Bridge structure health monitoring method and system based on distributed optical fiber sensing system
By using a distributed optical fiber sensing system and ensemble empirical mode decomposition method to preprocess and perform parallel computation on bridge vibration data, the limitations of traditional methods in processing nonlinear and non-stationary signals are overcome, and efficient, accurate and timely monitoring of bridge structural health is achieved.
Patent Information
- Application Number
- CN202511040800.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-28
- Publication Date
- 2025-11-11
AI Technical Summary
Traditional bridge health monitoring methods are difficult to effectively handle nonlinear and non-stationary vibration signals, and the EEMD method suffers from noise interference and high computational complexity.
A distributed fiber optic sensing system was adopted. The bridge vibration data was preprocessed and divided into multiple sub-data blocks. Parallel computation was performed and combined with the ensemble empirical mode decomposition method to perform time series splicing and frequency component matching, construct a health state reference model, and use similarity calculation to judge changes in bridge structure.
It improves the accuracy and timeliness of bridge structural health monitoring, reduces processing time, optimizes the use of computing resources, can run smoothly on large-scale datasets, and issues timely warnings.
Smart Images
Figure CN120930012A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of bridge structural health monitoring technology, and in particular to a bridge structural health monitoring method and system based on a distributed optical fiber sensing system. Background Technology
[0002] Ensemble Empirical Mode Decomposition (EEMD) is a signal analysis method commonly used in bridge health monitoring, particularly suitable for handling nonlinear and non-stationary vibration signals. With the increasing severity of infrastructure aging problems, bridge health monitoring and structural safety assessment have become important research areas. Traditional bridge health monitoring methods, such as vibration analysis based on accelerometers or strain gauges, while effectively capturing the vibration characteristics of bridges, often face many challenges, such as significant susceptibility to noise interference and difficulty in handling complex nonlinear and non-stationary signals. Traditional signal processing methods, such as Fourier transform and wavelet transform, typically assume that the signal is linear and steady-state, which limits their application in processing bridge vibration signals. To overcome these problems, EEMD, as an emerging adaptive signal decomposition method, provides a powerful tool.
[0003] However, while the EEMD method can effectively handle the nonlinear and nonstationary characteristics in vibration, earthquake and meteorological sensor data, it also faces some challenges, such as noise, high computational complexity and difficulty in real-time processing. Summary of the Invention
[0004] To address the technical problems existing in the background art, this invention proposes a method and system for monitoring the health of bridge structures based on a distributed optical fiber sensing system.
[0005] In a first aspect, the present invention proposes a bridge structural health monitoring method based on a distributed optical fiber sensing system, comprising:
[0006] Acquire bridge vibration data collected by a distributed fiber optic sensing system;
[0007] Preprocess the bridge vibration data;
[0008] Based on a preset time window or spatial segment, the preprocessed bridge vibration data is divided into multiple independently processable sub-data blocks;
[0009] The ensemble empirical mode decomposition method is used to perform parallel computation on multiple sub-data blocks to obtain the decomposition results of multiple sub-data blocks;
[0010] The decomposition results of multiple sub-data blocks are spliced together in time series and frequency component matching to obtain the final decomposition result;
[0011] Based on the final decomposition results, determine whether the bridge structure has changed;
[0012] If so, an early warning will be issued.
[0013] Preferably, the bridge vibration data is preprocessed, specifically including:
[0014] Median removal processing was performed on the bridge vibration data;
[0015] Linear detrending processing is performed on the bridge vibration data after median removal;
[0016] High-frequency noise and low-frequency interference in bridge vibration data that has undergone linear detrending processing are removed by bandpass filtering.
[0017] Preferably, after preprocessing the bridge vibration data, the method further includes:
[0018] Normalize or standardize the preprocessed bridge vibration data.
[0019] Preferably, after normalizing or standardizing the preprocessed bridge vibration data, the method further includes:
[0020] Normalized or standardized bridge vibration data are corrected using interpolation completion, sliding window filtering, or outlier detection and removal methods.
[0021] Preferably, the decomposition result of each sub-data block is:
[0022]
[0023] In the formula, IMF k (t) is the k-th eigenmode function of the t-th sub-data block, i.e., the decomposition result of the t-th sub-data block; IMF i (t) is the intrinsic mode function of the i-th computation node of the t-th sub-data block, and N is the number of computation nodes of the t-th sub-data block.
