Machine vision-based structural vibration and strain tracking analysis method and system

Abstract

The present application belongs to the technical field of structural vibration analysis, and specifically discloses a structural vibration and strain tracking analysis method and system based on machine vision. According to the obtained high-precision multi-structure vibration target synchronous structural analysis result, in the multi-point synchronous structural vibration and strain tracking analysis application scene, compared with the installation requirements of multiple laser displacement sensor devices, the detection unit of a single millimeter wave radar can be further reduced in the process of structural vibration and strain tracking analysis. The structural vibration and strain tracking analysis system can conveniently and reliably meet the multi-point synchronous vibration test demand and the vibration displacement structural vibration and strain tracking analysis precision requirement in engineering practice, so that the whole method improves the analysis speed in the tracking analysis process, and the accuracy of structural vibration and strain tracking analysis is also more accurate.

Description

Technical Field

[0001] This invention belongs to the field of structural vibration analysis technology, specifically relating to a method and system for structural vibration and strain tracking analysis based on machine vision. Background Technology

[0002] Structural vibration analysis methods require a detailed discussion of all aspects of the structure to gain a comprehensive understanding before drawing conclusions. This embodies the principle of not only knowing what happens but also why it happens. This paper discusses linear vibration analysis methods for structures, including analysis methods for free and forced vibrations of single-degree-of-freedom systems, modal analysis methods for natural vibrations and responses of two-degree-of-freedom and multi-degree-of-freedom systems, modal analysis methods for free vibrations and responses of elastic bodies, and approximate vibration analysis methods such as the Rayleigh energy method and Lütz analysis.

[0003] However, common analytical methods require direct contact analysis with the equipment during the analysis process, and the unit precision in the analysis and detection process is low, which in turn results in low detection accuracy. Summary of the Invention

[0004] The purpose of this invention is to overcome the above-mentioned shortcomings and provide a method and system for structural vibration and strain tracking analysis based on machine vision.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] In a first aspect, the present invention provides a machine vision-based method for structural vibration and strain tracking analysis, comprising:

[0007] The I / Q signal parameters of the multi-scanning periodic baseband that the millimeter-wave radar can sense are obtained. The spectrum analysis of the I / Q signal parameters of the multi-scanning periodic baseband is performed to obtain the peak distribution of the amplitude spectrum. The range image parameters and the target positioning parameters are obtained based on the peak distribution of the amplitude spectrum.

[0008] The peak distribution of the amplitude spectrum is searched and discriminated to obtain all the measured targets in the range image parameters. The frequency deviation compensation correction method based on Fourier series fitting is used to process the measured targets, thereby suppressing the coupling interference of neighboring components.

[0009] The processed data is searched for and located using time-frequency ridge peaks to obtain individual target components;

[0010] A phase estimation algorithm based on approximate maximum likelihood estimation is used to process each target component, resulting in an estimated time series of the interferometric phase evolution for each target.

[0011] Based on the estimated time series of the interference phase evolution for each target, the structural vibration parameters of each target are inverted and calculated.

[0012] Based on the structural vibration parameters, structural strain analysis is performed to obtain the corresponding linear vibration characteristics of the structure.

[0013] Based on the linear vibration characteristics of the structure, the logical relationship between the relative distance between the corresponding vibrating structure and the millimeter-wave radar during time period t is obtained.

[0014] Acquire the echo signal generated when the millimeter-wave radar first detects the corresponding structure and the instantaneous phase of the echo signal within that period;

[0015] Based on the linear frequency modulation form of the triangular wave, the corresponding structural strain value detected by the millimeter-wave radar is obtained from the echo signal and the instantaneous phase.

[0016] A further improvement of this invention lies in the following specific method for Fourier series approximation fitting:

[0017]

[0018] Where L and F0 are the order and fundamental frequency of the Fourier series fitting of the structural vibration, respectively, and are used to control the bandwidth of the signal parameter decomposition generated during the structural vibration process. A m，i f m，i and θ m，i Let be the amplitude, frequency, and phase of the m-th component signal parameters in the i-th sweep cycle, respectively. Due to the modulation of the vibration motion, It is a nonlinear time-varying frequency. For a rough estimate of the frequency of this component, the model parameters of the structural vibration are... and It is obtained through joint least squares optimization estimation of multiple components.

