Full-field-domain structure high-precision deformation and vibration measurement system based on microwave interference radar

Abstract

The invention discloses a full-field-domain structure high-precision deformation and vibration measurement system based on a microwave interference radar, and the system comprises an observation configuration module which is used for obtaining radar observation parameters and completing the signal transmission and reception; the signal processing module is used for preprocessing the echo signal and constructing an interference phase matrix; the phase unwrapping module is used for generating a continuous phase sequence and calculating the displacement of a reflection point; the frequency spectrum extraction module is used for extracting the frequency component and the vibration amplitude of each reflection point; the response clustering module is used for extracting reflection point clusters with similar frequency responses and constructing a vibration distribution diagram; the modal identification module is used for generating a vibration modal feature vector and marking a difference region; and the track generation module is used for generating a dynamic deformation track diagram of the full field domain structure. According to the invention, high-precision deformation and vibration mode synchronous measurement of a complex structure under full-field-domain and non-contact conditions is realized, and the precision, efficiency and intelligent level of structure monitoring are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of microwave interferometric radar technology, and in particular to a high-precision full-field structural deformation and vibration measurement system based on microwave interferometric radar. Background Technology

[0002] In the operational safety management of large civil structures, infrastructure, industrial equipment, and tall structures, structural health monitoring (SHM) technology has become an important means of achieving real-time condition awareness and disaster early warning. Existing structural monitoring methods mainly include contact or semi-contact methods such as strain gauges, accelerometers, fiber optic gratings, laser rangefinders, and video image monitoring. Although these methods have a certain degree of measurement accuracy, they usually suffer from problems such as complex deployment, sparse points, susceptibility to environmental interference, and difficulty in achieving large-scale synchronous measurement, which limits their practicality in complex structures or dynamic environments.

[0003] In recent years, microwave interferometric radar technology has been widely used in the deformation and vibration monitoring of structures such as bridges, tunnels, and dams due to its advantages of being non-contact, all-weather, and high-resolution. This technology can achieve millimeter or even sub-millimeter displacement detection at long distances by measuring the phase difference of electromagnetic wave reflections. However, existing microwave interferometric radars are mostly point-based, with limited spatial resolution, and are prone to signal aliasing in multi-target scenarios. Furthermore, traditional methods generally lack unwrapping mechanisms for the temporal phase changes of the echo signal, leading to jump errors in the measurement results. In addition, existing technologies rely heavily on external analysis software or manual interpretation for frequency domain feature extraction, vibration mode fitting, and modal anomaly identification, lacking a unified high-precision calculation process, making it difficult to meet the application requirements of full-field, multi-target, and dynamic response measurement.

[0004] Especially in the field of vibration monitoring, most current methods cannot simultaneously achieve frequency resolution and spatial continuity, resulting in discontinuities, incoherence, or blurred modal features in vibration distribution maps. Identification of structural anomaly regions relies on human experience, making it difficult to quantify and assess modal differences, and hindering efficient full-field modal zoning and anomaly detection.

[0005] Therefore, how to provide a high-precision deformation and vibration measurement system for the entire field of structures based on microwave interferometric radar is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0006] One objective of this invention is to propose a high-precision deformation and vibration measurement system for structures across the entire field based on microwave interferometric radar. This invention fully utilizes the principles of microwave interferometry, time-series phase unwrapping algorithms, sliding window spectrum extraction, and mode shape fitting modeling techniques. It details the entire process of realizing structural vibration modal analysis and deformation trajectory reconstruction under non-contact, wide-angle, and multi-target conditions. It has the advantages of wide coverage, high measurement accuracy, strong anti-interference capability, and high degree of intelligent modal recognition.

[0007] A high-precision full-field structural deformation and vibration measurement system based on microwave interferometric radar according to an embodiment of the present invention includes: The observation configuration module is used to acquire radar observation parameters of the coverage area, transmit linear frequency modulated continuous wave signals and receive corresponding echo signals. The signal processing module is used to preprocess the echo signal, divide the signal interval according to the range gate, extract the phase information of the target reflection point within each range gate, and construct the interference phase matrix; The phase unwrapping module is used to perform time-series phase unwrapping operations on the interference phase matrix, remove phase jump points and generate a continuous phase sequence, and calculate the displacement sequence of each reflection point on the structure surface. The spectrum extraction module is used to divide the displacement sequence into fixed time windows and use sliding window Fourier transform to extract the frequency components and vibration amplitude of each reflection point in each time window; The response clustering module is used to perform clustering operations based on frequency similarity and spatial proximity, extract clusters of reflection points with similar frequency responses, and construct a vibration distribution map of the entire field structure. The modal recognition module is used to perform dominant frequency calibration and mode shape profile fitting on each cluster in the vibration distribution map, generate vibration mode feature vectors of structural sub-regions, and mark local areas with large modal differences; The trajectory generation module is used to generate a dynamic deformation trajectory diagram of the structure in the entire field during the measurement period based on the vibration mode feature vector and the corresponding displacement sequence.

