A perception system-based floating breakwater state monitoring method and system

Abstract

The application discloses a kind of floating breakwater state monitoring method and system based on perception system, belong to floating breakwater technical field, comprising: by distributed sensor network real-time acquisition pressure, strain, acceleration, gyroscope and other multi-source data, dynamic evaluation parameter matrix is formed after fusion;Accordingly identify abnormal node, diagnose its health level and give credibility weight, then compensate abnormal data;Again, use the modified data to construct anchoring system three-dimensional stress model and trigger overrun early warning;After secondary verification, determine the safety level of structure;Combining the dynamic update of marine environment change bearing capacity and stability index, and self-adaptive adjustment monitoring frequency and network topology, realize system continuous and stable operation.The application realizes the intelligent real-time monitoring and early warning of breakwater structure health state, significantly improves the reliability and accuracy of marine engineering structure safety monitoring.

Description

Technical Field

[0001] This invention belongs to the field of floating breakwater technology, and particularly relates to a floating breakwater status monitoring method and system based on a sensing system. Background Technology

[0002] Floating breakwaters, as key infrastructure in marine engineering, play a vital role in port construction, coastal protection, and marine resource development. With the rapid development of the marine economy and the frequent occurrence of extreme marine environments, the need for precise monitoring of the operational status of floating breakwaters is becoming increasingly urgent.

[0003] Current methods for monitoring floating breakwaters primarily rely on periodic manual inspections and single-point parameter measurements. These methods exhibit significant shortcomings when facing complex and ever-changing marine environments. Traditional monitoring methods struggle to capture real-time changes in the breakwater's condition under dynamic sea conditions, fail to establish correlations between various monitoring parameters, and struggle to guarantee the reliability and continuity of monitoring data under harsh marine conditions. Existing technologies lack the ability to comprehensively assess the overall performance of the breakwater system, often providing only partial information that fails to reflect the structure's overall health.

[0004] Therefore, the core challenge in monitoring the condition of floating breakwaters stems from the complexity of multi-physics coupling effects in the marine environment. Under the combined dynamic loads of waves, tides, and wind, breakwaters exhibit strong mutual coupling characteristics in their load-bearing state, attitude changes, anchoring stress, and structural response. This coupling effect further leads to the problem of multi-source heterogeneity of monitoring data, meaning that monitoring data from different locations and different types of sensors differ significantly in spatiotemporal scale, physical meaning, and data format. For example, when a breakwater undergoes drastic attitude changes under the impact of large waves, it not only causes a redistribution of the structural load-bearing state but also exacerbates stress concentration in the anchoring system. Simultaneously, the communication quality between sensor nodes fluctuates due to structural deformation, ultimately making it difficult to form a unified and reliable condition assessment result from data obtained from different monitoring subsystems.

[0005] Therefore, how to construct a comprehensive sensing system that can effectively handle the coupling effects of multiple physical fields and multi-source heterogeneous monitoring data, and realize accurate real-time monitoring of the operating status of floating breakwaters, has become a key issue that urgently needs to be addressed. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention provides a method and system for monitoring the condition of a floating breakwater based on a sensing system. Specifically, the method for monitoring the condition of a floating breakwater based on a sensing system includes: The raw monitoring data of various sensors in the distributed sensor network of the floating breakwater are acquired, and the raw monitoring data are fused to obtain the dynamic evaluation parameter matrix of the overall structure of the breakwater. Based on the dynamic evaluation parameter matrix, identify abnormal sensor nodes; Determine the health status level and data reliability weight coefficient of the abnormal sensor node; Based on the health status level and data reliability weight coefficient, the data of abnormal sensor nodes are compensated to obtain the corrected load-bearing status and stress accumulation monitoring values. Based on the corrected stress accumulation monitoring values, a stress distribution model of the breakwater anchoring system is established, and a safety warning is triggered when the stress exceeds the fatigue limit; after secondary verification of the warning information, the overall safety status level of the structure is determined. Based on the aforementioned safety status level and combined with data on changes in marine environmental conditions, the load-bearing capacity and stability assessment indicators of the breakwater are updated. Based on the changing trends of the stability assessment indicators, the monitoring frequency and data acquisition strategy of the sensor network are dynamically adjusted to monitor the status of the floating breakwater in real time and obtain monitoring results.

