Method for evaluating concrete performance degradation based on acoustic emission signal processing

Abstract

The present invention relates to a method for evaluating the performance degradation of concrete based on acoustic emission signal processing. The technical solution of the present invention is a method for evaluating the performance degradation of concrete based on acoustic emission signal processing, characterized in that: acquiring acoustic emission signal samples generated during the flexural tension loading of concrete specimens, and extracting the characteristic quantities of the acoustic emission signal samples; based on the characteristic quantity threshold of the normal state of the specimens, collecting samples in the normal state and the abnormal state, and calculating the Mahalanobis distance between the normal state samples and the abnormal state samples; calculating the mean signal-to-noise ratio based on the Mahalanobis distance of the abnormal state samples, and judging whether a certain characteristic quantity is effective according to the change of the mean signal-to-noise ratio when a certain characteristic quantity is not included; representing the samples with effective characteristic quantities, denoising the samples, and successively adding the Mahalanobis distance deviations of the samples above the threshold over time to obtain a concrete degradation index. The present invention is applicable to the field of concrete performance evaluation.

Description

Technical Field

[0001] The present invention relates to a method for evaluating the performance degradation of concrete based on acoustic emission signal processing, which is applicable to the field of concrete performance evaluation. Background Art

[0002] In the process of detecting concrete cracks by existing acoustic emission technologies, it is difficult to accurately select the parameters most sensitive to crack propagation from the characteristic quantities of ultrasonic waves for analysis, and it is difficult to determine the initial state of cracks in the identification process of acoustic emission characteristic quantities. Problems such as difficulty in identifying early cracks and inability to predict the performance degradation of concrete during the entire life cycle will occur. Summary of the Invention

[0003] The technical problem to be solved by the present invention is: in view of the above problems, to provide a method for evaluating the performance degradation of concrete based on acoustic emission signal processing.

[0004] The technical solution adopted by the present invention is: a method for evaluating the performance degradation of concrete based on acoustic emission signal processing, characterized in that:

[0005] Obtain the acoustic emission signal samples generated during the flexural-tensile loading of concrete specimens, and extract the characteristic quantities of the acoustic emission signal samples;

[0006] Based on the characteristic quantity threshold of the normal state of the specimen, collect samples in the normal state and the abnormal state, and calculate the Mahalanobis distance between the normal state samples and the abnormal state samples;

[0007] Calculate the mean signal-to-noise ratio based on the Mahalanobis distance of the abnormal state samples, and judge whether a certain characteristic quantity is effective according to the change of the mean signal-to-noise ratio when a certain characteristic quantity is not included;

[0008] Represent the samples with effective characteristic quantities, denoise the samples, and add up the Mahalanobis distance deviations of the samples above the threshold over time in sequence to obtain the concrete degradation index.

[0009] The extraction of the characteristic quantities of the acoustic emission signal samples includes extracting the peak amplitude, rise time, decay time, acoustic emission count, and acoustic emission energy of the acoustic emission signal.

[0010] The calculation of the Mahalanobis distance of the normal state samples includes:

[0011]

[0012] Where N i =[n i1 ,...,n ip T , n ij ​is the normalized value of the j-th feature quantity of the i-th sample, p is the number of feature quantities, and C is the covariance matrix of the samples.

[0013] The said n ij is the normalized value of the j-th feature quantity of the i-th sample, including:

[0014]

[0015] where f ij represents the value of the j-th feature quantity of the i-th sample, is the average value of the j-th feature quantity, and σ j is the standard deviation of the j-th feature quantity.

[0016] The said calculation of the Mahalanobis distance of the abnormal state samples includes:

[0017] The feature quantities in the abnormal state samples are standardized using the mean and standard deviation of the corresponding feature quantities in the normal state samples, and the Mahalanobis distance of the abnormal state samples is calculated using the covariance matrix corresponding to the normal state.

[0018] The said calculation of the mean signal-to-noise ratio based on the Mahalanobis distance of the abnormal state samples includes:

[0019]

[0020] where SNR is the mean signal-to-noise ratio, MD t is the Mahalanobis distance of the t-th abnormal state sample, and m is the number of abnormal state samples.