[0008] Optionally, modules can be integrated using the following methods: S1. Obtain radar observation parameters for the coverage area, set microwave transmission frequency range, modulation method and beam scanning path, transmit linear frequency modulated continuous wave signal and receive corresponding echo signal; S2. Preprocess the received echo signal, divide the echo signal interval according to the range gate, extract the phase information of the target reflection point in each range gate, and construct the interference phase matrix; S3. Perform a time-series phase unwrapping operation on the interference phase matrix, remove phase jump points and generate a continuous phase sequence, and calculate the displacement sequence of each reflection point on the corresponding structural surface. S4. Divide the displacement sequence into fixed time windows and use the sliding window Fourier transform method to extract the frequency components and vibration amplitude of each reflection point in each time window. S5. Based on the frequency components and vibration amplitude, perform clustering operations based on frequency similarity and spatial proximity to extract clusters of reflection points with similar frequency responses and construct a vibration distribution map of the entire field structure. S6. Perform dominant frequency calibration and mode shape profile fitting operations on each cluster in the vibration distribution map to generate vibration mode feature vectors for each structural sub-region and mark local regions with large modal differences. S7. Based on the vibration mode feature vectors and corresponding displacement sequences within each time window, generate the dynamic deformation trajectory diagram of the entire field structure within the measurement period.

[0009] Optionally, the echo signal is the reflected signal generated by the structure under test under microwave irradiation, which contains amplitude information and phase information. It is collected in time sequence and converted from analog to digital to form a complex signal sequence that can be used for distance resolution and phase interference calculation.

[0010] Optionally, the preprocessing includes sequentially performing filtering, sampling, demodulation, denoising, time synchronization, and gain correction operations on the received microwave echo signal.

[0011] Optionally, the fixed time window adopts an adaptive partitioning strategy based on the structural vibration characteristics. The initial window length is set based on the prior frequency distribution and historical response data. The time window length is dynamically adjusted by combining the frequency change rate and vibration amplitude fluctuation of each reflection point in a continuous time period. When a high-frequency vibration trend is detected, the window is shortened to enhance the time resolution. When a low-frequency deformation or vibration stability state is detected, the window is extended to improve the frequency resolution and continuous expression capability. Finally, an adaptive window set covering all frequency response intervals is generated in each observation cycle.

[0012] Optionally, S3 specifically includes: S31. Construct a two-dimensional time series structure for the interference phase matrix according to the distance gate index and the time frame index, which is used to describe the phase change of each reflection point under different time frames. S32. Perform a time-series phase unwrapping operation on the interference phase matrix, that is, perform point-by-point difference on the interference phase values ​​in adjacent time frames to obtain the phase increment matrix of each reflection point; S33. Detect whether the phase increment exceeds the phase jump threshold. If it does, eliminate the phase jump effect through compensation to obtain a continuously changing phase sequence. To avoid misjudgments caused by strong electromagnetic interference or signal aliasing, a reliable phase change range is constructed by extracting the median of the phase increments at multiple adjacent time points as a reference. If the current increment deviates from the reference range by more than a set deviation threshold, it is identified as an error point, and the median of the reference value is used to replace the outlier to complete the phase correction and ensure the accuracy of the continuous phase sequence. S34. Based on the relationship between the continuous phase sequence and the wavelength of the microwave signal, calculate the absolute displacement value of each reflection point in each time frame; S35. Organize the absolute displacement values ​​of all reflection points according to the distance gate and time frame to form a three-dimensional displacement sequence for vibration spectrum analysis and modal feature extraction.

[0013] Optionally, S4 specifically includes: S41. Divide the three-dimensional displacement sequence into multiple sliding time windows according to a fixed number of frames, and extract the local displacement sequence of the corresponding reflection point in each window; S42. Perform a discrete Fourier transform on the local displacement sequence within each time window and calculate the corresponding frequency components and frequency amplitude distribution. S43. Extract the dominant frequency index with the largest amplitude from the frequency amplitude distribution of each time window, and obtain the corresponding vibration amplitude based on the dominant frequency index for vibration response clustering and modal analysis.

[0014] Optionally, S5 specifically includes: S51. Construct a multi-dimensional feature vector containing the dominant frequency value, vibration amplitude and spatial coordinates of each reflection point to describe the frequency response and spatial location attributes; S52. Calculate the frequency difference between any two reflection points to measure the similarity of their vibration frequencies; S53. Calculate the spatial Euclidean distance between any two reflection points to measure their proximity. S54. Construct a joint similarity function, weighted combination of frequency similarity and spatial proximity, and set a joint distance threshold for clustering discrimination; S55. Perform density clustering operation within each time window to extract all clusters of reflection points that satisfy the joint distance condition; S56. Integrate the clustering results under each time window according to the time dimension to construct a complete vibration response cluster map, and map the vibration distribution map of the entire field structure according to spatial coordinates.

