μ N H (x H,j ) (7)

[0031] Among them, μ T δ (x δ,j ) is the small trapezoidal distribution membership function of the relative error δ of the clustering distance, μ N δ (x δ,j ) is the membership function of the small normal distribution with the Hurst index H.

[0032] 5) Through regression analysis, the influence of the four filter parameters on its filtering effect is obtained, and the optimal parameter combination of the adaptive line enhancer is optimized by genetic algorithm. In the genetic algorithm optimization, the fitness function is formula (7).

[0033] The present invention has the following advantages: (1) The Hurst exponent is introduced into the adaptive spectral line enhancer to accelerate the convergence speed of the filter. (2) The filter parameter combination is optimized by using multi-parameter fuzzy evaluation and genetic algorithm to improve the filtering effect of the adaptive spectral line enhancer of acoustic emission signals based on the Hurst exponent. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 This is the structure diagram of the adaptive line enhancer based on the Hurst index.

[0035] Figure 2 A process is established to evaluate the parameter model for comprehensive fuzzy signal extraction effect.

[0036] Figure 3 Fiber Bragg grating acoustic emission detection system.

[0037] Figure 4 is the reflection spectrum of the fiber Bragg grating sensor.

[0038] Figure 5 Schematic diagram of the large-incident-angle acoustic emission positioning experiment.

[0039] Figure 6 The leads S1 and S2 receive signals when A1 is disconnected.

[0040] Figure 7 This is the time reversal result of the A1-S1 signal.

[0041] Figure 8 is the binarization result of the amplitude ratio difference.

[0042] Fig. 9 This is the incident angle acoustic emission positioning image result and local magnified image. DETAILED DESCRIPTION

[0043] The present invention will be further described below in conjunction with specific experiments:

[0044] Combination Figures 1 to 9 , using the fiber Bragg grating acoustic emission detection system, taking the acoustic emission source positioning of the broken lead signal on the board with large incident angle as an example, the improved adaptive line spectrum enhancement method based on Hurst exponent is explained in detail, which is suitable for the extraction of acoustic emission signals under strong noise conditions.

[0045] The specific implementation steps are given below.

[0046] 1. According to Figure 3The system block diagram shown in the figure is used to build a fiber Bragg grating acoustic emission detection system, in which the fiber Bragg grating is attached to the aluminum plate by remote bonding, and the remote bonding method is to attach the optical fiber near the grating area to the aluminum plate. The instruments and devices used in the detection system are shown in Table 1.

[0047] Table 1 Instruments and devices used in the detection system

[0048]

[0049] The reflection spectrum of the fiber Bragg grating sensor used is as follows: Figure 1 As shown, the central wavelengths of the two fiber grating sensors are 1539.36 nm and 1539.67 nm respectively, and the -3dB bandwidths are both 0.7 nm.

[0050] 2. To prove that the improved adaptive line spectrum enhancement method based on Hurst exponent is suitable for extracting acoustic emission signals under strong noise conditions, a large incident angle lead breaking experiment was carried out on the board. The experimental schematic diagram is shown in Figure 5 As shown, the received signal diagram is as follows Figure 6 The experimental configuration is shown in Table 2.

[0051] Table 2 Location information of large incident angle acoustic emission positioning experiment

[0052]

[0053] 3. Use the orthogonal experimental method to analyze the influence of the four characteristic parameters of the filter (filter order, Lagrange multiplier and two time offsets) on the filtering effect of the adaptive spectral line enhancer.

[0054] The orthogonal experiment was designed using the factor levels in Table 3. Table 4 gives the specific values ​​of the four influencing factors for different experimental combinations.

[0055] Table 3 Different factor levels in orthogonal experiment

[0056]

[0057]

[0058] Table 4 Combinations of different factors in orthogonal test

[0059]

[0060] Based on fuzzy theory, a fuzzy evaluation model including two characteristic indicators, the relative error of the time-reversal acoustic emission signal convergence distance and the Hurst index, was established. The comprehensive fuzzy evaluation model including the two evaluation indicators is:

[0061]

[0062] The comprehensive evaluation index values ​​of each group of orthogonal experiments are shown in Table 5.

[0063] Table 5 Comprehensive evaluation index values ​​of each group of orthogonal test

[0064]

[0065] 4. In order to achieve the best signal extraction effect, the genetic algorithm is used to optimize the four parameters proposed previously. Through regression analysis, the influence of the four filter parameters on its filtering effect is obtained, and the genetic algorithm is used to optimize the best parameter combination of the adaptive line enhancer. The feasible domain of the parameters to be optimized is shown in Table 6, the relevant parameter settings of the genetic algorithm are shown in Table 7, and the global optimal parameter combination obtained after the genetic algorithm parameter optimization is shown in Table 8.

[0066] Table 6 Genetic algorithm optimization parameters

[0067]

[0068] Table 7 Genetic algorithm parameter configuration

[0069]

[0070]

[0071] Table 8 Best parameter combination after optimization

[0072]

[0073] 5. The proposed adaptive spectral line enhancer based on Hurst index and the optimal parameter combination are used to process and extract the typical strong noise acoustic emission signal, and the extracted signal is applied to the acoustic emission signal time reversal positioning method based on the directional characteristics of fiber Bragg grating. The time reversal result of A1-S1 signal is shown in Figure 7 shown. Figure 7 The time-reversed wave field shown in a) is extracted by extracting the peak value of the reversal signal corresponding to each reversal distance in the time-reversed wave field to obtain Figure 7 b) The result is, Figure 7 The horizontal coordinate corresponding to the highest point of the curve in b) is the time reversal convergence distance. The reversal result corresponding to this distance is as follows: Figure 7 As shown in c), Figure 7 The horizontal coordinate corresponding to the peak value in c) is the time reversal convergence moment. The spatial distribution of the amplitude at the time reversal convergence moment is extracted as follows: Figure 7 As shown in d).

