Processing method and device for laser shock composite back surface particle velocity signal

Abstract

The application discloses a processing method of a laser shock composite back surface particle velocity signal, and comprises the following steps: low-pass filtering and reducing sampling of a to-be-processed signal; zero-drift processing of the processed signal; S transformation of the processed signal to obtain a time-frequency graph, and extraction of a time-frequency ridge line to obtain a time-frequency feature of the signal; and processing of the signal by using a demodulation time-frequency analysis method in combination with the time-frequency feature to obtain a time-frequency feature of the signal. The application further discloses a processing device of the laser shock composite back surface particle velocity signal, and comprises a preprocessing module, a zero-drift reduction module, a transformation module and a time-frequency analysis module. Compared with the prior art, the time-frequency resolution is improved; and the time-frequency dynamic characteristics of the signal are more deeply and accurately understood through the combination of forward demodulation, backward modulation, short-time Fourier transformation and feature enhancement of the signal.

Description

Technical Field

[0001] This invention belongs to the field of signal processing technology, and relates to a method for processing particle velocity signals on the back side of laser-impacted composite materials, as well as a device for processing particle velocity signals on the back side of laser-impacted composite materials. Background Technology

[0002] LBI (Laser Bond Injection) utilizes the propagation of laser-induced tensile waves in materials and the principle of delamination. It employs high-energy laser shock to carbon fiber reinforced polymer (CFRP) laminates to effectively detect the interfacial bonding strength of CFRP laminates without obvious physical defects.

[0003] To achieve LBI (Laser-Induced Plasma) technology, it is necessary to obtain the particle velocity on the back surface of composite materials under laser-induced plasma shock wave irradiation. Currently, the main method used is to measure the velocity of free particles on the back surface of the material using a photonic Doppler velocimetry (PDV) system. There are many methods for processing back-side particle signals, including the fringe method, short-time Fourier transform, and wavelet transform. However, for signals with significant non-stationary characteristics, such as the velocity of free surfaces of composite materials subjected to laser shock, the above methods struggle to effectively capture and analyze the complex and rapid changes in time and frequency, resulting in insufficient resolution for accurate measurement and detailed analysis of non-stationary velocity signals of composite materials. Generally, the time resolution of PDV is on the order of 110 ns, making it difficult to obtain satisfactory results using only the instrument itself. Therefore, more advanced and adaptable signal processing methods are needed for composite material data to improve signal resolution and obtain details of the velocity evolution process at the rising front. Current research mainly focuses on PDV signal processing methods for metallic materials, with less attention paid to composite materials. This is because composite materials are anisotropic and have a large number of interfaces, which leads to a very complex stress wave distribution within the composite material. Factors such as the different material properties on both sides of the interface, acoustic impedance mismatch, and coupling of tensile and compressive waves need to be considered, resulting in very weak original PDV signal characteristics obtained from the test. At the same time, LBI technology often uses nanosecond pulsed lasers, and the shock wave duration is extremely short, ultimately resulting in insufficient time and frequency resolution of the acquired PDV signal, thus making it impossible to accurately obtain the velocity of particles on the back side. Summary of the Invention

[0004] The purpose of this invention is to provide a method for processing particle velocity signals on the back side of laser-shocked composite materials, which solves the problem of insufficient time-frequency resolution of the acquired PDV signals in the prior art.

[0005] The technical solution adopted in this invention is a method for processing particle velocity signals on the back side of laser-shocked composite materials, comprising the following steps:

[0006] Step 1: Perform low-pass filtering and downsampling on the signal to be processed;

[0007] Step 2: Perform zero-drift processing on the signal obtained in Step 1;

[0008] Step 3: Perform an S-transform on the signal obtained in Step 2 to obtain a time-frequency diagram, and extract the time-frequency ridges to obtain the time-frequency characteristics of the signal;

[0009] Step 4: Using the demodulation time-frequency analysis method, combined with the time-frequency characteristics obtained in Step 3, process the signal obtained in Step 2 to obtain the time-frequency characteristics of the signal.

