Acoustic emission signal denoising method based on multi-method fusion

Abstract

The invention belongs to the technical field of signal processing, and particularly discloses an acoustic emission signal denoising method based on multi-method fusion, and the method comprises the steps: collecting an acoustic emission signal through an acoustic emission sensor; a band-pass frequency range is determined by adjusting the frequency range based on a plurality of evaluation indexes, and the evaluation indexes are determined by analyzing the difference between the first signal sample and the second signal sample; the first signal sample and the second signal sample are samples collected when no cavitation phenomenon exists and when the cavitation phenomenon exists respectively; wavelet threshold de-noising is carried out, and a signal after background noise is removed is obtained; through independent component analysis, a target signal and an interference signal are separated. According to the invention, by combining a plurality of denoising technologies, accurate suppression of different types of noise is realized, the denoising effect can be ensured in a complex environment, the signal-to-noise ratio of the acoustic emission signal is effectively improved, and the quality and reliability of the signal are ensured.

Description

Technical Field

[0001] The present application belongs to the field of signal processing technology, and more specifically, relates to an acoustic emission signal denoising method based on multi-method fusion. Background Art

[0002] Acoustic emission (AE) technology is an important method for structural health monitoring. It detects defects such as cracks and cavitation by detecting ultrasonic signals generated by materials or equipment during operation. However, AE signals are often interfered with by various noises during acquisition, including environmental noise, electromagnetic noise, and mechanical vibration noise. This noise degrades signal quality and affects the accuracy of fault detection. Therefore, how to effectively remove noise while retaining the useful signal is a key issue in AE signal processing.

[0003] Existing denoising methods mainly include frequency domain filtering, wavelet transform, singular value decomposition (SVD), independent component analysis (ICA), etc. However, a single method often has limitations and is difficult to guarantee denoising effects in complex environments. Summary of the Invention

[0004] In response to the defects of the existing technology, the purpose of this application is to provide an acoustic emission signal denoising method based on multi-method fusion, aiming to solve the problem that a single method in the existing technology often has limitations and is difficult to ensure denoising effect in complex environments.

[0005] To achieve the above objectives, in a first aspect, the present application provides an acoustic emission signal denoising method based on multi-method fusion, the method comprising: Acquiring acoustic emission signals through an acoustic emission sensor; The collected acoustic emission signal is filtered through a bandpass filter to obtain a filtered signal, wherein the bandpass frequency range of the bandpass filter is determined by adjusting the frequency range based on multiple evaluation indicators, and the evaluation indicators are determined by analyzing the difference between a first signal sample and a second signal sample, wherein the first signal sample is an acoustic emission signal sample collected when there is no cavitation phenomenon, and the second signal sample is an acoustic emission signal sample collected when there is cavitation phenomenon; Based on the filtered signal, wavelet threshold denoising is performed to obtain the signal after removing the background noise; Based on the signal after removing the background noise, the target signal and the interference signal are separated through independent component analysis.

[0006] It can be understood that band-pass filtering can remove low-frequency and high-frequency irrelevant signals. Since the components of most frequency bands in the mixed signal are suppressed or removed, and the energy of the removed components accounts for a large proportion, the crack signal is enhanced and the signal-to-noise ratio is improved; then the wavelet threshold method can be used to remove background random noise. In the wavelet domain, the coefficient corresponding to the effective signal is large, while the coefficient corresponding to the noise is small. The coefficient corresponding to the noise in the wavelet domain still satisfies the Gaussian white noise distribution, so by processing the coefficients, the denoising effect can be achieved; then independent component analysis (ICA) is used to perform signal blind source separation (separating the target signal from the interference signal). The ICA method is not affected by the aliasing of frequency bands between source signals, nor is it affected by the interference of different source signal strengths.

[0007] Therefore, this application combines multiple denoising techniques, including frequency-domain filtering, wavelet threshold denoising, and independent component analysis (ICA), to achieve precise suppression of different noise types. This approach ensures effective denoising even in complex environments, effectively improving the signal-to-noise ratio of acoustic emission signals and ensuring signal quality and reliability. This approach is suitable for health monitoring of equipment such as hydro turbines, bridge structures, and pressure vessels.