[0074] 6. The spatial distribution of the amplitude at the moment of signal convergence is superimposed in the positioning area, and the difference of the directional characteristic amplitude ratio is calculated and binarized to obtain Figure 8 Results shown.

[0075] 7. Figure 8 The result in is multiplied by the acoustic emission positioning result based on the directional characteristics to obtain Fig. 9 The image result is shown. Fig. 9 It can be seen that after the acoustic emission signal is extracted using the improved adaptive line spectrum enhancement method based on the Hurst exponent, not only does the highlight area after amplitude superposition appear at the correct position, but the directional characteristics of the vertical position also make the highlight area in the image result more concentrated.

[0076] 8. The point with the highest amplitude in the image is taken as the positioning result. The positioning results of the three acoustic emission sources are shown in Table 9. The distance ΔL between the positioning coordinates and the actual coordinates is taken as the positioning error.

[0077] Table 9 Statistics of acoustic emission positioning results at large incident angles

[0078]

[0079] From the positioning results, it can be seen that the large-incident-angle acoustic emission signal processed by the improved adaptive line spectrum enhancement method based on the Hurst exponent can not only complete the acoustic emission positioning, but also the distance error of the positioning result is within 15 mm, which proves that the improved adaptive line spectrum enhancement method based on the Hurst exponent is suitable for the extraction of acoustic emission signals under strong noise conditions.

[0080] The above is a typical application of the present invention, and the application of the present invention is not limited thereto.

[0081] References

[0082] [1]Li D,Kuang K,Chan G K.Rail crack monitoring based on Tsallissynchrosqueezed wavelet entropy of acoustic emission signals:A field study[J].

[0083] Structural health monitoring,2018,17(6):1410-1424.

[0084] [2]Xin Z,Cui Y,Yan W,et al.An improved AE detection method of raildefect based on multi-level ANC with VSS-LMS[J].Mechanical Systems&SignalProcessing,2018,99(15):420-433.

Claims

1. A method for adaptive spectral line enhancement of acoustic emission signals based on Hurst exponent, characterized in that: The following steps are involved: 1) Analyze the noise fractal characteristics of acoustic emission signals; Calculate the mean and standard deviation of the noise signal at different scale increments to determine whether it conforms to fractal Brownian motion, that is, the increment Y of the noise signal m is a zero-mean Gaussian signal, and Y m The Gaussian statistical characteristics of are basically consistent with the normal distribution; If the noise signal conforms to Brownian motion, calculate its Hurst exponent Among them, α1 is the time shift 1 of the acoustic emission signal, and α2 is the time shift 2 of the acoustic emission signal; 2) The Hurst exponent is introduced into the cost function of the adaptive filter, and an adaptive spectral line enhancement method based on the Hurst exponent is proposed; the constructed cost function Q w for: Q w =E{e 2 (n)}+λH (2) Among them, λ is the Lagrange multiplier, H is the Hurst index of the noise signal of the acoustic emission signal with a large incident angle; The cost function is solved using the stochastic gradient descent method: Wherein, λ is the Lagrange multiplier, i1 is the acoustic emission signal offset 1, and i2 is the acoustic emission signal offset 2; The final cost function recursive equation is: Wherein, μ is the iteration step, i1 is the acoustic emission signal offset 1, and i2 is the acoustic emission signal offset 2; The main part of the adaptive line enhancer based on Hurst index includes a delay K and an adaptive finite impulse response FIR filter. By delaying the input data samples of the adaptive line enhancer in time, the broadband noise components in the data samples before and after the delay are made uncorrelated, while the line spectrum components remain correlated. The adaptive iteration of weights is performed by minimizing the mean square error, Hurst index and Lagrange multiplier, so as to achieve the purpose of rapid adaptive convergence and signal enhancement. 3) The orthogonal test method is used to analyze the influence of the four characteristic parameters of the filter, namely the filter order, Lagrange multiplier and two time offsets, on the filtering effect of the adaptive spectral line enhancer; based on fuzzy theory, a fuzzy evaluation model including two characteristic indicators, the relative error of the convergence distance of the time-reversed acoustic emission signal and the Hurst index, is established; 4) Through regression analysis, the influence of the four filter parameters on its filtering effect is obtained, and the best parameter combination of the adaptive spectral line enhancer is optimized by using genetic algorithm; 5) The proposed adaptive line enhancer based on Hurst exponent and the optimal parameter combination are used to process typical strong noise acoustic emission signals and evaluate the filtering effect.

2. The method for adaptive spectral line enhancement of acoustic emission signals based on Hurst exponent according to claim 1, characterized in that: The Hurst exponent is introduced into the adaptive spectral line enhancer, and the iterative update of the adaptive weights is realized by minimizing the Hurst exponent, so as to achieve the purpose of fast adaptive convergence and signal enhancement.

3. The method for adaptive spectral line enhancement of acoustic emission signals based on Hurst exponent according to claim 1, characterized in that: The best parameter combination of the filter is selected by using a multi-parameter fuzzy evaluation model and a genetic algorithm. The adaptive line enhancer with optimized parameters is applied to the processing of typical strong-noise acoustic emission signals.

4. The method for adaptive spectral line enhancement of acoustic emission signals based on Hurst exponent according to claim 1, characterized in that: The Hurst exponent is introduced into the adaptive spectral line enhancer to accelerate the convergence speed of the filter. At the same time, multi-parameter fuzzy evaluation and genetic algorithm are used to optimize the filter parameter combination to improve its filtering effect.

Images