[0019] A further improvement of this invention lies in the following specific method for processing each target component using a phase estimation algorithm based on approximate maximum likelihood estimation:

[0020]

[0021] Where arg[] is the complex phase operation, S m (iT+nT s Let T be the discrete signal parameter of the m-th component during the i-th sweep frequency period. s Where N is the sampling time interval, and N is the total number of discrete points of the signal parameters. This is the estimated beat frequency of the m-th component.

[0022] A further improvement of this invention lies in the following specific method for performing structural strain analysis:

[0023]

[0024] Where f0 is the carrier frequency, λ is the frequency modulation slope, and B and These are the effective bandwidth and effective time width during analysis, respectively. Let be the structural linear vibration characteristics of the millimeter-wave radar signal in the k-th period.

[0025] A further improvement of this invention lies in the following logical relationship between the relative distance between the corresponding vibrating structure and the millimeter-wave radar during time period t:

[0026] R(t) = R0 + v i c+f d

[0027] Where R0 is the initial distance between the corresponding vibrating structure and the millimeter-wave radar, c is the propagation speed of electromagnetic waves in the air, and v i This represents the transmission rate of the millimeter-wave radar in a stationary state.

[0028] A further improvement of this invention is that the echo signal generated when the millimeter-wave radar detects the corresponding structure for the first time is:

[0029]

[0030] Where B is the effective bandwidth during analysis, and τ(t) is the instantaneous echo delay of the corresponding structure in time period t. This is the reflection coefficient.

[0031] A further improvement of this invention is that the instantaneous phase of the echo signal within this period is:

[0032]

[0033] Where λ is the frequency modulation slope. This refers to the effective timeframe during analysis.

[0034] A further improvement of this invention lies in the following method for calculating the corresponding structural strain values ​​detected by millimeter-wave radar:

[0035]

[0036] in, The echo signal generated when the millimeter-wave radar first detects the corresponding structure. Let be the instantaneous phase of the echo signal within the period, B be the effective bandwidth during analysis, τ(t) be the instantaneous echo delay of the corresponding structure in time period t, and λ be the frequency modulation slope. Where c is the effective time span for analysis, and c is the propagation speed of electromagnetic waves in the air.

[0037] Secondly, the present invention provides a machine vision-based structural vibration and strain tracking analysis system, comprising:

[0038] The baseband signal parameter processing module is used to acquire the I / Q signal parameters of the multi-scanning period baseband that can be sensed by the millimeter-wave radar, perform spectral analysis on the I / Q signal parameters of the multi-scanning period baseband to obtain the peak distribution of the amplitude spectrum, and obtain the range image parameters and target positioning parameters based on the peak distribution of the amplitude spectrum.

[0039] The target processing module is used to search and identify the peak distribution of the amplitude spectrum to obtain all the targets under test in the range image parameters. It uses a frequency deviation compensation correction method based on Fourier series fitting to process the targets under test, thereby suppressing the coupling interference of neighboring components.

[0040] The target component acquisition module is used to search for and locate the time-frequency ridge peaks in the processed data to obtain individual target components.

[0041] The phase interferometry module is used to process each target component using a phase estimation algorithm with approximate maximum likelihood estimation, and obtain the time series of the estimated interferometric phase evolution for each target.

[0042] The vibration parameter inversion module is used to invert and calculate the structural vibration parameters of each target based on the estimated time series of the interference phase evolution for each target.

[0043] The structural linear vibration characteristic acquisition module is used to perform structural strain analysis based on structural vibration parameters and obtain the corresponding structural linear vibration characteristics.

[0044] The logical relationship acquisition module is used to obtain the logical relationship between the relative distance between the vibrating structure and the millimeter-wave radar during time period t, based on the linear vibration characteristics of the structure.

[0045] The echo phase acquisition module acquires the echo signal generated when the millimeter-wave radar detects the corresponding structure for the first time and the instantaneous phase of the echo signal within that period.