[0015] Optionally, the vibration response cluster map consists of multiple reflection point clusters, each with a unified dominant frequency characteristic and similar spatial distribution characteristics. The cluster structure within each time period is constructed using a time series sliding window method, and the cluster distribution area is represented in a coordinate mapping form in a two-dimensional space. The clusters in the map are partitioned and identified according to the frequency center value, the average vibration amplitude, and the spatial location boundary, in order to present the vibration response characteristics and frequency consistency of the structure in different regions.

[0016] Optionally, S6 specifically includes: S61. For each time window, collect the dominant frequency data of all reflection points in the extracted vibration response cluster. S62. Take the average of the main frequency values ​​within each cluster and use it as the center main frequency value of the structural sub-region. S63. Pair the vibration amplitude of the reflection point in each cluster with the spatial coordinates, and reconstruct the mode shape profile in the two-dimensional coordinate plane using the surface fitting method. When the structural region has a regular shape, the least squares surface fitting is used to generate the mode shape profile function; if the fitting residual exceeds the set threshold, the current region is judged to be an irregular structure, and traditional surfaces are difficult to model accurately. In this case, the method is switched to a nonparametric fitting method based on radial basis functions. Using multiple center points as references, the continuous reconstruction of irregular mode shapes is achieved by constructing a local response model. S64. Construct a modal feature vector composed of the dominant frequency, the mean amplitude, the amplitude fluctuation degree, and the fitted mode shape, and calculate the difference in dominant frequency and amplitude between any two adjacent modal feature vectors as the basis for measuring modal differences. S65. Mark modal feature regions where the difference measurement exceeds the set threshold as local regions with large modal differences for structural response analysis and key monitoring and location.

[0017] Optionally, S7 specifically includes: S71. During the measurement period, extract the modal feature vectors of all structural sub-regions and the absolute displacement values ​​of the corresponding reflection points for each time window; S72. Based on the mode shape profile function and central dominant frequency value corresponding to each cluster, calculate the vibration deformation at each time point in each time frame; S73. The absolute displacement values ​​of each reflection point are superimposed with their corresponding vibration deformations to obtain the displacement trajectory sequence in the current time frame. S74. Map the displacement trajectory sequence of all reflection points throughout the entire measurement cycle to two-dimensional spatial coordinates, and construct a frame diagram for each frame, ultimately generating a dynamic deformation trajectory diagram composed of continuous frames.

[0018] Optionally, the vibration deformation is a relative displacement calculated based on the mode shape profile function and the central dominant frequency value, reflecting the local dynamic offset of the reflection point relative to the average state of each cluster. The absolute displacement value is a global displacement obtained through continuous phase accumulation and wavelength conversion, expressing the actual spatial change trajectory of the reflection point in the measurement coordinate system. To avoid repetition and deviation in superposition calculation, before performing the fusion processing of relative deformation and absolute displacement, the spatial reference coordinate system of the two is unified, and the direction and amplitude scale of the relative displacement in the global coordinate system are calibrated by projection normalization.

[0019] The beneficial effects of this invention are: First, this invention introduces a non-contact linear frequency modulated continuous wave radar observation method, combined with a multi-range gate parallel phase extraction strategy, to achieve real-time coverage of reflection points on a large range of structural surfaces, significantly improving the spatial and temporal resolution of structural response monitoring.

[0020] Secondly, this invention innovatively designs a displacement calculation method based on time-series phase unwrapping and a sliding window Fourier spectrum analysis technique, which not only effectively solves the measurement error problem caused by phase jump in complex environments in traditional interferometric radar, but also achieves high-precision extraction of vibration frequency components and relative amplitudes, providing a solid data foundation for subsequent vibration mode identification and structural stability assessment.

[0021] Finally, this invention extracts clusters of reflection points with consistent response characteristics in the structure by combining a clustering algorithm based on frequency similarity and spatial proximity. Furthermore, it reconstructs the mode shape profile by combining a surface fitting method, thus constructing a structural modal map with spatiotemporal continuity. This enables automatic identification and visualization of modal difference regions, improving the intelligence and automation level of structural anomaly early warning. Attached Figure Description

[0022] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a block diagram of the high-precision deformation and vibration measurement system for the entire field structure based on microwave interferometric radar proposed in this invention. Figure 2 This is a flowchart of the method for a high-precision full-field structural deformation and vibration measurement system based on microwave interferometric radar proposed in this invention. Figure 3 This is a flowchart of the vibration mode feature extraction and anomaly identification of the full-field structural high-precision deformation and vibration measurement system based on microwave interferometric radar proposed in this invention. Detailed Implementation

