A gust alleviation or flutter suppression test system and method for visual imaging measurements

Abstract

The application provides a gust alleviation or flutter suppression test system and method for visual imaging measurement, and relates to the field of wind tunnel test. The system solves the problems of model damage, limited measurement accuracy, poor safety and inability to measure the whole field caused by installing sensors on the model in traditional wind tunnel test. The system comprises a gust generator, a test model, a support device, a video measurement camera, a light source and a data processing and control system. The method is as follows: installing and debugging the gust generator, the model and the support device; pasting mark points on the model and installing a calibration camera; laying and debugging the light source; after the wind tunnel and the gust generator reach a predetermined state, the camera collects data; after digital image stabilization, motion estimation and noise suppression, the data is fed back to the control system to drive the rudder deflection to achieve load alleviation. The flutter suppression test does not use a gust generator. The application does not need to lay sensors, can improve measurement safety and efficiency, and can realize whole field vibration monitoring of the structure.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of wind tunnel test, in particular to a gust alleviation or buffet suppression test system and method based on visual imaging measurement. BACKGROUND

[0002] The atmospheric environment in which an aircraft flies is not always smooth airflow, but there are various disturbances. "Gust" (also known as "sudden wind") is a form of atmospheric disturbance, which refers to the sudden change of wind speed and direction in a short time. This change is usually caused by weather conditions (such as thunderstorms, fronts, etc.) or topographic effects (such as mountains, buildings, etc.).

[0003] It is inevitable that an aircraft flying in the atmosphere will be affected by gusts. Gusts can suddenly change the airspeed and angle of attack of the aircraft, causing changes in lift distribution, thereby bringing additional aerodynamic loads; gusts can cause the aircraft to pitch, roll and yaw, leading to loss of control of flight attitude; as an external excitation, gusts usually excite the elastic modes of the aircraft structure, causing additional elastic vibration and structural load, reducing the fatigue life of the wing and fuselage structure. Large aspect ratio aircraft wings have smaller wing loads and lower structural elastic frequencies, and are particularly susceptible to gusts. In addition, the buffeting caused by gusts will reduce the ride comfort of the personnel on board, and in severe cases will cause injury to the personnel, and will affect the pilot's control, endangering flight safety. Studies have shown that humans are more sensitive to low-frequency vibrations of about 1 Hz, and when the vertical vibration overload exceeds 0.2g, it is difficult to interpret the instruments; when it exceeds 0.5g and lasts for several minutes, the pilot will feel restless and difficult to control, endangering flight safety. In recent years, aviation accidents and incidents caused by gusts or atmospheric turbulence have occurred repeatedly.

[0004] In order to ensure flight safety and improve passenger comfort, it is necessary to carry out gust research on civil aircraft. Wind tunnel test is one of the effective ways to realize gust load alleviation technology. The wind tunnel gust response and alleviation control test method has been established through years of accumulation of gust design research. However, the previous gust load alleviation technology mainly arranges acceleration sensors and strain gauges at the wing tip and wing root of the model. The overload at the wing tip is fed back through the acceleration sensor, the wing root moment is fed back through the wing root strain gauge, and then the wing tip overload and wing root moment are fed back to the control system. The control system drives the rudder to deflect to redistribute the aerodynamic load, thereby achieving load alleviation. This method of arranging sensors inside or on the surface of the model to feed back gust load information has the following problems: first, the sensor cable arrangement is complex, and the model needs to have a wiring slot inside. This makes the model difficult to be used for other types of tests, and also causes high model cost, long processing cycle, delay of aircraft development process and other problems. Second, the sensors pasted on the surface of the model are easy to fall off or be damaged, especially in high-speed flow field. The wiring is easy to twist and the structure is easy to interfere, which not only affects the model modal measurement accuracy, but also brings safety risks. Third, this method can only obtain discrete point vibration information, and the spatial resolution is limited, which is difficult to reflect the real response state of the whole field structure. Finally, this method relies on single-point discrete time series data, which often cannot accurately identify weak modal components in the context of multi-modal coupling and low signal-to-noise ratio. The frequency extraction speed is slow, and the automation degree is low, which seriously affects the efficiency and reliability of wind tunnel test. SUMMARY

[0005] In order to solve the problems of model damage, limited measurement accuracy, poor safety and inability to measure the whole field in the traditional wind tunnel test caused by installing sensors on the model, the present application provides a gust alleviation or flutter suppression test system based on visual imaging measurement, which comprises:

[0006] A model support device for fixing the model;

[0007] The visual imaging measurement unit comprises a camera A, a camera B and a light source for illuminating the surface of the model, and is used for collecting video data;

[0008] The surface of the model is provided with a mark point for video measurement;

[0009] A data acquisition unit connected with the visual imaging measurement unit, used for receiving and processing video data;

[0010] A rudder control system connected with the data acquisition unit and the rudder of the model; the data acquisition unit sends the processed vibration displacement and frequency characteristics to the rudder control system;

[0011] The rudder control system drives the rudder to deflect according to the received vibration displacement and frequency characteristics, so as to realize gust load alleviation or flutter suppression.