[0024] Preferably, based on the final decomposition results, it is determined whether the bridge structure has changed, specifically including:
[0025] Historical fiber optic vibration data of a bridge structure in a healthy state are acquired using a distributed fiber optic sensing system.
[0026] Based on historical fiber vibration data, statistical modeling is performed on the eigenvectors of each eigenmode function of the bridge structure in a healthy state, constructing a modal feature space reference model for the healthy state, and forming a healthy state reference vector.
[0027] Extract the eigenvectors of each order of intrinsic mode functions from the final decomposition result;
[0028] The similarity distance is obtained by comparing the eigenvectors of each intrinsic mode function in the final decomposition result with the health status reference vector.
[0029] Based on the similarity distance and the confidence interval of the preset empirical threshold or reference distribution, determine whether there are modal changes in the bridge structure;
[0030] When the similarity distance exceeds the empirical threshold or the confidence interval of the reference distribution, it is determined that the bridge structure has modal changes.
[0031] When a bridge structure exhibits modal changes, it is determined that the bridge structure has potential damage or stiffness changes.
[0032] Preferably, the similarity calculation method includes Euclidean distance, cosine similarity, or Mahalanobis distance.
[0033] Preferably, issuing an early warning includes:
[0034] Obtain the mapping relationship between optical fibers and real-world locations in a distributed optical fiber sensing system;
[0035] Based on the mapping relationship between optical fibers and real-world locations, the location of bridge structures that have suffered damage or changes in stiffness can be determined.
[0036] Early warnings are issued based on the location of bridge structures that have suffered damage or experienced changes in stiffness.
[0037] Secondly, this invention also proposes a bridge structural health monitoring system based on a distributed optical fiber sensing system, comprising:
[0038] The acquisition module is used to acquire bridge vibration data collected by the distributed fiber optic sensing system.
[0039] The pre-decomposition and integration module is used to preprocess bridge vibration data;
[0040] The decomposition and integration module is used to divide the preprocessed bridge vibration data into multiple independently processable sub-data blocks according to a preset time window or spatial segment; to perform parallel calculations on the multiple sub-data blocks using the ensemble empirical mode decomposition method to obtain the decomposition results of the multiple sub-data blocks; and to perform time series splicing and frequency component matching on the decomposition results of the multiple sub-data blocks to obtain the final decomposition result.
[0041] The judgment module is used to determine whether the bridge structure has changed based on the final decomposition results;
[0042] The early warning module is used to issue an early warning when changes are detected in the bridge structure.
[0043] Preferably, the bridge vibration data is preprocessed, specifically including:
[0044] Median removal processing was performed on the bridge vibration data;
[0045] Linear detrending processing is performed on the bridge vibration data after median removal;
[0046] High-frequency noise and low-frequency interference in bridge vibration data that has undergone linear detrending processing are removed by bandpass filtering.
[0047] Preferably, after preprocessing the bridge vibration data, the method further includes:
[0048] Normalize or standardize the preprocessed bridge vibration data.
[0049] Preferably, after normalizing or standardizing the preprocessed bridge vibration data, the method further includes:
[0050] Normalized or standardized bridge vibration data are corrected using interpolation completion, sliding window filtering, or outlier detection and removal methods.
[0051] Preferably, the decomposition result of each sub-data block is:
[0052]
[0053] In the formula, IMF k (t) is the k-th eigenmode function of the t-th sub-data block, i.e., the decomposition result of the t-th sub-data block; IMF i (t) is the intrinsic mode function of the i-th computation node of the t-th sub-data block, and N is the number of computation nodes of the t-th sub-data block.
[0054] Preferably, based on the final decomposition results, it is determined whether the bridge structure has changed, specifically including:
[0055] Historical fiber optic vibration data of a bridge structure in a healthy state are acquired using a distributed fiber optic sensing system.
[0056] Based on historical fiber vibration data, statistical modeling is performed on the eigenvectors of each eigenmode function of the bridge structure in a healthy state, constructing a modal feature space reference model for the healthy state, and forming a healthy state reference vector.
[0057] Extract the eigenvectors of each order of intrinsic mode functions from the final decomposition result;
[0058] The similarity distance is obtained by comparing the eigenvectors of each intrinsic mode function in the final decomposition result with the health status reference vector.