[0023] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0024] refer to Figure 1 A high-precision full-field structural deformation and vibration measurement system based on microwave interferometric radar includes: The observation configuration module is used to acquire radar observation parameters of the coverage area, transmit linear frequency modulated continuous wave signals and receive corresponding echo signals. The signal processing module is used to preprocess the echo signal, divide the signal interval according to the range gate, extract the phase information of the target reflection point within each range gate, and construct the interference phase matrix; The phase unwrapping module is used to perform time-series phase unwrapping operations on the interference phase matrix, remove phase jump points and generate a continuous phase sequence, and calculate the displacement sequence of each reflection point on the structure surface. The spectrum extraction module is used to divide the displacement sequence into fixed time windows and use sliding window Fourier transform to extract the frequency components and vibration amplitude of each reflection point in each time window; The response clustering module is used to perform clustering operations based on frequency similarity and spatial proximity, extract clusters of reflection points with similar frequency responses, and construct a vibration distribution map of the entire field structure. The modal recognition module is used to perform dominant frequency calibration and mode shape profile fitting on each cluster in the vibration distribution map, generate vibration mode feature vectors of structural sub-regions, and mark local areas with large modal differences; The trajectory generation module is used to generate a dynamic deformation trajectory diagram of the structure in the entire field during the measurement period based on the vibration mode feature vector and the corresponding displacement sequence.

[0025] refer to Figure 2-3 In this embodiment, the modules are connected through the following method: S1. Obtain radar observation parameters for the coverage area, set microwave transmission frequency range, modulation method and beam scanning path, transmit linear frequency modulated continuous wave signal and receive corresponding echo signal; S2. Preprocess the received echo signal, divide the echo signal interval according to the range gate, extract the phase information of the target reflection point in each range gate, and construct the interference phase matrix; S3. Perform a time-series phase unwrapping operation on the interference phase matrix, remove phase jump points and generate a continuous phase sequence, and calculate the displacement sequence of each reflection point on the corresponding structural surface. S4. Divide the displacement sequence into fixed time windows and use the sliding window Fourier transform method to extract the frequency components and vibration amplitude of each reflection point in each time window. S5. Based on the frequency components and vibration amplitude, perform clustering operations based on frequency similarity and spatial proximity to extract clusters of reflection points with similar frequency responses and construct a vibration distribution map of the entire field structure. S6. Perform dominant frequency calibration and mode shape profile fitting operations on each cluster in the vibration distribution map to generate vibration mode feature vectors for each structural sub-region and mark local regions with large modal differences. S7. Based on the vibration mode feature vectors and corresponding displacement sequences within each time window, generate the dynamic deformation trajectory diagram of the entire field structure within the measurement period.

[0026] In this embodiment, during the process of acquiring radar observation parameters of the coverage area, a ranging interval and a transmit power threshold are set, the echo gain is dynamically adjusted according to the target distance, an environmental interference suppression mechanism is introduced, and interference characteristic signals such as rain, snow, and dust are monitored in real time. Unstructured reflections are suppressed through adaptive filtering and echo dynamic threshold strategy, and a multi-channel antenna array is used for beam enhancement and spatial filtering to improve signal stability and echo quality under long-distance and noisy conditions.

[0027] In this embodiment, the echo signal is formed by the reflection of the incident microwave signal by the surface of the structure under test. It contains amplitude information and phase information. After being received by the antenna and converted into an intermediate frequency signal by the radio frequency front end, it is sequentially converted into a complex signal sequence by analog-to-digital conversion. The real part represents the echo intensity and the imaginary part represents the instantaneous phase, which is used to perform distance resolution and interference phase extraction processing.

[0028] In this embodiment, the preprocessing includes the following steps: performing bandpass filtering on the received echo signal to retain the effective microwave components within a preset frequency band; performing uniform time sampling on the filtered analog signal to obtain a digital sampling point sequence; performing frequency mixing and demodulation on the digital sampling point sequence to extract the intermediate frequency component; performing noise reduction processing on the demodulated signal to filter out environmental noise and system interference; performing time synchronization on the processed signal to correct time deviations in different observation angles or repeated measurements; and performing gain correction on the synchronized signal to compensate for antenna directivity attenuation and propagation path loss.

[0029] In this embodiment, the fixed time window adopts an adaptive partitioning strategy based on the structural vibration characteristics. The initial window length is set based on the prior frequency distribution and historical response data. The time window length is dynamically adjusted by combining the frequency change rate and vibration amplitude fluctuation of each reflection point in a continuous time period. When a high-frequency vibration trend is detected, the window is shortened to enhance the time resolution. When a low-frequency deformation or vibration stability state is detected, the window is extended to improve the frequency resolution and continuous expression capability. Finally, an adaptive window set covering all frequency response intervals is generated in each observation period.