[0046] The structural strain value acquisition module is used to obtain the corresponding structural strain value detected by millimeter-wave radar based on the echo signal and instantaneous phase according to the linear frequency modulation form of the triangular wave.

[0047] A further improvement of this invention is that a structural vibration sensor is installed at the detection point to collect digital signals of structural vibration, obtain comparative values ​​of vibration detection, and use time-domain analysis to analyze the effectiveness and accuracy of structural vibration detection.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] Based on the high-precision synchronous structural analysis results of multiple structural vibration targets, this invention, in the application scenario of multi-point synchronous structural vibration and strain tracking analysis, compared with the requirement of installing multiple laser displacement sensor devices, allows a single millimeter-wave radar to further reduce the unit of detection during the structural vibration and strain tracking analysis process. The structural vibration and strain tracking analysis system can conveniently and reliably meet the needs of multi-point synchronous vibration testing and the accuracy requirements of vibration displacement structural vibration and strain tracking analysis in engineering practice. As a result, the analysis speed is improved during the tracking analysis process, and the accuracy of structural vibration and strain tracking analysis is also more accurate. Attached Figure Description

[0050] Figure 1 This is a flowchart of the present invention;

[0051] Figure 2 This is a system diagram of the present invention;

[0052] Figure 3 This is a flowchart of an embodiment of the present invention. Detailed Implementation

[0053] To further understand the content of this invention, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments are merely illustrative and not limiting of the invention.

[0054] See Figure 1 This invention provides a machine vision-based method for structural vibration and strain tracking analysis, including:

[0055] S1. Obtain the I / Q signal parameters of the multi-scan frequency periodic baseband that the millimeter-wave radar can sense. Perform spectrum analysis on the I / Q signal parameters of the multi-scan frequency periodic baseband to obtain the peak distribution of the amplitude spectrum. Obtain the range image parameters and the target positioning parameters based on the peak distribution of the amplitude spectrum.

[0056] S2 searches and identifies the peak distribution of the amplitude spectrum to obtain all the measured targets in the range image parameters. The measured targets are then processed using a frequency deviation compensation correction method based on Fourier series fitting, thereby suppressing the coupling interference of neighboring components.

[0057] S3 performs time-frequency ridge peak search and location on the processed data to obtain individual target components.

[0058] S4. The phase estimation algorithm based on approximate maximum likelihood estimation is used to process each target component to obtain the time series of the estimated interferometric phase evolution for each target.

[0059] S5. Based on the estimated time series of the interference phase evolution of each target, the structural vibration parameters of each target are inverted and calculated.

[0060] S6. Based on the structural vibration parameters, perform structural strain analysis to obtain the corresponding linear vibration characteristics of the structure.

[0061] S7. Based on the linear vibration characteristics of the structure, the logical relationship between the relative distance between the vibrating structure and the millimeter-wave radar during time period t is obtained.

[0062] S8, acquire the echo signal generated when the millimeter-wave radar detects the corresponding structure for the first time and the instantaneous phase of the echo signal within that period.

[0063] S9, based on the linear frequency modulation form of the triangular wave, obtains the corresponding structural strain value detected by the millimeter-wave radar from the echo signal and the instantaneous phase.

[0064] See Figure 2 This invention provides a machine vision-based structural vibration and strain tracking analysis system, comprising:

[0065] The baseband signal parameter processing module is used to acquire the I / Q signal parameters of the multi-scanning period baseband that can be sensed by the millimeter-wave radar, perform spectral analysis on the I / Q signal parameters of the multi-scanning period baseband to obtain the peak distribution of the amplitude spectrum, and obtain the range image parameters and target positioning parameters based on the peak distribution of the amplitude spectrum.

[0066] The target processing module is used to search and identify the peak distribution of the amplitude spectrum to obtain all the targets under test in the range image parameters. It uses a frequency deviation compensation correction method based on Fourier series fitting to process the targets under test, thereby suppressing the coupling interference of neighboring components.

[0067] The target component acquisition module is used to search for and locate the time-frequency ridge peaks in the processed data to obtain individual target components.

