CN113970419A - Shock tunnel force measurement balance signal data processing method based on time-frequency transformation - Google Patents
Shock tunnel force measurement balance signal data processing method based on time-frequency transformation Download PDFInfo
- Publication number
- CN113970419A CN113970419A CN202111192645.6A CN202111192645A CN113970419A CN 113970419 A CN113970419 A CN 113970419A CN 202111192645 A CN202111192645 A CN 202111192645A CN 113970419 A CN113970419 A CN 113970419A
- Authority
- CN
- China
- Prior art keywords
- frequency
- signal
- frequency coefficient
- balance
- low
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M9/00—Aerodynamic testing; Arrangements in or on wind tunnels
- G01M9/06—Measuring arrangements specially adapted for aerodynamic testing
- G01M9/065—Measuring arrangements specially adapted for aerodynamic testing dealing with flow
Landscapes
- Physics & Mathematics (AREA)
- Fluid Mechanics (AREA)
- General Physics & Mathematics (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention discloses a shock tunnel force measurement balance signal data processing method based on time-frequency transformation, which comprises the steps of building a shock tunnel aerodynamic force measurement system of a hypersonic aircraft model, collecting balance step signals of the shock tunnel aerodynamic force measurement system by adopting an SVDC (space vector direct current) technology, and constructing real aerodynamic force signals output by an ideal step signal simulation balance; performing wavelet decomposition on the balance step signal by adopting a wavelet threshold denoising method to obtain a sub-signal and filtering a high-frequency noise signal to obtain an effective characteristic signal; performing Hilbert-yellow transformation on the effective characteristic signal subjected to the high-frequency noise interference filtering, and obtaining a plurality of inherent modal functions and residual components by adopting an empirical mode decomposition method; performing Hilbert spectrum analysis on each inherent modal function, and filtering interference signals in effective characteristic signals to obtain effective aerodynamic force signals; the method effectively identifies different interference components in the balance signal and outputs a reliable aerodynamic force result.
Description
Technical Field
The invention relates to the technical field of shock tunnel force measurement tests, in particular to a shock tunnel force measurement balance signal data processing method based on time-frequency transformation.
Background
The pulse type strain balance is a measuring device which can measure the voltage change of a strain gauge by quickly responding the deformation of a model under the action of an impact load and further reflect the model load, and is widely applied to force measurement tests of hypersonic aircrafts and the like due to the characteristics of large integral structure rigidity, low interference among components, high output sensitivity, strong stability, high precision and the like. The output signal of the balance comprises an aerodynamic force signal, an inertial vibration interference signal and other interference signals (such as interference signals caused by structure high-order modal vibration, unsteady aerodynamic load or other flow field interference factors).
When a force measurement test is carried out in a shock tunnel, a force measurement system is subjected to impact excitation at the moment of starting a flow field, inertial interference is generated on output signals of a balance, particularly, in the extremely short effective test time (millisecond level), the regularity of the output signals of the balance superposed with dynamic aerodynamic force and inertial vibration can not be directly distinguished, and a larger error is generated between a signal processing result and real aerodynamic force, even data is unavailable.
Due to the structural complexity of the model force-measuring balance system, part of high-frequency components (high-frequency interference caused by structural high-order modal vibration, unsteady aerodynamic load or other flow field interference and other factors) in balance signals of the existing shock tunnel force-measuring balance system can not be completely attenuated in effective test time, and at the moment, the signals are directly subjected to traditional filtering processing and fast Fourier transform analysis, but the error of the processing result is possibly increased.
Disclosure of Invention
The invention aims to provide a method for processing signal data of a shock tunnel force measuring balance based on time-frequency transformation, which aims to solve the technical problem in the prior art.
In order to solve the technical problems, the invention specifically provides the following technical scheme:
a shock tunnel force measurement balance signal data processing method based on time-frequency transformation comprises the following steps:
As a preferred aspect of the present invention, in step 200, the wavelet threshold denoising method performs approximate decomposition and detail decomposition on the balance step signal, and decomposes the balance step signal to obtain a low-frequency coefficient and a high-frequency coefficient, where the low-frequency coefficient is used to show a trend of the whole balance step signal, and the high-frequency coefficient is used to show a detail component of the whole balance step signal, and the specific implementation manner of the wavelet threshold denoising method performing wavelet decomposition and filtering high-frequency noise is as follows:
step 201, performing multi-level wavelet decomposition on the denoised balance step signal by using one-dimensional discrete wavelet transform to obtain a low-frequency coefficient and a high-frequency coefficient of the balance step signal subjected to the multi-level wavelet decomposition;
step 202, performing fast Fourier transform on the low-frequency coefficient, the high-frequency coefficient and the ideal step signal, and calculating 99% of occupied bandwidth of the low-frequency coefficient, the high-frequency coefficient and the ideal step signal after the fast Fourier transform;
and 203, comparing the 99% occupied bandwidth of the high-frequency coefficient and the low-frequency coefficient with the reference by taking the 99% occupied bandwidth of the ideal step signal as the reference, and filtering the high-frequency coefficient and the low-frequency coefficient which are completely inconsistent with the 99% occupied bandwidth of the ideal step signal.