[0030] In this embodiment, S3 specifically includes: S31. The constructed interference phase matrix Arranged according to distance gate index and time frame index, where Indicates the first The distance gate is at the first Interference phase value of the frame; S32, to Perform the timing phase unwrapping operation, i.e., the interpolation operation between adjacent frames, to obtain the phase increment matrix. The calculation formula is: ; in, Indicates the first The distance gate is at the first Interference phase value of the frame; S33. Under each distance gate index, the phase increment matrix Phase transition detection is performed, and the threshold for judgment is... When satisfied When a condition is met, if the record is a transition point, then a phase compensation operation is performed to construct the phase compensation function: ; in, Indicates the first The distance gate is at the first The continuous phase values ​​of the frame, Indicates the first The distance gate is at the first The continuous phase values ​​of the frame, Pi is a constant. This represents the rounding function; To eliminate unreal transitions caused by strong electromagnetic interference and signals from multiple reflection points, a multi-window statistical filtering method is introduced to construct a reliable set of phase increments: ; in, Indicates the first Each reflection point in time The reference phase increment at that point is taken as the median value of adjacent frames. These represent the phase increment matrices in different frames. This represents the median function, used for smoothing to prevent interference. when When this phase increment is considered an error point, a jump correction operation is performed, and the corrected phase recursive formula is: ; in, The threshold for determining phase deviation; By continuously performing sliding window updates and compensation replacements, a robust and continuous phase sequence is constructed. S34. Based on the phase compensation function, combined with microwave wavelength Calculate the first The distance gate is at the first absolute displacement value corresponding to the frame The calculation formula is: ; S35, All The three-dimensional displacement sequence is reconstructed using the distance gate index and time frame index, and used for spectrum extraction and vibration mode calculation.

[0031] In this embodiment, S4 specifically includes: S41. Divide the three-dimensional displacement sequence into segments of length based on the time frame index. The sliding time window sequence is defined as: ; in, Indicates the first The distance gate is at the first The first window, the Local displacement values ​​at each frame position The sliding step size, The number of frames for each time window, Indicates the first The distance gate is at the first The absolute displacement value corresponding to the frame; S42. For each sliding time window sequence Perform a discrete Fourier transform operation and calculate the corresponding frequency components and frequency amplitude distribution. The calculation formula is as follows: ; ; in, Indicates the first The distance gate is at the first The frequency index within each time window is , frequency components, This represents a complex exponential function, used to map time-domain signals to the frequency domain. For frequency index The frequency amplitude represents the first... The distance gate is at the first Frequency response intensity over a time window The imaginary unit satisfies ; S43, Statistics The distance gate is at the first The primary frequency index within a time window is defined as: ; in, The frequency index representing the maximum value among the frequency amplitudes. This indicates the location of the main frequency; S44, Based on the clock frequency index Calculate the corresponding vibration amplitude, defined as follows: ; in, Indicates the first The distance gate is at the first The vibration amplitude within a time window.

[0032] In this embodiment, S5 specifically includes: S51. Define the index of each reflection point in the sliding window. The eigenvectors below Its composition is as follows: ; in, Represents the main frequency value, satisfying , Indicates the sampling frequency. Indicates the vibration amplitude. They represent the first The spatial coordinates of each reflection point in a two-dimensional plane; S52. Constructing a frequency-distance metric function Defined as: ; in, Indicates the first With the The reflection point at the th... Frequency difference measurement within a time window Indicates different clock frequency values; S53. Constructing a spatial distance metric function Defined as: ; S54. Constructing the joint similarity function This is used to describe the similarity between the combined frequency and spatial distance, and is defined as: ; in, As a weighting factor, ; S55, Set threshold For all that satisfy Similarity connections are established between pairs of reflection points, and density clustering is used to extract clusters of reflection points under each time window; S56. Perform temporal unification on the clustering results within all time windows to construct a complete vibration response cluster map, and map the vibration distribution map of the entire field structure according to spatial coordinates.

[0033] In this embodiment, the vibration response cluster map is constructed based on the clustering results of multiple sliding time windows. Within each time window, a set of reflection points with similar dominant frequency values ​​and spatial proximity is extracted to form multiple frequency response clusters. Each cluster is numbered and stored according to the corresponding time window index. Then, under a unified coordinate system, the clusters of each time window are aligned on the time axis and the spatially overlapping areas are merged to form the evolution trajectory of the reflection point clusters covering the entire observation period. Finally, a two-dimensional map is generated. In the two-dimensional map, each cluster is labeled with the central dominant frequency value, average vibration amplitude, spatial boundary box, and time interval. Similar frequency response regions are identified by color coding or region filling, reflecting the distribution of regions with similar dynamic responses in the entire field of the structure.