[0059] Based on the similarity distance and the confidence interval of the preset empirical threshold or reference distribution, determine whether there are modal changes in the bridge structure;
[0060] When the similarity distance exceeds the empirical threshold or the confidence interval of the reference distribution, it is determined that the bridge structure has modal changes.
[0061] When a bridge structure exhibits modal changes, it is determined that the bridge structure has potential damage or stiffness changes.
[0062] Preferably, the similarity calculation method includes Euclidean distance, cosine similarity, or Mahalanobis distance.
[0063] Preferably, issuing an early warning includes:
[0064] Obtain the mapping relationship between optical fibers and real-world locations in a distributed optical fiber sensing system;
[0065] Based on the mapping relationship between optical fibers and real-world locations, the location of bridge structures that have suffered damage or changes in stiffness can be determined.
[0066] Early warnings are issued based on the location of bridge structures that have suffered damage or experienced changes in stiffness.
[0067] The bridge structural health monitoring method and system proposed in this invention, based on a distributed optical fiber sensing system, effectively removes noise components from the bridge vibration data collected by the DAS (distributed optical fiber sensing system), thereby optimizing signal quality; and
[0068] Based on a preset time window or spatial segment, the preprocessed bridge vibration data is divided into multiple independently processable sub-data blocks. The ensemble empirical mode decomposition method is used to perform parallel computation on the multiple sub-data blocks to obtain the decomposition results of the multiple sub-data blocks. The decomposition results of the multiple sub-data blocks are then spliced together in time series and frequency component matching to obtain the final decomposition result. Based on the final decomposition result, it is determined whether the bridge structure has changed, that is, the health status of the bridge is monitored, and an early warning is issued in a timely manner when structural changes occur in the bridge structure.
[0069] This invention effectively improves the accuracy of bridge structural health monitoring results by combining DAS and ensemble empirical mode decomposition methods, and by using parallel computation of ensemble empirical mode decomposition methods. By decomposing the signal processing task onto multiple computing units, it not only reduces processing time and achieves high efficiency in data processing, but also optimizes the use of computing resources through multi-core processors or GPU acceleration technology, enabling EEMD to run smoothly on large-scale datasets and effectively improving the timeliness of bridge structural health monitoring results. Attached Figure Description
[0070] Figure 1 This is a flowchart illustrating a bridge structure health monitoring method based on a distributed optical fiber sensing system, as proposed in one embodiment of the present invention.
[0071] Figure 2 This is a schematic diagram of the final decomposition result in one embodiment of the present invention.
[0072] Figure 3 This is a schematic diagram of the final decomposition result in another embodiment of the present invention. Detailed Implementation
[0073] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.
[0074] Reference Figure 1 The present invention proposes a method for monitoring the structural health of bridges based on a distributed optical fiber sensing system, comprising:
[0075] Acquire bridge vibration data collected by a distributed fiber optic sensing system;
[0076] Preprocess the bridge vibration data;
[0077] Based on a preset time window or spatial segment, the preprocessed bridge vibration data is divided into multiple independently processable sub-data blocks;
[0078] The ensemble empirical mode decomposition method is used to perform parallel computation on multiple sub-data blocks to obtain the decomposition results of multiple sub-data blocks;
[0079] The decomposition results of multiple sub-data blocks are spliced together in time series and frequency component matching to obtain the final decomposition result;
[0080] Based on the final decomposition results, determine whether the bridge structure has changed;
[0081] If so, a warning will be issued.
[0082] During signal acquisition, bridge vibration data collected by DAS (Distributed Fiber Optic Sensing System) is often affected by environmental noise, equipment noise, and fiber optic transmission interference, directly impacting the accuracy of subsequent EEMD decomposition processing. This invention addresses the complex vibration signals commonly encountered in bridge health monitoring by preprocessing the bridge vibration data collected by DAS to effectively remove noise components, optimize signal quality, and improve the accuracy of subsequent bridge structural health monitoring results. Based on preset time windows or spatial segments, the preprocessed bridge vibration data is divided into multiple independently processable sub-data blocks. The ensemble empirical mode decomposition method is used to perform parallel computation on these sub-data blocks, yielding decomposition results. These results are then time-series concatenated and frequency component matched to obtain the final decomposition result. In other words, the signal processing task is decomposed across multiple computational units through parallel computation. This parallel computation not only reduces processing time and achieves high data processing efficiency but also further optimizes the use of computational resources through multi-core processors or GPU acceleration technology, enabling EEMD to run smoothly on large-scale datasets. Moreover, based on the rapidly obtained final decomposition results, this invention can quickly and accurately determine whether the bridge structure has changed, that is, monitor the health status of the bridge, and issue timely warnings when structural changes occur in the bridge structure, effectively improving the timeliness of bridge structure monitoring results.