[0008] In a possible implementation, the multiple evaluation indicators include: spectrum overlap, frequency center of gravity difference, mutual information, and spectrum energy difference.

[0009] In one possible implementation, the bandpass frequency range of the bandpass filter is determined by the following steps: Based on the initial band-pass frequency range, gradually narrow the band-pass frequency range until the comprehensive evaluation index tends to be stable; Among them, the comprehensive evaluation index is determined by weighted summation based on spectrum overlap, frequency center of gravity difference, mutual information and spectrum energy difference.

[0010] In one possible implementation, the wavelet threshold denoising includes: Based on the filtered signal, discrete wavelet transform is performed to obtain low-frequency wavelet coefficients and high-frequency wavelet coefficients; Based on the wavelet threshold, the high-frequency wavelet coefficients are subjected to threshold processing to obtain the high-frequency wavelet coefficients after threshold processing; Based on the low-frequency wavelet coefficients and the high-frequency wavelet coefficients after threshold processing, the signal after removing the background noise is obtained through inverse wavelet transform.

[0011] In one possible implementation, the wavelet threshold of each decomposition layer is determined by the following formula: ; in, Indicates the The wavelet threshold of the decomposition layer, represents the root mean square error of the wavelet decomposition coefficients, Indicates the length of the wavelet decomposition coefficients.

[0012] In one possible implementation, the above-mentioned separation of the target signal and the interference signal by independent component analysis based on the signal after background noise removal includes: Based on the signal after removing the background noise, multiple independent components are obtained through independent component analysis; Determining an energy threshold based on the energy of each independent component; Based on the energy threshold, the target component is screened out from multiple independent components; Based on the filtered target components, a target signal is constructed.

[0013] In one possible implementation, the energy threshold is determined by the following formula: ; in, represents the energy threshold, is the preset coefficient, represents the energy of each independent component, Indicates the mean.

[0014] In a second aspect, the present application provides an acoustic emission signal denoising device based on multi-method fusion, comprising: An acquisition module, used for acquiring acoustic emission signals through an acoustic emission sensor; a filtering module for filtering the collected acoustic emission signal through a bandpass filter to obtain a filtered signal, wherein the bandpass frequency range of the bandpass filter is determined by adjusting the frequency range based on multiple evaluation indicators, and the evaluation indicators are determined by analyzing the difference between a first signal sample and a second signal sample, wherein the first signal sample is an acoustic emission signal sample collected when there is no cavitation phenomenon, and the second signal sample is an acoustic emission signal sample collected when there is cavitation phenomenon; A background noise removal module is used to perform wavelet threshold denoising based on the filtered signal to obtain a signal after background noise removal; The signal blind source separation module is used to separate the target signal and the interference signal through independent component analysis based on the signal after removing the background noise.

[0015] In a third aspect, the present application provides an electronic device comprising: at least one memory for storing programs; and at least one processor for executing the programs stored in the memory. When the program stored in the memory is executed, the processor is used to execute the method described in the first aspect or any possible implementation of the first aspect.

[0016] In a fourth aspect, the present application provides a computer-readable storage medium, which stores a computer program. When the computer program runs on a processor, the processor executes the method described in the first aspect or any possible implementation of the first aspect.

[0017] In general, the above technical solutions conceived by this application have the following beneficial effects compared with the existing technologies: This application combines multiple denoising techniques, including frequency-domain filtering, wavelet threshold denoising, and independent component analysis (ICA), to precisely suppress different types of noise. It maintains denoising effectiveness even in complex environments, effectively improving the signal-to-noise ratio of acoustic emission signals and ensuring signal quality and reliability. It is suitable for health monitoring of equipment such as hydro turbines, bridge structures, and pressure vessels. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 Schematic diagram of a flow chart of an acoustic emission signal denoising method based on multi-method fusion provided in an embodiment of the present application; Figure 2 Schematic diagram of the structure of an acoustic emission signal denoising device based on multi-method fusion provided in an embodiment of the present application; Figure 3 It is a structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0020] In the specification and claims of this application, the terms "first" and "second" are used to distinguish different objects, rather than to describe a specific order of objects. For example, the terms "first signal sample" and "second signal sample" are used to distinguish different signal samples, rather than to describe a specific order of signal samples.