As a preferred aspect of the present invention, a spectrogram is generated after performing fast fourier transform on a low-frequency coefficient, a high-frequency coefficient and an ideal step signal, the high-frequency coefficient exceeding 99% of an occupied bandwidth of the ideal step signal is filtered, and a main frequency of each high-frequency coefficient is determined according to an amplitude of the high-frequency coefficient and a corresponding frequency.
As a preferable aspect of the present invention, if the 99% occupied bandwidth of the high-frequency coefficient and the low-frequency coefficient does not intersect with the 99% occupied bandwidth of the ideal step signal, it indicates that the 99% occupied bandwidth of the high-frequency coefficient and the low-frequency coefficient does not completely meet the 99% occupied bandwidth of the ideal step signal, and the high-frequency coefficient and the low-frequency coefficient that do not meet the requirement are filtered out.
As a preferable scheme of the present invention, the 99% occupied bandwidth is that the signal power of the low-frequency coefficient, the high-frequency coefficient, and the ideal step signal after the fast fourier transform accounts for 99% of the total signal power.
As a preferable scheme of the invention, the high-frequency coefficient and the low-frequency coefficient which are completely inconsistent are filtered and then subjected to inverse Fourier transform, and a proper coefficient of the high-frequency coefficient and the low-frequency coefficient which are completely inconsistent is selected for filtering and is subjected to signal reconstruction, so that an effective characteristic signal after a high-frequency noise signal is filtered is obtained.
As a preferred embodiment of the present invention, in step 300, the data processing method for the effective characteristic signal by using the hilbert-yellow transform is: performing multi-stage processing on the low-frequency coefficient in the effective characteristic signal by using an empirical mode decomposition method of Hilbert-Huang transform to obtain a multi-stage inherent mode function and a multi-stage processed residual component;
wherein the natural mode function of the previous stage in the empirical mode decomposition is a sum of a natural mode function and a residual component of the next stage.
As a preferred embodiment of the present invention, in step 400, Hilbert spectrum analysis is performed on the obtained natural mode functions of each stage by using a Hilbert spectrum analysis method to obtain an instantaneous frequency, an instantaneous amplitude and a Hilbert spectrum of each natural mode function, and interference signals of the natural mode functions are filtered to obtain an effective aerodynamic signal, which is implemented by the following steps:
acquiring a time domain diagram of a multi-stage natural mode function and a residual component of the low-frequency signal;
acquiring Hilbert spectrums of the multi-level intrinsic mode functions of the low-frequency signals, obtaining instantaneous frequencies of the intrinsic mode functions of each level, and calculating 99% occupied bandwidth of the intrinsic mode functions and residual components of each level;
the remaining components and the natural mode functions that fully fit 99% of the occupied bandwidth of the ideal step signal are retained as valid aerodynamic signals.
In a preferred embodiment of the present invention, the primary frequencies of the original balance step signal are obtained from the Hilbert spectrum of the natural mode functions in multiple stages, and the primary frequencies of the balance step signal and the spectrogram of the natural mode functions having the same waveform and the largest amplitude ratio are used as the inertial vibration frequencies.
Compared with the prior art, the invention has the following beneficial effects:
according to the method, wavelet transformation and Hilbert-Huang transformation are adopted, noise reduction and time-frequency transformation analysis processing are carried out on shock tunnel balance signals of the hypersonic aircraft model, different interference components in the balance signals are effectively identified, a reliable aerodynamic force result is output, errors of inertial characteristic frequencies of a subsequent fast Fourier transform filtering system are reduced, and the precision of the whole shock tunnel aerodynamic force measurement system is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below. It should be apparent that the drawings in the following description are merely exemplary, and that other embodiments can be derived from the drawings provided by those of ordinary skill in the art without inventive effort.
FIG. 1 is a schematic flow chart of a method for processing signal data of a force measuring balance according to an embodiment of the present invention;
FIG. 2 is a time domain diagram of a step signal of the balance provided by an embodiment of the invention;
FIG. 3 is a time domain signal diagram of low frequency coefficients of a wavelet decomposition provided by an embodiment of the present invention;
FIG. 4 is a time domain signal diagram of the high frequency coefficients of the wavelet decomposition provided by the embodiment of the present invention;
FIG. 5 is a diagram illustrating a time-domain comparison result of a wavelet de-dried significant feature signal and a low-frequency coefficient according to an embodiment of the present invention;
FIG. 6 is a diagram illustrating the comparison result of the frequency spectrums of the wavelet dessicated significant feature signals and the low frequency coefficients according to the embodiment of the present invention;
FIG. 7 is a time domain diagram of an empirical mode decomposition signal provided by an embodiment of the present invention;
FIG. 8 is an instantaneous frequency plot of a multi-stage natural mode function provided by an embodiment of the present invention;
FIG. 9 is a diagram illustrating the time domain comparison result of the residual component and the significant feature signal after wavelet de-weighting according to an embodiment of the present invention;
fig. 10 is a graph showing the comparison result of the frequency spectrums of the wavelet dessicated significant characteristic signal and the residual component according to the embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1, the invention provides a method for processing data of a shock tunnel force measurement balance signal based on time-frequency transformation, and the embodiment adopts wavelet transformation and hilbert-yellow transformation to perform noise reduction and time-frequency transformation analysis processing on the shock tunnel balance signal of a hypersonic aircraft model, effectively identify different interference components in the balance signal, and output a reliable aerodynamic force result.