[0083] To ensure the purity of the bridge vibration signal when subsequently processed in parallel using the ensemble empirical mode decomposition method, in one specific embodiment, the bridge vibration data is preprocessed, specifically including:
[0084] Median-reducing processing was performed on the bridge vibration data to eliminate sensor baseline offset;
[0085] Linear detrending processing is performed on the median-devalued bridge vibration data to suppress low-frequency trend terms caused by temperature changes or slow structural changes.
[0086] High-frequency noise and low-frequency interference in bridge vibration data that has undergone linear detrending processing are removed by bandpass filtering, leaving only the main frequency bands of the structural response (e.g., 0.5Hz to 50Hz).
[0087] In a further embodiment, after preprocessing the bridge vibration data, the method further includes:
[0088] The preprocessed bridge vibration data is normalized or standardized to further improve the consistency and comparability of subsequent EEMD decomposition.
[0089] To address issues such as data abrupt changes, missing data, or glitch signals during the acquisition process, a further embodiment includes, after normalizing or standardizing the preprocessed bridge vibration data, the following steps are also taken:
[0090] The normalized or standardized bridge vibration data is corrected by interpolation completion, sliding window filtering, or outlier detection and removal methods, thereby ensuring the stability and reliability of the subsequent decomposition process.
[0091] It's important to understand that the EEMD method decomposes the raw bridge vibration data into a series of intrinsic mode functions (IMFs) and residual terms; its formula is as follows:
[0092]
[0093] Where x(t) is the original bridge vibration data, IMF k r(t) is the k-th intrinsic mode function, r(t) is the residual term, and K is the number of intrinsic mode functions.
[0094] However, in the parallel computing process of this embodiment, it is assumed that the original bridge vibration signal data is divided into N sub-data blocks, and the decomposition result of each sub-data block is as follows:
[0095]
[0096] In the formula, IMF k (t) is the k-th eigenmode function of the t-th sub-data block, i.e., the decomposition result of the t-th sub-data block; IMF i (t) is the intrinsic mode function of the i-th computation node of the t-th sub-data block, and N is the number of computation nodes of the t-th sub-data block.
[0097] To accurately determine whether the bridge structure has changed, this embodiment determines whether the bridge structure has changed based on the final decomposition results, specifically including:
[0098] Historical fiber optic vibration data of a bridge structure in a healthy state are acquired using a distributed fiber optic sensing system.
[0099] Based on historical fiber vibration data, statistical modeling is performed on the eigenvectors of each eigenmode function of the bridge structure in a healthy state, constructing a modal feature space reference model for the healthy state, and forming a healthy state reference vector.
[0100] Extract the eigenvectors of each order of intrinsic mode functions from the final decomposition result;
[0101] The similarity distance is obtained by comparing the eigenvectors of each intrinsic mode function in the final decomposition result with the health status reference vector.
[0102] Based on the similarity distance and the confidence interval of the preset empirical threshold or reference distribution, determine whether there are modal changes in the bridge structure;
[0103] When the similarity distance exceeds the empirical threshold or the confidence interval of the reference distribution, it is determined that the bridge structure has modal changes.
[0104] When a bridge structure exhibits modal changes, it is determined that the bridge structure has potential damage or stiffness changes.
[0105] The similarity calculation methods in this embodiment include Euclidean distance, cosine similarity, or Mahalanobis distance.
[0106] Among them, the health status reference vector refers to the stable feature representation extracted by statistically analyzing the modal feature vectors obtained from multiple time periods or multiple observations under the health status, such as the feature space, so as to construct a standard modal representation, which is convenient for similarity matching or deviation analysis with the current feature vector in subsequent state change detection.
[0107] In this embodiment, the early warning is specifically implemented as follows:
[0108] Obtain the mapping relationship between optical fibers and actual locations in a distributed optical fiber sensing system; locate the position of a bridge structure that has been damaged or has changed stiffness based on the mapping relationship between optical fibers and actual locations.