[0007] Preferably, the process of obtaining the dynamic evaluation parameter matrix of the overall structure of the breakwater includes: Acquire raw monitoring data from pressure sensors, strain sensors, accelerometers, and gyroscopes; The raw monitoring data is filtered and standardized to obtain a sensor dataset in a unified format; The Kalman filter algorithm was used to perform time-series fusion on the sensor dataset to obtain the load-bearing status assessment results of each part of the breakwater; Based on the correlation between the load-bearing state assessment results and acceleration and gyroscope data, an attitude perception model of the breakwater is established. Based on the attitude perception model and the load-bearing state data, a dynamic response feature vector of the breakwater structure is constructed. Principal component analysis was performed on the dynamic response feature vector of the breakwater structure to extract key evaluation parameters and obtain the dynamic evaluation parameter matrix.

[0008] Preferably, the process of identifying abnormal sensor nodes based on the dynamic evaluation parameter matrix includes: Based on the stress accumulation characteristic values ​​and attitude change amplitude in the dynamic evaluation parameter matrix; The stress accumulation characteristic value and the attitude change amplitude are compared with preset thresholds respectively. If either exceeds the threshold, the anomaly detection process is triggered. The support vector machine algorithm is used to analyze all sensor data at the current moment to obtain the anomaly probability distribution matrix; Based on the anomaly probability distribution matrix, cross-validation is performed on adjacent sensor data to obtain a validation result vector; If the data deviation of a certain sensor in the verification result vector exceeds a preset value and the correlation coefficient is lower than a preset value, then the corresponding sensor will be marked as an abnormal state.

[0009] Preferably, the process of determining the health status level and data reliability weighting coefficient of the abnormal sensor node includes: Analyze the data drift patterns of sensor nodes marked as abnormal to identify the drift type; Based on the drift type, the fault type is classified as minor drift or severe drift; Based on the fault type, the sensor's health status level is set; Calculate the basic data reliability score based on the health status level; The basic scores are dynamically adjusted using a Bayesian network algorithm to obtain the final credibility weight coefficient.

[0010] Preferably, the process of obtaining the corrected load-bearing state and stress accumulation monitoring values ​​includes: Acquire real-time monitoring data from abnormal sensor nodes and determine the specific time period and degree of deviation of the abnormal data; Based on the time period, normal data of adjacent nodes within the same time period are obtained, and the theoretical monitoring value is calculated using a spatial interpolation algorithm. Extract reference correction values ​​for similar operating conditions from historical data of abnormal sensors; The theoretical monitoring value and the reference correction value are weighted and fused to obtain the data compensation value; Based on the data compensation value, the corrected bearing state and stress accumulation monitoring value are calculated.

[0011] Preferably, the process of establishing a stress distribution model for the breakwater anchoring system based on the corrected stress accumulation monitoring values, and triggering a safety warning when the stress exceeds the fatigue limit, includes: Based on the spatial coordinates and stress values ​​in the corrected stress accumulation monitoring values, a three-dimensional model of stress distribution in the breakwater anchoring system is constructed. The gradient calculation method is used to identify stress concentration regions and extract stress peak data from these regions. The peak stress data is compared with the material fatigue limit data. If the fatigue limit is exceeded, a structural safety early warning mechanism is triggered.

[0012] Preferably, the process of determining the overall safety status level of the structure after secondary verification of the early warning information includes: Extract the warning information records to be verified from the warning information database, and retrieve the original monitoring data and historical warning records within the corresponding time period to construct a verification dataset; The cross-validation algorithm is used to perform consistency analysis on the validation dataset to filter out reliable early warning information. Obtain real-time operating status parameters of each component of the breakwater and construct a collaborative performance evaluation matrix; Based on the collaborative work performance evaluation matrix, identify key risk sources; Based on the key risk sources and collaborative performance indicators, the overall safety status level of the structure is assessed.