[0034] In this embodiment, S6 specifically includes: S61, in the Within a time window, let the cluster number of the reflection points be... ,in Indicates the first The first time window Each cluster, extracting clusters. The set of dominant frequency values ​​of all reflection points within: S62. Calculate the cluster based on the set of main frequency values. center frequency value The average clock frequency is defined as follows: ; in, Indicates the clock frequency value. Represents a cluster Number of internal reflection points; S63, Cluster vibration amplitude of all reflection points According to spatial coordinates Mapped to a two-dimensional plane, the mode shape profile function is generated using least-squares surface fitting. The mode shape profile function is defined as follows: ; in, Indicates the first Within the first time window, the first Two-dimensional coordinates on a cluster The vibration displacement value, The fitting coefficients represent the constant, linear, and quadratic parameters in the fitted surface. When the fitting residuals satisfy: ; The current structure is considered to be an irregular region. A nonparametric fitting method based on radial basis functions is used to reconstruct the mode shape distribution, and the mode shape expression of the radial basis functions is constructed as follows: ; in, This indicates the number of center points involved in the fitting. Indicates the first The coordinates of the radial center point This represents the weighting coefficient of the corresponding radial basis. The radial basis functions are represented using the Gaussian kernel function form. This represents the mean squared residuals of the fit. This represents the residual tolerance threshold; exceeding this threshold triggers nonparametric fitting. S64. Constructing vibration modal feature vectors Defined as: ; in, For the first The cluster in the th The average vibration amplitude within a time window, For the first The standard deviation of the vibration amplitude of a cluster represents the degree of amplitude fluctuation. S65. Calculate the difference measure between any two modal eigenvectors. Used to evaluate the first Within the first time window The cluster and the first The modal feature differences between clusters are defined as: ; in, This represents the central master frequency value corresponding to different clusters. This represents the average vibration amplitude corresponding to different clusters; S66. Set the modal difference threshold. ,when At that time, the first The cluster and the first The regions where each cluster is located are marked as local regions with significant modal differences, among which It is a constant.

[0035] In this embodiment, S7 specifically includes: S71, Assume the measurement period includes Using a sliding time window, the vibration mode feature vectors of all clusters within each window are extracted. absolute displacement value of the corresponding reflection point ,in , , , This is the sliding step size; S72. Based on the mode profile function corresponding to each cluster. With the center main frequency value At each point in time Calculate vibration deformation The expression is: ; in, Sampling frequency, Represents a hyperbolic sine function; S73, Combining the absolute displacement value of each reflection point within the measurement period. relative deformation of its cluster Calculate the displacement trajectory sequence after dynamic superposition. : ; S74. Sequence of displacement trajectories of all reflection points throughout the entire measurement period. Mapping to two-dimensional space coordinates The system constructs a frame graph for each frame, ultimately generating a dynamic deformation trajectory graph composed of consecutive frames.

[0036] In this embodiment, the vibration deformation Based on mode profile function With the center main frequency value The calculated relative displacement reflects the local dynamic offset of the reflection point relative to the average state of each cluster. The absolute displacement value is the global displacement obtained by continuous phase accumulation and wavelength conversion, which expresses the actual spatial change trajectory of the reflection point in the measurement coordinate system. To avoid repetition and deviation in the superposition calculation, the spatial reference coordinate system of the two is unified before performing the fusion processing of relative deformation and absolute displacement. The direction and amplitude scale of the relative displacement in the global coordinate system are calibrated by projection normalization, thereby ensuring that the superimposed trajectory sequence has spatiotemporal continuity and geometric consistency, and meets the accuracy requirements of dynamic trajectory reconstruction.

[0037] Example 1: To verify the feasibility of this invention in practice, it was applied to the operational status monitoring of a large steel structure bridge. This bridge has a large span and complex structure, with numerous cantilever sections, welded joints, and flexible connection areas. During long-term operation, it is prone to minor deformations and uneven local vibration attenuation due to wind loads, traffic loads, and temperature and humidity changes. Traditional monitoring methods mainly rely on embedded strain gauges and accelerometers for contact measurements at bridge nodes, which suffers from difficulties in deployment, sparse measurement points, and the inability to simultaneously cover the entire structure. Furthermore, in multi-target areas, traditional methods struggle to obtain the differential characteristics between local modes, making it impossible to accurately determine the early degradation trend of the structure.

[0038] In this monitoring scenario, this invention deploys a microwave interferometric radar array to conduct non-contact observation of the entire bridge structure. The radar emits a linear frequency modulated continuous wave signal, forming multiple reflection point areas on the bridge deck and the outer wall of the main beam. Echo phase information for each reflection point is obtained through time-series acquisition. Through signal preprocessing and range-gating, the system extracts the interferometric phase matrix at 150 observation points and calculates millimeter-level continuous displacement data using an improved phase unwrapping algorithm. For vibration analysis, a sliding window Fourier transform method is used to extract the dominant frequency and vibration amplitude of each reflection point within each 2-second time window, thereby obtaining a frequency distribution spectrum.

[0039] Using frequency similarity and spatial proximity clustering algorithms, this invention identified seven frequency response clusters from 150 points across the entire structure and performed surface fitting on their mode shape profiles to construct a complete vibration modal feature map. Radar monitoring data revealed a frequency abrupt change in the middle section of the main beam, with its modal frequency 0.87 Hz higher than adjacent clusters, and the vibration amplitude variance increased to 0.22 mm², while the average variance in the normal structural area was only 0.06 mm². The system automatically marked regions of modal difference, enabling rapid identification of weak sections of the bridge.