In order to facilitate data acquisition and analysis, in the shock tunnel aerodynamic force measurement system, a step load signal (abbreviated as a balance step signal) is used for simulating an excitation effect of a force measurement system in a wind tunnel test, a constructed ideal step load signal (abbreviated as an ideal step signal) is used for simulating a real aerodynamic force signal output by a balance, and then time-frequency conversion analysis processing is carried out on the balance step signal for verifying the reliability of the time-frequency method.
When a shock tunnel test is carried out, the force measuring system is subjected to the instant impact action of incoming flow to generate a sudden change signal, and the system is subjected to the excitation action of a wind tunnel flow field, so that a balance signal is possibly unstable. In the effective test time, the balance signal is superposed with a plurality of signals with the frequency changing along with the time, and is a typical nonlinear and non-stable abrupt signal.
Aiming at the characteristics of the impulse vibration of the pulse wind tunnel force measuring balance signal, the embodiment adopts a wavelet threshold denoising method and a Hilbert-Huang transform method to perform time-frequency transform analysis processing on the shock wind tunnel balance signal.
The method specifically comprises the following steps:
It should be added that, in the SVDC technique, a stable load is applied to the system by suspending the steel wire, and an impact step load is applied to the system by instantaneously shearing the steel wire.
The collecting device of the shock tunnel aerodynamic force measuring system is composed of a standard pointed cone model with a half cone angle of 10 degrees and a length of 750mm, a three-component pulse type strain antenna and a bent blade supporting structure. The device is adopted to collect a proper amount of step load signals for the force measuring system, and the signals reflect the characteristics of wind tunnel force measuring signals in the actual wind tunnel test to a certain extent.
And selecting any one collected balance step signal, and drawing a time domain waveform diagram of the balance step signal into a diagram shown in fig. 2. The dashed line in fig. 2 is used to simulate the response signal of the force measuring system when subjected to the transient impulse excitation from the flow field initiation, and the solid line represents the ideal step signal constructed to simulate a simplified steady state pneumatic force signal. The balance step signal is used for simulating a simplified wind tunnel force measurement signal, the sampling time is 150ms, the balance step signal comprises a zero-value signal of the first 50ms and a pneumatic signal of 100ms after loading, the signal jumps at the 50 th ms, the edge trigger time is ignored, the ideal step signal is changed from 0N to about 2.5N in a sudden change mode, and the ideal step signal is kept stable and unchanged in the last 100 ms.
The wavelet threshold denoising method carries out approximate decomposition and detail decomposition on the balance step signal to obtain a low-frequency coefficient and a high-frequency coefficient through decomposition, wherein the low-frequency coefficient is used for showing the trend of the whole balance step signal, the high-frequency coefficient is used for showing the detail components of the whole balance step signal, and the specific implementation mode of the wavelet threshold denoising method for carrying out wavelet decomposition and filtering high-frequency noise is as follows:
and step 201, performing multi-level wavelet decomposition on the denoised balance step signal by using one-dimensional discrete wavelet transform to obtain a low-frequency coefficient and a high-frequency coefficient of the balance step signal subjected to the multi-level wavelet decomposition.
The wavelet decomposition can be generally expressed as a pair of complementary low-pass filter and high-pass filter, the balance step signal is subjected to the first-level wavelet decomposition to obtain a low-frequency coefficient and a high-frequency coefficient, and at the moment, the low-frequency coefficient can be continuously decomposed to improve the signal filtering precision.
In the present embodiment, 8-level wavelet decomposition is performed on the noise-reduced scale step signal by using one-dimensional discrete wavelet transform, the obtained low-frequency coefficients a1 to A8 are compared with the high-frequency coefficients D1 to D8, and the time domain has a difference in the time domain according to the principle of wavelet decomposition
S＝(A1+D1)
＝(A2+D2)+D1
＝(A3+D3)+D2+D1
＝…
＝(A8+D8)+D7+D6+…+D1；
In the formula, S is a balance step signal subjected to wavelet de-noising and deburring, and A1-A8 and D1-D8 are respectively a low-frequency coefficient and a high-frequency coefficient obtained through 8-level wavelet decomposition.
The low frequency coefficients a 1-a 5 have substantially the same waveform as the original signal, and show the approximate information of the original signal, while the high frequency coefficients reflect the detailed information of the original signal, and the specific time domain diagrams of the low frequency coefficients and the high frequency coefficients are shown in fig. 3 and fig. 4.