[0021] The said judgment of whether a certain feature quantity is effective according to the change of the mean signal-to-noise ratio when a certain feature quantity is not included includes:

[0022] Δ = SNR + -SNR -

[0023] where SNR + is the mean signal-to-noise ratio calculated based on all feature quantities; SNR - is the mean signal-to-noise ratio calculated when a certain feature quantity is not included;

[0024] When Δ > 0, it means that the certain feature quantity is effective; when Δ ≤ 0, it means that the certain feature quantity is ineffective.

[0025] The said representation of samples using effective feature quantities, denoising of samples, and sequential addition of the Mahalanobis distance deviation of samples above the threshold over time to obtain the concrete degradation index includes:

[0026] DI(n) = DI(n - 1) + DI t (n)

[0027] Among them, DI(n) represents the concrete degradation index corresponding to the nth sample; DI t (n) represents the Mahalanobis distance deviation of the nth sample;

[0028] When the deviation value of the Mahalanobis distance of the sample from the average value μ of the Mahalanobis distance of the samples in the normal state i is the threshold value ε i or the threshold value ε i or less, the Mahalanobis distance deviation of this sample is set to 0;

[0029] When the deviation value of the Mahalanobis distance of the sample from the average value μ of the Mahalanobis distance of the samples in the normal state i is above the threshold value ε i or more, the Mahalanobis distance deviation of this sample is set to

[0030] DI t (n) = MD(n) - (μ i + ε i )

[0031] Among them, MD(n) is the Mahalanobis distance of the nth sample, and ε i is the deviation value from the average value μ of the Mahalanobis distance of the samples in the normal state when changing from the normal state to the abnormal state, and μ i is the average value of the Mahalanobis distance of the samples in the normal state. i The normal state is the state where the concrete is not cracked; the abnormal state is the state where the concrete is cracked.

[0032] A concrete performance degradation evaluation device based on acoustic emission signal processing, characterized by comprising:

[0033] A signal acquisition module, configured to acquire acoustic emission signal samples generated during the flexural tension loading of a concrete specimen, and extract the characteristic quantities of the acoustic emission signal samples;

[0034] A sample collection module, configured to collect samples in the normal state and the abnormal state based on the characteristic quantity threshold of the specimen in the normal state, and calculate the Mahalanobis distance of the samples in the normal state and the abnormal state;

[0035] An effective judgment module, configured to calculate the mean signal-to-noise ratio based on the Mahalanobis distance of the samples in the abnormal state, and judge whether a certain characteristic quantity is effective according to the change of the mean signal-to-noise ratio when a certain characteristic quantity is not included;

[0036] An index calculation module, configured to represent the samples with effective characteristic quantities, denoise the samples, and sequentially add the Mahalanobis distance deviations of the samples above the threshold value over time to obtain the concrete degradation index.

[0037]

[0038] ​A storage medium stores a computer program executable by a processor, characterized in that: when the computer program is executed, the steps of the concrete performance degradation evaluation method based on acoustic emission signal processing are realized.

[0039] A computer device has a memory and a processor, and the memory stores a computer program executable by the processor, characterized in that: when the computer program is executed, the steps of the concrete performance degradation evaluation method based on acoustic emission signal processing are realized.

[0040] The beneficial effect of the present invention is that the present invention successively adds the sample Mahalanobis distance deviations over time to obtain a concrete degradation index, which can display the time course of concrete performance degradation and reflect the degradation development process of concrete from crack formation to failure.

[0041] The present invention establishes a reference Mahalanobis distance of characteristic quantities based on the normal state of concrete, so as to accurately judge the occurrence of early cracks. The present invention screens effective characteristic quantities according to the signal-to-noise ratio and eliminates ineffective characteristic quantities, which can reduce subsequent operations and improve the accuracy of the degradation index. Description of the Drawings

[0042] Figure 1 It is a flowchart of the embodiment. Detailed Embodiment

[0043] This embodiment is a concrete performance degradation evaluation method based on acoustic emission signal processing, and the specific steps are as follows:

[0044] S1. Obtain an acoustic emission signal sample generated when a concrete specimen is subjected to flexural tension loading, and extract the characteristic quantities of the acoustic emission signal sample. The characteristic quantities of the acoustic emission signal include peak amplitude, rise time, decay time, acoustic emission count, and acoustic emission energy.