[0013] Preferably, the process of updating the load-bearing capacity and stability assessment indicators of the breakwater includes: Acquire data on changes in marine environmental conditions and breakwater attitude sensing parameters to construct a multi-dimensional environmental parameter dataset; The Kalman filter algorithm is used to perform time-series fusion on the environmental parameter dataset to calculate the correlation coefficient between environmental changes and structural response. Based on the correlation coefficients, an environment-structure correlation feature matrix is ​​constructed, and key environmental variables are identified; Based on the historical change sequence of the key environmental variables, an environmental change trend prediction model is established and predictions are made to obtain trend prediction results. Based on the trend prediction results, update the parameters of the breakwater's dynamic safety assessment model and calculate the dynamic bearing capacity assessment value. Based on the dynamic bearing capacity assessment values, update the bearing capacity and stability assessment indicators of the breakwater.

[0014] Preferably, the process of dynamically adjusting the monitoring frequency and data acquisition strategy of the sensor network based on the changing trend of the stability evaluation index includes: Acquire real-time monitoring data from each sensor node and calculate a sequence of performance evaluation metrics; Based on the performance evaluation index sequence, analyze the changing trend parameters to determine the stability of the network operation status; If the change trend parameter exceeds the threshold, the monitoring frequency adjustment coefficient is determined based on the change magnitude; The acquisition strategy of the sensor nodes is updated according to the monitoring frequency adjustment coefficient. Monitor the communication quality between nodes. If there is an abnormal node, trigger the network self-healing function and use the neighboring node data fusion method to compensate for the missing data. Based on the feedback from the system's operational status, the network topology is adjusted to monitor the status of the floating breakwater in real time and obtain monitoring results.

[0015] The present invention also provides a floating breakwater condition monitoring system based on a sensing system, comprising: The data acquisition module is used to acquire raw monitoring data from various sensors in the distributed sensor network of the floating breakwater; The data fusion processing module is used to fuse the original monitoring data to obtain a dynamic evaluation parameter matrix of the overall structure of the breakwater; An anomaly detection module is used to identify abnormal sensor nodes based on the dynamic evaluation parameter matrix; The fault diagnosis module is used to determine the health status level and data reliability weight coefficient of the abnormal sensor node; The data compensation module is used to compensate the data of abnormal sensor nodes according to the health status level and data reliability weight coefficient to obtain the corrected bearing status and stress accumulation monitoring values. The model building module is used to establish a stress distribution model of the breakwater anchoring system based on the corrected stress accumulation monitoring values, and to trigger a safety warning when the stress exceeds the fatigue limit; after secondary verification of the warning information, the overall safety status level of the structure is determined. The dynamic adjustment module is used to update the bearing capacity and stability assessment indicators of the breakwater based on the safety status level and combined with the data on changes in marine environmental conditions; and to dynamically adjust the monitoring frequency and data acquisition strategy of the sensor network according to the changing trend of the stability assessment indicators, so as to monitor the status of the floating breakwater in real time and obtain monitoring results.

[0016] Compared with the prior art, the present invention has the following advantages and technical effects: This invention addresses the core problem of abnormal sensor data and difficulty in fault diagnosis in the safety monitoring of breakwater structures in marine environments. It utilizes a multi-sensor data fusion algorithm to process raw data from pressure sensors, strain sensors, accelerometers, and gyroscopes, constructing a dynamic evaluation parameter matrix for the overall breakwater structure. An abnormal data identification algorithm performs cross-validation analysis on the monitoring data, automatically identifying abnormal states when sensor data deviations exceed preset thresholds. A sensor fault diagnosis algorithm analyzes data drift patterns to determine the health status level. An automatic error correction mechanism compensates for abnormal data using normal data from adjacent sensors and historical data patterns. A three-dimensional model of anchorage stress distribution is established to achieve structural safety early warning. A dynamic safety assessment model is built by combining data on changes in marine environmental conditions. The monitoring frequency and data acquisition strategy are automatically adjusted according to the changing trends of assessment indicators, achieving intelligent real-time monitoring and early warning of the breakwater structure's health status, significantly improving the reliability and accuracy of marine engineering structure safety monitoring. Attached Figure Description