[0040] To verify the advantages of this invention in high-precision deformation and modal identification, it was compared with traditional contact monitoring methods. During the comparison, the dominant frequency error, mode shape fitting residual, and low-frequency micro-deformation capture capability of the same structure under different loading conditions within a 24-hour operating cycle were recorded. As shown in Table 1, this invention still achieves extremely high frequency identification accuracy and structural anomaly localization capability in non-contact measurement mode.

[0041] Table 1 Comparison of the effects of the present invention and traditional methods in structural monitoring

[0042] The implementation results show that this invention significantly outperforms existing contact-based methods in terms of accuracy, timeliness, and coverage, and can achieve high-frequency dynamic monitoring without altering the original structural state. Especially in complex structures and multi-target simultaneous detection scenarios, this invention possesses comprehensive capabilities such as high-resolution modal recognition, automatic anomaly labeling, and dynamic trajectory construction, demonstrating promising engineering application prospects and practical value.

[0043] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A high-precision full-field structural deformation and vibration measurement system based on microwave interferometric radar, characterized in that, include: The observation configuration module is used to acquire radar observation parameters of the coverage area, transmit linear frequency modulated continuous wave signals and receive corresponding echo signals. The signal processing module is used to preprocess the echo signal, divide the signal interval according to the range gate, extract the phase information of the target reflection point within each range gate, and construct the interference phase matrix; The phase unwrapping module is used to perform time-series phase unwrapping operations on the interference phase matrix, remove phase jump points and generate a continuous phase sequence, and calculate the displacement sequence of each reflection point on the structure surface. The spectrum extraction module is used to divide the displacement sequence into fixed time windows and use sliding window Fourier transform to extract the frequency components and vibration amplitude of each reflection point in each time window; The response clustering module is used to perform clustering operations based on frequency similarity and spatial proximity, extract clusters of reflection points with similar frequency responses, and construct a vibration distribution map of the entire field structure. The modal recognition module is used to perform dominant frequency calibration and mode shape profile fitting on each cluster in the vibration distribution map, generate vibration mode feature vectors of structural sub-regions, and mark local areas with large modal differences; The trajectory generation module is used to generate a dynamic deformation trajectory diagram of the structure in the entire field during the measurement period based on the vibration mode feature vector and the corresponding displacement sequence.

2. The high-precision deformation and vibration measurement system for full-field structures based on microwave interferometric radar according to claim 1, characterized in that, The modules are connected in the following way: S1. Obtain radar observation parameters for the coverage area, set microwave transmission frequency range, modulation method and beam scanning path, transmit linear frequency modulated continuous wave signal and receive corresponding echo signal; S2. Preprocess the received echo signal, divide the echo signal interval according to the range gate, extract the phase information of the target reflection point in each range gate, and construct the interference phase matrix; S3. Perform a time-series phase unwrapping operation on the interference phase matrix, remove phase jump points and generate a continuous phase sequence, and calculate the displacement sequence of each reflection point on the corresponding structural surface. S4. Divide the displacement sequence into fixed time windows and use the sliding window Fourier transform method to extract the frequency components and vibration amplitude of each reflection point in each time window. S5. Based on the frequency components and vibration amplitude, perform clustering operations based on frequency similarity and spatial proximity to extract clusters of reflection points with similar frequency responses and construct a vibration distribution map of the entire field structure. S6. Perform dominant frequency calibration and mode shape profile fitting operations on each cluster in the vibration distribution map to generate vibration mode feature vectors for each structural sub-region and mark local regions with large modal differences. S7. Based on the vibration mode feature vectors and corresponding displacement sequences within each time window, generate the dynamic deformation trajectory diagram of the entire field structure within the measurement period.

3. The high-precision deformation and vibration measurement system for full-field structures based on microwave interferometric radar according to claim 2, characterized in that, The echo signal is the reflected signal generated by the structure under test under microwave irradiation. It contains amplitude and phase information. It is collected in time sequence and converted from analog to digital to form a complex signal sequence that can be used for distance resolution and phase interference calculation.

4. The high-precision deformation and vibration measurement system for full-field structures based on microwave interferometric radar according to claim 2, characterized in that, The preprocessing includes sequentially filtering, sampling, demodulating, denoising, time synchronization, and gain correction operations on the received microwave echo signal.

5. The high-precision deformation and vibration measurement system for full-field structures based on microwave interferometric radar according to claim 2, characterized in that, S3 specifically includes: S31. Construct a two-dimensional time series structure for the interference phase matrix according to the distance gate index and the time frame index, which is used to describe the phase change of each reflection point under different time frames. S32. Perform a time-series phase unwrapping operation on the interference phase matrix, that is, perform point-by-point difference on the interference phase values ​​in adjacent time frames to obtain the phase increment matrix of each reflection point; S33. Detect whether the phase increment exceeds the phase jump threshold. If it does, it is identified as a jump point. Then, the phase jump effect is eliminated by compensation to obtain a continuously changing phase sequence. To avoid misjudgments caused by strong electromagnetic interference or signal aliasing, a reliable phase change range is constructed by extracting the median of the phase increments at multiple adjacent time points as a reference. If the current increment deviates from the reference range by more than a set deviation threshold, it is identified as an error point, and the median of the reference value is used to replace the outlier to complete the phase correction and ensure the accuracy of the continuous phase sequence. S34. Based on the relationship between the continuous phase sequence and the wavelength of the microwave signal, calculate the absolute displacement value of each reflection point in each time frame; S35. Organize the absolute displacement values ​​of all reflection points according to the distance gate and time frame to form a three-dimensional displacement sequence for vibration spectrum analysis and modal feature extraction.