[0109] Early warnings are issued based on the location of bridge structures that have suffered damage or changes in stiffness, in order to facilitate accurate inspection and repair.
[0110] Among them, the fiber optic deployment and channel spatial mapping are achieved by using the signal from tapping the fiber to locate the channel and correspond it to a physical point on the bridge, thus forming a mapping relationship between the fiber and the actual location. It can be assumed that each channel of the DAS represents a spatial location (e.g., a distribution with 1m intervals). Based on the fiber vibration data, the DAS of the bridge structure that has been damaged or has changed stiffness is determined, thereby determining the spatial location of the bridge structure that has been damaged or has changed stiffness.
[0111] Secondly, this invention also proposes a bridge structural health monitoring system based on a distributed optical fiber sensing system, comprising:
[0112] The acquisition module is used to acquire bridge vibration data collected by the distributed fiber optic sensing system.
[0113] The preprocessing module is used to preprocess bridge vibration data;
[0114] The decomposition and integration module is used to divide the preprocessed bridge vibration data into multiple independently processable sub-data blocks according to a preset time window or spatial segment; to perform parallel calculations on the multiple sub-data blocks using the ensemble empirical mode decomposition method to obtain the decomposition results of the multiple sub-data blocks; and to perform time series splicing and frequency component matching on the decomposition results of the multiple sub-data blocks to obtain the final decomposition result.
[0115] The judgment module is used to determine whether the bridge structure has changed based on the final decomposition results;
[0116] The early warning module is used to issue an early warning when changes are detected in the bridge structure.
[0117] In this embodiment, the bridge vibration data is preprocessed, specifically including:
[0118] Median removal processing was performed on the bridge vibration data;
[0119] Linear detrending processing is performed on the bridge vibration data after median removal;
[0120] High-frequency noise and low-frequency interference in bridge vibration data that has undergone linear detrending processing are removed by bandpass filtering.
[0121] In this embodiment, after preprocessing the bridge vibration data, the following steps are also included:
[0122] Normalize or standardize the preprocessed bridge vibration data.
[0123] In this embodiment, after normalizing or standardizing the preprocessed bridge vibration data, the method further includes:
[0124] Normalized or standardized bridge vibration data are corrected using interpolation completion, sliding window filtering, or outlier detection and removal methods.
[0125] In this embodiment, the decomposition result of each sub-data block is:
[0126]
[0127] In the formula, IMF k (t) is the k-th eigenmode function of the t-th sub-data block, i.e., the decomposition result of the t-th sub-data block; IMF i (t) is the intrinsic mode function of the i-th computation node of the t-th sub-data block, and N is the number of computation nodes of the t-th sub-data block.
[0128] In this embodiment, determining whether the bridge structure has changed based on the final decomposition results specifically includes:
[0129] Historical fiber optic vibration data of a bridge structure in a healthy state are acquired using a distributed fiber optic sensing system.
[0130] Based on historical fiber vibration data, statistical modeling is performed on the eigenvectors of each eigenmode function of the bridge structure in a healthy state, constructing a modal feature space reference model for the healthy state, and forming a healthy state reference vector.
[0131] Extract the eigenvectors of each order of intrinsic mode functions from the final decomposition result;
[0132] The similarity distance is obtained by comparing the eigenvectors of each intrinsic mode function in the final decomposition result with the health status reference vector.
[0133] Based on the similarity distance and the confidence interval of the preset empirical threshold or reference distribution, determine whether there are modal changes in the bridge structure;
[0134] When the similarity distance exceeds the empirical threshold or the confidence interval of the reference distribution, it is determined that the bridge structure has modal changes.
[0135] When a bridge structure exhibits modal changes, it is determined that the bridge structure has potential damage or stiffness changes.
[0136] The similarity calculation methods in this embodiment include Euclidean distance, cosine similarity, or Mahalanobis distance;
[0137] In this embodiment, the early warning is specifically implemented as follows:
[0138] Obtain the mapping relationship between optical fibers and actual locations in a distributed optical fiber sensing system; locate the position of a bridge structure that has been damaged or has changed stiffness based on the mapping relationship between optical fibers and actual locations.