Step 202, performing fast fourier transform on the low-frequency coefficient, the high-frequency coefficient and the ideal step signal, and calculating 99% of occupied bandwidth of the low-frequency coefficient, the high-frequency coefficient and the ideal step signal after the fast fourier transform.
And 203, comparing the 99% occupied bandwidth of the high-frequency coefficient and the low-frequency coefficient with the reference by taking the 99% occupied bandwidth of the ideal step signal as the reference, and filtering the high-frequency coefficient and the low-frequency coefficient which are completely inconsistent with the 99% occupied bandwidth of the ideal step signal.
Preferably, the low-frequency coefficient, the high-frequency coefficient and the ideal step signal are subjected to fast fourier transform to generate a spectrogram, the high-frequency coefficient exceeding 99% of the occupied bandwidth of the ideal step signal is filtered, and the main frequency of each high-frequency coefficient is determined according to the amplitude and the corresponding frequency of the high-frequency coefficient.
And if the 99% occupied bandwidth of the high-frequency coefficient and the low-frequency coefficient is not intersected with the 99% occupied bandwidth of the ideal step signal, the 99% occupied bandwidth of the high-frequency coefficient and the low-frequency coefficient is completely inconsistent with the 99% occupied bandwidth of the ideal step signal, and the high-frequency coefficient and the low-frequency coefficient which are completely inconsistent are filtered and removed. The 99% occupied bandwidth is that the signal power of the low-frequency coefficient, the high-frequency coefficient and the ideal step signal after fast Fourier transform accounts for 99% of the total signal power.
The results of the low-frequency coefficient, the high-frequency coefficient and the 99% occupied bandwidth of the ideal step signal are shown in the following table 1, and the 99% occupied bandwidth of the ideal step signal is 0.02-205.07 Hz. The high-frequency coefficients D1-D5 are small in amplitude, almost approach to zero and keep stable, the frequencies of the high-frequency coefficients are very high and are all above 240.60Hz, and the high-frequency coefficients can be used as high-frequency noise interference to be filtered relative to original signals. D6 and D7 occupy the dominant position of high frequency coefficient, the amplitude of the high frequency coefficient is equivalent to that of an original signal, the main frequency of the high frequency coefficient is 379.89Hz, and the high frequency coefficient can be considered as the inertial vibration frequency of the shock tunnel aerodynamic force measurement system. The 99% occupied bandwidth of D8 is 74.96-243.03 Hz, and has an overlapping part with the ideal step signal, so D8 cannot be completely filtered as an interference signal.
99% occupied bandwidth of signal after table 18 level wavelet decomposition
The low-frequency coefficient a7 after the high-frequency coefficients D1-D7 are filtered out preliminarily reflects the characteristics of the ideal step load, the frequency of the ideal step load conforms to 99% of the occupied bandwidth of the ideal step signal, and the ideal step load is compared with the low-frequency coefficient a7, as shown in fig. 5 and 6. Fig. 5 and 6 show a comparison between a time domain waveform diagram and a frequency domain amplitude spectrogram, respectively, wherein a dotted line is a balance step signal after wavelet de-noising and deburring, a dotted line is an ideal step signal, and a solid line is a7 th-level low-frequency coefficient a7 after 8-level wavelet decomposition. Compared with a balance step signal S, A7 has removed most vibration interference signals, wherein the main frequency of S is about 379.89Hz, and the inertial vibration frequency of the force measuring system is 379.89 Hz.
And performing inverse Fourier transform on the high-frequency coefficients (D1-D5) and the low-frequency coefficients (A1-A4) which are completely inconsistent with each other after filtering, and selecting proper coefficients (D6 and A5-A8) of the high-frequency coefficients and the low-frequency coefficients which are completely inconsistent with each other after filtering to perform signal reconstruction to obtain effective characteristic signals after high-frequency noise signals are filtered.
And 300, performing Hilbert-yellow transformation on the effective characteristic signal subjected to the high-frequency noise interference filtering, and obtaining a plurality of inherent mode functions and residual components by adopting an empirical mode decomposition method.
The data processing mode of the effective characteristic signal by using Hilbert-Huang transform is as follows: performing multi-stage processing on the low-frequency coefficient in the effective characteristic signal by using an empirical mode decomposition method of Hilbert-Huang transform to obtain a multi-stage inherent mode function and a multi-stage processed residual component;
wherein the natural mode function of the previous stage in the empirical mode decomposition is a sum of a natural mode function and a residual component of the next stage.
Since the Hilbert-Huang transform method is easily affected by high-frequency noise, the balance step signal is first subjected to high-frequency noise reduction, the low-frequency coefficient subjected to wavelet decomposition and noise reduction is filtered to remove frequencies above 724.06Hz, and the Hilbert-Huang transform can be used for processing and analysis.
