Dam structure intelligent monitoring method based on distributed optical fiber vibration sensing technology

Abstract

The invention discloses an intelligent dam structure monitoring method based on a distributed optical fiber vibration sensing technology, and relates to the technical field of monitoring, and the method comprises the steps: obtaining an echo response signal of each optical fiber channel based on a plurality of optical fibers arranged along a dam, calculating the response inertia time of each channel, and constructing a space-response inertia point set; judging whether the plurality of channels belong to homologous coupling response or not; if it is judged that the response is homologous coupling response, the abnormal degree value of each channel is calculated to determine an abnormal source channel; and performing abnormal positioning analysis on the echo response signal of the abnormal source channel to determine the abnormal position of the dam structure. According to the method, when the dam is monitored to be abnormal through the optical fiber vibration sensing technology, whether the abnormal condition monitored by the multiple optical fiber channels is true or caused by response coupling of the other optical fiber channels can be judged, it is ensured that the monitored dam abnormal condition is true, and the problems of false alarm amplification, abnormal positioning deviation or data fusion failure are reduced.

Description

Technical Field

[0001] This invention relates to the field of monitoring technology, specifically to an intelligent monitoring method for dam structures based on distributed fiber optic vibration sensing technology. Background Technology

[0002] In dam structural health monitoring, distributed fiber optic vibration sensing technology has gradually become an important online monitoring method due to its high sensitivity, high spatial resolution, and long-distance coverage. This technology typically involves deploying continuous single-mode optical fibers along the interior or surface of the dam, utilizing the fiber's sensitivity to minute strain and vibration changes. When the monitoring host periodically injects laser pulses into the fiber, the Rayleigh scattering echo signal in the fiber undergoes subtle phase changes in response to structural stress or vibration variations. By demodulating the time delay and intensity changes of the echo signal, the system can achieve meter-level spatial positioning and identify structural hazards such as abnormal strain accumulation and vibration surges, thus enabling the monitoring of high-risk evolution processes such as leakage, piping, and collapse.

[0003] To further enhance spatial coverage density and structural detail perception capabilities, multiple optical fibers are often deployed side-by-side in dam monitoring during practical engineering projects. For example, a double-layered deployment can cover structures at different depths, or a three-fiber deployment can enhance lateral anomaly detection. This parallel deployment significantly improves the spatial resolution of the monitoring system and its three-dimensional anomaly identification capabilities, facilitating the early detection of complex issues such as localized slippage, asymmetric settlement, and deep seepage. Simultaneously, data from multiple fiber channels can form a mutually referencing response network, providing higher confidence and fault tolerance for anomaly identification.

[0004] However, while parallel deployment of multiple optical fibers improves resolution, it also brings new technical challenges. Since the optical fibers in each channel often coexist in the same soil medium, the same structural disturbance (such as localized collapse or piping) may simultaneously or nearly simultaneously excite multiple channels to produce similar responses. This makes it difficult for the system to determine whether a single anomaly source triggers a multi-point response or whether anomalies occur at multiple structural locations separately. Without an effective multi-channel response coupling identification mechanism and anomaly origin point determination method, the system is highly susceptible to false alarm amplification, anomaly location misalignment, or data fusion failure, thereby affecting the accuracy of monitoring results and the reliability of engineering decisions. Summary of the Invention

[0005] The purpose of this invention is to solve the problems mentioned above and provide an intelligent monitoring method for dam structures based on distributed optical fiber vibration sensing technology.

[0006] This invention proposes an intelligent monitoring method for dam structures based on distributed fiber optic vibration sensing technology, the method comprising: Based on multiple optical fibers laid along the embankment, the echo response signals of each optical fiber channel are obtained. Based on the echo response signals of each fiber optic channel, calculate the response inertia time of each channel and construct a space-response inertia point set; determine whether multiple channels belong to the same source coupling response based on the space-response inertia point set. If it is determined to be a response of the same source, then based on the echo response signal of each channel, the abnormality degree value of each channel is calculated, and the abnormality source channel is determined according to the abnormality degree value. Anomaly localization analysis is performed on the echo response signal of the abnormal source channel to determine the abnormal location of the dam structure.

[0007] Optionally, the steps for calculating the response inertial time of each fiber optic channel and constructing a space-response inertial point set based on the echo response signal of each fiber optic channel are as follows: For each optical fiber channel deployed in the dam structure, the echo response signal of each channel within the current preset monitoring period is extracted. For each fiber optic channel The rise segment of the echo response signal is analyzed to identify its "fast response start" time and "maximum response" time; the response inertia time is obtained by subtracting the "fast response start" time from the "maximum response" time. Obtain the physical location of each fiber channel and combine it with the corresponding response inertia time to form a corresponding space-response pair. Use all space-response pairs as a set of space-response inertia points.

[0008] 3. The intelligent monitoring method for dam structures based on distributed optical fiber vibration sensing technology according to claim 1, characterized in that the step of determining whether multiple channels belong to the same source coupling response based on the space-response inertial point set is as follows: After arranging the spatial-response inertial point set in ascending order of spatial location, the response inertial slope between every two adjacent points is calculated sequentially to obtain the slope sequence. By traversing the slope sequence using a sliding window, a set of local perturbation trend pattern vectors is constructed. Calculate the similarity between any two sets of trend vectors pairwise and construct a trend consistency matrix; The mean of the similarity of all off-diagonal lines in the statistical trend consistency matrix is ​​used as the trend consistency. The trend consistency is used to determine whether multiple channels belong to the same source coupled response.

[0009] Optionally, the steps for determining whether multiple channels belong to the same source coupling response based on trend consistency are as follows: Compare the trend consistency with the preset trend consistency threshold. If the trend consistency is not less than the preset trend consistency threshold, then multiple channels belong to the same source coupled response. If the trend consistency is less than the preset trend consistency threshold, then multiple channels do not belong to the same source coupling response; the echo response signals of each channel are analyzed independently, and the abnormal location of the dam structure is monitored and located based on the analysis results of the echo response signals of each channel.

[0010] Optionally, the steps for calculating the anomaly level value of each channel based on the echo response signal of each channel are as follows: The entropy reversal rate and local disturbance driving intensity of each channel are calculated based on the echo response signal of each channel. The entropy reversal rate and local disturbance driving intensity are then added together to obtain the anomaly degree value of each channel.