[0010] The invention is further characterized by:

[0011] The zero-drift reduction process in step 2 is as follows: After low-pass filtering the signal to be processed in step 1, the trend term of the signal is obtained. The trend term is then subtracted from the downsampled signal to complete the zero-drift reduction.

[0012] Step 3S transformation process is as follows:

[0013] First, the Gaussian window function is introduced into the S-transform formula for the signal. After performing the S-transform, we get:

[0014]

[0015]

[0016] In the formula, S(τ,f) represents the characteristics of the signal in the time and frequency domains; |f| is the absolute value of the frequency f. Here, is a constant used for normalization; A(τ) is the amplitude of the signal at time τ; This is the phase term, which contains the phase information of the signal; Let be the Fourier transform of the window function, representing the characteristics of the window function in the frequency domain.

[0017] Step 3, obtaining the time-frequency ridge line, involves connecting the coordinates corresponding to the values ​​at the points of maximum energy represented by the time-frequency in the time-frequency diagram to obtain the time-frequency ridge line.

[0018] Step 4 includes the following steps:

[0019] Step 4.1, send signal Introducing forward demodulation operator The instantaneous frequency of the signal is obtained by calling the ridge function in step 3, thus eliminating the modulation source. The effect is to demodulate the signal into a pure carrier signal:

[0020]

[0021] Step 4.2: Introduce a backward modulation operator into the signal obtained in Step 4.1. The instantaneous frequency is restored to its original position, correcting any frequency shift that may have been introduced during demodulation.

[0022]

[0023] Step 4.3: The signal z obtained in step 4.1 d Substituting (t,u) into the short-time Fourier transform, we obtain the time-frequency characteristics of the signal:

[0024]

[0025] In the formula, u represents the sliding of the window on the time axis, ξ represents the frequency of the Fourier transform, and g σ (tu) is a Gaussian window or other window function, σ determines the window's support width, and e -iξ(t-u) Perform frequency domain analysis on the signal.

[0026] Another object of the present invention is to provide a device for processing particle velocity signals on the back side of a laser-shocked composite material.

[0027] Another technical solution adopted in this invention is a processing device for particle velocity signals on the back side of a laser-shocked composite material, comprising:

[0028] The preprocessing module is used to perform low-pass filtering and downsampling on the signal to be processed;

[0029] The zero-drift reduction module is used to perform zero-drift reduction processing on the preprocessed signal;

[0030] The transformation module is used to perform S-transform on the zero-drifted signal to obtain a time-frequency diagram, and extract the time-frequency ridge to obtain the time-frequency characteristics of the signal;

[0031] The time-frequency analysis module is used to process the zero-drifted signal by using demodulation time-frequency analysis methods and combining the time-frequency characteristics of the signal to obtain the time-frequency characteristics of the signal.

[0032] The beneficial effects of this invention are as follows: The method for processing particle velocity signals on the back side of laser-shocked composite materials employs S-transform to convert the time-domain signal to the complex frequency domain, making frequency domain analysis, system modeling, and signal processing more convenient and intuitive. Furthermore, by extracting time-frequency ridges, it can accurately capture the trajectory of major frequency changes from complex signals, making it suitable for analyzing time-frequency characteristics and robust to noise. The introduction of DTFA combines new features with the processed signal through time-frequency matching, and integrates the time-frequency features obtained from the S-transform, improving time-frequency resolution compared to existing methods. Finally, through a combination of forward demodulation, backward modulation, short-time Fourier transform, and feature enhancement, the time-frequency dynamic characteristics of the signal can be understood more deeply and accurately. Attached Figure Description

[0033] Figure 1 This is a flowchart of the method for processing particle velocity signals on the back side of the laser-shocked composite material according to the present invention;