[0068] The phase interferometry module is used to process each target component using a phase estimation algorithm with approximate maximum likelihood estimation, and obtain the time series of the estimated interferometric phase evolution for each target.

[0069] The vibration parameter inversion module is used to invert and calculate the structural vibration parameters of each target based on the estimated time series of the interference phase evolution for each target.

[0070] The structural linear vibration characteristic acquisition module is used to perform structural strain analysis based on structural vibration parameters and obtain the corresponding structural linear vibration characteristics.

[0071] The logical relationship acquisition module is used to obtain the logical relationship between the relative distance between the vibrating structure and the millimeter-wave radar during time period t, based on the linear vibration characteristics of the structure.

[0072] The echo phase acquisition module acquires the echo signal generated when the millimeter-wave radar detects the corresponding structure for the first time and the instantaneous phase of the echo signal within that period.

[0073] The structural strain value acquisition module is used to obtain the corresponding structural strain value detected by millimeter-wave radar based on the echo signal and instantaneous phase according to the linear frequency modulation form of the triangular wave.

[0074] Example:

[0075] See Figure 3 A machine vision-based method for structural vibration and strain tracking analysis includes the following steps:

[0076] S1: First, a millimeter-wave radar can be installed to obtain the I / Q signal parameters of the multi-scan frequency periodic baseband that the millimeter-wave radar can sense. Then, the spectrum of the multi-scan frequency periodic baseband signal parameters can be analyzed. Based on the peak distribution of the amplitude spectrum obtained after analysis and detection, the range image parameters and the target positioning parameters can be obtained.

[0077] The structure exists in different states under different stress conditions: the natural state of the structure without external forces, the static state of the structure under dead load, and the vibration state. The vibration state includes deformation in three directions: the width direction, the length direction, and the perpendicular direction.

[0078] In the process of detecting structural vibration, the single-cycle detection signal parameters of millimeter-wave radar are affected by clutter and noise in the external environment, which can cause deviations in the detection results. Therefore, by sending millimeter-wave radar signal parameters for multiple cycles, the echo signal parameters of the corresponding cycles near the structure can be obtained, thereby increasing the signal-to-noise ratio and signal-to-clutter ratio of the signal parameters, completing the detection and reducing the detection error.

[0079] S2: By peak search and discrimination, it can be determined from the distance image parameters whether there are multiple targets being measured. After the initial estimation of the beat frequency of the components, in order to avoid the coupling interference generated by the components of the neighboring target in the multi-target measurement scenario, the machine vision-based structural vibration and strain tracking analysis method in this application uses a frequency deviation compensation correction method based on Fourier series fitting to suppress the coupling interference of the neighboring components.

[0080] Taking the I-channel signal parameters as an example, its signal parameter model can be written as:

[0081]

[0082] Among them, A m，i f m，i and θ m，i Let be the amplitude, frequency, and phase of the m-th component signal parameters in the i-th sweep cycle, respectively. Due to the modulation of the vibration motion, It is often a nonlinear time-varying frequency. This provides a rough estimate of the frequency of the component.

[0083] S3: By searching and locating the peak of the time-frequency ridge, the time-varying amplitude envelope caused by the frequency estimation error in the above equation is obtained, and the Fourier series is used to approximate and fit it as follows:

[0084]

[0085] Where L and F0 are the order and fundamental frequency of the Fourier series fitting of the structural vibration, respectively, and can be used to control the bandwidth of the signal parameter decomposition generated during the structural vibration process, and the model parameters of the structural vibration. and The target component signal parameters are obtained through multi-component joint least squares optimization estimation, and the final step is to reconstruct the target component signal parameters based on the estimated structural vibration model parameters.

[0086] S4: After obtaining each individual target component using the method in step S3, it is necessary to estimate the interferometric phase evolution parameters spanning multiple frequency sweep cycles.