[0017] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a schematic diagram of the method flow according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the system structure according to an embodiment of the present invention. Detailed Implementation

[0018] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0019] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0020] like Figure 1 As shown, this embodiment provides a method for monitoring the status of a floating breakwater based on a sensing system, including: The raw monitoring data of various sensors in the distributed sensor network of the floating breakwater are acquired and fused to obtain the dynamic evaluation parameter matrix of the overall structure of the breakwater. Based on the dynamic evaluation parameter matrix, identify abnormal sensor nodes; Determine the health status level and data reliability weighting coefficient of abnormal sensor nodes; Based on the health status level and data reliability weight coefficient, the data of abnormal sensor nodes are compensated to obtain the corrected load-bearing status and stress accumulation monitoring values. Based on the corrected stress accumulation monitoring values, a stress distribution model of the breakwater anchoring system is established, and a safety warning is triggered when the stress exceeds the fatigue limit; after secondary verification of the warning information, the overall safety status level of the structure is determined. Based on the safety status level and combined with data on changes in marine environmental conditions, the load-bearing capacity and stability assessment indicators of the breakwater are updated. Based on the changing trends of stability assessment indicators, the monitoring frequency and data acquisition strategy of the sensor network are dynamically adjusted to monitor the status of the floating breakwater in real time and obtain monitoring results.

[0021] Furthermore, the process of obtaining the dynamic evaluation parameter matrix of the overall breakwater structure includes: Acquire raw monitoring data from pressure sensors, strain sensors, accelerometers, and gyroscopes; The raw monitoring data is filtered and standardized to obtain a sensor dataset in a unified format; The Kalman filter algorithm was used to perform time-series fusion of the sensor dataset to obtain the load-bearing status assessment results of each part of the breakwater; Based on the correlation between the load-bearing state assessment results and acceleration and gyroscope data, an attitude perception model of the breakwater is established. Based on the attitude perception model and load-bearing state data, a dynamic response feature vector of the breakwater structure is constructed. Principal component analysis was performed on the dynamic response feature vector of the breakwater structure to extract key evaluation parameters and obtain the dynamic evaluation parameter matrix.

[0022] Furthermore, this embodiment acquires the raw data streams from pressure sensors, strain sensors, accelerometers, and gyroscopes distributed at various monitoring points on the floating breakwater. After filtering and standardizing the sensor data, a sensor dataset with a unified format is obtained. Based on the numerical variation characteristics of the pressure and strain data in the sensor dataset, a Kalman filter algorithm is used to perform time-series fusion processing on the multi-source sensor data. If the pressure data fluctuation exceeds a preset threshold, the load-bearing state anomaly is identified by combining it with the strain data, resulting in an assessment of the load-bearing state of each part of the breakwater. Through correlation analysis between the load-bearing state assessment results and the acceleration and gyroscope values, a breakwater attitude perception model is established. If the acceleration value shows a continuous shift in the vertical direction, it is determined that the breakwater has tilted, and real-time attitude angle parameters are obtained. By coupling the attitude angle parameters with the load-bearing state data, a dynamic response feature vector of the breakwater structure is constructed. Based on the weight distribution of each component in the feature vector, the structural stability index is determined, obtaining the basic data for the dynamic assessment of the structure. By assessing the spatiotemporal distribution characteristics of the structural dynamic evaluation basic data, principal component analysis is used to extract key evaluation parameters. If the evaluation parameters at a certain monitoring point deviate from the normal range, that area is marked as a key monitoring area, thus obtaining a set of core evaluation parameters after dimensionality reduction. Based on the numerical relationships of the core evaluation parameter set, a multidimensional evaluation parameter matrix of the overall structural state of the breakwater is established. The row vectors of the matrix represent the state parameters at different monitoring points, and the column vectors represent different types of evaluation indicators, resulting in a complete dynamic evaluation parameter matrix for the breakwater.

[0023] In one possible implementation, the multi-source sensor data fusion processing of floating breakwaters requires the establishment of a complete data acquisition and processing system.

[0024] Specifically, in this embodiment, pressure sensors are placed at key locations on the bottom and sidewalls of the breakwater to monitor changes in seawater pressure in real time. When the pressure suddenly rises from the normal 50 kPa to 80 kPa, the system immediately initiates anomaly detection procedures. Strain sensors are installed at the main beams and connection nodes of the structure to monitor structural deformation. Normal strain values ​​are maintained within 100 microstrains; values ​​exceeding this range indicate that the structure is under abnormal loads.

[0025] For example, the Kalman filter algorithm plays a crucial role in the multi-source data fusion of this embodiment, achieving optimal data estimation through two stages: prediction and updating. When the pressure data at a monitoring point fluctuates by more than 15%, the system automatically correlates it with the strain data at that point for comprehensive analysis. If the strain value simultaneously increases by more than 20%, the system determines that the load-bearing state of the area is abnormal. This fusion processing method can effectively reduce the false alarm rate of a single sensor and improve monitoring accuracy.

[0026] In one embodiment, the breakwater attitude perception model is constructed using accelerometer and gyroscope data. The accelerometer monitors acceleration changes along three axes. When the vertical acceleration value continuously deviates from the gravitational acceleration of 9.8 m / s² by more than 0.5 m / s², it indicates that the breakwater is tilting. The gyroscope provides angular velocity information, which, combined with the acceleration data, calculates the real-time attitude angle. An early warning is triggered when the roll angle exceeds 5 degrees or the pitch angle exceeds 3 degrees.

[0027] For example, in this embodiment, the structural dynamic response feature vector is constructed by coupling the load-bearing state data with attitude parameters. The feature vector includes multiple dimensions such as pressure components, strain components, and tilt angle components, and the weight of each component is determined according to its influence on structural stability. The weight of the pressure component is set to 0.4, the weight of the strain component is 0.3, and the weight of the attitude component is 0.3. The comprehensive stability index is obtained through weighted calculation.

[0028] In one possible implementation, principal component analysis (PCA) is used to reduce the dimensionality of the structural dynamic assessment data. The original assessment parameters contain 12 dimensions. PCA extracts the top 5 principal components, with a cumulative contribution rate exceeding 85%. When the value of the first principal component at a monitoring point deviates from the normal range by 2 standard deviations, that area is marked as a key monitoring area, achieving effective compression of data dimensions while preserving key information.

[0029] For example, the establishment of the multidimensional evaluation parameter matrix in this embodiment provides a data foundation for the overall status assessment of the breakwater. The matrix has a size of 8×6, representing 8 monitoring points and 6 types of evaluation indicators. The row vectors reflect the comprehensive status of each monitoring point, and the column vectors reflect the spatial distribution characteristics of different indicators. The sliding window mechanism updates the matrix data every 10 minutes. When more than 4 matrix elements show abnormal changes within 3 consecutive time windows, the system automatically triggers a structural safety warning, providing reliable protection for the safe operation of the breakwater.

[0030] Furthermore, based on the dynamic evaluation parameter matrix, the process of identifying anomalous sensor nodes includes: Based on the stress accumulation characteristic values ​​and attitude change amplitude in the dynamic evaluation parameter matrix; The stress accumulation characteristic value and the attitude change amplitude are compared with preset thresholds respectively. If either exceeds the threshold, the anomaly detection process is triggered. The support vector machine algorithm is used to analyze all sensor data at the current moment to obtain the anomaly probability distribution matrix; Based on the anomaly probability distribution matrix, cross-validation is performed on adjacent sensor data to obtain a validation result vector; If the data deviation of a certain sensor in the verification result vector exceeds the preset value and the correlation coefficient is lower than the preset value, the corresponding sensor will be marked as abnormal.