[0034] Figure 2 This is the original signal diagram of the first embodiment of the laser shock composite material back particle velocity signal processing method of the present invention;

[0035] Figure 3 This is a signal image obtained by downsampling in the first embodiment of the laser shock composite material back particle velocity signal processing method of the present invention;

[0036] Figure 4 This is a signal diagram obtained by subtracting zero drift in the first embodiment of the laser shock composite material back particle velocity signal processing method of the present invention;

[0037] Figure 5 This is a two-dimensional time-frequency distribution diagram obtained by S-transformation in the first embodiment of the laser shock composite material particle velocity signal processing method of the present invention;

[0038] Figure 6 This is a three-dimensional time-frequency distribution diagram obtained by S-transformation in the first embodiment of the laser shock composite material particle velocity signal processing method of the present invention;

[0039] Figure 7 This is a time-frequency diagram obtained by S-transformation in the first embodiment of the laser shock composite material particle velocity signal processing method of the present invention;

[0040] Figure 8 This is a time-frequency ridge diagram obtained by S-transformation in the first embodiment of the laser shock composite material back particle velocity signal processing method of the present invention;

[0041] Figure 9 This is the final time-frequency diagram obtained in the first embodiment of the laser shock composite material back particle velocity signal processing method of the present invention;

[0042] Figure 10 This is the final analysis result of the back-side particle velocity obtained in the first embodiment of the laser shock composite material back-side particle velocity signal processing method of the present invention;

[0043] Figure 11 This is the original signal diagram of the second embodiment of the laser shock composite material back particle velocity signal processing method of the present invention;

[0044] Figure 12 This is a signal image obtained by downsampling in the second embodiment of the laser shock composite material particle velocity signal processing method of the present invention;

[0045] Figure 13 This is a signal diagram obtained by subtracting zero drift in the second embodiment of the laser shock composite material back particle velocity signal processing method of the present invention;

[0046] Figure 14 This is a time-frequency diagram obtained by S-transformation in the second embodiment of the laser shock composite material back particle velocity signal processing method of the present invention;

[0047] Figure 15 This is a time-frequency ridge diagram obtained by S-transformation in the second embodiment of the laser shock composite material back particle velocity signal processing method of the present invention;

[0048] Figure 16 This is the final time-frequency diagram obtained from the second embodiment of the laser shock composite material back particle velocity signal processing method of the present invention.

[0049] Figure 17 This is the final analysis result of the back-side particle velocity obtained from the second embodiment of the laser-shock composite material back-side particle velocity signal processing method of the present invention. Detailed Implementation

[0050] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0051] Example 1

[0052] Methods for processing particle velocity signals on the back side of laser-shocked composite materials, such as... Figure 1 As shown, it includes the following steps:

[0053] Step 1: Perform low-pass filtering and downsampling on the signal to be processed;

[0054] Step 2: Perform zero-drift processing on the signal obtained in Step 1;

[0055] Step 3: Perform an S-transform on the signal obtained in Step 2 to obtain a time-frequency diagram, and extract the time-frequency ridges to obtain the time-frequency characteristics of the signal;

[0056] Step 4: Using the demodulation time-frequency analysis method, combined with the time-frequency characteristics obtained in Step 3, process the signal obtained in Step 2 to obtain the time-frequency characteristics of the signal.

[0057] Example 2

[0058] A method for processing particle velocity signals on the back side of laser-shocked composite materials includes the following steps:

[0059] Step 1: Perform low-pass filtering and downsampling on the signal to be processed; after low-pass filtering, the trend term of the signal is obtained; low-pass filtering can effectively suppress high-frequency noise and unnecessary details, thereby improving the clarity and distinguishability of the signal; by smoothing the signal, low-pass filtering can weaken unstable rapid changes and enhance the stability and predictability of the signal.

[0060] Step 2: Perform zero-drift reduction processing on the signal obtained in Step 1. The process is as follows: Subtract the trend term obtained by low-pass filtering from the downsampled signal to obtain the signal after eliminating baseline drift. Its advantage is that by removing the long-term variation part in the signal, eliminating the trend term can reduce the impact of noise on signal analysis, making short-term fluctuations more obvious and easier to capture.

[0061] Step 3: Perform an S-transform on the signal obtained in Step 2 to obtain a time-frequency plot, and extract the time-frequency ridges to obtain the time-frequency characteristics of the signal; using the S-transform, based on the input signal and the exponential decay function (e st Multiplying and integrating the components (where s is a complex number) achieves the conversion from the time domain to the complex frequency domain; it provides a mathematical tool for converting time-domain signals to the complex frequency domain, making frequency domain analysis, system modeling, and signal processing more convenient and intuitive. Time-frequency ridge extraction locates the main frequency components of a signal at different time points by finding the frequency with the largest amplitude in each column of the time-frequency plot. It can accurately capture the trajectory of the main frequency changes in complex signals, is suitable for analyzing time-frequency characteristics, and is robust to noise.

[0062] Specifically, in step 3.1, we assume the signal model is as follows: The result of its S-transform is:

[0063]

[0064] make Introducing the Gaussian window function into the S-transform formula simplifies the cumbersome integration part into a Gaussian function, making the mathematical expression clearer and easier to process; then, for the signal model... After performing the S-transform, we get:

[0065]

[0066]

[0067] In the formula, S(τ,f) represents the characteristics of the signal in the time and frequency domains; |f| is the absolute value of the frequency f. Here, is a constant used for normalization; A(τ) is the amplitude of the signal at time τ; This is the phase term, which contains the phase information of the signal; Let be the Fourier transform of the window function, representing the characteristics of the window function in the frequency domain.

[0068] Step 3.2: Since the Fourier transform of a Gaussian window is also a Gaussian window, the maximum energy of the S-transform is concentrated in... Nearby; connect the coordinates corresponding to the values ​​at the points of maximum energy represented by the time frequency in the time-frequency graph to obtain the time-frequency ridge line, and thus obtain the time-frequency characteristics.

[0069] The S-transform is used to obtain the time-frequency information of a signal, revealing its preliminary characteristics at different times and frequencies. However, due to the constraint of the Heisenberg uncertainty principle, the time and frequency spans of the window function cannot be minimized simultaneously, resulting in relatively coarse features. To improve time-frequency resolution, the Demodulation Time-Frequency Analysis (DTFA) method is introduced. DTFA further processes the features obtained from the S-transform by constructing new nonlinear odd functions, further enhancing the finer detail of the signal's time-frequency characteristics. This method combines the new features with the original signal through time-frequency matching, achieving significantly improved time-frequency analysis results. Therefore, the introduction of DTFA greatly improves the time-frequency resolution of traditional linear time-frequency analysis methods, providing an effective means to more accurately capture the time-frequency characteristics of signals.

[0070] Step 4: Using the demodulation time-frequency analysis method, combined with the time-frequency characteristics obtained in Step 3, process the signal obtained in Step 2 to obtain the time-frequency characteristics of the signal.

[0071] Step 4.1, send signal Introducing forward demodulation operator The instantaneous frequency of the signal is obtained by calling the ridge function in step 3, thus eliminating the modulation source. The effect is to demodulate the signal into a pure carrier signal:

[0072]

[0073] Step 4.2: Introduce a backward modulation operator into the signal obtained in Step 4.1. The instantaneous frequency is restored to its original position, correcting any frequency shift that may have been introduced during demodulation.

[0074]

[0075] Step 4.3: The signal z obtained in step 4.1 d Substituting (t,u) into the short-time Fourier transform, we obtain the time-frequency characteristics of the signal:

[0076]

[0077] In the formula, This is the result of the short-time Fourier transform, representing the signal characteristics in the time and frequency domains. u represents the sliding of the window along the time axis, ξ is the frequency of the Fourier transform, and z... d (t,u) represents the signal in the time domain, i.e., the window function g. σ (tu) weighted signal, g σ (tu) is a Gaussian window or other window function used to locally weight the signal; σ determines the support width of the window; e -iξ(t-u) Perform frequency domain analysis on the signal.

[0078] Example 3

[0079] A method for processing particle velocity signals on the back side of laser-shocked composite materials. Under the conditions of a laser pulse width of 20 ns and an energy of 0.7 J, the raw time-voltage signal acquired by the PDV acquisition system is shown below. Figure 2 As shown, the method of the present invention is used to process carbon fiber T300 / QY8911, including the following steps:

[0080] Step 1: Perform low-pass filtering and downsampling on the signal to be processed, reducing the number of sample points from 200,000 to 8,000. Figure 3 As shown;

[0081] Step 2: Perform zero-drift reduction processing on the signal obtained in Step 1, such as... Figure 4 As shown;

[0082] Step 3: Perform an S-transform on the signal obtained in Step 2 to obtain the time-frequency diagram, see... Figure 5-7 And extract the time-frequency ridge line, see Figure 8 The time-frequency characteristics of the signal are obtained.

[0083] Step 4: Using demodulation time-frequency analysis methods, combined with the time-frequency characteristics obtained in Step 3, process the signal obtained in Step 2 to obtain the signal's time-frequency characteristics, such as... Figure 9-10 Analysis of the figure shows that signal processing improves the time-frequency resolution, thus enabling more accurate acquisition of the back-side particle velocity.

[0084] Example 4

[0085] A method for processing particle velocity signals on the back side of laser-shocked composite materials. Under the conditions of a laser pulse width of 20 ns and an energy of 1.5 J, the raw time-voltage signal acquired by the PDV acquisition system is as follows: Figure 11 As shown, the method of the present invention is used to process carbon fiber T300 / QY8911, including the following steps:

[0086] Step 1: Perform low-pass filtering and downsampling on the signal to be processed, reducing the number of sample points from 200,000 to 8,000. Figure 12 As shown;

[0087] Step 2: Perform zero-drift reduction processing on the signal obtained in Step 1, such as... Figure 13 As shown;

[0088] Step 3: Perform an S-transform on the signal obtained in Step 2 to obtain the time-frequency diagram, as shown below. Figure 14 As shown, the time-frequency ridge line was extracted. Figure 15 The time-frequency characteristics of the signal are obtained.

[0089] Step 4: Using the demodulation time-frequency analysis method, combined with the time-frequency characteristics obtained in Step 3, process the signal obtained in Step 2. The resulting time-frequency diagram is shown below. Figure 16 As shown, its time-frequency resolution is higher than that of the S-transform, enabling more accurate acquisition of the back-side particle velocity, such as... Figure 17 As shown. Through the above methods, the processing method for back-side particle velocity signals of laser-shocked composite materials of the present invention employs S-transform to convert the time-domain signal to the complex frequency domain, making frequency domain analysis, system modeling, and signal processing more convenient and intuitive. Furthermore, by extracting time-frequency ridges, it can accurately capture the trajectory of major frequency changes from complex signals, making it suitable for analyzing time-frequency characteristics and robust to noise. The introduction of DTFA combines new features with the processed signal through time-frequency matching, and, combined with the time-frequency features obtained from the S-transform, yields the signal's time-frequency characteristics, improving time-frequency resolution and thus more accurately acquiring the back-side particle velocity. By combining forward demodulation, backward modulation, short-time Fourier transform, and feature enhancement, the time-frequency dynamic characteristics of the signal can be understood more deeply and accurately.