Therefore, the empirical mode decomposition method in hilbert-yellow transform is used to perform three-level processing on the low-frequency coefficient a5 to obtain three intrinsic mode functions IMFs and a Residual component Residual, which are:
in the formula, a5 is a low-frequency coefficient a5 of the balance step signal after wavelet decomposition and filtering of high-frequency coefficients D1-D5, and basically has no interference of high-frequency noise, IMFs 1-3 are inherent mode functions of the balance step signal after three-level empirical mode decomposition, R1-R3 are residual components of the balance step signal after three-level empirical mode decomposition, and a time domain diagram of an effective characteristic signal after the balance step signal is specifically processed by an empirical mode decomposition method is shown in fig. 7.
In fig. 7, the first-stage intrinsic mode function IMF1 is similar to the original balance step signal in waveform, basically reflects an approximate value of the original balance step signal, occupies a dominant position in the balance step signal, has a main frequency of 379.89Hz, and is the same as the inertial vibration frequency obtained by wavelet decomposition, the second-stage intrinsic mode function IMF2 and the third-stage intrinsic mode function IMF3 reflect the detailed values of the signal, and have a small occupancy ratio, the second-stage intrinsic mode function IMF2 has an amplitude only when the signal jumps, and the amplitude thereof remains stable and tends to zero after the signal jumps, and the third-stage intrinsic mode function IMF3 has fluctuations in the time period after the signal jumps.
In step 400, Hilbert spectrum analysis is performed on the obtained intrinsic mode functions of each stage by using a Hilbert spectrum analysis method to obtain an instantaneous frequency, an instantaneous amplitude and a Hilbert spectrum of each intrinsic mode function, and interference signals of the intrinsic mode functions are filtered to obtain effective aerodynamic signals, which is implemented by:
and acquiring Hilbert spectrums of the inherent modal functions of the multiple stages of the low-frequency signals to obtain the instantaneous frequency of each stage of the inherent modal function.
It should be added that Hilbert spectrum analysis is performed on the obtained multi-stage intrinsic mode functions by using a Hilbert spectrum analysis method in Hilbert-yellow transform to obtain the instantaneous frequency of each IMF, specifically, as shown in fig. 8, the instantaneous frequency of the third-stage intrinsic mode function IMF3 in fig. 8 is found to have a main frequency of about 31.30Hz and is completely located within 99% occupied bandwidth of an ideal step signal, so that the third-stage intrinsic mode function IMF3 cannot be directly filtered as a low-frequency interference signal.
Performing fast Fourier transform on all the inherent modal functions and the residual components, and calculating 99% occupied bandwidth of the inherent modal functions and the residual components;
the remaining components and the natural mode functions that fully fit 99% of the occupied bandwidth of the ideal step signal are retained as valid aerodynamic signals.
The bandwidth is occupied by the three-stage inherent mode function and 99% of the residual component, the result is collated to be shown in table 2,
TABLE 2 99% occupied bandwidth of HHT processed signals
According to the data in table 2, the residual component R2 after filtering out the intrinsic mode functions IMF1 and IMF2 completely conforms to 99% occupied bandwidth of the ideal step signal and substantially reflects the characteristics of the ideal step signal, and the comparison result of the distribution diagram of the residual component R2 in the time domain and the frequency domain with the effective characteristic signal after wavelet decomposition is collated as shown in fig. 9 and fig. 10.
And acquiring the original main frequency of the balance step signal according to Hilbert spectrums of the natural modal functions of multiple stages, and taking the main frequency with the same waveform and the largest amplitude ratio in the spectrogram of the balance step signal and the natural modal function as the inertial vibration frequency.
The high-frequency coefficients D1-D5 after wavelet decomposition are high-frequency low-amplitude noise interference components in the signals, and the low-frequency coefficient A5 after the noise interference is removed basically keeps the same waveform and amplitude as the original signals, but obvious burr signals are removed. The analysis of R2 after the wavelet decomposition and Hilbert-Huang transform method is applied to the balance step signal can be known. The high frequency coefficients D6, D7 and the eigenmode function IMF1 have the same waveform, the dominant frequency of which is 379.9Hz, reflecting the inertial frequency of the force measurement system vibrations. The high-frequency coefficient D8 and the second-stage intrinsic mode function IMF2 generate signal jump near 50ms, the action time is short, the action force is large, after the signal jump, the amplitude becomes zero, and the impact load action on a system generated at the moment of shearing a steel wire when the signal is collected is reflected.
According to the embodiment, wavelet transformation and Hilbert-Huang transformation are adopted to perform noise reduction and time-frequency transformation analysis processing on shock tunnel balance signals, different interference components in the balance signals are effectively identified, reliable aerodynamic force results are output, the output aerodynamic force results are linear and stable sudden change signals, linear and stable signals are provided for the follow-up utilization of the inertial characteristic frequency of the fast Fourier transform identification system, in addition, the embodiment can directly identify the inertial frequency characteristics, and the follow-up inertial characteristic frequency screening work is facilitated.