[0021] In the embodiments of this application, words such as "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of this application should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0022] In the description of the embodiments of the present application, unless otherwise specified, "multiple" means two or more, for example, multiple processing units means two or more processing units, etc.; multiple elements means two or more elements, etc.

[0023] The embodiments of the present application are described below in conjunction with the drawings in the embodiments of the present application.

[0024] Figure 1 This is a flow chart of the acoustic emission signal denoising method based on multi-method fusion provided in the embodiment of the present application, such as Figure 1 As shown, the method includes the following steps (1) to (4).

[0025] Step (1): Use acoustic emission sensors to collect signals.

[0026] In this application, the acoustic emission sensor used has a sampling frequency of 1 MHz, so it will collect acoustic emission signals with a wide frequency band. In addition to crack signals and cavitation signals, there are also water flow impact noise, vibration noise, unit rotation noise, etc.

[0027] Step (2): Remove low-frequency and high-frequency irrelevant signals through bandpass filtering.

[0028] To filter out "clutter," frequency domain filtering and bandpass filtering are used. Signals in the frequency bands where non-crack and cavitation signals reside are removed, and "point-clearing" of irrelevant signal frequencies is performed. By suppressing or removing components in most frequency bands of the mixed signal, and since the removed components have a large energy share, the crack signal is enhanced, improving the signal-to-noise ratio.

[0029] Step (3): Use the wavelet threshold method to remove background random noise.

[0030] Filter out background noise. Since signals have a certain degree of continuity in space (or time domain), the modulus of the wavelet coefficients generated by the effective signal in the wavelet domain is often large. However, Gaussian white noise has no continuity in space (or time domain). Therefore, after the wavelet transform, the noise still appears highly random in the wavelet domain and is generally still considered Gaussian white noise. Therefore, in the wavelet domain, the coefficients corresponding to the effective signal are large, while the coefficients corresponding to the noise are small. The coefficients corresponding to the noise in the wavelet domain still satisfy the Gaussian white noise distribution, so by processing the coefficients, the denoising effect can be achieved.

[0031] Step (4): Use independent component analysis (ICA) to separate the target signal from the interference signal.

[0032] Blind source separation of signals occurs because crack and cavitation frequency bands contain components that are close to or overlap with their own signal frequencies, thus interfering with the desired signal. Frequency domain filtering cannot be used to prevent crack signal removal. Therefore, to address the issues of overlapping signal bands and the low energy content of the signal to be extracted, ICA analysis is employed. This method is immune to the effects of inter-band aliasing between source signals and interference from varying source signal strengths.

[0033] The basic principle is: assume that the independent component ,in As a useful independent component, is the noise variable, that is, the number of mixed signals is k+l. is a set of observation signals (where ), Each component is composed of The independent signal sources in are linearly combined through the mixing matrix A, that is: ; The basic idea of ICA is to separate the source signals from the observation signal X by unmixing the coefficient matrix W when the coefficient matrix A and the source matrix S are unknown, so that the output Y is the optimal approximation of S, that is: ; Therefore, ICA is actually an optimization calculation under a certain independence criterion. Here A is a mixing matrix that combines the true independent components and the noise covariance structure. Therefore, find m (where ) directions with extreme non-Gaussianity can be used to estimate the true independent components. The remaining independent components cannot be estimated because they are noise variables and, in practice, do not need to be estimated. Finally, a (pseudo) inverse transform is performed on each independent component to obtain a signal containing only that component.

[0034] The following is an example to illustrate the acoustic emission signal denoising method based on multi-method fusion provided by this application.