Claims

1. A method for processing particle velocity signals on the back side of a laser-shocked composite material, characterized in that, Includes the following steps: Step 1: Perform low-pass filtering and downsampling on the signal to be processed; Step 2: Perform zero-drift processing on the signal obtained in Step 1; Step 3: Perform an S-transform on the signal obtained in Step 2 to obtain a time-frequency diagram, and extract the time-frequency ridges to obtain the time-frequency characteristics of the signal; Step 4: Using the demodulation time-frequency analysis method, combined with the time-frequency characteristics obtained in Step 3, process the signal obtained in Step 2 to obtain the time-frequency characteristics of the signal; Step 4 includes the following steps: Step 4.1, send signal Introducing forward demodulation operator The instantaneous frequency of the signal is obtained by calling the ridge function in step 3, thus eliminating the modulation source. The effect is to demodulate the signal into a pure carrier signal: Step 4.2: Introduce a backward modulation operator into the signal obtained in Step 4.

1. This restores the instantaneous frequency to its original position, correcting any frequency shifts that may be introduced during demodulation. in, Let be the amplitude of the signal at time τ; Step 4.3: The signal obtained in step 4.2 Substituting into the short-time Fourier transform, we obtain the time-frequency characteristics of the signal: In the formula, For the sliding of the window on the timeline, The frequency of the Fourier transform. For Gaussian windows or other window functions, σ For the support width of the window, Perform frequency domain analysis on the signal.

2. The method for processing particle velocity signals on the back side of a laser-shocked composite material according to claim 1, characterized in that, The zero-drift reduction process described in step 2 is as follows: After performing low-pass filtering on the signal to be processed in step 1, the trend term of the signal is obtained. The trend term is then subtracted from the downsampled signal to complete the zero-drift reduction.

3. The method for processing particle velocity signals on the back side of a laser-shocked composite material according to claim 1, characterized in that, The S-transform process described in step 3 is as follows: First, the Gaussian window function is introduced into the S-transform formula for the signal. After performing the S-transform, we get: In the formula, These are the characteristics of the signal in the time and frequency domains. For frequency The absolute value; Here is a constant used for normalization; This is the phase term, which contains the phase information of the signal; Let be the Fourier transform of the window function, representing the characteristics of the window function in the frequency domain.

4. The method for processing particle velocity signals on the back side of a laser-shocked composite material according to claim 1, characterized in that, The process of obtaining the time-frequency ridge line in step 3 is as follows: connect the coordinates corresponding to the values ​​of the points where the time-frequency represents the maximum energy in the time-frequency diagram to obtain the time-frequency ridge line.

5. A device for processing particle velocity signals on the back side of a laser-shocked composite material, characterized in that, By using demodulation time-frequency analysis methods and combining time-frequency characteristics, the signal is processed to obtain its time-frequency features; this includes the following steps: First, to the signal Introducing forward demodulation operator The instantaneous frequency of the signal is obtained by calling the ridge function, thus eliminating the modulation source. The effect is to demodulate the signal into a pure carrier signal: Secondly, a backward modulation operator is introduced into the obtained pure carrier signal. This restores the instantaneous frequency to its original position, correcting any frequency shifts that may be introduced during demodulation. Let be the amplitude of the signal at time τ; Finally, the obtained signal Substituting into the short-time Fourier transform, we obtain the time-frequency characteristics of the signal: In the formula, For the sliding of the window on the timeline, The frequency of the Fourier transform. For Gaussian windows or other window functions, σ For the support width of the window, Perform frequency domain analysis on the signal; It includes: a preprocessing module for low-pass filtering and downsampling of the signal to be processed; a zero-drift reduction module for zero-drift reduction of the preprocessed signal; a transformation module for performing an S-transform on the zero-drifted signal to obtain a time-frequency graph and extracting time-frequency ridges to obtain the time-frequency characteristics of the signal; and a time-frequency analysis module for processing the zero-drifted signal using demodulation time-frequency analysis methods, combined with the time-frequency characteristics of the signal, to obtain the time-frequency characteristics of the signal.

Images