[0087] S5: To accurately extract the initial phase parameters of the target component in each frequency sweep cycle, a phase estimation algorithm based on approximate maximum likelihood estimation is proposed. The initial phase of the m-th component signal parameter in the i-th frequency sweep cycle can then be estimated as follows:

[0088]

[0089] Where arg[] is the complex phase operation, Sm(iT+nT) s Let T be the discrete signal parameter of the m-th component during the i-th sweep frequency period. s Where N is the sampling time interval, and N is the total number of discrete points of the signal parameters. The estimated beat frequency of the m-th component is given. Since the range of values ​​for complex phase operations is limited to [-π, π], the phase evolution of adjacent sweep cycles in the structural vibration and strain tracking analysis process will undergo abnormal jumps. Therefore, it is necessary to perform phase unwinding processing on the estimated phase evolution time series of the structural vibration and strain tracking analysis.

[0090] S6: Based on the estimated time series of the interferometric phase evolution for each target, the structural vibration parameters of each target are calculated by formula inversion.

[0091] The formula for inverting and calculating the structural vibration parameters of each target is as follows: Where Δφ(t) and Δx(t) are the changes in phase parameters of the baseband signal and vibration displacement in adjacent sweep cycles, respectively, and φ is the angle between the target vibration direction and the line of sight of the millimeter-wave radar beam. The time-domain parameters of the target vibration displacement are then inverted from the displacement changes. The interference phase parameters of the structural vibration band signal are linearly related to the time-domain parameters of the target object's vibration displacement. Therefore, the changes in the time-domain parameters of vibration displacement in adjacent sweep cycles can be obtained by tracking the interference phase evolution of the structural vibration baseband signal.

[0092] Structural vibration sensors are installed at the detection points to collect digital signals of structural vibration and obtain comparative values ​​of the vibration detection. To intuitively reflect the relationship between vibration amplitude and time, time-domain analysis is used to analyze the effectiveness and accuracy of structural vibration detection.

[0093] S7: Perform strain analysis on the structure. Considering the corresponding linear vibration characteristics of the structure, let the carrier frequency of the transmitted signal be f0, the frequency modulation slope be λ, and the effective bandwidth and effective time width during analysis be B, respectively. The time-domain characteristics of the millimeter-wave radar signal in the k-th period are:

[0094]

[0095] By utilizing quadratic functions, the automatic control stage of the structural vibration strain analysis model is made more consistent with the linear characteristics of structural vibration. Therefore, the maximum vibration amplitude at each monitoring point on the structure in different directions is significantly reduced proportionally and effectively controlled within ±0.1m.

[0096] S8: Assuming the Doppler frequency of the corresponding structure is fd, and the initial distance between the corresponding vibrating structure and the millimeter-wave radar is R0, and the propagation speed of electromagnetic waves in the air is c, derive the logical relationship between the relative distance R between the corresponding vibrating structure and the millimeter-wave radar during time interval t as follows:

[0097] R(t) = R0 + v i c+f d

[0098] v i This refers to the transmission rate of millimeter-wave radar in a stationary state.

[0099] S9: Assuming the instantaneous echo delay of the corresponding structure in time period t is τ(t) and the reflection coefficient is φ, the echo signal generated when the millimeter-wave radar detects the corresponding structure for the first time is:

[0100]

[0101] The instantaneous phase of the echo signal within this period is:

[0102]

[0103] During the measurement process, the vibration amplitude of the structure was controlled. Vibration signals before and after control were collected at 10 detection points by sensors, and the maximum vibration amplitude in each direction collected within 10 minutes was selected.

[0104] S10: Mix the transmitted signal according to the triangular wave linear frequency modulation method. With the initial echo signal Without considering mixing losses, a low-pass filter is used to remove high-frequency components, yielding the single-frequency beat signal of the echo signal. Assume the instantaneous phase of the beat signal is The corresponding structural strain value detected by millimeter-wave radar

[0105]

[0106] To verify the accuracy of the vibration and strain tracking analysis method for millimeter-wave structures, two alloy plates mounted on two slide rails were used as vibration targets to simulate a multi-point synchronous vibration test scenario. The two sliders were spaced approximately 30 cm apart, and the two slides were programmed to reciprocate in different vibration patterns.