[0011] Optionally, the calculation steps for the entropy reversal rate value are as follows: The first The echo response signal sequence of the fiber optic channel is denoted as , Indicates channel At the point of time The echo response signal value; An adaptive morphological change algorithm was used to identify clamped segments with a rapid descent-minimum-rapid ascent pattern in the echo response signal sequence of each fiber channel, and extreme value clamped segments were constructed. gather : In the formula, Indicates the first The start time of each clamped segment Indicates the first The end time of each closed segment; in Memory at a minimum point ,satisfy: , ; This represents the total number of closed segments in the set of identified closed segments; For each clamped segment, the signal corresponding to the minimum point is subtracted from the signal corresponding to the start time to obtain the corresponding compression amplitude, and the signal corresponding to the end time is subtracted from the signal corresponding to the minimum point to obtain the corresponding release amplitude. Divide the release amplitude of each clamped segment by the compression amplitude to obtain the inversion amplification ratio of the corresponding clamped segment.

[0012] Optionally, the calculation steps for the entropy reversal rate value are as follows: For each clamped segment, a normalized inversion path function is constructed, and the normalized inversion path value at each time point is calculated to obtain the inversion path curve. Calculate the release path steepness value for each clamped segment based on the reverse path curve; Compare the inversion amplification ratio of each clamped segment with the preset inversion amplification ratio threshold, and compare the release path steepness value with the preset release path steepness value threshold. The clamped segments with an inversion amplification ratio less than the preset inversion amplification ratio threshold and a release path steepness value less than the preset release path steepness value threshold are recorded as entropy inversion segments. Divide the total number of entropy inversion segments by the total number of closed segments in the closed segment set to obtain the entropy inversion rate value of each channel.

[0013] Optionally, the calculation steps for the local disturbance driving intensity value are as follows: For the echo response signal sequence of each fiber channel, the derivative of the echo response signal sequence is obtained to obtain the instantaneous rate of change of the signal at each time; the instantaneous rate of change of the channel signal at all times is taken as the instantaneous rate of change sequence; Identify abrupt events in the instantaneous rate of change sequence and construct a local impulse trigger function. In the formula, This is a pulse trigger function. Indicates the fiber optic cable number; For a small time interval; when the instantaneous rate of change of the signal changes from negative to positive, it is considered that there is a pulse trigger at that moment; All pulse triggers are grouped into a set, and the time interval between adjacent pulse triggers is calculated; Subtract the minimum adjacent interval from the maximum adjacent interval, and use the result of the subtraction as the numerator. Add the minimum adjacent interval to the maximum adjacent interval, and use the result of the addition as the denominator. Divide the numerator by the denominator to obtain the pulse timing asymmetry. .

[0014] Optionally, the calculation step of the local disturbance driving intensity value further includes: For each pulse trigger moment, calculate the integral before and after the trigger. The formula for calculation is: , ， In the formula, The preset short-time observation window length; and These represent the integral of the rate of change of the signal before the pulse and the integral of the rate of change of the signal after the pulse, respectively. Integral based on the rate of change of the signal before the pulse Integral of the rate of change of the signal after the pulse Calculate pulse deflection The calculation formula is: In the formula, express The degree of asymmetry in the changes before and after each pulse trigger; Calculate the ripple closure rate based on all pulse deflections. The calculation formula is: Subtract the fluctuation closure rate from the value 1 The closed stability is obtained. , ; Indicates the stability of the drive closure; Based on pulse timing asymmetry and closure stability The formula for calculating the local driving dominance coefficient is as follows: In the formula, For the first The local driving dominance coefficient of a fiber optic channel; Subtract the local driving dominance coefficient from the value of 1, and use the result of the subtraction as the local disturbance driving strength value.

[0015] Optionally, the anomaly origination channel can be determined based on the anomaly severity value, including: The fiber optic channel with the highest anomaly score is identified as the source channel of the anomaly. The beneficial effects of this invention are: This invention proposes an intelligent monitoring method for dam structures based on distributed optical fiber vibration sensing technology. First, multiple optical fibers deployed along the dam are used to acquire the echo response signals of each fiber channel. Then, based on the echo response signals of each fiber channel, the response inertia time of each channel is calculated, and a space-response inertial point set is constructed. The space-response inertial point set is used to determine whether multiple channels exhibit co-source coupling responses. If co-source coupling responses are identified, the anomaly severity value of each channel is calculated based on its echo response signals, and the anomaly originating channel is determined according to the anomaly severity value. Anomaly localization analysis is performed on the echo response signals of the anomaly originating channel to determine the location of the anomaly in the dam structure. Through this method, it is possible to determine whether the anomaly detected by multiple optical fiber channels is a genuine anomaly or caused by response coupling of other optical fiber channels when anomalies are detected by optical fiber vibration sensing technology. This ensures that the monitored dam anomalies are real, reduces the risk of false alarm amplification, anomaly location misalignment, or data fusion failure, and ensures the accuracy of dam structure anomaly monitoring results and the reliability of engineering decisions. Attached Figure Description

[0016] The invention will now be further described with reference to the accompanying drawings.

[0017] Figure 1 This is a flowchart of an intelligent monitoring method for dam structures based on distributed optical fiber vibration sensing technology. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] This invention provides an intelligent monitoring method for dam structures based on distributed fiber optic vibration sensing technology. See also... Figure 1 , Figure 1 A flowchart illustrating an intelligent monitoring method for dam structures based on distributed fiber optic vibration sensing technology, provided in an embodiment of the present invention. The method includes the following steps: S1: Based on multiple optical fibers laid along the embankment, obtain the echo response signal of each optical fiber channel; S2: Calculate the response inertia time of each fiber channel based on the echo response signal of each fiber channel and construct a space-response inertia point set; determine whether multiple channels belong to the same source coupling response based on the space-response inertia point set; S3: If it is determined to be a response of the same source, then based on the echo response signal of each channel, calculate the abnormality value of each channel, and determine the abnormality source channel according to the abnormality value. S4: Perform anomaly localization analysis on the echo response signal of the abnormal source channel to determine the abnormal location of the dam structure.

[0020] Based on the intelligent monitoring method for dam structures based on distributed optical fiber vibration sensing technology provided in this invention, the above-mentioned method can determine whether the anomaly detected by multiple optical fiber channels is a genuine anomaly or caused by the response coupling of other optical fiber channels when an anomaly is detected by optical fiber vibration sensing technology. This ensures that the monitored dam anomaly is real, reduces the problems of false alarm amplification, anomaly positioning offset or data fusion failure, and ensures the accuracy of dam structure anomaly monitoring results and the reliability of engineering decisions.