The above embodiments are only exemplary embodiments of the present application, and are not intended to limit the present application, and the protection scope of the present application is defined by the claims. Various modifications and equivalents may be made by those skilled in the art within the spirit and scope of the present application and such modifications and equivalents should also be considered to be within the scope of the present application.
Claims (9)
1. A shock tunnel force measurement balance signal data processing method based on time-frequency transformation is characterized by comprising the following steps:
step 100, constructing a shock tunnel aerodynamic force measurement system of a hypersonic aircraft model, collecting balance step signals of the shock tunnel aerodynamic force measurement system by adopting an SVDC (singular value decomposition direct current) technology, and constructing real aerodynamic force signals output by an ideal step signal simulation balance;
200, performing wavelet decomposition on the balance step signal by adopting a wavelet threshold denoising method to obtain a sub-signal, performing correlation analysis on the sub-signal, and filtering a high-frequency noise signal to obtain an effective characteristic signal;
step 300, performing Hilbert-Huang transformation on the effective characteristic signal after the high-frequency noise interference is filtered, and obtaining a plurality of inherent modal functions and residual components by adopting an empirical mode decomposition method;
step 400, performing Hilbert spectrum analysis on each inherent mode function to obtain corresponding instantaneous frequency, instantaneous amplitude and Hilbert spectrum, and filtering interference signals in the effective characteristic signals to obtain effective aerodynamic signals.
2. The method for processing the signal data of the shock tunnel force measuring balance based on the time-frequency transformation according to claim 1, wherein in step 200, the wavelet threshold denoising method performs approximate decomposition and detail decomposition on the balance step signal to obtain a low-frequency coefficient and a high-frequency coefficient, wherein the low-frequency coefficient is used for showing the trend of the whole balance step signal, the high-frequency coefficient is used for showing the detail components of the whole balance step signal, and the specific implementation manner of performing the wavelet decomposition and filtering the high-frequency noise by the wavelet threshold denoising method is as follows:
step 201, performing multi-level wavelet decomposition on the denoised balance step signal by using one-dimensional discrete wavelet transform to obtain a low-frequency coefficient and a high-frequency coefficient of the balance step signal subjected to the multi-level wavelet decomposition;
step 202, performing fast Fourier transform on the low-frequency coefficient, the high-frequency coefficient and the ideal step signal, and calculating 99% of occupied bandwidth of the low-frequency coefficient, the high-frequency coefficient and the ideal step signal after the fast Fourier transform;
and 203, comparing the 99% occupied bandwidth of the high-frequency coefficient and the low-frequency coefficient with the reference by taking the 99% occupied bandwidth of the ideal step signal as the reference, and filtering the high-frequency coefficient and the low-frequency coefficient which are completely inconsistent with the 99% occupied bandwidth of the ideal step signal.
3. The method for processing the signal data of the shock tunnel force measuring balance based on the time-frequency transformation according to claim 2, which is characterized in that: the method comprises the steps of carrying out fast Fourier transform on low-frequency coefficients, high-frequency coefficients and an ideal step signal to generate a spectrogram, filtering the high-frequency coefficients which exceed 99% of the occupied bandwidth of the ideal step signal, and determining the main frequency of each high-frequency coefficient according to the amplitude and the corresponding frequency of the high-frequency coefficients.
4. The method for processing the signal data of the shock tunnel force measuring balance based on the time-frequency transformation according to claim 2, which is characterized in that: and if the 99% occupied bandwidth of the high-frequency coefficient and the low-frequency coefficient is not intersected with the 99% occupied bandwidth of the ideal step signal, the 99% occupied bandwidth of the high-frequency coefficient and the low-frequency coefficient is completely inconsistent with the 99% occupied bandwidth of the ideal step signal, and the high-frequency coefficient and the low-frequency coefficient which are completely inconsistent are filtered and removed.
5. The method for processing the signal data of the shock tunnel force measuring balance based on the time-frequency transformation according to claim 4, wherein the method comprises the following steps: the 99% occupied bandwidth is that the signal power of the low-frequency coefficient, the high-frequency coefficient and the ideal step signal after fast Fourier transform accounts for 99% of the total signal power.
6. The method for processing the signal data of the shock tunnel force measuring balance based on the time-frequency transformation according to claim 2, which is characterized in that: and performing inverse Fourier transform after filtering the high-frequency coefficient and the low-frequency coefficient which are completely inconsistent, and selecting a proper coefficient of the high-frequency coefficient and the low-frequency coefficient which are completely inconsistent for signal reconstruction to obtain an effective characteristic signal after filtering the high-frequency noise signal.
7. The method for processing the signal data of the shock tunnel force measuring balance based on the time-frequency transform as claimed in claim 1, wherein in step 300, the data processing mode of the effective characteristic signal by using the hilbert-yellow transform is as follows: performing multi-stage processing on the low-frequency coefficient in the effective characteristic signal by using an empirical mode decomposition method of Hilbert-Huang transform to obtain a multi-stage inherent mode function and a multi-stage processed residual component;
wherein the natural mode function of the previous stage in the empirical mode decomposition is a sum of a natural mode function and a residual component of the next stage.