[0139] Early warnings are issued based on the location of bridge structures that have suffered damage or changes in stiffness, in order to facilitate accurate inspection and repair.
[0140] In one specific embodiment, vibration data of the Haiwen Bridge is collected using a DAS device, and the bridge structural health monitoring method based on a distributed optical fiber sensing system proposed in this embodiment is used to perform two tests on the structural health status of the Haiwen Bridge. The final decomposition result is as follows: Figure 2 and Figure 3 As shown.
[0141] The EEMD method is used to process the data. Results show that the signal decomposed by EEMD can accurately reflect the health status of the bridge and provide timely warnings when structural changes occur. For example... Figure 2 As shown.
[0142] exist Figure 2 and Figure 3 In the graph, the horizontal axis represents time, indicating the position of each data point on the time axis; the vertical axis represents the signal amplitude, indicating the value of each intrinsic mode function (IMF) at each time point. The amplitude reflects the intensity of the signal variation at various frequency components.
[0143] exist Figure 2 and Figure 3 In the diagram, the amplitude curves from top to bottom represent the original signal, IMF1 to IMF16, corresponding to the low-frequency to high-frequency decomposed signals. This is achieved through comparison with... Figure 2 Comparing the amplitudes of the original signals on the left side, it is concluded that IMF1 to IMF3 are closest in shape to the original signals, indicating that these three modes contain the main information components of the main bridge's vibration response, reflecting that the main component of the main bridge's vibration signal is a low-frequency signal. This identification result, as a basis for modal selection, can be further used in feature extraction, structural state discrimination, and other processing steps, enhancing the accuracy and specificity of the damage identification process.
[0144] 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 method for monitoring the structural health of bridges based on a distributed optical fiber sensing system, characterized in that, include: Acquire bridge vibration data collected by a distributed fiber optic sensing system; Preprocess the bridge vibration data; Based on a preset time window or spatial segment, the preprocessed bridge vibration data is divided into multiple independently processable sub-data blocks; The ensemble empirical mode decomposition method is used to perform parallel computation on multiple sub-data blocks to obtain the decomposition results of multiple sub-data blocks; The decomposition results of multiple sub-data blocks are spliced together in time series and frequency component matching to obtain the final decomposition result; Based on the final decomposition results, determine whether the bridge structure has changed; If so, an early warning will be issued.
2. The bridge structural health monitoring method based on a distributed optical fiber sensing system according to claim 1, characterized in that, Preprocessing of bridge vibration data includes: Median removal processing was performed on the bridge vibration data; Linear detrending processing is performed on the bridge vibration data after median removal; High-frequency noise and low-frequency interference in bridge vibration data that has undergone linear detrending processing are removed by bandpass filtering.
3. The bridge structural health monitoring method based on a distributed optical fiber sensing system according to claim 2, characterized in that, After preprocessing the bridge vibration data, the following steps are also included: Normalize or standardize the preprocessed bridge vibration data; Preferably, after normalizing or standardizing the preprocessed bridge vibration data, the method further includes: Normalized or standardized bridge vibration data are corrected using interpolation completion, sliding window filtering, or outlier detection and removal methods.
4. The bridge structural health monitoring method based on a distributed optical fiber sensing system according to claim 1, characterized in that, The decomposition result of each sub-data block is In the formula, IMF k (t) is the k-th eigenmode function of the t-th sub-data block, i.e., the decomposition result of the t-th sub-data block; IMF i (t) is the intrinsic mode function of the i-th computation node of the t-th sub-data block, and N is the number of computation nodes of the t-th sub-data block.
5. The bridge structural health monitoring method based on a distributed optical fiber sensing system according to claim 1, characterized in that, Based on the final decomposition results, determine whether the bridge structure has changed, specifically including: Historical fiber optic vibration data of a bridge structure in a healthy state are acquired using a distributed fiber optic sensing system. Based on historical fiber vibration data, statistical modeling is performed on the eigenvectors of each eigenmode function of the bridge structure in a healthy state, constructing a modal feature space reference model for the healthy state, and forming a healthy state reference vector. Extract the eigenvectors of each order of intrinsic mode functions from the final decomposition result; The similarity distance is obtained by comparing the eigenvectors of each intrinsic mode function in the final decomposition result with the health status reference vector. Based on the similarity distance and the confidence interval of the preset empirical threshold or reference distribution, determine whether there are modal changes in the bridge structure; When the similarity distance exceeds the empirical threshold or the confidence interval of the reference distribution, it is determined that the bridge structure has modal changes. When a bridge structure exhibits modal changes, it is determined that the bridge structure has potential damage or stiffness changes. Preferably, the similarity calculation method includes Euclidean distance, cosine similarity, or Mahalanobis distance; Preferably, issuing an early warning includes: Obtain the mapping relationship between optical fibers and real-world locations in a distributed optical fiber sensing system; Based on the mapping relationship between optical fibers and real-world locations, the location of bridge structures that have suffered damage or changes in stiffness can be determined. Early warnings are issued based on the location of bridge structures that have suffered damage or experienced changes in stiffness.