[0012] A visual imaging measurement method for gust mitigation or flutter suppression is also provided, implemented using the test system described above, including:

[0013] S1. Install and debug the gust generator, model and model support device;

[0014] S2. Mark points are placed on the model surface, and cameras A and B are installed and calibrated.

[0015] S3. Adjust the light source to ensure that it meets the acquisition requirements of camera A and camera B;

[0016] S4. Start the wind tunnel. After the incoming flow stabilizes, start the gust generator and control camera A and camera B to collect data and obtain video data.

[0017] S5. Transmit the video data to the data acquisition unit and process the video data to obtain vibration displacement and frequency characteristics;

[0018] S5 specifically includes:

[0019] S51. Process the video data, eliminate camera shake through motion estimation and compensation, and output a stable video sequence to achieve digital image stabilization.

[0020] S52. Based on a stable video sequence, construct a monocular video analytical signal and extract the phase from the analytical signal. Solve for the phase difference signal and establish a model between the phase difference and the vibration displacement signal.

[0021] S53. Perform noise reduction and frequency domain transformation on the vibration displacement signal, extract the vibration frequency features of the model, and identify abnormal states according to the preset threshold.

[0022] S6. The processed vibration displacement and frequency characteristics are sent to the rudder control system to drive the rudder deflection, thereby reducing gust load or suppressing flutter.

[0023] Furthermore, in S51, the formula for eliminating camera shake through motion estimation and compensation to output a stable video sequence is:

[0024]

[0025] In the formula, S represents the scaling factor, which controls the degree of scaling of the image during the transformation process. When S is greater than 1, the image is enlarged; when S is less than 1, the image is shrunk. Indicates the rotation angle, used to control the rotation of the image, in radians; This represents the amount of translation in the x-direction, that is, the distance the image moves in the horizontal direction; This represents the amount of translation in the y-direction, which is the distance the image moves in the vertical direction; with is the coordinate of the point on the original image; with is the coordinate of the corresponding point after transformation.

[0026] Further, in S52, constructing monocular video analysis signal and extracting phase from the analysis signal, solving phase difference signal specifically includes:

[0027] S521, constructing analysis signal I(x,y,t) with monocular video frame gray value I(x,y,t) as real part and its first derivative as imaginary part, wherein the derivative term is calculated by Sobel operator convolution: DA

[0028]

[0029] In the formula: I(x,y,t) represents the gray value of the video at spatial coordinates (x,y) and time t, i is the imaginary unit, m0 is the measured model displacement number, is the wavelength;

[0030] S522, obtaining pixel phase information by phase solving on the analysis signal:

[0031]

[0032] In the formula, is the pixel phase information;

[0033] S523, calculating the difference between phases at different times:

[0034]

[0035] In the formula, δ is the displacement signal, indicating that the phase difference is in linear proportion to the displacement, and the vibration of the model can be analyzed from the phase difference, is the difference between phases at different times, is the initial time, is the time interval.

[0036] Further, in S53, denoising and frequency domain transformation are performed on the vibration displacement signal to extract the vibration frequency characteristics of the model, specifically including:

[0037] S531, according to the data length of the detail wavelet coefficient, taking the absolute value of the numerical value of each discrete point in ascending order to obtain an ordered sequence and a corresponding square sequence, and generating a deviation sequence of the current value and the historical mean value:

[0038]

[0039] In the formula, ​represents the degree of deviation of the current value from the historical average, a square sequence corresponding to the ordered sequence obtained by arranging the absolute values of the numerical values of each discrete point in ascending order, is the measurement average;

[0040] S532, construct a risk function:

[0041]

[0042] where S k reflects the cumulative effect of the martingale difference, and N is the data length, is the final risk threshold, is the first prior term, is the second prior term;

[0043] The processed detail wavelet coefficients and the wavelet coefficients composed of the processed detail wavelet coefficients and the approximation coefficients are combined with the mixed components after noise reduction based on the cumulative martingale difference threshold rule and the useful components obtained by adaptive optimization decomposition to obtain the final phase difference signal after noise reduction.

[0044] S533, perform fast Fourier transform on the phase difference signal to extract the vibration frequency characteristics of the model:

[0045]

[0046] wherein, is a frequency domain function, is a rotation factor, is a time domain phase difference signal, is the total integration time, is the frequency, t is the time variable;

[0047] By peak detection on the transformation result, the main aircraft model vibration frequency f peak is determined.

[0048] The beneficial effects of the present application are:

[0049] 1. Ensure measurement safety and accuracy, reduce error interference. No need to arrange or paste sensors on the surface of the test model and lay cables, on the one hand, avoid the disturbance of sensor cables to the wind tunnel air flow field, on the other hand, eliminate the measurement error caused by the additional mass of the sensor, especially suitable for safe measurement in flexible large structure and high wind speed aerodynamic environment, effectively improve the credibility of the measurement results.

[0050] 2. Reduce testing costs, shorten testing cycles, and improve model utilization. The test model does not require slotting and wiring, simplifying the model processing flow, which not only reduces model costs but also shortens the manufacturing cycle. At the same time, the model does not require structural modifications due to sensor placement, and can be flexibly used for other types of wind tunnel tests, enhancing the model's reusability and indirectly reducing the overall testing investment.

[0051] 3. Overcoming spatial resolution limitations to achieve full-field visual monitoring. Based on ordinary industrial cameras, video measurement can achieve large field of view and high spatial resolution measurement, and supports arbitrary measurement point deployment. Compared with the limitations of traditional point sensors that can only acquire vibration information at discrete points, it can visualize and dynamically reconstruct the vibration response of the structure throughout the entire field, and more comprehensively reflect the true response state of the structure.

[0052] 4. Enhance adaptability to complex scenarios and data processing capabilities, improving experimental efficiency and reliability. Video measurement methods can be combined with advanced algorithms such as phase derivative optical flow, image texture enhancement, modal decoupling recognition, and motion magnification rendering to reconstruct high-precision dynamic displacement and modal features from two-dimensional images, achieving efficient mapping from "image data" to "structural state." Simultaneously, video data contains complete temporal and spatial information, enabling simultaneous frequency extraction, mode shape recognition, and energy distribution analysis. In scenarios with high modal complexity, severe spectral aliasing, or difficulty in detecting weak modes, it exhibits stronger adaptability and robustness than traditional methods, effectively improving wind tunnel testing efficiency and modal recognition reliability.

[0053] 5. Possesses the ability to identify and warn of abnormal states, ensuring the controllability of the test process. By extracting vibration frequency characteristics through fast Fourier transform of the phase difference signal, and setting frequency deviation thresholds and displacement amplitude thresholds as abnormal judgment criteria, an early warning mechanism can be triggered when the monitored value exceeds the threshold, realizing rapid identification and feedback of abnormal states such as aircraft gust reduction or flutter suppression, ensuring the safety and controllability of the test process. Attached Figure Description

[0054] Figure 1 This is a system diagram of the present invention;

[0055] Figure 2 This is a flowchart of the method of the present invention;

[0056] Figure 3 The time-displacement diagram of the model measured in this invention;

[0057] Figure 4 This is the frequency-amplitude diagram measured by the present invention.

[0058] 1-Model support device, 2-Marker point, 3-Model, 4-Camera A, 5-Camera B, 6-Light source, 7-Rudder surface, 8-Rudder surface control system, 9-Data acquisition unit. DETAILED DESCRIPTION

[0059] The technical solutions of the present application are further described below in conjunction with the examples, but are not limited thereto. Any modification or equivalent replacement of the technical solutions of the present application without departing from the spirit and scope of the present application shall be covered in the protection scope of the present application. The process equipment or device not specifically mentioned in the following examples is the conventional equipment or device in the art. If not specifically mentioned, the raw materials used in the examples of the present application are commercially available. If not specifically mentioned, the technical means used in the examples of the present application is the conventional means known to those skilled in the art.

[0060] Example 1, in conjunction with Figure 1 The present example is a gust alleviation or flutter suppression test system for visual imaging measurement, comprising:

[0061] A model support device 1 is used to fix the model 3;

[0062] The visual imaging measurement unit comprises a camera A 4, a camera B 5 and a light source 6 for illuminating the surface of the model 3, for collecting video data;

[0063] The surface of the model 3 is provided with a mark point 2 for video measurement;

[0064] A data acquisition unit 9 is connected with the visual imaging measurement unit, for receiving and processing video data;

[0065] A control system 8 of the control surface is connected with the data acquisition unit 9 and the control surface 7 of the model 3; the data acquisition unit 9 sends the processed vibration displacement and frequency characteristics to the control system 8 of the control surface;

[0066] The control system 8 of the control surface drives the deflection of the control surface 7 according to the received vibration displacement and frequency characteristics, so as to achieve gust load alleviation or flutter suppression.

[0067] Specifically, the system shoots the vibration video of the mark point on the surface of the model by the camera in a non-contact manner, calculates the vibration displacement and frequency characteristics of the model in real time through the data acquisition unit, and feeds back to the control system of the control surface according to the vibration displacement and frequency characteristics, so as to drive the active deflection of the control surface to redistribute the aerodynamic load, so as to achieve the purpose of suppressing vibration, alleviating gust load or suppressing flutter.

[0068] Example 2, in conjunction with Figure 2 The present example is a gust alleviation or flutter suppression test method for visual imaging measurement, which is realized by using the test system as described in Example 1, comprising:

[0069] S1, installing and debugging the gust generator, the model 3 and the model support device 1;

[0070] S2, arranging mark points 2 on the surface of model 3, installing and calibrating camera A4 and camera B5;

[0071] S3, debugging light source 6, ensuring that the acquisition requirements of camera A4 and camera B5 are met;

[0072] S4, starting the wind tunnel, after the incoming flow is stable, starting the gust generator, and controlling camera A4 and camera B5 to perform data acquisition to obtain video data;

[0073] S5, transmitting the video data to the data acquisition unit 9, and processing the video data to obtain the vibration displacement and frequency characteristics;

[0074] S5 specifically includes:

[0075] S51, processing the video data, eliminating camera shaking through motion estimation and compensation, outputting a stable video sequence, and realizing digital image stabilization;

[0076] S52, based on the stable video sequence, constructing a monocular video analysis signal and extracting the phase from the analysis signal, solving the phase difference signal, and establishing a model between the phase difference and the vibration displacement signal;

[0077] S53, denoising and frequency domain transforming the vibration displacement signal, extracting the vibration frequency characteristics of the model, and identifying abnormal states according to a preset threshold;

[0078] S6, sending the vibration displacement and frequency characteristics obtained by processing to the rudder control system 8, driving the rudder 7 to deflect, and realizing gust load mitigation or flutter suppression.

[0079] Specifically, Figure 3 The displacement curve in the time domain directly reflects the change law of the vibration amplitude, and Figure 4 The amplitude spectrum in the frequency domain accurately identifies the dominant vibration frequency of the system, and the two together verify that the measurement scheme can provide accurate and reliable displacement and frequency input parameters for gust mitigation and flutter suppression control.

[0080] In S51, the formula for eliminating camera shaking through motion estimation and compensation and outputting a stable video sequence is:

[0081]

[0082] In the formula, S represents a scaling factor, which is used to control the scaling degree of the image in the transformation process, and S is greater than 1 when the image is enlarged, and less than 1 when the image is reduced; represents the rotation angle, which is used to control the rotation of the image, and is in radians; represents the translation in the x direction, that is, the movement distance of the image in the horizontal direction; represents the translation amount in the y direction, that is, the moving distance of the image in the vertical direction; with is the coordinate of the point on the original image; with is the coordinate of the corresponding point obtained after transformation.

[0083] Specifically, in the digital stabilization process, first, the motion information between adjacent frames in the video sequence is analyzed to determine the translation, rotation and other motion parameters of the camera in the image or video. By detecting and matching the feature points in the image, the motion is then estimated according to the displacement of the feature points in the consecutive frames. The result of motion estimation is usually represented in the form of a transformation matrix, such as a translation matrix, a rotation matrix or an affine transformation matrix, which reflects the relative motion of the camera during shooting.

[0084] Further, the motion estimation result is used to correct each frame of the image to remove the image offset caused by the camera motion. For complex camera motion, affine transformation can handle translation, rotation, scaling and other deformations at the same time. Motion compensation performs affine transformation on the image to maintain the geometric shape of the image.

[0085] Finally, the image after motion compensation is repaired to fill in the blank area or black border generated during the compensation process. This process usually involves image cropping, interpolation and splicing operations to make the image as natural as possible, reduce visual discomfort caused by compensation, and maintain the quality of the image.