8. The method for processing the signal data of the shock tunnel force measuring balance based on the time-frequency transformation according to claim 3, which is characterized in that: in step 400, Hilbert spectrum analysis is performed on the obtained intrinsic mode functions of each stage by using a Hilbert spectrum analysis method to obtain an instantaneous frequency, an instantaneous amplitude and a Hilbert spectrum of each intrinsic mode function, and interference signals of the intrinsic mode functions are filtered to obtain effective aerodynamic signals, which is implemented by:
acquiring a time domain diagram of a multi-stage natural mode function and a residual component of the low-frequency signal;
acquiring Hilbert spectrums of the multi-level intrinsic mode functions of the low-frequency signals, obtaining instantaneous frequencies of the intrinsic mode functions of each level, and calculating 99% occupied bandwidth of the intrinsic mode functions and residual components of each level;
the remaining components and the natural mode functions that fully fit 99% of the occupied bandwidth of the ideal step signal are retained as valid aerodynamic signals.
9. The method for processing the signal data of the shock tunnel force measuring balance based on the time-frequency transformation according to claim 8, wherein the method comprises the following steps: and acquiring the original main frequency of the balance step signal according to Hilbert spectrums of the natural modal functions of multiple stages, and taking the main frequency with the same waveform and the largest amplitude ratio in the spectrogram of the balance step signal and the natural modal function as the inertial vibration frequency.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111192645.6A CN113970419B (en) | 2021-10-13 | 2021-10-13 | Shock tunnel force measurement balance signal data processing method based on time-frequency transformation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111192645.6A CN113970419B (en) | 2021-10-13 | 2021-10-13 | Shock tunnel force measurement balance signal data processing method based on time-frequency transformation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113970419A true CN113970419A (en) | 2022-01-25 |
| CN113970419B CN113970419B (en) | 2022-05-13 |
Family
ID=79587364
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111192645.6A Active CN113970419B (en) | 2021-10-13 | 2021-10-13 | Shock tunnel force measurement balance signal data processing method based on time-frequency transformation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113970419B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114993606A (en) * | 2022-05-31 | 2022-09-02 | 中国科学院力学研究所 | Wind tunnel test result processing method for unsteady pressure and aerodynamic data |
| CN115795282A (en) * | 2023-01-30 | 2023-03-14 | 武汉工程大学 | Shock tube dynamic pressure reconstruction method and device, electronic equipment and storage medium |
| CN117288418A (en) * | 2023-09-08 | 2023-12-26 | 西南交通大学 | Hypersonic wind tunnel aerodynamic force intelligent identification method based on wavelet decomposition |
| CN117909659A (en) * | 2024-03-20 | 2024-04-19 | 中国空气动力研究与发展中心超高速空气动力研究所 | Method and device for processing aerodynamic force signals of balance in wind tunnel dynamic test of variant aircraft |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104111108A (en) * | 2014-07-16 | 2014-10-22 | 云南电力试验研究院（集团）有限公司电力研究院 | Torsional vibration impact signal characteristic extracting method for rotating mechanism |
| WO2015065873A2 (en) * | 2013-10-28 | 2015-05-07 | University Of Massachusetts | Structural health monitoring of wind turbine blades using wireless acoustic sensing |
| KR101589568B1 (en) * | 2014-08-27 | 2016-01-28 | 국방과학연구소 | Sensor type Wind Experimental Model Setting System |
| CN106371002A (en) * | 2015-07-24 | 2017-02-01 | 国网四川省电力公司眉山供电公司 | Circuit breaker fault diagnosis method based on Hilbert-Huang transform algorithm |
| CN108731945A (en) * | 2018-08-02 | 2018-11-02 | 南昌航空大学 | A kind of extracting method of aircraft engine rotor system fault-signal characteristic information |
| CN109684898A (en) * | 2017-10-18 | 2019-04-26 | 中国航发商用航空发动机有限责任公司 | Aero-engine and its vibration signal blind separating method and device |
| CN110501158A (en) * | 2019-05-20 | 2019-11-26 | 山东理工大学 | Needing machine transmission mechanism vibration signal characteristics extracting method |
| CN111473947A (en) * | 2020-05-07 | 2020-07-31 | 中国科学院力学研究所 | Single-vector dynamic calibration method for pulse wind tunnel force measurement system |
-
2021
- 2021-10-13 CN CN202111192645.6A patent/CN113970419B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015065873A2 (en) * | 2013-10-28 | 2015-05-07 | University Of Massachusetts | Structural health monitoring of wind turbine blades using wireless acoustic sensing |
| CN104111108A (en) * | 2014-07-16 | 2014-10-22 | 云南电力试验研究院（集团）有限公司电力研究院 | Torsional vibration impact signal characteristic extracting method for rotating mechanism |