[0021] In one embodiment, S1: Based on multiple optical fibers laid along the dam, the echo response signal of each optical fiber channel is obtained; Specifically, multiple continuous single-mode optical fibers are deployed along key locations of the dam structure. These fibers can be deployed in a double-layered or multi-channel side-by-side configuration, depending on monitoring requirements. For example, a layer of fiber can be deployed on both the upstream and downstream sides of the dam, or three fibers can be deployed on the same cross-section to cover different lateral areas, thereby achieving multi-dimensional response sensing of the dam structure in the vertical, lateral, and longitudinal directions. The optical fibers are typically laid close to the structure using surface-mounted or buried installation methods, and form a closed monitoring loop with the fiber optic signal backbone through fusion splices or connectors. The deployment length can cover the entire dam body and even extend to upstream and downstream areas. After the optical fibers are deployed and connected to the monitoring host, the system periodically injects laser pulses into the fiber channels at set time intervals. During propagation in the fiber, Rayleigh scattering signals are generated due to the microstructural inhomogeneities of the fiber. A portion of these signals propagates backward along the fiber and is received by the host. If minor vibrations, strains, or deformations occur in the structural area where the fiber is located, the phase or amplitude of the scattered signal will change slightly. The monitoring host demodulates the echo signals and records the echo response sequence at all monitoring locations on each fiber optic channel in chronological order, forming a multidimensional response dataset indexed by "time-location-channel". For example, if a section of a dam is equipped with 3 optical fibers, each with a spatial resolution of 1 meter and a total length of 300 meters, and data is collected once per minute, then 3 × 300 = 900 spatial point response records will be generated per minute for subsequent analysis.

[0022] In one embodiment, S2: Calculate the response inertia time of each fiber channel based on the echo response signal of each fiber channel and construct a space-response inertia point set; determine whether multiple channels belong to the same source coupling response based on the space-response inertia point set; In one implementation, the steps of calculating the response inertial time of each fiber optic channel and constructing a space-response inertial point set based on the echo response signal of each fiber optic channel are as follows: For each optical fiber channel deployed in the dam structure, let its number be . Extract each channel The echo response signal within the current preset monitoring period is denoted as , In the formula, For the first Fiber optic channels in time The echo response signal value at a given time (e.g., phase difference or intensity value). Set the start time for the preset monitoring window; Set the preset monitoring window end time; For each fiber optic channel echo response signal Perform an ascending segment analysis to identify its "rapid response start" time. With "maximum response" time , For the first The time point at which the signal first shows a sudden increase in a single fiber optic channel; For the first The point in time when the signal reaches its maximum amplitude in a single fiber optic channel; The response inertia time is defined as the time interval between the rise of the signal and its peak value, and the formula for calculation is: In the formula, Indicates the first The response inertia time of a fiber optic channel represents the duration it takes for the channel to go from its initial fast response to reaching its maximum response; the smaller the value, the faster the response. The physical location of each fiber optic channel is denoted as... This forms a set of space-response pairs. , ,in, Indicates the first The spatial location of a fiber optic cable can be the dam axis coordinates, station number, or physical distance measurement location; All space-response pairs are treated as a space-response inertial point set. This set of points constitutes a scatter set in the "space-inertial-time" coordinate system, which is used for subsequent analysis of the structural distribution characteristics of the inter-fiber response.

[0023] It should be noted that the echo response signal of each fiber optic channel involved is acquired by a distributed fiber optic vibration monitoring host within a preset time window. Specifically, the system injects laser pulse signals into each fiber optic channel at fixed time intervals (e.g., every 5 seconds or every 10 seconds), and collects the backscattered echoes at different locations along the fiber's length using the Rayleigh scattering principle. The system then demodulates the minute phase changes or intensity fluctuations of that fiber segment at each moment, thus constructing a continuous sequence of response signals. The data for each channel actually originates from the corresponding entire fiber optic cable, and each fiber optic cable itself is physically deployed along the dam's axis, its spatial position... This typically refers to the actual starting point or representative midpoint of the fiber optic cable within the dam, expressed in meters (m), station numbers (e.g., 237+40), or mileage coordinates based on the dam axis. For example, in a typical dam structure, if three fiber optic cables are deployed on the upstream slope, the dam axis center, and the downstream slope, with each cable extending 300 meters from the dam's starting point to its endpoint, these three cables can be numbered as follows: Its physical location These can be set to the dam axis coordinates of 0m, 2.5m, and 5m, respectively. It is important to note that this location does not represent the specific spatial points of all sampling points on the optical fiber, but rather the physical location of the overall response of the channel. It is used to form the "spatial" coordinate portion of the spatial-response inertial structure point set. Together with the response inertial time calculated by the channel, it forms a spatial-inertial pair, providing basic data support for subsequent construction of the inertial compression structure and determining whether the responses between channels are from the same source.

[0024] In one implementation, the steps for determining whether multiple channels belong to the same source coupled response based on the space-response inertial point set are as follows: Space-response inertial point set By spatial location After sorting in ascending order, calculate the response inertia slope between every two adjacent points: , In the formula, Indicates the first The first channel and the first The slope of the response inertia change between the channels; Indicates the first The response inertia time of a fiber optic channel, Indicates the first The physical location of each fiber optic channel; With a length of A sliding window traversal slope sequence with a window width (e.g., 3). Construct a set of local perturbation trend pattern vectors: , ; Indicates the first Group of local disturbance trend pattern vectors; Calculate the similarity between any two pairs of trend vectors to construct a trend consistency matrix; the formula for calculating the similarity is: , , They represent the first The group of local disturbance trend pattern vectors represents the first The vector dot product of the local disturbance trend pattern vectors. and They represent the first The vector magnitude of the group of local disturbance trend pattern vectors represents the first... The vector magnitude of the local disturbance trend pattern vector; Indicates the angular similarity between local trends (the closer the value is to 1, the more similar the trends are). The trend consistency score is calculated by taking the mean similarity of all off-diagonal lines in the trend consistency matrix. This score is then used to determine whether multiple channels belong to the same source coupled response. Specifically, the formula for calculating trend consistency is: In the formula, This represents the degree of trend consistency.

[0025] It should be noted that in the analysis of the space-response inertial point set, the so-called ascending order of spatial location means arranging all fiber optic channels according to their physical deployment position in the dam structure, from near to far, from left to right, or from downstream to upstream, so that fiber optic channels with smaller numbers are located in the front section of the dam, and fiber optic channels with larger numbers are located in the rear section of the dam, thus ensuring that the sorting sequence has spatial continuity and physical relevance. The core purpose of this sorting is to ensure that adjacent channels have spatial adjacency when constructing the response change trend between channels, so as to capture the directionality and hierarchy of structural disturbance propagation in space. For example, in a typical three-channel deployment scenario, if the three optical fibers are located on the upstream slope, the center of the dam axis, and the downstream slope of the dam, respectively, and their physical deployment representative positions are 0m, 2.5m, and 5m, then the channel order arranged in ascending order of spatial location should be upstream slope (channel 1), dam axis center (channel 2), and downstream slope (channel 3). This sorting method conforms to the physical deployment logic of the dam's transverse structure and is also the basis for subsequent judgment on whether the abnormal response is a single source propagating in space.

[0026] It's important to note that the reason for using trend consistency to determine whether multiple fiber optic channels exhibit homogeneous coupling responses is that this method captures the inherent correlation between multiple channels in their spatial responses from the perspective of "structural evolution trends," rather than merely focusing on superficial comparisons of single-point response times or amplitudes. Specifically, when structural anomalies occur within a dam (such as seepage, erosion, or collapse), the physical disturbances diffuse outwards through the soil medium in a certain direction and at a certain speed. If multiple fiber optic channels are deployed within the dam, these disturbances will be transmitted to adjacent or relatively distant channels with a certain time delay and attenuation intensity, thus triggering synchronous or near-synchronous responses across multiple channels. However, this response is not chaotic but possesses a certain directionality, continuity, and inertial trend. Specifically, the response inertial time exhibits a stable trend in space, such as gradual delay, gradual increase, or gradual decrease. The aforementioned method for calculating trend consistency is designed based on this physical propagation law. Its core idea is to transform the inertial changes in response between adjacent channels into a slope sequence, construct local trend vectors using a sliding window approach, and then analyze whether all local trends maintain a high degree of consistency using cosine similarity. This allows for the determination of whether the entire system exhibits a unified and continuous perturbation evolution path. If all trend vectors have high similarity, it indicates a consistent response trend, most likely caused by a single anomalous source in spatial propagation. If the trend vectors differ significantly, it indicates a split response path, suggesting they do not originate from the same source.

[0027] In contrast, traditional judgment methods often rely on several common dimensions, such as the difference in response time, response amplitude, and response inertia between channels, the degree of synchronization of the maximum echo signal value, or even analyzing the fitting residuals after linear fitting of all response points. However, these methods have several key drawbacks: First, they mostly rely on single-point features and cannot capture the continuous evolution of the response in space, making them susceptible to local anomalies or noise, leading to misjudgments; second, simply calculating the difference between response time or amplitude ignores the trend path formed by the propagation of disturbances in space, and lacks the ability to identify the "propagation structure"; third, linear fitting methods essentially assume that the disturbance trend changes linearly, making it difficult to handle the response distortion caused by nonlinear propagation, uneven geological zoning, or structural heterogeneity in real dams; fourth, the judgment thresholds of these methods are usually based on experience, resulting in poor adaptability and weak universality, making it difficult to support high-reliability judgments in complex engineering environments.

[0028] In contrast, the trend consistency method has the following advantages: it constructs a "trend behavior map" based on local slope vectors and measures the uniformity of trend direction through cosine similarity, thus possessing stronger structural expressive power; it does not judge whether two channels are close, but rather whether the trends of the entire group of channels remain coordinated, belonging to a global stability analysis method, resulting in more robust results; it avoids the potential offset interference of response peaks or absolute time, instead focusing on the intrinsic laws of response evolution, thus possessing higher physical rationality; at the same time, it can adapt to the natural differences in response speed and intensity caused by factors such as different laying depths and path impedances in dams, therefore it can more accurately identify whether the response pattern is caused by the diffusion of a real anomaly source. In summary, the trend consistency judgment mechanism not only has extremely high engineering adaptability and anti-interference ability, but also demonstrates strong technical innovation and theoretical rigor. It is an innovative judgment method that transcends the traditional point-to-point judgment thinking and shifts to a trend pattern recognition paradigm, making it particularly suitable for intelligent discrimination and coupling identification of complex structural disturbance responses in multi-fiber parallel deployments.

[0029] In one implementation, the steps for determining whether multiple channels belong to the same source coupled response based on trend consistency are as follows: Compare the trend consistency with the preset trend consistency threshold. If the trend consistency is not less than the preset trend consistency threshold, then multiple channels belong to the same source coupled response. If the trend consistency is less than the preset trend consistency threshold, then multiple channels do not belong to the same source coupling response; the echo response signals of each channel are analyzed independently, and the abnormal location of the dam structure is monitored and located based on the analysis results of the echo response signals of each channel.

[0030] It should be noted that the trend consistency degree is compared with a preset trend consistency threshold to determine whether the responses of multiple fiber optic channels exhibit a high degree of trend consistency, thereby determining whether they belong to the same source coupling response. If the trend consistency degree is greater than or equal to the preset threshold, it indicates that the disturbance response trends among multiple channels are highly similar, possessing a unified direction and continuity. The system can determine this as a same source coupling response, meaning that multiple channels are affected by the diffusion of disturbances from the same abnormal source. In this case, the system can subsequently enter the unified source channel identification and anomaly localization process. However, if the trend consistency degree is less than the preset threshold, it indicates that the response trends among the channels differ significantly and cannot form a consistent propagation structure. The system will determine this as a non-same source response, meaning that each channel may independently exhibit abnormal or interference behavior. In this case, the system no longer analyzes the multi-channel data as a whole coordinated response, but instead analyzes and processes the echo response signal of each fiber optic channel independently. Specifically, the system can perform amplitude fluctuation judgment, frequency domain feature extraction, and response inertia change analysis on the signal of each channel to identify whether there are abnormal changes in each channel, and further, in conjunction with the physical location of the fiber optic deployment, locate the anomaly point. For example, if the trend characteristics of channel A are completely different from those of channels B and C among the three optical fibers, the system will perform anomaly localization on each of the three optical fibers A, B, and C respectively. Ultimately, it may identify two independent anomaly points on the upstream and downstream slopes respectively, rather than misjudging them as a single centralized anomaly source, thereby improving the accuracy of anomaly identification and avoiding misleading response aggregation.

[0031] In one embodiment, S3: If it is determined to be a co-source coupling response, then the anomaly degree value of each channel is calculated based on the echo response signal of each channel; based on the anomaly degree value, the optical fiber channel with the largest anomaly degree value is determined as the anomaly source channel. In one implementation, the steps for calculating the anomaly level value of each channel based on the echo response signal of each channel are as follows: The entropy reversal rate and local disturbance driving intensity of each channel are calculated based on the echo response signal of each channel. The entropy reversal rate and local disturbance driving intensity are then added together to obtain the anomaly degree value of each channel.

[0032] In one implementation, the steps for calculating the entropy reversal rate value are as follows: The first The echo response signal sequence of the fiber optic channel is denoted as , Indicates channel At the point of time The echo response signal value; An adaptive morphological transformation algorithm was used to identify the echo response signal sequence. There exist clamping segments with a pattern of rapid decline—minimum—rapid rise; construct a set of extreme value clamping segments. : In the formula, Indicates the first The start time (compression start point) of each clamped segment. Indicates the first The end time (release endpoint) of each clamped segment; in Memory at a minimum point ,satisfy: , ; This represents the total number of closed segments in the set of identified closed segments; For each clamped segment, the start time is... corresponding signal Subtract the minimum point corresponding signal To obtain the corresponding compression amplitude , End time corresponding signal Subtract the minimum point corresponding signal To obtain the corresponding release amplitude , ; Release amplitude of each clamped segment Divide by compression amplitude Obtain the inversion amplification ratio of the corresponding clamped segment. ; For each clamped segment Construct a normalized inversion path function, calculate the normalized inversion path value at each time point, and obtain the inversion path curve. The calculation formula is as follows: , In the formula, This represents the normalized inversion path curve within the segment interval, used to analyze the structural characteristics of the signal release process; The release path steepness value for each clamped segment is calculated using the following formula: In the formula, Indicates the first The steepness value of the release path of each clamped segment; the larger the value, the more rapid the signal release in that segment, with obvious reversal burst characteristics; Compare the inversion amplification ratio of each clamped segment with the preset inversion amplification ratio threshold, and compare the release path steepness value with the preset release path steepness value threshold. The clamped segments with an inversion amplification ratio less than the preset inversion amplification ratio threshold and a release path steepness value less than the preset release path steepness value threshold are recorded as entropy inversion segments. Dividing the total number of entropy-inverted segments by the total number of clipped segments in the clipped segment set yields the entropy inversion rate value for each channel. This value quantifies the significance of the "compression-burst" inversion phenomenon in the echo response signal of that channel; a larger value indicates a more pronounced entropy inversion characteristic.

[0033] It should be noted that in the calculation of the entropy reversal rate, the data involved in each step comes from the real-time acquisition and analysis of the fiber optic channels by the distributed fiber optic vibration monitoring host. First, when constructing the echo response signal sequence of the u-th fiber optic channel, the required raw signal data is obtained by the host injecting laser pulse signals into the fiber at fixed time intervals (e.g., every 5 seconds or every 10 seconds) within a preset time window. The system uses the Rayleigh scattering principle to receive the scattered signals reflected back along the fiber and demodulates the phase change or light intensity change at each time point, thereby forming a continuous response signal sequence. For example, on channel A of a dam, if the monitoring host collects 600 sampling points within one minute, then... This is a sequence containing the echo response values ​​at these 600 time points. Secondly, an adaptive morphological change algorithm is used to identify... When encountering a "rapid descent-minimum-rapid ascent" clamping segment, the algorithm automatically detects abrupt changes in the signal slope and curvature within the acquired time series to identify the compression segment, minimum point, and release segment. For example, if the signal rapidly decreases to the minimum point within 2.0 to 3.5 seconds and then rapidly rises after 4.0 seconds, the system automatically identifies this interval as a clamping segment. And record the start time. Minimum point With end time Secondly, in calculating the compression amplitude... and release amplitude At that time, the one used , , All of these are directly derived from the echo response signal values ​​recorded by the host, for example, if =0.82、 =0.60、 =0.95, then we can get =0.22、 =0.35. Subsequently, in constructing the normalized inversion path function... At the same time, the signal at each time point τ in the calculation All signals are taken from the original response sequence within the same time period. The signals continuously collected by the host form a complete curve trajectory. For example, the value of τ ranges from... =2.0 seconds to arrive Between 4.0 seconds, each sampling point has a corresponding The value. Next, the steepness value of the release path is calculated. At that time, the system based on The time-series data of the curve is numerically differentiated, and the overall slope intensity of the curve within the release segment is obtained through integration. This data is supported by high-frequency sampling points collected by the monitoring host. For example, if the signal changes drastically within the release segment, the average derivative value is high. The values ​​increase accordingly. Finally, when comparing the inversion amplification ratio with the preset threshold and the kurtosis value with the preset threshold, the thresholds used are all fixed values ​​set by the system through historical data sample analysis or engineering experience. The final calculated entropy inversion rate value is automatically completed by the monitoring platform through a statistical program. For example, when a channel identifies 8 clamped segments, and 3 of these segments simultaneously meet the inversion amplification ratio threshold and the kurtosis threshold conditions, the entropy inversion rate value of that channel is 3 ÷ 8 = 0.375. Through the above method, the signal data, time points, and amplitudes required for each step in the calculation process are directly derived from the original sampling results of the distributed fiber optic vibration monitoring host, ensuring that the entire value calculation process is real-time, traceable, and engineering feasible.

[0034] It should be noted that the entropy reversal rate is an indicator used to measure whether there is a significant "information compression-burst" structural evolution characteristic in the echo response signal of an optical fiber channel. Its core purpose is to identify whether a typical anomalous trend of "early energy accumulation - mid-term minimum convergence - late rapid release" appears in the signal sequence. The larger the value, the higher the frequency and the more significant the characteristics of this anomalous evolution pattern in the channel, meaning that the physical disturbance or structural response sensed by the channel is more concentrated and explosive. In dam structure monitoring, optical fiber channels are deployed along the structure to sense minute vibrations, displacements, or strain changes along the line with high precision. When there are structural anomalies (such as local seepage, collapse precursors, or interface slippage), a significant "disturbance convergence-rapid release" signal behavior is usually generated first in the optical fiber channel near the source of the anomaly, forming an asymmetric structural feature of signal fluctuation. These features are precisely captured by the entropy reversal rate. For example, if a local structural disturbance occurs within a section of a dam, channel A, located near the disturbance source, might initially show a relatively stable signal, but then suddenly drop to an extremely low point before rapidly rebounding, forming a typical "squeezed-down segment." Channels B and C, located further away from the disturbance source, might only receive secondary disturbances that are attenuated by coupling after propagation through the medium. Their signal fluctuations are smaller and they lack the squeezing-down characteristic, resulting in entropy reversal rates significantly lower than channel A. Therefore, when multiple channels are identified as having a common source of coupling response, the signal trends among the channels may exhibit some synchronicity. However, only the channel closest to the physical disturbance source is most likely to simultaneously display frequent squeezing-down segments, significant release amplification, and a steep rebound path, thus resulting in a significantly higher entropy reversal rate than other channels. Based on this, it can be inferred that this channel is most likely located in the anomaly origin region and can therefore be selected as a priority for location. This judgment method, based on differences in structural signal morphology rather than amplitude or frequency characteristics, has stronger anti-interference capabilities and structural discrimination ability, and is particularly suitable for refined diagnosis of local anomaly sources in multi-fiber parallel deployments.

[0035] It's important to note that the greatest advantage of calculating the entropy reversal rate using the above method is that it doesn't rely on conventional statistical characteristics like signal amplitude, mean, standard deviation, or first-order difference. Instead, it starts from the perspective of "signal morphological evolution structure," accurately capturing the "asymmetric clamping fluctuations" in the echo response signal—that is, a pattern of rapid decline followed by a rapid release after reaching a minimum. This pattern is precisely a typical manifestation of the transmission of local physical disturbances at the structural level in a dam, often implying real physical aggregation and release behaviors, and possessing strong interpretative and locational value. In contrast, other methods often either rely too heavily on overall trends (such as moving averages), are insensitive to local structures (such as Fourier analysis), or are easily affected by noise (such as conventional gradient / slope analysis), making it difficult to effectively identify the "compression-minimum-burst" process in the signal. This method constructs a complex but stable judgment framework through multiple mechanisms such as "clamping segment identification + compression / release amplitude + steepness calculation + threshold discrimination". It not only has good noise robustness, but also improves the comparability between channels and time periods through the calculation of normalized structural quantities. It is suitable for anomaly source identification scenarios under multi-fiber parallel monitoring and is an innovative structural evolution value with morphological perception capabilities.

[0036] In one implementation, the calculation steps for the local disturbance driving intensity value are as follows: The first The echo response signal sequence of the fiber optic channel is denoted as ,

[0037] Indicates channel At the point of time The echo response signal value; right Differentiating the signal, we obtain the instantaneous rate of change at each moment. The formula for calculation is: ,in, Indicates the first Fiber optic channels in time Instantaneous rate of change at any given time; the instantaneous rate of change of all channel signals at any given time is taken as an instantaneous rate of change sequence; Identify abrupt events in the instantaneous rate of change sequence and construct a local impulse trigger function. In the formula, This is a pulse trigger function; A tiny time interval is used to detect abrupt boundary changes; when the instantaneous rate of change of the signal changes from negative to positive, it is considered that a pulse trigger exists at that moment. All pulse triggers are grouped into a set: ,in, Indicates the first Each pulse trigger moment; Indicates the total number of pulse triggers identified; Calculate the time interval between adjacent pulse triggers The calculation formula is: ; Indicates the first With the The time interval between each pulse trigger; Calculate pulse timing asymmetry based on adjacent intervals. The calculation formula is: In the formula, Indicates the first The degree of asymmetry in the pulse time distribution of a fiber optic channel, ranging from 0 to 1; For each pulse trigger moment, calculate the integral before and after the trigger. The formula for calculation is: , ， In the formula, The preset short-time observation window length; and These represent the integral of the rate of change of the signal before the pulse and the integral of the rate of change of the signal after the pulse, respectively. Integral based on the rate of change of the signal before the pulse Integral of the rate of change of the signal after the pulse Calculate pulse deflection The calculation formula is: In the formula, express The greater the asymmetry of the changes before and after a pulse is triggered, the more unidirectional the driving force of the pulse is. Calculate the ripple closure rate based on all pulse deflections. The calculation formula is: Subtract the fluctuation closure rate from the value 1 The closed stability is obtained. , ; This indicates the stability of the drive closure; a larger value indicates a more stable and regular drive mode. Based on pulse timing asymmetry and closure stability The formula for calculating the local driving dominance coefficient is as follows: In the formula, For the first The local drive dominance coefficient of a fiber optic channel; used to reflect the relative degree of difference between pulse time imbalance and regularity; Subtract the local driving dominance coefficient from the value of 1, and use the result as the local disturbance driving strength value. The larger the value, the stronger the instantaneous driving capability and the higher the probability of disturbance origination in the same-source coupling response.

[0038] It should be noted that in the above calculation process, This is a very short time interval, specifically referring to the time interval used to detect signal transitions, capturing the moment when the signal's instantaneous rate of change changes from negative to positive. Typically, this time interval is very short, usually set as a small proportion of the signal sampling frequency, for example, in the case of 1000 samples per second. It can be set to 0.001 seconds to ensure accurate capture of the instantaneous characteristics of signal jumps. For example, if the signal is in time... The value suddenly changes from negative to positive, and this jump occurs during... Within the time window, it is considered that... A pulse trigger event occurred at that moment. The preset short-time observation window length refers to the time length used to define the observation range when calculating the integral of the signal change rate before and after a pulse. For example, if δ = 0.05 seconds, it means that when calculating a certain pulse, we will focus on the signal change within 0.05 seconds before and after the pulse to evaluate whether the pulse has significant unidirectional driving force. It can be flexibly configured according to different signal characteristics and monitoring needs to ensure an immediate response to signal changes.

[0039] It should be noted that the local disturbance drive strength value is used to measure whether fiber optic channels, when deployed in parallel with multiple fibers to monitor dam structures, play a driving role in the overall system's abnormal response within a short period of time. Specifically, it reflects whether a particular fiber optic channel exhibits significant "disturbance-induced behavior" at the moment of signal response. A larger local disturbance drive strength value indicates that the channel's signal exhibits stronger unidirectional pulse behavior in time, and the pulse appears earlier than other channels, demonstrating a stronger active guiding effect. This is because when a local anomaly occurs in the dam (such as cracks or seepage), the fiber optic channels near the anomaly source will be the first to detect the abnormal signal, and these signals typically manifest as rapid and powerful disturbance "pulses"—this is the so-called "disturbance drive." Other channels, due to coupling effects, usually respond with a lag, exhibiting a reverse rebound or a weaker response. This driving behavior can be identified and quantified through the local disturbance drive strength value; the fiber optic channel with a larger local disturbance drive strength value is usually the "source channel" that detects the anomaly earliest. For example, if a dam experiences a localized collapse, fiber A, located near the site, will respond earliest, generating a significant pulse signal. Fibers B and C, located further away from the anomaly source, may only sense the transmitted disturbance, responding more slowly and with smaller fluctuations. By calculating the driving intensity of the local disturbance, the channel with the fastest and strongest signal response can be identified as the anomaly source channel, thus helping to accurately pinpoint the origin of the anomaly in the dam structure.

[0040] It should be noted that, The greatest advantage of calculating the local disturbance driving strength value using the above method is that it accurately captures the "initiating effect" of the anomaly source channel on system disturbances, starting from the asymmetry of signal timing and the time difference of pulse triggering. Unlike traditional signal amplitude and frequency analysis methods, the local disturbance driving strength value focuses on analyzing the driving effect of instantaneous signal jumps, that is, whether a certain channel shows a significant "pulse response" before other channels, and whether this response is unidirectional, rapid, and regular. This calculation method has high physical rationality and can reveal the source of "active disturbance" from the evolution of the signal time series, rather than just the degree of signal response, which is especially important for multi-fiber coupled systems. By capturing these instantaneous, local disturbance driving signals, the local disturbance driving strength value can effectively distinguish the channel that first senses the anomaly and generates an "active response," differentiating it from other channels that are only affected by coupling transmission. Therefore, the local disturbance driving strength value can accurately indicate the location of the anomaly source channel, and is not easily affected by external noise or signals far from the anomaly source. Compared to traditional calculation methods based on amplitude or energy, the local disturbance driving intensity value has stronger sensitivity and temporal sequence, and can identify the most critical anomaly source channels in the dam monitoring system, thereby improving the accuracy and precision of anomaly location, and ultimately achieving efficient and accurate monitoring of dam structural anomalies.

[0041] In one embodiment, determining the anomaly origin channel based on the anomaly severity value includes: The fiber channel with the highest anomaly level value is taken as the source channel of the anomaly.

[0042] It should be noted that the process of determining the anomaly source channel based on the anomaly severity value involves calculating the anomaly severity value of each fiber channel, combining it with the entropy reversal rate value and the local disturbance driving strength value, to evaluate the response strength and driving effect of each channel to the anomaly signal during monitoring, thereby quantifying the anomaly characteristics of each channel when monitoring the dam structure. When multiple channels exhibit co-current coupling responses, signal synchronization or near-synchronization often occurs. However, the strongest anomaly signal often originates from the location closest to the anomaly source. By comparing the anomaly severity values ​​of all channels and selecting the fiber channel with the highest anomaly severity value as the anomaly source channel, the source channel causing the anomaly can be effectively identified. The advantage of this approach is that it allows for precise location of the anomaly source through quantitative methods, rather than relying solely on simple comparisons of amplitude or frequency. This enables the system to more sensitively and accurately identify potential anomaly sources in monitoring scenarios with multiple fibers deployed in parallel, reducing false alarms and improving location accuracy. This method has high robustness and reliability, and is particularly suitable for complex monitoring environments with multiple channels in parallel. It can effectively identify which channel senses the anomaly and generates a driving response earliest when multiple channels exhibit co-current coupling responses.

[0043] In one embodiment, S4: Perform anomaly localization analysis on the echo response signal of the abnormal source channel to determine the abnormal location of the dam structure.

[0044] It should be noted that anomaly localization analysis of the echo response signal of an anomaly source channel first requires a detailed spatiotemporal analysis based on the echo signal sequence of that channel and its physical location within the dam. Once a fiber optic channel is identified as the anomaly source channel, the signal analysis begins from that channel and proceeds along the dam structure. Specifically, by calculating the correlation between the anomalous signal of this channel and the signals of surrounding fiber optic channels, as well as the time propagation delay of the signal, and combining this with the actual physical structure layout of the dam, the approximate location of the anomaly source within the dam can be inferred. For example, suppose three fiber optic channels are deployed on the dam, and one of them is identified as the anomaly source channel, with strong anomalous signal fluctuations in the echo response of this channel. By comparing the signal of this channel with other channels and combining time delay analysis, it can be determined that the anomalous signal first appears in a certain section of the dam, and then the specific anomaly area can be determined by locating the spatial position of the fiber optic cable. For example, if the signal fluctuations in a particular channel show significant anomalies at a specific point in time, and the response signals from other surrounding channels exhibit a delayed or gradually increasing trend, the location of the anomaly source within the dam can be inferred using this spatiotemporal data. This method not only provides a quantitative assessment of anomalous signals but also improves the accuracy and sensitivity of dam monitoring systems in locating structural anomalies, thereby providing clear guidance for subsequent on-site inspections and maintenance.

[0045] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A method for intelligent monitoring of dam structures based on distributed optical fiber vibration sensing technology, characterized in that, Includes the following steps: Based on multiple optical fibers laid along the embankment, the echo response signals of each optical fiber channel are obtained. Based on the echo response signals of each fiber optic channel, calculate the response inertia time of each channel and construct a space-response inertia point set; determine whether multiple channels belong to the same source coupling response based on the space-response inertia point set. If it is determined to be a response of the same source, then based on the echo response signal of each channel, the abnormality degree value of each channel is calculated, and the abnormality source channel is determined according to the abnormality degree value. Anomaly localization analysis is performed on the echo response signal of the abnormal source channel to determine the abnormal location of the dam structure.

2. The intelligent monitoring method for dam structures based on distributed optical fiber vibration sensing technology according to claim 1, characterized in that, The steps for calculating the response inertial time of each fiber optic channel and constructing a space-response inertial point set based on the echo response signals of each channel are as follows: For each optical fiber channel deployed in the dam structure, the echo response signal of each channel within the current preset monitoring period is extracted. For each fiber optic channel The rise segment of the echo response signal is analyzed to identify its "fast response start" time and "maximum response" time; the response inertia time is obtained by subtracting the "fast response start" time from the "maximum response" time. Obtain the physical location of each fiber channel and combine it with the corresponding response inertia time to form a corresponding space-response pair. Use all space-response pairs as a set of space-response inertia points.

3. The intelligent monitoring method for dam structures based on distributed optical fiber vibration sensing technology according to claim 1, characterized in that, The steps for determining whether multiple channels belong to the same source coupled response based on the space-response inertial point set are as follows: After arranging the spatial-response inertial point set in ascending order of spatial location, the response inertial slope between every two adjacent points is calculated sequentially to obtain the slope sequence. By traversing the slope sequence using a sliding window, a set of local perturbation trend pattern vectors is constructed. Calculate the similarity between any two sets of trend vectors pairwise and construct a trend consistency matrix; The mean of the similarity of all off-diagonal lines in the statistical trend consistency matrix is ​​used as the trend consistency. The trend consistency is used to determine whether multiple channels belong to the same source coupled response.

4. The intelligent monitoring method for dam structures based on distributed optical fiber vibration sensing technology according to claim 3, characterized in that, The steps to determine whether multiple channels belong to the same source coupled response based on trend consistency are as follows: Compare the trend consistency with the preset trend consistency threshold. If the trend consistency is not less than the preset trend consistency threshold, then multiple channels belong to the same source coupled response. If the trend consistency is less than the preset trend consistency threshold, then multiple channels do not belong to the same source coupled response; The echo response signals of each channel are analyzed independently, and the abnormal locations of the dam structure are monitored and located based on the analysis results of the echo response signals of each channel.

5. The intelligent monitoring method for dam structures based on distributed optical fiber vibration sensing technology according to claim 1, characterized in that, The steps for calculating the anomaly level value of each channel based on the echo response signal of each channel are as follows: The entropy reversal rate and local disturbance driving intensity of each channel are calculated based on the echo response signal of each channel. The entropy reversal rate and local disturbance driving intensity are then added together to obtain the anomaly degree value of each channel.

6. The intelligent monitoring method for dam structures based on distributed optical fiber vibration sensing technology according to claim 5, characterized in that, The steps for calculating the entropy reversal rate value are as follows: The first The echo response signal sequence of the fiber optic channel is denoted as , Indicates channel At the point of time The echo response signal value; An adaptive morphological change algorithm was used to identify clamped segments with a rapid descent-minimum-rapid ascent pattern in the echo response signal sequence of each fiber channel, and a set of extreme value clamped segments was constructed. In the formula, Indicates the first The start time of each clamped segment Indicates the first The end time of each closed segment; in Memory at a minimum point ,satisfy: ; This represents the total number of closed segments in the set of identified closed segments; For each clamped segment, the signal corresponding to the minimum point is subtracted from the signal corresponding to the start time to obtain the corresponding compression amplitude, and the signal corresponding to the end time is subtracted from the signal corresponding to the minimum point to obtain the corresponding release amplitude. Divide the release amplitude of each clamped segment by the compression amplitude to obtain the inversion amplification ratio of the corresponding clamped segment.

7. The intelligent monitoring method for dam structures based on distributed optical fiber vibration sensing technology according to claim 6, characterized in that, The steps for calculating the entropy reversal rate value are as follows: For each clamped segment, a normalized inversion path function is constructed, and the normalized inversion path value at each time point is calculated to obtain the inversion path curve. Calculate the release path steepness value for each clamped segment based on the reverse path curve; Compare the inversion amplification ratio of each clamped segment with the preset inversion amplification ratio threshold, and compare the release path steepness value with the preset release path steepness value threshold. The clamped segments with an inversion amplification ratio less than the preset inversion amplification ratio threshold and a release path steepness value less than the preset release path steepness value threshold are recorded as entropy inversion segments. Divide the total number of entropy inversion segments by the total number of closed segments in the closed segment set to obtain the entropy inversion rate value of each channel.

8. The intelligent monitoring method for dam structures based on distributed optical fiber vibration sensing technology according to claim 5, characterized in that, The calculation steps for the local disturbance driving intensity value are as follows: For the echo response signal sequence of each fiber channel, the derivative of the echo response signal sequence is obtained to obtain the instantaneous rate of change of the signal at each time; the instantaneous rate of change of the channel signals at all times is taken as the instantaneous rate of change sequence; Identify abrupt events in the instantaneous rate of change sequence and construct a local impulse trigger function. In the formula, This is a pulse trigger function. Indicates the fiber optic cable number; For a small time interval; when the instantaneous rate of change of the signal changes from negative to positive, it is considered that there is a pulse trigger at that moment; All pulse triggers are grouped into a set, and the time interval between adjacent pulse triggers is calculated; Subtract the minimum adjacent interval from the maximum adjacent interval, and use the result of the subtraction as the numerator. Add the minimum adjacent interval to the maximum adjacent interval, and use the result of the addition as the denominator. Divide the numerator by the denominator to obtain the pulse timing asymmetry. .

9. The intelligent monitoring method for dam structures based on distributed optical fiber vibration sensing technology according to claim 8, characterized in that, The calculation steps for the local disturbance driving intensity value also include: For each pulse trigger moment, calculate the integral before and after the trigger. The formula for calculation is: , ， In the formula, The preset short-time observation window length; and These represent the integral of the rate of change of the signal before the pulse and the integral of the rate of change of the signal after the pulse, respectively. Integral based on the rate of change of the signal before the pulse Integral of the rate of change of the signal after the pulse Calculate pulse deflection The calculation formula is: In the formula, express The degree of asymmetry in the changes before and after each pulse trigger; Calculate the ripple closure rate based on all pulse deflections. The calculation formula is: Subtract the fluctuation closure rate from the value 1 The closed stability is obtained. , ; Indicates the stability of the drive closure; Based on pulse timing asymmetry and closure stability The formula for calculating the local driving dominance coefficient is as follows: In the formula, For the first The local driving dominance coefficient of a fiber optic channel; Subtract the local driving dominance coefficient from the value of 1, and use the result of the subtraction as the local disturbance driving strength value.

10. The intelligent monitoring method for dam structures based on distributed optical fiber vibration sensing technology according to claim 1, characterized in that, Based on the anomaly severity value, the anomaly origin channel is determined to include: The fiber optic channel with the highest anomaly level value is selected as the source channel of the anomaly.

Images