[0035] Step (1): Use an acoustic emission sensor to collect the acoustic emission signal of the turbine unit. The sampling frequency of the acoustic emission sensor is The following relationship must be satisfied: ; in is the acoustic emission frequency generated when a fault occurs in the turbine unit. Generally between 20kHz~400kHZ At least 1000kHZ.

[0036] Step (2), removing irrelevant frequency band signals by bandpass filtering, includes the following sub-steps: step (2-1) to step (2-9).

[0037] Step (2-1): Collect the signal when the turbine is operating normally (without cavitation) and the signal when cavitation occurs , and record the sampling frequency .

[0038] Step (2-2): Fourier transform is performed to obtain spectrum information. Fourier transform is performed on the two types of signals to obtain spectrum information: ; in, represents the Fourier transform.

[0039] Steps (2-3): Manually select the initial upper and lower bandpass limits.

[0040] Select the widest range: Based on the Fourier transformed spectrum, manually select the widest initial bandpass range to cover all frequency bands that may contain cavitation differences. This range should be as wide as possible to avoid missing any potential difference frequency bands.

[0041] For example, based on the spectrum comparison difference: by visually observing the spectrum graph of the two types of signals, select a frequency band so that the spectrum difference within the frequency band is more obvious. For example, if the energy of the cavitation signal in a certain frequency band is significantly higher than that of the normal signal, or the frequency distribution is significantly different, then this frequency band can be used as the initial bandpass range. Assuming that the spectrum difference is mainly concentrated between 200Hz and 800Hz, you can choose and as the initial range.

[0042] Steps (2-4): Calculate evaluation indicators.

[0043] The following four evaluation indicators are calculated, dimensionlessized and standardized.

[0044] (1) Spectrum overlap: ; The smaller the spectrum overlap, the more significant the signal difference.

[0045] (2) Frequency center of gravity difference : Calculate the frequency centroid: ; Frequency center of gravity difference: .

[0046] (3) Mutual Information I: ; in, is the joint probability density, and is the marginal probability density, given by , Normalized. The smaller the mutual information I, the more significant the signal difference.

[0047] (4) Spectral energy difference : Calculate the spectrum energy: ; Energy Difference: .

[0048] Steps (2-5): Dimensionless processing and standardization.

[0049] Normalize the four indicator values to the range [0,1]: ; in, Represents any indicator, and are the minimum and maximum values of the indicator respectively.

[0050] Step (2-6): Calculate comprehensive evaluation indicators.

[0051] The weight of each indicator is 0.25, and the comprehensive evaluation index is calculated: ; Steps (2-7): Iteratively adjust the upper and lower limits of the frequency domain.

[0052] (1) Adjustment upper limit: From the initial upper limit Initially, gradually reduce the upper limit by one step each time (e.g., △f = 10 Hz) and calculate the comprehensive evaluation index S. If S increases, continue to reduce the upper limit; otherwise, stop reducing the upper limit and record the current optimal upper limit.

[0053] (2) Adjust the lower limit: From the initial lower limit Initially, gradually increase the lower limit by one step each time (e.g., △f = 10 Hz) and calculate the comprehensive evaluation index S. If S increases, continue increasing the lower limit; otherwise, stop increasing the lower limit and record the current optimal lower limit.

[0054] (3) Adjustment strategy: When adjusting the upper and lower limits, alternate between decreasing the upper limit and increasing the lower limit to ensure the optimal range is not missed. For example, first decrease the upper limit, then increase the lower limit, alternating between them. Since S changes linearly, the adjustment is considered appropriate when S no longer increases significantly. After each adjustment, recalculate the spectrum information and evaluation index until S no longer changes significantly.

[0055] Steps (2-8): Repeat the optimization process.

[0056] Reduce the step size: After finding a better upper and lower limit range, reduce the step size (for example, from △f=10Hz to △f=5Hz) to further optimize the upper and lower limits.

[0057] Steps (2-9): Repeat the optimization process.

[0058] When the comprehensive evaluation index S no longer increases significantly, that is, the change of the comprehensive evaluation index S is within a preset range, it can be understood that the comprehensive evaluation index S tends to be stable at this time, and the final upper and lower limits are recorded. and , is the lower limit of the bandpass frequency range allowed to pass by the bandpass filter, The upper limit of the bandpass frequency range that the bandpass filter allows to pass.

[0059] Step (3): Perform wavelet threshold denoising on the signal finally obtained in step (2), including the following sub-steps, step (3-1) to step (3-3).

[0060] Step (3-1): Perform discrete wavelet transform (DWT) on the noisy signal f(t) (the signal obtained after filtering in step (2)) to decompose it into low-frequency information (approximate components) and high-frequency information (detail components) of different scales.

[0061] Step (3-2): Threshold processing, threshold processing is performed on the high-frequency wavelet coefficients (D, or detail coefficients), and the coefficients greater than the threshold are selected. To remove noise.

[0062] This application adopts an effective improvement method, in which different thresholds are set for each layer of wavelet coefficients to improve signal fidelity. The threshold function uses the Garrote threshold function, and the calculation formula is: ; Where: is the threshold, is the number of decomposition layers, is the root mean square error of the wavelet decomposition coefficients, and N is the length of the wavelet decomposition coefficients.

[0063] ; Where: is a symbolic function; is the regulating factor, is the detail coefficient (high-frequency wavelet coefficient) of wavelet decomposition, specifically The first decomposition layer The detail coefficient, Represents the detail coefficient after threshold processing.

[0064] Step (3-3): Inverse wavelet transform.

[0065] The high-frequency wavelet coefficients after threshold processing Perform inverse wavelet transform (IDWT) on the unprocessed low-frequency wavelet coefficient A to reconstruct the denoised signal or image: .

[0066] Step (4): For the signal after wavelet threshold denoising, ICA is performed to separate the target signal (clean signal) and the noise signal to eliminate the influence of frequency band overlap, including the following sub-steps, step (4-1) to step (4-3).

[0067] Step (4-1): Assume that the signal processed in step (3) is the input signal, and the input signal matrix is: .

[0068] Step (4-2): Perform independent component analysis (ICA) and use the FastICA algorithm to decompose the signal into independent components. The steps of this method are as follows.

[0069] (1) Establishing ICA model: ICA assumes that the observed signal is a linear mixture of multiple independent signals: ; in is the observed noisy signal matrix, is the source signal matrix (target signal + noise), is the unknown mixing matrix.

[0070] The goal of ICA is to find a separation matrix W (or unmixing coefficient matrix) such that: ; Make the portion The statistical independence between them is maximized. The separation matrix W is iteratively estimated by the ICA algorithm according to the independence maximization criterion (negative entropy).

[0071] (2) Optimization of independence: Independent components are separated by maximizing non-Gaussianity or minimizing mutual information. FastICA uses negentropy as the independence measure: ; where represents a hypothesized Gaussian distributed source signal, which is used to measure and compare the non-Gaussianity of the true source signal S. H(S) is the entropy of the signal, and the entropy of a Gaussian signal is the largest. Therefore, the goal is to find independent signals far from the Gaussian distribution.

[0072] Step (4-3): Select useful independent components and perform denoising using the energy threshold method. The steps of this method are as follows.

[0073] After ICA decomposition, each independent component has different energies: ; where represents the total energy of the i-th independent component is the amplitude of this component at time t.

[0074] Assume that: Signal components usually have higher energies because they carry the main useful information, and noise components usually have lower energies because the randomness of noise makes its power lower.

[0075] (1) Calculate the energy of each independent component: .

[0076] (2) Set the energy threshold T: The threshold can be set according to the mean or percentile of the energy distribution: ; where is an empirical coefficient (such as 0.1 - 0.3). It can also take a certain percentile (such as 20%) of all ; If < T, then this component is considered noise.

[0077] (3) Screen independent components: Only retain the components with ≥ T, and set the components with energy lower than the threshold to zero to remove noise.

[0078] The following describes the acoustic emission signal denoising device based on multi-method fusion provided by the present application. The acoustic emission signal denoising device based on multi-method fusion described below and the acoustic emission signal denoising method based on multi-method fusion described above can refer to each other.

[0079] Figure 2 Schematic diagram of the structure of the acoustic emission signal denoising device based on multi-method fusion provided in the embodiment of the present application. Figure 2 As shown, the device includes: an acquisition module 10, a filtering module 20, a background noise removal module 30 and a signal blind source separation module 40. Among them: Acquisition module 10, used for collecting acoustic emission signals through acoustic emission sensors; A filtering module 20 is configured to filter the collected acoustic emission signal using a bandpass filter to obtain a filtered signal, wherein the bandpass frequency range of the bandpass filter is determined by adjusting the frequency range based on multiple evaluation indicators, and the evaluation indicators are determined by analyzing the difference between a first signal sample and a second signal sample, wherein the first signal sample is an acoustic emission signal sample collected when there is no cavitation phenomenon, and the second signal sample is an acoustic emission signal sample collected when there is cavitation phenomenon; The background noise removal module 30 is used to perform wavelet threshold denoising based on the filtered signal to obtain a signal after removing the background noise; The signal blind source separation module 40 is used to separate the target signal from the interference signal through independent component analysis based on the signal after the background noise is removed.

[0080] It is understandable that the detailed functional implementation of each of the above units / modules can be found in the introduction of the aforementioned method embodiment, and will not be repeated here.

[0081] It should be understood that the above-mentioned device is used to execute the method in the above-mentioned embodiment. The implementation principle and technical effect of the corresponding program module in the device are similar to those described in the above-mentioned method. The working process of the device can refer to the corresponding process in the above-mentioned method and will not be repeated here.

[0082] Based on the method in the above embodiment, an embodiment of the present application provides an electronic device, Figure 3 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application, such as Figure 3 As shown, the electronic device may include: a processor (Processor) 810, a communication interface (Communications Interface) 820, a memory (Memory) 830, and a communication bus 840. The processor 810, the communication interface 820, and the memory 830 communicate with each other via the communication bus 840. The processor 810 may call the logic instructions in the memory 830 to execute the method in the above embodiment.

[0083] In addition, the logic instructions in the aforementioned memory 830 can be implemented in the form of a software functional unit and, when sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the portion that contributes to the prior art, or the portion of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application.

[0084] Based on the method in the above embodiment, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program runs on a processor, the processor executes the method in the above embodiment.

[0085] Based on the method in the above embodiment, an embodiment of the present application provides a computer program product. When the computer program product runs on a processor, the processor executes the method in the above embodiment.

[0086] It is understood that the processor in the embodiments of the present application may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA), other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. The general-purpose processor may be a microprocessor or any conventional processor.

[0087] The method steps in the embodiments of the present application can be implemented by hardware or by a processor executing software instructions. The software instructions can be composed of corresponding software modules, which can be stored in random access memory (RAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, mobile hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be an integral part of the processor. The processor and storage medium can be located in an ASIC.

[0088] The above embodiments can be implemented in whole or in part using software, hardware, firmware, or any combination thereof. When implemented using software, they can be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions. When loaded and executed on a computer, the computer program instructions fully or partially produce the processes or functions described in the embodiments of this application. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted via the computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible by a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium can be magnetic media (e.g., floppy disk, hard disk, tape), optical media (e.g., DVD), or semiconductor media (e.g., solid-state drive (SSD)).

[0089] It will be understood that the various numerical numbers involved in the embodiments of the present application are merely distinctions for the convenience of description and are not intended to limit the scope of the embodiments of the present application.

[0090] It is easy for those skilled in the art to understand that the above is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the scope of protection of the present application.

Claims

1. A method for denoising acoustic emission signals based on multi-method fusion, characterized in that: include: Acquiring acoustic emission signals through an acoustic emission sensor; The collected acoustic emission signal is filtered through a bandpass filter to obtain a filtered signal, wherein the bandpass frequency range of the bandpass filter is determined by adjusting the frequency range based on multiple evaluation indicators, and the evaluation indicators are determined by analyzing the difference between a first signal sample and a second signal sample, wherein the first signal sample is an acoustic emission signal sample collected when there is no cavitation phenomenon, and the second signal sample is an acoustic emission signal sample collected when there is cavitation phenomenon; Based on the filtered signal, wavelet threshold denoising is performed to obtain the signal after removing the background noise; Based on the signal after removing the background noise, the target signal and the interference signal are separated through independent component analysis.

2. The acoustic emission signal denoising method based on multi-method fusion according to claim 1 is characterized in that: Multiple evaluation indicators include: spectrum overlap, frequency center of gravity difference, mutual information and spectrum energy difference.

3. The acoustic emission signal denoising method based on multi-method fusion according to claim 2 is characterized in that: The passband frequency range of a bandpass filter is determined by the following steps: Based on the initial band-pass frequency range, gradually narrow the band-pass frequency range until the comprehensive evaluation index tends to be stable; Among them, the comprehensive evaluation index is determined by weighted summation based on spectrum overlap, frequency center of gravity difference, mutual information and spectrum energy difference.

4. The acoustic emission signal denoising method based on multi-method fusion according to claim 1, characterized in that: The wavelet threshold denoising includes: Based on the filtered signal, discrete wavelet transform is performed to obtain low-frequency wavelet coefficients and high-frequency wavelet coefficients; Based on the wavelet threshold, the high-frequency wavelet coefficients are subjected to threshold processing to obtain the high-frequency wavelet coefficients after threshold processing; Based on the low-frequency wavelet coefficients and the high-frequency wavelet coefficients after threshold processing, the signal after removing the background noise is obtained through inverse wavelet transform.

5. The method for denoising acoustic emission signals based on multi-method fusion according to claim 4, characterized in that: The wavelet threshold of each decomposition layer is determined by the following formula: ； in, Indicates the The wavelet threshold of the decomposition layer, represents the root mean square error of the wavelet decomposition coefficients, Indicates the length of the wavelet decomposition coefficients.

6. The method for denoising acoustic emission signals based on multi-method fusion according to claim 1, characterized in that: The method of separating the target signal from the interference signal by independent component analysis based on the signal after removing the background noise includes: Based on the signal after removing the background noise, multiple independent components are obtained through independent component analysis; Determining an energy threshold based on the energy of each independent component; Based on the energy threshold, the target component is screened out from multiple independent components; Based on the filtered target components, a target signal is constructed.

7. The method for denoising acoustic emission signals based on multi-method fusion according to claim 6, characterized in that: The energy threshold is determined by the following formula: ； in, represents the energy threshold, is the preset coefficient, represents the energy of each independent component, Indicates the mean.

8. An acoustic emission signal denoising device based on multi-method fusion, characterized in that: include: An acquisition module, used for acquiring acoustic emission signals through an acoustic emission sensor; a filtering module for filtering the collected acoustic emission signal through a bandpass filter to obtain a filtered signal, wherein the bandpass frequency range of the bandpass filter is determined by adjusting the frequency range based on multiple evaluation indicators, and the evaluation indicators are determined by analyzing the difference between a first signal sample and a second signal sample, wherein the first signal sample is an acoustic emission signal sample collected when there is no cavitation phenomenon, and the second signal sample is an acoustic emission signal sample collected when there is cavitation phenomenon; A background noise removal module is used to perform wavelet threshold denoising based on the filtered signal to obtain a signal after background noise removal; The signal blind source separation module is used to separate the target signal and the interference signal through independent component analysis based on the signal after removing the background noise.

9. An electronic device, characterized in that: include: at least one memory for storing a computer program; At least one processor is used to execute the program stored in the memory. When the program stored in the memory is executed, the processor is used to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed on a processor, the processor is caused to execute the method according to any one of claims 1 to 7.

Images