| KR101589568B1 (en) * | 2014-08-27 | 2016-01-28 | 국방과학연구소 | Sensor type Wind Experimental Model Setting System |
| CN106371002A (en) * | 2015-07-24 | 2017-02-01 | 国网四川省电力公司眉山供电公司 | Circuit breaker fault diagnosis method based on Hilbert-Huang transform algorithm |
| CN109684898A (en) * | 2017-10-18 | 2019-04-26 | 中国航发商用航空发动机有限责任公司 | Aero-engine and its vibration signal blind separating method and device |
| CN108731945A (en) * | 2018-08-02 | 2018-11-02 | 南昌航空大学 | A kind of extracting method of aircraft engine rotor system fault-signal characteristic information |
| CN110501158A (en) * | 2019-05-20 | 2019-11-26 | 山东理工大学 | Needing machine transmission mechanism vibration signal characteristics extracting method |
| CN111473947A (en) * | 2020-05-07 | 2020-07-31 | 中国科学院力学研究所 | Single-vector dynamic calibration method for pulse wind tunnel force measurement system |
Non-Patent Citations (3)
| Title |
|---|
| 张鹏等: "一种基于新阈值函数的小波降噪方法在高速风洞中的应用", 《兵工自动化》 * |
| 汪运鹏等: "基于深度学习技术的激波风洞智能测力系统研究", 《力学学报》 * |
| 汪运鹏等: "长试验时间激波风洞测力技术研究", 《力学学报》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114993606A (en) * | 2022-05-31 | 2022-09-02 | 中国科学院力学研究所 | Wind tunnel test result processing method for unsteady pressure and aerodynamic data |
| CN115795282A (en) * | 2023-01-30 | 2023-03-14 | 武汉工程大学 | Shock tube dynamic pressure reconstruction method and device, electronic equipment and storage medium |
| CN117288418A (en) * | 2023-09-08 | 2023-12-26 | 西南交通大学 | Hypersonic wind tunnel aerodynamic force intelligent identification method based on wavelet decomposition |
| CN117288418B (en) * | 2023-09-08 | 2024-05-07 | 西南交通大学 | Hypersonic wind tunnel aerodynamic force intelligent identification method based on wavelet decomposition |
| CN117909659A (en) * | 2024-03-20 | 2024-04-19 | 中国空气动力研究与发展中心超高速空气动力研究所 | Method and device for processing aerodynamic force signals of balance in wind tunnel dynamic test of variant aircraft |
| CN117909659B (en) * | 2024-03-20 | 2024-05-14 | 中国空气动力研究与发展中心超高速空气动力研究所 | Method and device for processing aerodynamic force signals of balance in wind tunnel dynamic test of variant aircraft |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113970419B (en) | 2022-05-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113970419B (en) | Shock tunnel force measurement balance signal data processing method based on time-frequency transformation | |
| CN113970420B (en) | Deep learning-based shock tunnel force measurement signal frequency domain analysis method | |
| CN102539150B (en) | Self-adaptive failure diagnosis method of rotary mechanical component based on continuous wavelet transformation | |
| CN112101245B (en) | Short-time Fourier transform mechanical impact feature extraction method based on frequency domain window function | |
| CN106771598B (en) | A kind of Adaptive spectra kurtosis signal processing method | |
| CN114199364A (en) | Vibration monitoring system of aircraft engine | |
| CN112326245A (en) | Rolling bearing fault diagnosis method based on variational Hilbert-Huang transform | |
| CN104869519A (en) | Method and system for testing background noise of microphone | |
| Paliwal | Estimation of noise variance from the noisy AR signal and its application in speech enhancement | |
| CN107564512A (en) | Voice activity detection method and device | |
| Zhou et al. | Application of IPSO-MCKD-IVMD-CAF in the compound fault diagnosis of rolling bearing | |
| Melgoza et al. | Comparing radar receiver pulse deinterleaving performance of differing window functions for bandpass FIR filter design | |
| EP2877820B1 (en) | Method of extracting zero crossing data from full spectrum signals | |
| CN110689901B (en) | Voice noise reduction method and device, electronic equipment and readable storage medium | |
| CN108181486A (en) | The processing method and processing device of acceleration signal | |
| KR100727488B1 (en) | Waveform synthesis methods for shock response spectrum over a short time interval, a digital recursive filter for a shock response history and its applications | |
| CN111505375A (en) | Frequency measurement method and device | |
| CN107224284A (en) | The noise detecting method and system of a kind of digital electrocardiosignal | |
| Zaitsev | Elimination on power line interference from ECG signal using combined bidirectional narrow-band notch filter | |
| CN111641411B (en) | Signal acquisition method and system | |
| CN109682958B (en) | Acceleration sensor signal compensation method for thrombelastogram instrument | |
| Yang et al. | A fast method for robust estimation of gearbox instantaneous speed | |
| CN209296808U (en) | A kind of anti-interference different-frequency measure device based on wave cancellation | |
| CN112540386B (en) | Narrow-band anti-interference suppression method | |
| CN107449580B (en) | A kind of the harmonic characteristic parameter extracting method and system of the impulse response signal of power transmission tower |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |