System and method for monitoring real-time displacement of hydraulic structure

Abstract

The invention discloses a hydraulic structure real-time displacement monitoring system and method, and belongs to the technical field of hydraulic structure displacement monitoring, and the method comprises the steps: obtaining preprocessed Beidou GNSS displacement monitoring data and machine vision displacement monitoring data; performing time sequence reconstruction on the preprocessed Beidou GNSS displacement monitoring data to obtain Beidou GNSS displacement monitoring data after time sequence reconstruction; constructing a dual-source displacement monitoring data fusion model based on the machine vision displacement monitoring data and the Beidou GNSS displacement monitoring data after the time sequence reconstruction; and performing space-time registration fusion on the machine vision displacement monitoring data and the Beidou GNSS displacement monitoring data after time sequence reconstruction through the double-source displacement monitoring data fusion model, correcting a fusion result, and outputting a real-time joint measurement displacement monitoring value of the hydraulic structure. According to the monitoring method, a more accurate hydraulic structure displacement monitoring value can be obtained.

Description

Technical Field

[0001] This invention relates to the field of hydraulic structure displacement monitoring technology, and in particular to a real-time displacement monitoring system and method for hydraulic structures. Background Technology

[0002] With the continuous improvement of displacement monitoring technology for hydraulic structures, acquiring spatiotemporal displacement information that characterizes the real-time operational status of dams through the deployment of visual instruments such as cameras has become an emerging intelligent means for safety monitoring of large-scale hydraulic structures. However, for complex and diverse hydraulic engineering structures, more comprehensive and reliable monitoring data is required. The measurement results obtained by camera-based machine vision displacement measurement methods are too simplistic. Therefore, machine vision displacement measurement methods, which are currently dominated by single cameras, are gradually shifting towards joint and synchronous monitoring using new monitoring methods, such as BeiDou GNSS (Global Navigation Satellite System) and UAV monitoring, resulting in denser image sequences and richer monitoring information.

[0003] However, BeiDou GNSS has limitations in measurement accuracy and frequency, and its positioning accuracy is often affected by orbital errors and ionospheric delay errors. Machine vision monitoring, with its short-range, high-frequency characteristics, can effectively compensate for these deficiencies. Furthermore, machine vision monitoring suffers from a lack of absolute displacement and low long-term measurement stability, which can be effectively corrected by BeiDou GNSS. Therefore, this invention proposes a monitoring system and method for real-time displacement multi-source information fusion of hydraulic structures based on BeiDou GNSS and machine vision. Summary of the Invention

[0004] The purpose of this invention is to provide a real-time displacement monitoring system and method for hydraulic structures. This system utilizes a BeiDou GNSS-machine vision integrated monitoring device to monitor the displacement of hydraulic structures in real time. Simultaneously, by analyzing and processing the collected data through data analysis and fusion techniques, a more accurate displacement estimate of the hydraulic structure can be obtained. This invention is achieved through the following technical solutions.

[0005] In a first aspect, the present invention provides a real-time displacement monitoring system for hydraulic structures, comprising: a BeiDou GNSS-machine vision joint displacement monitoring device, a data acquisition terminal, and a data processing terminal connected in sequence.

[0006] The data acquisition terminal acquires and stores dual-source displacement monitoring data from the BeiDou GNSS-machine vision joint displacement monitoring device; the data processing terminal acquires the dual-source displacement monitoring data stored in the data acquisition terminal and performs collaborative analysis. The dual-source displacement monitoring data includes BeiDou GNSS displacement monitoring data and machine vision displacement monitoring data.

[0007] Optionally, the BeiDou GNSS-machine vision joint monitoring device includes a base and a pole arranged on the part of the hydraulic structure to be measured, a data acquisition electrical cabinet, a machine vision measuring instrument and a visual monitoring target, a camera and a reflector set on the pole, and a BeiDou GNSS receiving antenna and a BeiDou GNSS receiver set on the top of the pole.

[0008] The machine vision measuring instrument and visual monitoring target are used to collect machine vision displacement monitoring data of hydraulic structures and store it in the data acquisition electrical cabinet; the Beidou GNSS receiving antenna and Beidou GNSS receiver are used to collect Beidou GNSS displacement monitoring data of hydraulic structures and store it in the data acquisition electrical cabinet; the data acquisition electrical cabinet is also used to power the machine vision monitoring instrument and the Beidou GNSS monitoring instrument.

[0009] The camera is used to acquire direct images of the visual monitoring target, a virtual image formed by mirror reflection, and scale lines engraved on the visual monitoring target, and stores the acquisition results in the data acquisition electrical cabinet.

[0010] In practical applications, the BeiDou GNSS receiving antenna receives and amplifies satellite signals, and the BeiDou GNSS receiver processes and calculates the positioning of the signals received by the BeiDou GNSS receiving antenna, thereby collecting BeiDou GNSS displacement monitoring data of hydraulic structures.

[0011] Optionally, the data processing terminal is configured with a dual-source displacement monitoring data fusion and analysis module; the dual-source displacement monitoring data fusion and analysis module includes a data input unit, a data preprocessing unit, a data fusion unit, and a data output unit connected in sequence;

[0012] The data input unit is used to receive dual-source displacement monitoring data stored in the data acquisition terminal; the data preprocessing unit is used to preprocess the dual-source displacement monitoring data to obtain preprocessed dual-source displacement monitoring data; the data fusion unit is used to fuse the preprocessed dual-source displacement monitoring data to obtain fused dual-source displacement monitoring data; and the data output unit is used to correct the fused dual-source displacement monitoring data and output the real-time joint displacement monitoring value of the hydraulic structure.

[0013] Optionally, the real-time displacement monitoring value of the hydraulic structure is calculated using the following formula:

[0014] ,

[0015] In the formula, The real-time displacement monitoring value of the hydraulic structure at time t is... Let be the machine vision displacement monitoring value at time t. The value is the BeiDou GNSS displacement monitoring value at time t.

[0016] Optionally, when t=t+1, the increments of the machine vision displacement monitoring value and the BeiDou GNSS displacement monitoring value are as follows:

[0017] ,

[0018] ,

[0019] In the formula, and These represent the increments of machine vision monitoring values ​​and BeiDou GNSS monitoring values, respectively.

[0020] At time t+1, the real-time displacement monitoring value of the hydraulic structure is calculated using the following formula:

[0021] ,

[0022] In the formula, The displacement monitoring value of the hydraulic structure at time t+1 is the real-time joint measurement value. This represents the cumulative average of the measured values ​​from BeiDou GNSS displacement monitoring data up to time t+1. The weights of the machine vision displacement monitoring values. The weights of the BeiDou GNSS displacement monitoring values ​​are calculated using the following formulas:

[0023] ,

[0024] .

[0025] In a second aspect, the present invention provides a method for real-time displacement monitoring of hydraulic structures, employing the real-time displacement monitoring system for structures as described in the first aspect, comprising the following steps:

[0026] The pre-processed BeiDou GNSS displacement monitoring data and machine vision displacement monitoring data are obtained through the data processing terminal.

[0027] The preprocessed BeiDou GNSS displacement monitoring data is reconstructed in time to obtain reconstructed BeiDou GNSS displacement monitoring data; this is used to ensure that the time and frequency are consistent with the machine vision displacement monitoring data.

[0028] A dual-source displacement monitoring data fusion model is constructed based on the machine vision displacement monitoring data and the time-series reconstructed BeiDou GNSS displacement monitoring data.

[0029] The dual-source displacement monitoring data fusion model performs spatiotemporal registration and fusion of the machine vision displacement monitoring data and the time-series reconstructed BeiDou GNSS displacement monitoring data, and corrects the fusion results to output real-time joint measurement displacement monitoring values ​​of hydraulic structures.

[0030] Optionally, the machine vision displacement monitoring data employs offset angle correction. Specifically, this involves substituting the real-time offset angle parameters of the machine vision displacement monitoring data into the displacement calculation model of the machine vision monitoring instrument to perform dynamic geometric correction on the visual displacement monitoring data. This ensures that the machine vision displacement monitoring data is stably monitored over a long period.

[0031] Optionally, the step of correcting the fusion result includes: using a sliding window optimization framework to eliminate recursion errors and dynamically correcting the fusion result;

[0032] Under the sliding window optimization framework, the optimization objective is simplified to the weighted minimization of two residuals, namely the visual relative displacement residual and the GNSS absolute displacement constraint residual.

[0033] These are used to ensure that the displacement difference between adjacent time points is consistent with the visual measurement value, and to ensure that the displacement at the end of the window is consistent with the GNSS measurement value, respectively.

[0034] Optionally, the real-time displacement monitoring value of the hydraulic structure is calculated using the following formula:

[0035] ,

[0036] In the formula, The real-time displacement monitoring value of the hydraulic structure at time t is... Let be the machine vision displacement monitoring value at time t. The value is the BeiDou GNSS displacement monitoring value at time t.

[0037] Optionally, when t=t+1, the increments of the machine vision displacement monitoring value and the BeiDou GNSS displacement monitoring value are as follows:

[0038] ,

[0039] ,

[0040] In the formula, and These represent the increments of machine vision monitoring values ​​and BeiDou GNSS monitoring values, respectively.

[0041] At time t+1, the real-time displacement monitoring value of the hydraulic structure is calculated using the following formula:

[0042] ,

[0043] In the formula, The displacement monitoring value of the hydraulic structure at time t+1 is the real-time joint measurement value. This represents the cumulative average of the measured values ​​from BeiDou GNSS displacement monitoring data up to time t+1. and The weights for machine vision displacement monitoring values ​​and BeiDou GNSS displacement monitoring values ​​are calculated using the following formulas:

[0044] ,

[0045] .

[0046] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0047] (1) The real-time displacement monitoring system for hydraulic structures provided by the present invention overcomes the limitations of the existing technology, such as the single nature of the monitoring content of machine vision monitoring instruments being relative deformation and the narrow monitoring range, as well as the defects of Beidou GNSS monitoring instruments being unable to meet the accuracy requirements of the current water conservancy project monitoring specifications and being affected by calculation and satellite blind spots, by setting up a Beidou GNSS-machine vision joint displacement monitoring device.

[0048] (2) The real-time displacement monitoring method for hydraulic structures provided by the present invention integrates Beidou GNSS displacement monitoring data and machine vision displacement monitoring data, thereby improving the real-time monitoring accuracy of Beidou GNSS monitoring instruments from 5-8mm to within 2mm, which can meet the current requirements for the accuracy and real-time performance of health monitoring of water conservancy projects and obtain more accurate displacement estimation of hydraulic structures. Attached Figure Description

[0049] Figure 1 The diagram shown is a flowchart of a real-time displacement monitoring method for hydraulic structures according to one embodiment of the present invention.

[0050] Figure 2 The diagram shown is a schematic representation of the BeiDou GNSS-machine vision combined displacement monitoring device in one embodiment of the present invention.

[0051] Figure 3 The diagram shown is a side view of the BeiDou GNSS-machine vision combined displacement monitoring device in one embodiment of the present invention.

[0052] Figure 4 The diagram shown is a side view of the BeiDou GNSS-machine vision combined displacement monitoring device in one embodiment of the present invention.

[0053] In the diagram: 1-base, 2-pole, 3-data acquisition electrical cabinet, 4-machine vision measuring instrument, 5-visual monitoring target, 6-BeiDou GNSS receiving antenna, 7-BeiDou GNSS receiver, 8-camera, 9-reflector. Detailed Implementation

[0054] The following description, in conjunction with the accompanying drawings and specific embodiments, provides further details. In this description, it should be understood that the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature.

[0055] Example 1

[0056] This embodiment introduces a real-time displacement monitoring system for hydraulic structures, including: a BeiDou GNSS-machine vision joint displacement monitoring device, a data acquisition terminal, and a data processing terminal connected in sequence;

[0057] The data acquisition terminal acquires and stores dual-source displacement monitoring data from the BeiDou GNSS-machine vision joint displacement monitoring device; the data processing terminal acquires the dual-source displacement monitoring data stored in the data acquisition terminal and performs collaborative analysis; wherein, the dual-source displacement monitoring data includes BeiDou GNSS displacement monitoring data and machine vision displacement monitoring data.

[0058] In one specific embodiment of the present invention, the BeiDou GNSS-machine vision joint monitoring device is as follows: Figures 2-4 As shown, it includes a base 1 and a pole 2 arranged on the part of the hydraulic structure to be measured, a data acquisition electrical cabinet 3, a machine vision measuring instrument 4, a visual monitoring target 5, a camera 8 and a reflector 9 set on the pole 2, and a Beidou GNSS receiving antenna 6 and a Beidou GNSS receiver 7 set on the top of the pole 2.

[0059] Among them, the machine vision measuring instrument 4 and the visual monitoring target 5 are used to collect machine vision displacement monitoring data of hydraulic structures and store it in the data acquisition electrical cabinet 3; the Beidou GNSS receiving antenna 6 and the Beidou GNSS receiver are used to collect Beidou GNSS displacement monitoring data of hydraulic structures and store it in the data acquisition electrical cabinet 3; the data acquisition electrical cabinet 3 is also used to power the machine vision monitoring instrument and the Beidou GNSS monitoring instrument.

[0060] The camera 8 is used to acquire the direct image of the visual monitoring target 5, the visual virtual image formed by the mirror reflection of the mirror 9, and the precisely engraved scale lines on the visual monitoring target 5, and to store the acquisition results in the data acquisition electrical cabinet 3.

[0061] In practical applications, the BeiDou GNSS receiving antenna receives and amplifies satellite signals, and the BeiDou GNSS receiver processes and calculates the positioning of the signals received by the BeiDou GNSS receiving antenna, thereby collecting BeiDou GNSS displacement monitoring data of hydraulic structures.

[0062] In one specific embodiment of the present invention, the data processing terminal is configured with a dual-source displacement monitoring data fusion and analysis module; the dual-source displacement monitoring data fusion and analysis module includes a data input unit, a data preprocessing unit, a data fusion unit, and a data output unit connected in sequence;

[0063] The system includes a data input unit for receiving dual-source displacement monitoring data stored in the data acquisition terminal; a data preprocessing unit for preprocessing the dual-source displacement monitoring data to obtain preprocessed dual-source displacement monitoring data; a data fusion unit for fusing the preprocessed dual-source displacement monitoring data to obtain fused dual-source displacement monitoring data; and a data output unit for correcting the fused dual-source displacement monitoring data and outputting real-time displacement monitoring values ​​of the hydraulic structure.

[0064] Example 2

[0065] This embodiment introduces a method for real-time displacement monitoring of hydraulic structures, such as... Figure 1 As shown, it specifically includes the following:

[0066] Acquire preprocessed BeiDou GNSS displacement monitoring data and machine vision displacement monitoring data;

[0067] The preprocessed BeiDou GNSS displacement monitoring data is reconstructed in time series to obtain reconstructed BeiDou GNSS displacement monitoring data; this is used to ensure time-frequency consistency with machine vision displacement monitoring data.

[0068] A dual-source displacement monitoring data fusion model is constructed based on machine vision displacement monitoring data and the time-series reconstructed BeiDou GNSS displacement monitoring data.

[0069] The machine vision displacement monitoring data and the time-series reconstructed BeiDou GNSS displacement monitoring data are spatiotemporally registered and fused using a dual-source displacement monitoring data fusion model. The fusion results are then corrected to output real-time joint measurement displacement monitoring values ​​for hydraulic structures.

[0070] In one specific embodiment of the present invention, the machine vision displacement monitoring data is corrected by offset angle. Specifically, the real-time offset angle parameter of the machine vision displacement monitoring data is substituted into the displacement calculation model of the machine vision monitoring instrument to perform dynamic geometric correction on the visual displacement monitoring data. This ensures that the machine vision displacement monitoring data is stably monitored over a long period of time.

[0071] In practical applications, comparative analysis of BeiDou GNSS displacement monitoring data and machine vision displacement monitoring data reveals a systematic difference between the two on short timescales: the displacement change obtained by the former is usually greater than that obtained by the latter at the same time. However, on long timescales, the cumulative displacement difference presented by machine vision displacement monitoring data often surpasses the monitoring value of BeiDou satellites. In-depth analysis shows that the inherent geometric configuration of the machine vision monitoring instrument is the key factor causing this phenomenon. Specifically, the offset angle between the optical axis of camera 8 and the plane normal of the machine vision target 5 introduces a significant cumulative effect of systematic errors in long-term monitoring. This error amplifies with the increase of monitoring time, causing the apparent displacement value of the machine vision monitoring instrument to deviate from the true value.

[0072] To effectively suppress or eliminate the coupled influence of baseline distance and offset angle on the accuracy of machine vision monitoring, an innovative geometric error compensation scheme was proposed and implemented: a high-precision reflector 9 was installed on the pole 2. The installation height of the reflector 9 was strictly calibrated to ensure that its reflective surface was precisely at the same horizontal level as the center point of the visual monitoring target 5, forming a common viewing reference plane. Through this reflector, the camera 8 can simultaneously capture two key pieces of information: the direct image of the visual monitoring target 5 and the virtual image of the visual monitoring target 5 formed by the mirror reflection of the reflector 9, along with the precisely engraved scale lines on it.

[0073] The core of this scheme lies in using the geometric characteristics of the reflected light path of the mirror to dynamically measure the offset angle. In the reflected image acquired by camera 8, the position of the virtual image of the visual monitoring target 5's scale line will move due to the change in its relative geometric relationship with the mirror. By calculating the pixel-level offset of the target scale line virtual image in the reflected image relative to its preset zero-position reference line (0 scale line) in real time, and combining it with the known baseline distance and mirror position parameters, the real-time optical axis offset angle can be accurately calculated using the principles of geometric optics and triangulation.

[0074] By substituting the calculated real-time offset angle parameters into the displacement calculation model of the machine vision monitoring instrument, dynamic geometric correction can be performed on the original visual monitoring data. This process effectively compensates for the displacement measurement error introduced by the viewing angle deviation, significantly improving the accuracy and reliability of displacement change measurement in long-term monitoring of the machine vision system, making it closer to the high-precision absolute displacement reference provided by the BeiDou satellite navigation system.

[0075] In one specific embodiment of the present invention, correcting the fusion result includes: using a sliding window optimization framework to eliminate recursive errors and dynamically correcting the fusion result;

[0076] Within the sliding window optimization framework, the optimization objective simplifies to the weighted minimization of two residuals: the visual relative displacement residual and the GNSS absolute displacement constraint residual. These are used to ensure that the displacement difference between adjacent time points is consistent with the visual measurement value, and to ensure that the displacement at the end of the window is consistent with the GNSS measurement value, respectively.

[0077] In one specific embodiment of the present invention, the real-time displacement monitoring value of the hydraulic structure is obtained through the following steps:

[0078] 1. Data Preprocessing

[0079] 1.1 Preprocessing of BeiDou GNSS displacement monitoring data

[0080] First, variational mode decomposition is performed on the original BeiDou GNSS displacement monitoring data G(t) to obtain K intrinsic mode functions. This is done to separate different frequency components and reduce high-frequency noise. Then, for each intrinsic mode function... Applying wavelet thresholding for denoising further suppresses residual noise, yielding the denoised intrinsic mode functions. Finally, the denoised intrinsic mode function The noise-reduced BeiDou GNSS displacement monitoring data is obtained by summing the results. .

[0081] The original BeiDou GNSS displacement monitoring data G(t) is subjected to variational mode decomposition using the following formula:

[0082] ,

[0083] ,

[0084] In the formula, k is the index of the intrinsic mode function, which takes values ​​from 1 to K, and K is the total number of intrinsic mode functions. Let k be the intrinsic mode function after decomposition. Its corresponding center frequency. Numbering by time Let j be the Dirac function, and j be the wavelet transform level. This is a secondary exponential signal. The k-th intrinsic mode function after decomposition Furthermore, L2 regularization constraints are introduced to suppress over-decomposition of high-frequency noise; and the energy ratio termination criterion is used ( <1%), to avoid modal redundancy. The k-th intrinsic mode function component after decomposition energy, This represents the total energy of the original BeiDou GNSS displacement monitoring data.

[0085] Denoising is performed on each intrinsic mode function after decomposition using a translation-invariant wavelet threshold based on SURE theory, resulting in the denoised intrinsic mode functions. and BeiDou GNSS displacement monitoring data after noise reduction processing. The translation-invariant wavelet threshold based on SURE theory is used for noise reduction using the following formula:

[0086] ,

[0087] ,

[0088] In the formula, j represents the wavelet transform level, n represents the wavelet position index, and N represents the decomposed intrinsic mode function. Signal length, For the wavelet coefficients at the j-th layer and the n-th position, The absolute values ​​of the wavelet coefficients at the j-th layer and the n-th position reflect the amplitude of the intrinsic mode function signal. Let be the noise standard deviation of the wavelet coefficients at the j-th level;

[0089] Let J be the intrinsic mode function after denoising at the j-th layer and the n-th position. For BeiDou GNSS displacement monitoring data after noise reduction processing, the intrinsic mode function is obtained after noise reduction. The sum is obtained.

[0090] The sign function, preserving the phase information of the coefficients, is defined as:

[0091] .

[0092] 1.2 Preprocessing of machine vision displacement monitoring data

[0093] Based on smoothing filtering of machine vision displacement monitoring data Preprocessing is performed to improve the robustness and accuracy of the optimization, resulting in denoised machine vision displacement monitoring data. Let the original visual displacement measurement sequence be... The exponentially smoothed sequence is First-order exponential smoothing filters out high-frequency noise by weighting the current measurement with the smoothed value from the previous time step.

[0094] ,

[0095] ,

[0096] ,

[0097] In the formula, i is the sequence number. The smoothing coefficient determines the relative weight of the filter on new observations and historical smoothed values. When When the value is large, the filter response is more sensitive and can quickly track changes in the real signal; when When the filter is smaller, its reliance on historical data increases, allowing it to suppress high-frequency noise more effectively. (Machine vision displacement monitoring data) From the original visual displacement measurement sequence Composition, noise-reduced machine vision displacement monitoring data The sequence after exponential smoothing is constitute.

[0098] 2. Time-series reconstruction of BeiDou GNSS displacement monitoring data

[0099] Due to technological limitations, there are often differences in the data acquisition frequency between BeiDou GNSS displacement monitoring data and machine vision displacement monitoring data, with BeiDou GNSS displacement monitoring data typically having a lower acquisition frequency. To ensure the temporal consistency between BeiDou GNSS displacement monitoring data and machine vision high-frequency data sets, it is necessary to interpolate and reconstruct the BeiDou GNSS displacement monitoring data.

[0100] First, the noise-reduced BeiDou GNSS displacement monitoring data and machine vision displacement monitoring data The data was integrated and arranged chronologically to identify the time difference between BeiDou GNSS displacement monitoring data and machine vision displacement monitoring data, thus determining the number of missing BeiDou GNSS displacement monitoring data and the corresponding time points based on the machine vision displacement monitoring data. Secondly, for two adjacent GNSS monitoring times, a piecewise cubic Hermite interpolation polynomial was constructed, and the interpolation results were calculated to obtain the reconstructed BeiDou GNSS displacement monitoring data. Through reconstructed BeiDou GNSS displacement monitoring data It can achieve time-series alignment with machine vision displacement monitoring data.

[0101] 3. Fusion of dual-source displacement monitoring data

[0102] Noise-reduced machine vision displacement monitoring data and the time-reconstructed BeiDou GNSS displacement monitoring data The fusion process is performed, meaning the real-time displacement monitoring values ​​of the hydraulic structure are calculated using the following fusion formula:

[0103] ,

[0104] In the formula, The real-time displacement monitoring value of the hydraulic structure at time t is... Let be the machine vision displacement monitoring value at time t. The value is the BeiDou GNSS displacement monitoring value at time t.

[0105] When t=t+1, the increments of the machine vision displacement monitoring value and the BeiDou GNSS displacement monitoring value are as follows:

[0106] ,

[0107] ,

[0108] In the formula, and These represent the increments of machine vision monitoring values ​​and BeiDou GNSS monitoring values, respectively.

[0109] At time t+1, the real-time displacement monitoring value of the hydraulic structure is calculated using the following formula:

[0110] ,

[0111] In the formula, The displacement monitoring value of the hydraulic structure at time t+1 is the real-time joint measurement value. This represents the cumulative average of the measured values ​​from BeiDou GNSS displacement monitoring data up to time t+1. and The weights for machine vision displacement monitoring values ​​and BeiDou GNSS displacement monitoring values ​​are calculated using the following formulas:

[0112] ,

[0113] .

[0114] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A real-time displacement monitoring system for hydraulic structures, characterized in that, This includes a BeiDou GNSS-machine vision joint displacement monitoring device, a data acquisition terminal, and a data processing terminal connected in sequence. The data acquisition terminal acquires and stores dual-source displacement monitoring data from the BeiDou GNSS-machine vision joint displacement monitoring device; the data processing terminal acquires the dual-source displacement monitoring data stored in the data acquisition terminal and performs collaborative analysis; wherein the dual-source displacement monitoring data includes BeiDou GNSS displacement monitoring data and machine vision displacement monitoring data.

2. The multi-source fusion monitoring system for real-time displacement of hydraulic structures according to claim 1, characterized in that, The Beidou GNSS-machine vision joint monitoring device includes a base (1) and a pole (2) arranged on the part of the hydraulic structure to be measured, a data acquisition electrical cabinet (3), a machine vision measuring instrument (4), a visual monitoring target (5), a camera (8) and a reflector (9) set on the pole (2), and a Beidou GNSS receiving antenna (6) and a Beidou GNSS receiver (7) set on the top of the pole (2). The machine vision measuring instrument (4) and the visual monitoring target (5) are used to collect machine vision displacement monitoring data of hydraulic structures and store it in the data acquisition electrical cabinet (3); the Beidou GNSS receiving antenna (6) and the Beidou GNSS receiver are used to collect Beidou GNSS displacement monitoring data of hydraulic structures and store it in the data acquisition electrical cabinet (3); the data acquisition electrical cabinet (3) is also used to power the machine vision monitoring instrument and the Beidou GNSS monitoring instrument. The camera (8) is used to collect the direct image of the visual monitoring target (5), the visual virtual image formed by the mirror reflection of the mirror (9), and the scale lines engraved on the visual monitoring target (5), and to store the collection results in the data acquisition electrical cabinet (3).

3. The multi-source fusion monitoring system for real-time displacement of hydraulic structures according to claim 1, characterized in that, The data processing terminal is equipped with a dual-source displacement monitoring data fusion and analysis module; the dual-source displacement monitoring data fusion and analysis module includes a data input unit, a data preprocessing unit, a data fusion unit, and a data output unit connected in sequence; The data input unit is used to receive dual-source displacement monitoring data stored in the data acquisition terminal; the data preprocessing unit is used to preprocess the dual-source displacement monitoring data to obtain preprocessed dual-source displacement monitoring data; the data fusion unit is used to fuse the preprocessed dual-source displacement monitoring data to obtain fused dual-source displacement monitoring data; and the data output unit is used to correct the fused dual-source displacement monitoring data and output the real-time joint displacement monitoring value of the hydraulic structure.