6. A bridge structural health monitoring system based on a distributed optical fiber sensing system, characterized in that, include: The acquisition module is used to acquire bridge vibration data collected by the distributed fiber optic sensing system. The pre-decomposition and integration module is used to preprocess bridge vibration data; The decomposition and integration module is used to divide the preprocessed bridge vibration data into multiple independently processable sub-data blocks according to a preset time window or spatial segment; and to perform parallel calculations on the multiple sub-data blocks using the ensemble empirical mode decomposition method to obtain the decomposition results of the multiple sub-data blocks. The decomposition results of multiple sub-data blocks are concatenated into time series and frequency component matched to obtain the final decomposition result. The judgment module is used to determine whether the bridge structure has changed based on the final decomposition results; The early warning module is used to issue an early warning when changes are detected in the bridge structure.
7. The bridge structural health monitoring system based on a distributed optical fiber sensing system according to claim 6, characterized in that, Preprocessing of bridge vibration data includes: Median removal processing was performed on the bridge vibration data; Linear detrending processing is performed on the bridge vibration data after median removal; High-frequency noise and low-frequency interference in bridge vibration data that has undergone linear detrending processing are removed by bandpass filtering.
8. The bridge structural health monitoring system based on a distributed optical fiber sensing system according to claim 1, characterized in that, After preprocessing the bridge vibration data, the following steps are also included: Normalize or standardize the preprocessed bridge vibration data; Preferably, after normalizing or standardizing the preprocessed bridge vibration data, the method further includes: Normalized or standardized bridge vibration data are corrected using interpolation completion, sliding window filtering, or outlier detection and removal methods.
9. The bridge structural health monitoring system based on a distributed optical fiber sensing system according to claim 1, characterized in that, The decomposition result of each sub-data block is as follows: In the formula, IMF k (t) is the eigenmode function of the t-th sub-data block, i.e., the decomposition result of the t-th sub-data block; IMF i (t) is the intrinsic mode function of the i-th computation node of the t-th sub-data block, and N is the number of computation nodes of the t-th sub-data block.
10. The bridge structural health monitoring system based on a distributed optical fiber sensing system according to claim 1, characterized in that, Based on the final decomposition results, determine whether the bridge structure has changed, specifically including: Historical fiber optic vibration data of a bridge structure in a healthy state are acquired using a distributed fiber optic sensing system. Based on historical fiber vibration data, statistical modeling is performed on the eigenvectors of each eigenmode function of the bridge structure in a healthy state, constructing a modal feature space reference model for the healthy state, and forming a healthy state reference vector. Extract the eigenvectors of each order of intrinsic mode functions from the final decomposition result; The similarity distance is obtained by comparing the eigenvectors of each intrinsic mode function in the final decomposition result with the health status reference vector. Based on the similarity distance and the confidence interval of the preset empirical threshold or reference distribution, determine whether there are modal changes in the bridge structure; When the similarity distance exceeds the empirical threshold or the confidence interval of the reference distribution, it is determined that the bridge structure has modal changes. When a bridge structure exhibits modal changes, it is determined that the bridge structure has potential damage or stiffness changes. Preferably, the similarity calculation method includes Euclidean distance, cosine similarity, or Mahalanobis distance; Preferably, issuing an early warning includes: Obtain the mapping relationship between optical fibers and real-world locations in a distributed optical fiber sensing system; Based on the mapping relationship between optical fibers and real-world locations, the location of bridge structures that have suffered damage or changes in stiffness can be determined. Early warnings are issued based on the location of bridge structures that have suffered damage or experienced changes in stiffness.