6. The high-precision deformation and vibration measurement system for full-field structures based on microwave interferometric radar according to claim 2, characterized in that, S4 specifically includes: S41. Divide the three-dimensional displacement sequence into multiple sliding time windows according to a fixed number of frames, and extract the local displacement sequence of the corresponding reflection point in each window; S42. Perform a discrete Fourier transform on the local displacement sequence within each time window and calculate the corresponding frequency components and frequency amplitude distribution. S43. Extract the dominant frequency index with the largest amplitude from the frequency amplitude distribution of each time window, and obtain the corresponding vibration amplitude based on the dominant frequency index for vibration response clustering and modal analysis.

7. The high-precision deformation and vibration measurement system for full-field structures based on microwave interferometric radar according to claim 2, characterized in that, S5 specifically includes: S51. Construct a multi-dimensional feature vector containing the dominant frequency value, vibration amplitude and spatial coordinates of each reflection point to describe the frequency response and spatial location attributes; S52. Calculate the frequency difference between any two reflection points to measure the similarity of their vibration frequencies; S53. Calculate the spatial Euclidean distance between any two reflection points to measure their proximity. S54. Construct a joint similarity function, weighted combination of frequency similarity and spatial proximity, and set a joint distance threshold for clustering discrimination; S55. Perform density clustering operation within each time window to extract all clusters of reflection points that satisfy the joint distance condition; S56. Integrate the clustering results under each time window according to the time dimension to construct a complete vibration response cluster map, and map the vibration distribution map of the entire field structure according to spatial coordinates.

8. The high-precision deformation and vibration measurement system for full-field structures based on microwave interferometric radar according to claim 7, characterized in that, The vibration response cluster map consists of multiple reflection point clusters. Each reflection point cluster has a unified dominant frequency characteristic and similar spatial distribution characteristics. The cluster structure in each time period is constructed by a time series sliding window method, and the cluster distribution area is represented in a coordinate mapping form in a two-dimensional space. The clusters in the map are divided into zones based on the frequency center value, the average vibration amplitude, and the spatial location boundary, which is used to present the vibration response characteristics and frequency consistency of the structure in different regions.

9. The high-precision deformation and vibration measurement system for full-field structures based on microwave interferometric radar according to claim 2, characterized in that, S6 specifically includes: S61. For each time window, collect the dominant frequency data of all reflection points in the extracted vibration response cluster. S62. Take the average of the main frequency values ​​within each cluster and use it as the center main frequency value of the structural sub-region. S63. Pair the vibration amplitude of the reflection point in each cluster with the spatial coordinates, and reconstruct the mode shape profile in the two-dimensional coordinate plane using the surface fitting method. When the structural region has a regular shape, the least squares surface fitting is used to generate the mode shape profile function; if the fitting residual exceeds the set threshold, the current region is judged to be an irregular structure, and traditional surfaces are difficult to model accurately. In this case, the method is switched to a nonparametric fitting method based on radial basis functions. Using multiple center points as references, the continuous reconstruction of irregular mode shapes is achieved by constructing a local response model. S64. Construct a modal feature vector composed of the dominant frequency, the mean amplitude, the amplitude fluctuation degree, and the fitted mode shape, and calculate the difference in dominant frequency and amplitude between any two adjacent modal feature vectors as the basis for measuring modal differences. S65. Mark modal feature regions where the difference measurement exceeds the set threshold as local regions with large modal differences for structural response analysis and key monitoring and location.

10. The high-precision deformation and vibration measurement system for full-field structures based on microwave interferometric radar according to claim 2, characterized in that, Specifically, S7 includes: S71. During the measurement period, extract the modal feature vectors of all structural sub-regions and the absolute displacement values ​​of the corresponding reflection points for each time window; S72. Based on the mode shape profile function and central dominant frequency value corresponding to each cluster, calculate the vibration deformation at each time point in each time frame; S73. The absolute displacement values ​​of each reflection point are superimposed with their corresponding vibration deformations to obtain the displacement trajectory sequence in the current time frame. S74. Map the displacement trajectory sequence of all reflection points throughout the entire measurement cycle to two-dimensional spatial coordinates, and construct a frame diagram for each frame, ultimately generating a dynamic deformation trajectory diagram composed of continuous frames.

Images