[0107] S11: Store the data obtained in step S10 into the cloud platform to end the entire machine vision-based structural vibration and strain tracking analysis process.

[0108] In this invention, the vibration characteristic description equation is used to obtain the vibration characteristics of the structure of the device under test in and out of the plane, providing reliable evidence for millimeter-wave radar measurement. By using millimeter-wave radar, the unit that can be detected is further reduced. The millimeter-wave radar receives and can feed back signals by transmitting multiple cycles, which increases the signal-to-noise ratio and signal-to-clutter ratio. Therefore, the changes in the structural vibration detection results and the vibration sensor acquisition results over time in multiple directions are not only basically consistent with the overall trend, but also have a high degree of curve fitting, which further increases the overall accuracy in the detection and analysis process.

[0109] In this invention, the millimeter-wave radar machine vision-based structural vibration and strain tracking analysis method can obtain high-precision synchronous structural analysis results for multiple structural vibration targets. In multi-point synchronous structural vibration and strain tracking analysis applications, compared to the requirement of installing multiple laser displacement sensor devices, a single millimeter-wave radar can further reduce the unit of detection during structural vibration and strain tracking analysis. The structural vibration and strain tracking analysis system can conveniently and reliably meet the needs of multi-point synchronous vibration testing and the accuracy requirements of vibration displacement structural vibration and strain tracking analysis in practical engineering. This improves the analysis speed and accuracy of the entire method during the tracking analysis process.

[0110] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A machine vision-based method for structural vibration and strain tracking analysis, characterized in that, include: The I / Q signal parameters of the multi-scanning periodic baseband that the millimeter-wave radar can sense are obtained. The spectrum analysis of the I / Q signal parameters of the multi-scanning periodic baseband is performed to obtain the peak distribution of the amplitude spectrum. The range image parameters and the target positioning parameters are obtained based on the peak distribution of the amplitude spectrum. The peak distribution of the amplitude spectrum is searched and discriminated to obtain all the measured targets in the range image parameters. The frequency deviation compensation correction method based on Fourier series fitting is used to process the measured targets, thereby suppressing the coupling interference of neighboring components. The processed data is searched for and located using time-frequency ridge peaks to obtain individual target components; A phase estimation algorithm based on approximate maximum likelihood estimation is used to process each target component, resulting in an estimated time series of the interferometric phase evolution for each target. Based on the estimated time series of the interference phase evolution for each target, the structural vibration parameters of each target are inverted and calculated. Based on the structural vibration parameters, structural strain analysis is performed to obtain the corresponding linear vibration characteristics of the structure. Based on the linear vibration characteristics of the structure, the logical relationship between the relative distance between the corresponding vibrating structure and the millimeter-wave radar during time period t is obtained. Acquire the echo signal generated when the millimeter-wave radar first detects the corresponding structure and the instantaneous phase of the echo signal within that period; Based on the linear frequency modulation form of the triangular wave, the corresponding structural strain value detected by the millimeter-wave radar is obtained from the echo signal and the instantaneous phase.

2. The machine vision-based structural vibration and strain tracking analysis method according to claim 1, characterized in that, The specific method for Fourier series approximation fitting is as follows: Where L and F0 are the order and fundamental frequency of the Fourier series fitting of the structural vibration, respectively, and are used to control the bandwidth of the signal parameter decomposition generated during the structural vibration process. A m,i f m,i and θ m,i Let be the amplitude, frequency, and phase of the m-th component signal parameters in the i-th sweep cycle, respectively. Due to the modulation of the vibration motion, It is a nonlinear time-varying frequency. For a rough estimate of the frequency of this component, the model parameters of the structural vibration are... and It is obtained through joint least squares optimization estimation of multiple components.

3. The machine vision-based structural vibration and strain tracking analysis method according to claim 1, characterized in that, The specific method for processing each target component using the phase estimation algorithm with approximate maximum likelihood estimation is as follows: Where arg[] is the complex phase operation, S m (iT+nT s Let T be the discrete signal parameter of the m-th component during the i-th sweep frequency period. s Where N is the sampling time interval, and N is the total number of discrete points of the signal parameters. This is the estimated beat frequency of the m-th component.