[0086] In S52, the monocular video analysis signal is constructed and the phase is extracted from the analysis signal to solve the phase difference signal, which specifically includes:

[0087] S521, the monochrome video frame gray value I(x, y, t) is taken as the real part, and its first derivative is taken as the imaginary part to construct the analysis signal I DA (x, y, t), wherein the derivative term is calculated by Sobel operator convolution:

[0088]

[0089] In the formula: I(x, y, t) represents the gray value of the video at spatial coordinates (x, y) and time t, i is the imaginary unit, m0 is the measured model displacement number, is the wavelength;

[0090] S522, the phase of the analysis signal is solved to obtain the pixel phase information:

[0091]

[0092] wherein, is the pixel phase information;

[0093] S523, calculate the difference between phases at different times:

[0094]

[0095] wherein δ is a displacement signal, indicating that the phase difference is in linear proportion to the displacement, and the vibration of the aircraft model can be analyzed from the phase difference, is the difference between phases at different times, is the initial time, is the time interval.

[0096] Specifically, this step is based on the brightness change characteristics of the monocular image to extract the phase difference signal of the structural micro-vibration.

[0097] In S53, the vibration displacement signal is denoised and frequency domain transformed to extract the vibration frequency characteristics of the model, which specifically includes:

[0098] S531, according to the data length of the detail wavelet coefficient of each discrete point, the absolute value is sorted in ascending order to obtain an ordered sequence and a corresponding square sequence, and the deviation degree sequence of the current value and the historical mean value is generated:

[0099]

[0100] wherein, represents the deviation degree of the current value and the historical mean value, is the square sequence corresponding to the ordered sequence obtained by sorting the absolute value of each discrete point in ascending order, is the measurement mean value;

[0101] S532, construct a risk function:

[0102]

[0103] wherein, S k reflects the cumulative effect of martingale difference, N is the data length, is the final risk threshold, is the first prior term, is the second prior term;

[0104] The soft threshold function is used to process the detail wavelet coefficient, and the processed wavelet coefficient composed of the detail wavelet coefficient and the approximation coefficient is combined with the useful component obtained by adaptive optimization decomposition to obtain the final denoised phase difference signal;

[0105] S533, perform fast Fourier transform on the phase difference signal to extract the vibration frequency characteristics of the model:

[0106]

[0107] wherein, is a frequency domain function, is a rotation factor, is a time domain phase difference signal, is the total integration time, is a frequency, t is a time argument;

[0108] By peak detection on the transform result, the structural main aircraft model vibration frequency f peak is determined, which provides a frequency domain characteristic parameter for vibration modal analysis.

[0109] Specifically, a frequency deviation threshold ε f is set, and a displacement amplitude threshold ε s is set as an abnormality determination criterion, and when the monitoring value satisfies |f-f ref |>ε f or s>ε s , a pre-warning mechanism is triggered, wherein f ref is a reference frequency, realizing rapid identification and feedback on aircraft gust alleviation or flutter suppression abnormal state.

[0110] Then, the half-power method (half-power bandwidth method) is used to identify the damping parameter, and the core formula is based on the amplitude-frequency characteristics of the frequency response function, and the damping ratio is ;

[0111] In the formula: ζ is the modal damping ratio of the structure (dimensionless, usually expressed in percentage, such as ζ=0.02, which is 2%); Δf is the frequency bandwidth (unit: Hz) corresponding to the peak amplitude times of the amplitude-frequency curve of the frequency response function, also known as the "half-power bandwidth"; is the natural frequency (unit: Hz) of the modal, that is, the frequency corresponding to the peak of the amplitude-frequency curve of the frequency response function.

Claims

1. A gust alleviation or flutter suppression test method of visual imaging measurement, implemented by a gust alleviation or flutter suppression test system of visual imaging measurement, the system comprising: a model support device (1) for fixing a model (3) ; a visual imaging measurement unit comprising a camera A (4), a camera B (5) and a light source (6) for illuminating the surface of the model (3), for collecting video data; the surface of the model (3) is provided with mark points (2) for video measurement; a data acquisition unit (9) connected with the visual imaging measurement unit, for receiving and processing video data; a control system (8) of a control surface connected with the data acquisition unit (9) and the control surface (7) of the model (3) ; the data acquisition unit (9) sends the processed vibration displacement and frequency characteristics to the control system (8) of the control surface; the control system (8) of the control surface drives the control surface (7) to deflect according to the received vibration displacement and frequency characteristics, so as to realize gust load alleviation or flutter suppression; characterized in that the method comprises: S1, installing and debugging the gust generator, the model (3) and the model support device (1) ; S2, arranging mark points (2) on the surface of the model (3), installing and calibrating the camera A (4) and the camera B (5) ; S3, debugging the light source (6) to ensure that the acquisition requirements of the camera A (4) and the camera B (5) are met; S4, starting the wind tunnel, after the flow is stable, starting the gust generator, and controlling the camera A (4) and the camera B (5) to collect data, to obtain video data; S5, transmitting the video data to the data acquisition unit (9), and processing the video data to obtain vibration displacement and frequency characteristics; S5 specifically comprises: S51, processing the video data, eliminating camera shaking through motion estimation and compensation, outputting stable video sequences, and realizing digital image stabilization; S52, based on the stable video sequences, constructing monocular video analysis signals and extracting phases from the analysis signals, solving phase difference signals, and establishing a model between the phase difference and the vibration displacement signal; S53, denoising and frequency domain transforming the vibration displacement signal, extracting the vibration frequency characteristics of the model, and identifying abnormal states according to a preset threshold; S6, sending the processed vibration displacement and frequency characteristics to the control system (8) of the control surface, driving the control surface (7) to deflect, and realizing gust load alleviation or flutter suppression. In S51, the formula for eliminating camera shaking through motion estimation and compensation and outputting stable video sequences is:

2. A gust alleviation or shimmy suppression test method for visual imaging measurements according to claim 1, characterized in that, In S52, constructing monocular video analysis signals and extracting phases from the analysis signals, and solving phase difference signals specifically comprises: ； In the formula, S represents a scaling factor, used to control the scaling degree of the image in the transformation process, and S is greater than 1, the image is enlarged, and less than 1, the image is reduced; represents a rotation angle, used to control the rotation of the image, in radians; represents the translation amount in the x direction, that is, the moving distance of the image in the horizontal direction; represents the translation amount in the y direction, that is, the moving distance of the image in the vertical direction; and is the coordinate of the point on the original image; and is the coordinate of the corresponding point obtained after the transformation.

3. A gust alleviation or shimmy suppression test method for visual imaging measurements according to claim 2, characterized in that, S522, obtaining pixel phase information by solving the phase of the analysis signal: S521, construct an analytic signal I(x, y, t) with the monocular video frame gray value I(x, y, t) as the real part and its first derivative as the imaginary part DA (x, y, t), wherein the derivative term is calculated by Sobel operator convolution: ； where I(x, y, t) represents the gray value of the video at spatial coordinates (x, y) and time t, i is the imaginary unit, m0is the measured model displacement number, is the wavelength; S523, calculating the difference between phases at different times: ； wherein, is the pixel phase information; 4.A gust alleviation or flutter suppression test method of visual imaging measurement according to claim 3, characterized in that ； Wherein, the displacement signal δ indicates that the phase difference is linearly proportional to the displacement, and the phase difference can be analyzed to analyze the vibration of the aircraft model, is the difference between the phases at different times, is the initial time, is the time interval. in S53, denoising and frequency domain transforming the vibration displacement signal, and extracting the vibration frequency characteristics of the model specifically comprises: ​ S531, according to the data length of the detail wavelet coefficient of each discrete point, the absolute value is arranged in ascending order to obtain an ordered sequence and a corresponding square sequence, and a deviation sequence of the current value and the historical mean value is generated: ； wherein, represents the degree of deviation of the current value from the historical average value, is the square sequence corresponding to the ordered sequence obtained by arranging the absolute values of the numerical values of each discrete point in ascending order, is the measurement average value; S532, construct a risk function: ； wherein S k reflecting the cumulative effect of martingale difference, N is the data length, is the final risk threshold, is the first prior term, is the second prior term; The soft threshold function is used to process the detail wavelet coefficient, and the processed wavelet coefficient composed of the approximate coefficient is combined with the mixed component after denoising based on the cumulative martingale difference threshold rule and the useful component obtained by adaptive optimization decomposition to obtain the final denoised phase difference signal; S533, the phase difference signal is subjected to fast Fourier transform, and the vibration frequency characteristics of the model are extracted: ； wherein is a frequency domain function, is a rotation factor, is a time domain phase difference signal, is a total integration time, is a frequency, t is a time argument; By peak detection of the transform result, the structural main airplane model vibration frequency f is determined peak to provide frequency domain characteristic parameters for vibration modal analysis.

Images