[0045] S2. Based on the characteristic quantity threshold of the normal state of the specimen (selecting the maximum value of the characteristic quantity of the specimen in the unloaded state), collect samples in the normal state and the abnormal state, and calculate the Mahalanobis distance between the normal state sample and the abnormal state sample, where the normal state is the uncracked state of the concrete, and the abnormal state is the cracked state of the concrete.

[0046] In this example, the calculation method of the Mahalanobis distance of the normal state sample includes:

[0047]

[0048] Among them, N i =[n i1 ,...,n ip T ,n ij ​is the normalized value of the j-th feature quantity of the i-th sample, p is the number of feature quantities, and C is the covariance matrix of the samples.

[0049] The calculation method of the normalized value of the j-th feature quantity of the i-th sample in this embodiment includes:

[0050]

[0051] where f ij represents the value of the j-th feature quantity of the i-th sample, is the average value of the j-th feature quantity, and σ j is the standard deviation of the j-th feature quantity.

[0052] The calculation method of the Mahalanobis distance of the abnormal state samples in this embodiment includes: using the Mahalanobis distance calculation formula of the normal state samples; performing standardization using the mean and standard deviation of the corresponding features in the normal state; calculating the Mahalanobis distance of the abnormal state samples using the covariance matrix corresponding to the normal state samples.

[0053] S3. Calculate the mean value of the signal-to-noise ratio based on the Mahalanobis distance of the abnormal state samples, and judge whether a certain feature quantity is effective according to the change of the mean value of the signal-to-noise ratio when a certain feature quantity is not included.

[0054] The calculation method of the mean value of the signal-to-noise ratio in this embodiment includes:

[0055]

[0056] where SNR is the mean value of the signal-to-noise ratio, MD t is the Mahalanobis distance of the t-th abnormal state sample, and m is the number of abnormal state samples.

[0057] In this example, the change Δ of the mean value of the signal-to-noise ratio when a certain feature quantity is not included is calculated by the following method:

[0058] Δ = SNR + -SNR -

[0059] where SNR + is the mean value of the signal-to-noise ratio calculated based on all feature quantities; SNR - is the mean value of the signal-to-noise ratio calculated when a certain feature quantity is not included.

[0060] In this embodiment, when Δ > 0, it means that a certain feature quantity is effective; when Δ ≤ 0, it means that a certain feature quantity is invalid.

[0061] S4. Represent the samples using the effective feature quantities, denoise the samples, and add up the deviations of the Mahalanobis distances of the samples above the threshold over time to obtain the concrete degradation index.

[0062] The denoising method uses Chebyshev's inequality P(MD i -u i >ε)≤σ i / ε 2 , where MD i is the Mahalanobis distance calculated from the effective feature quantities of the normal state samples, μ i and σ i are the mean and variance of the Mahalanobis distances of the normal state samples respectively, and P represents the probability that the Mahalanobis distance of the sample deviates from the average value μ i of the Mahalanobis distances of the normal state samples by ε.

[0063] In this embodiment, the denoising method based on Chebyshev's inequality is used to process the Mahalanobis distance of the acoustic emission signal, which can make the degradation curve smooth and monotonic, and better reflect the severity of component degradation.

[0064] In this example, the deviation value from the average value μ i of the Mahalanobis distances of the normal state samples when changing from the normal state to the abnormal state is defined as ε i , that is, the concrete cracking threshold is ε i , and Chebyshev's inequality can calculate the denoising accuracy of this threshold.

[0065] In this embodiment, the calculation method of the concrete degradation index DI includes:

[0066] DI(n) = DI(n - 1) + DI t (n)

[0067] where DI(n) represents the concrete degradation index corresponding to the nth sample collected sequentially over time; DI t (n) represents the Mahalanobis distance deviation of the nth sample.

[0068] When the deviation value of the Mahalanobis distance of the sample from the average value μ i of the Mahalanobis distances of the normal state samples is the threshold ε i or below the threshold ε i , let the Mahalanobis distance deviation DI t of this sample be 0; when the deviation value of the Mahalanobis distance of the sample from the average value μ i is above the threshold ε i , let the Mahalanobis distance deviation of this sample be

[0069] DI t (n) = MD(n) - (μ i +ε i )

[0070] where MD(n) is the Mahalanobis distance of the nth sample, and ε iThe deviation value from the average value μ of the Mahalanobis distance of the normal state samples when changing from the normal state to the abnormal state. i of the deviation.

[0071] This embodiment also provides a device for evaluating the performance degradation of concrete based on acoustic emission signal processing, including: a signal acquisition module, a sample collection module, an effective judgment module, and an index calculation module.

[0072] In this example, the signal acquisition module is used to acquire the acoustic emission signal samples generated when the concrete specimen is subjected to flexural tension loading, and extract the characteristic quantities of the acoustic emission signal samples; the sample collection module is used to collect the samples in the normal state and the abnormal state based on the characteristic quantity threshold of the normal state of the specimen, and calculate the Mahalanobis distance between the normal state samples and the abnormal state samples; the effective judgment module is used to calculate the average signal-to-noise ratio based on the Mahalanobis distance of the abnormal state samples, and judge whether a certain characteristic quantity is effective according to the change of the average signal-to-noise ratio when a certain characteristic quantity is not included; the index calculation module is used to represent the samples with effective characteristic quantities, denoise the samples, and add up the Mahalanobis distance deviations of the samples above the threshold over time in sequence to obtain the concrete degradation index.

[0073] This embodiment also provides a storage medium, on which a computer program executable by a processor is stored, and when the computer program is executed, the steps of the method for evaluating the performance degradation of concrete based on acoustic emission signal processing in this example are realized.

[0074] This embodiment also provides a computer device, which has a memory and a processor, and a computer program executable by the processor is stored on the memory, and when the computer program is executed, the steps of the method for evaluating the performance degradation of concrete based on acoustic emission signal processing in this example are realized.

Claims

1. A method for evaluating the performance degradation of concrete based on acoustic emission signal processing, characterized in that: Obtain the acoustic emission signal samples generated during the flexural tensile loading of concrete specimens, and extract the characteristic quantities of the acoustic emission signal samples; Based on the characteristic quantity thresholds of the specimens in the normal state, collect samples in the normal state and the abnormal state, and calculate the Mahalanobis distances of the normal state samples and the abnormal state samples; Calculate the mean signal-to-noise ratio based on the Mahalanobis distances of the abnormal state samples, and judge whether a certain characteristic quantity is effective according to the change of the mean signal-to-noise ratio when a certain characteristic quantity is not included; Represent the samples with effective characteristic quantities, denoise the samples, and successively add up the Mahalanobis distance deviations of the samples above the threshold over time to obtain the concrete degradation index; The calculation of the Mahalanobis distance of the normal state samples includes: where N i = [n i1 ,..., n ip T , n ij is the normalized value of the j-th feature quantity of the i-th sample, p is the number of feature quantities, and C is the covariance matrix of the samples;​ The said n ij is the normalized value of the j-th feature quantity of the i-th sample, including: Among them, f ij represents the value of the j-th feature quantity of the i-th sample, is the average value of the j-th feature quantity, and σ j is the standard deviation of the j-th feature quantity; The calculation of the Mahalanobis distance of the abnormal state samples includes: The characteristic quantities in the abnormal state samples are standardized using the mean and standard deviation of the corresponding characteristic quantities in the normal state samples, and the Mahalanobis distance of the abnormal state samples is calculated using the covariance matrix corresponding to the normal state; The calculation of the mean signal-to-noise ratio based on the Mahalanobis distances of the abnormal state samples includes: where SNR is the mean signal-to-noise ratio, MD t is the Mahalanobis distance of the t-th abnormal state sample, and m is the number of abnormal state samples.

2. The method for evaluating the performance degradation of concrete based on acoustic emission signal processing according to claim 1, characterized in that: The extraction of the characteristic quantities of the acoustic emission signal samples includes extracting the peak amplitude, rise time, decay time, acoustic emission count, and acoustic emission energy of the acoustic emission signal.

3. The method for evaluating the performance degradation of concrete based on acoustic emission signal processing according to claim 1, characterized in that, The judgment of whether a certain characteristic quantity is effective according to the change of the mean signal-to-noise ratio when a certain characteristic quantity is not included includes: Δ = SNR + -SNR - Among them, SNR + is the mean signal-to-noise ratio calculated based on all feature quantities; SNR - is the mean signal-to-noise ratio calculated when a certain feature quantity is not included; When Δ>0, it means that a certain characteristic quantity is effective; when Δ≤0, it means that a certain characteristic quantity is ineffective.

4. The method for evaluating the performance degradation of concrete based on acoustic emission signal processing according to claim 1, characterized in that, The representation of the samples with effective characteristic quantities, the denoising of the samples, and the successive addition of the Mahalanobis distance deviations of the samples above the threshold over time to obtain the concrete degradation index includes: DI(n) = DI(n - 1) + DI t (n) Among them, DI(n) represents the concrete degradation index corresponding to the nth sample; DI t (n) represents the Mahalanobis distance deviation of the nth sample; When the Mahalanobis distance of a sample deviates from the average value μ of the Mahalanobis distances of normal state samples i and the deviation value is the threshold ε i or the threshold ε i is below, set the Mahalanobis distance deviation of this sample to 0; When the Mahalanobis distance of a sample deviates from the average value μ of the Mahalanobis distances of normal state samples i and the deviation value is above the threshold ε i then the Mahalanobis distance deviation of this sample is set to DI t I(n) = MD(n) - (μ i + ε i ) Among them, MD(n) is the Mahalanobis distance of the nth sample, and ε i is the deviation value from the average value μ i of the Mahalanobis distance of the normal state samples when changing from the normal state to the abnormal state, and μ i is the average value of the Mahalanobis distance of the normal state samples.

5. The method for evaluating the performance degradation of concrete based on acoustic emission signal processing according to claim 1, characterized in that: The normal state is the uncracked state of the concrete; the abnormal state is the cracked state of the concrete.

6. An apparatus for evaluating the performance degradation of concrete based on acoustic emission signal processing, characterized in that, Includes: A signal acquisition module for obtaining the acoustic emission signal samples generated during the flexural tensile loading of concrete specimens and extracting the characteristic quantities of the acoustic emission signal samples; A sample collection module for collecting samples in the normal state and the abnormal state based on the characteristic quantity thresholds of the specimens in the normal state and calculating the Mahalanobis distances of the normal state samples and the abnormal state samples; An effective judgment module for calculating the mean signal-to-noise ratio based on the Mahalanobis distances of the abnormal state samples and judging whether a certain characteristic quantity is effective according to the change of the mean signal-to-noise ratio when a certain characteristic quantity is not included; An index calculation module for representing the samples with effective characteristic quantities, denoising the samples, and successively adding up the Mahalanobis distance deviations of the samples above the threshold over time to obtain the concrete degradation index; The calculation of the Mahalanobis distance of the normal state samples includes: Among them, N i = [n i1 ,..., n ip T , where n ij is the normalized value of the j-th feature quantity of the i-th sample, p is the number of feature quantities, and C is the covariance matrix of the samples;​ The said n ij is the normalized value of the j-th feature quantity of the i-th sample, including: where, f ij represents the value of the j-th feature quantity of the i-th sample, is the average value of the j-th feature quantity, and σ j is the standard deviation of the j-th feature quantity; The calculation of the Mahalanobis distance of the abnormal state samples includes: The characteristic quantities in the abnormal state samples are standardized using the mean and standard deviation of the corresponding characteristic quantities in the normal state samples, and the Mahalanobis distance of the abnormal state samples is calculated using the covariance matrix corresponding to the normal state; The calculation of the mean signal-to-noise ratio based on the Mahalanobis distances of the abnormal state samples includes: where SNR is the mean signal-to-noise ratio, MD t is the Mahalanobis distance of the t-th abnormal state sample, and m is the number of abnormal state samples.

7. A storage medium having a computer program stored thereon that can be executed by a processor, characterized in that: When the computer program is executed, it implements the steps of the concrete performance degradation evaluation method based on acoustic emission signal processing described in any one of claims 1 to 5.

8. A computer device having a memory and a processor, with a computer program stored on the memory that can be executed by the processor, characterized in that: When the computer program is executed, it implements the steps of the method for evaluating the degradation of concrete performance based on acoustic emission signal processing according to any one of claims 1 to 5.

Images