4. The machine vision-based structural vibration and strain tracking analysis method according to claim 1, characterized in that, The specific methods for performing structural strain analysis are as follows: Where f0 is the carrier frequency, λ is the frequency modulation slope, and B and These are the effective bandwidth and effective time width during analysis, respectively. Let be the structural linear vibration characteristics of the millimeter-wave radar signal in the k-th period.

5. The machine vision-based structural vibration and strain tracking analysis method according to claim 1, characterized in that, The logical relationship between the relative distance between the corresponding vibrating structure and the millimeter-wave radar during time period t is as follows: R(t)=R0+v i c+f d Where R0 is the initial distance between the corresponding vibrating structure and the millimeter-wave radar, c is the propagation speed of electromagnetic waves in the air, and v i This represents the transmission rate of the millimeter-wave radar in a stationary state.

6. The machine vision-based structural vibration and strain tracking analysis method according to claim 1, characterized in that, The echo signal generated when the millimeter-wave radar first detects the corresponding structure is: Where B is the effective bandwidth during analysis, and τ(t) is the instantaneous echo delay of the corresponding structure in time period t. This is the reflection coefficient.

7. The machine vision-based structural vibration and strain tracking analysis method according to claim 6, characterized in that, The instantaneous phase of the echo signal during this period is: Where λ is the frequency modulation slope. This refers to the effective timeframe during analysis.

8. The machine vision-based structural vibration and strain tracking analysis method according to claim 1, characterized in that, The calculation method for the corresponding structural strain values ​​detected by millimeter-wave radar is as follows: in, The echo signal generated when the millimeter-wave radar first detects the corresponding structure. Let be the instantaneous phase of the echo signal within the period, B be the effective bandwidth during analysis, τ(t) be the instantaneous echo delay of the corresponding structure in time period t, and λ be the frequency modulation slope. Where c is the effective time span for analysis, and c is the propagation speed of electromagnetic waves in the air.

9. A machine vision-based structural vibration and strain tracking analysis system, characterized in that, include: The baseband signal parameter processing module is used to acquire the I / Q signal parameters of the multi-scanning period baseband that can be sensed by the millimeter-wave radar, perform spectral analysis on the I / Q signal parameters of the multi-scanning period baseband to obtain the peak distribution of the amplitude spectrum, and obtain the range image parameters and target positioning parameters based on the peak distribution of the amplitude spectrum. The target processing module is used to search and identify the peak distribution of the amplitude spectrum to obtain all the targets under test in the range image parameters. It uses a frequency deviation compensation correction method based on Fourier series fitting to process the targets under test, thereby suppressing the coupling interference of neighboring components. The target component acquisition module is used to search for and locate the time-frequency ridge peaks in the processed data to obtain individual target components. The phase interferometry module is used to process each target component using a phase estimation algorithm with approximate maximum likelihood estimation, and obtain the time series of the estimated interferometric phase evolution for each target. The vibration parameter inversion module is used to invert and calculate the structural vibration parameters of each target based on the estimated time series of the interference phase evolution for each target. The structural linear vibration characteristic acquisition module is used to perform structural strain analysis based on structural vibration parameters and obtain the corresponding structural linear vibration characteristics. The logical relationship acquisition module is used to obtain the logical relationship between the relative distance between the vibrating structure and the millimeter-wave radar during time period t, based on the linear vibration characteristics of the structure. The echo phase acquisition module acquires the echo signal generated when the millimeter-wave radar detects the corresponding structure for the first time and the instantaneous phase of the echo signal within that period. The structural strain value acquisition module is used to obtain the corresponding structural strain value detected by millimeter-wave radar based on the echo signal and instantaneous phase according to the linear frequency modulation form of the triangular wave.

10. The machine vision-based structural vibration and strain tracking analysis system according to claim 9, characterized in that, Structural vibration sensors are installed at the detection points to collect digital signals of structural vibration, obtain comparative values ​​of vibration detection, and use time-domain analysis to analyze the effectiveness and accuracy of structural vibration detection.

Images