4. The multi-source fusion monitoring system for real-time displacement of hydraulic structures according to claim 3, characterized in that, The real-time displacement monitoring value of the hydraulic structure is calculated using the following formula: ， In the formula, The real-time displacement monitoring value of the hydraulic structure at time t is... Let be the machine vision displacement monitoring value at time t. The value is the BeiDou GNSS displacement monitoring value at time t.

5. The multi-source fusion monitoring system for real-time displacement of hydraulic structures according to claim 4, characterized in that, When t=t+1, the increments of the machine vision displacement monitoring value and the BeiDou GNSS displacement monitoring value are as follows: ， ， In the formula, and These represent the increments of machine vision monitoring values ​​and BeiDou GNSS monitoring values, respectively. At time t+1, the real-time displacement monitoring value of the hydraulic structure is calculated using the following formula: ， In the formula, The displacement monitoring value of the hydraulic structure at time t+1 is the real-time joint measurement value. This represents the cumulative average of the measured values ​​from BeiDou GNSS displacement monitoring data up to time t+1. The weights of the machine vision displacement monitoring values. The weights of the BeiDou GNSS displacement monitoring values ​​are calculated using the following formulas: ， 。 6. A method for real-time displacement monitoring of hydraulic structures, characterized in that, The real-time structural displacement monitoring system as described in any one of claims 1-5 includes the following steps: The pre-processed BeiDou GNSS displacement monitoring data and machine vision displacement monitoring data are obtained through the data processing terminal. The preprocessed BeiDou GNSS displacement monitoring data is reconstructed in time series to obtain the reconstructed BeiDou GNSS displacement monitoring data. A dual-source displacement monitoring data fusion model is constructed based on the machine vision displacement monitoring data and the time-series reconstructed BeiDou GNSS displacement monitoring data. The dual-source displacement monitoring data fusion model performs spatiotemporal registration and fusion of the machine vision displacement monitoring data and the time-series reconstructed BeiDou GNSS displacement monitoring data, and corrects the fusion results to output real-time joint measurement displacement monitoring values ​​of hydraulic structures.

7. The real-time displacement monitoring method for hydraulic structures according to claim 6, characterized in that, The machine vision displacement monitoring data is corrected by offset angle. Specifically, the real-time offset angle parameter of the machine vision displacement monitoring data is substituted into the displacement calculation model of the machine vision monitoring instrument to perform dynamic geometric correction on the visual displacement monitoring data.

8. The method for real-time displacement monitoring of hydraulic structures according to claim 7, characterized in that, The correction of the fusion result includes: using a sliding window optimization framework to eliminate recursive errors and dynamically correcting the fusion result; Under the sliding window optimization framework, the optimization objective is simplified to the weighted minimization of two residuals, namely the visual relative displacement residual and the GNSS absolute displacement constraint residual.

9. The method for real-time displacement monitoring of hydraulic structures according to claim 6, characterized in that, The real-time displacement monitoring value of the hydraulic structure is calculated using the following formula: ， In the formula, The real-time displacement monitoring value of the hydraulic structure at time t is... Let be the machine vision displacement monitoring value at time t. The value is the BeiDou GNSS displacement monitoring value at time t.

10. The method for real-time displacement monitoring of hydraulic structures according to claim 9, characterized in that, When t=t+1, the increments of the machine vision displacement monitoring value and the BeiDou GNSS displacement monitoring value are as follows: ， ， In the formula, and These represent the increments of machine vision monitoring values ​​and BeiDou GNSS monitoring values, respectively. At time t+1, the real-time displacement monitoring value of the hydraulic structure is calculated using the following formula: ， In the formula, The displacement monitoring value of the hydraulic structure at time t+1 is the real-time joint measurement value. This represents the cumulative average of the measured values ​​from BeiDou GNSS displacement monitoring data up to time t+1. The weights of the machine vision displacement monitoring values. The weights of the BeiDou GNSS displacement monitoring values ​​are calculated using the following formulas: ， 。

Images