[0031] Furthermore, this embodiment uses an eigenvalue decomposition method to extract stress accumulation feature values ​​based on a dynamic evaluation parameter matrix, and calculates the change amplitude between adjacent time windows based on the attitude angle sequence, obtaining a stress feature value vector and an attitude change amplitude vector. If the maximum eigenvalue in the stress feature value vector exceeds a preset stress threshold or the maximum change amplitude in the attitude change amplitude vector exceeds a preset attitude threshold, an anomaly detection process is triggered, acquiring the data streams of all sensors at the current moment for in-depth analysis. An anomaly data identification model is constructed using a support vector machine algorithm, using the data stream as the input feature vector, calculating the distance between each sensor data point and the normal data distribution boundary, and obtaining an anomaly probability distribution matrix. Based on the anomaly probability values ​​corresponding to each sensor in the anomaly probability distribution matrix, a cross-validation method is used to verify the consistency of the monitoring data of adjacent sensors, calculating the data deviation and correlation coefficient, and obtaining a verification result vector. If the verification result vector shows that the data deviation of a certain sensor exceeds the dynamically adjusted threshold range and the correlation coefficient is lower than the preset correlation threshold, the system automatically generates an anomaly identification code and updates the status code of that sensor to an anomaly state.

[0032] This embodiment also employs time series analysis to retrospectively analyze historical data of sensors marked as abnormal states, identify the start time and duration of the abnormality, and update the weight coefficients of the corresponding sensors in the dynamic evaluation parameter matrix to obtain the corrected monitoring network configuration parameters.

[0033] In one possible implementation, the eigenvalue decomposition method of this embodiment plays a core role in the stress accumulation analysis of floating breakwaters. By performing eigenvalue decomposition on the 8×6-dimensional dynamic evaluation parameter matrix, six eigenvalues ​​are extracted, with the largest eigenvalue typically reflecting the main stress concentration direction. When the stress eigenvalue in a certain area rises from the normal 1.2 to 2.8, exceeding the preset stress threshold of 2.5, the system immediately identifies stress accumulation in that area. Simultaneously, attitude angle sequence analysis calculates the angle change between adjacent 10-minute time windows. When the roll angle change suddenly increases from 0.8 degrees to 3.2 degrees, exceeding the preset attitude threshold of 3.0 degrees, the anomaly detection process is triggered.

[0034] For example, in the construction of the support vector machine anomaly detection model, the system uses real-time data streams from 64 sensors as input feature vectors. Each data point contains multi-dimensional information such as pressure, strain, and acceleration values. The model generates a 64×1-dimensional anomaly probability distribution matrix by calculating the distance between each data point and the boundary of the normal data distribution. When the anomaly probability value of a certain pressure sensor reaches 0.85, it indicates that the sensor's monitoring data has significant abnormal characteristics and requires further verification and analysis.

[0035] In one embodiment, cross-validation is crucial for verifying the consistency of data from adjacent sensors. The system selects four adjacent sensors within a 5-meter radius of the abnormal sensor and calculates the data deviation and correlation coefficient. Under normal circumstances, the correlation coefficient between adjacent sensors should remain above 0.75. When the correlation coefficient between a sensor and its adjacent sensors drops to 0.42, and the data deviation exceeds the dynamic threshold of 15%, the system automatically generates an anomaly identifier code "ABNM_P07_20241201" and updates the sensor's status code to an abnormal state.

[0036] For example, time series backtracking analysis can accurately pinpoint the starting point of anomalies. The system backtracked 72 hours of historical data from sensors marked as having anomalies. Through sliding window analysis, it was found that the anomaly began 48 hours ago at 14:30 and lasted for 6.5 hours. Based on the backtracking analysis results, the system adjusted the weight coefficient of this sensor in the dynamic evaluation parameter matrix from 0.125 to 0.08 to prevent abnormal data from interfering with the overall evaluation results.

[0037] It should be noted that the dynamic update mechanism of the monitoring network configuration parameters ensures the system's adaptability. When more than three sensors are detected to be abnormal, the system automatically reallocates the weight coefficients of each monitoring point, prioritizing the increase in the weight ratio of normal sensors. The corrected configuration parameters are transmitted to each monitoring node through a real-time update algorithm, realizing intelligent adjustment of the monitoring network and effectively improving the reliability and accuracy of safety monitoring of floating breakwater structures.

[0038] Furthermore, the process of determining the health status level and data reliability weighting coefficient of abnormal sensor nodes includes: Analyze the data drift patterns of sensor nodes marked as abnormal to identify the drift type; Based on the type of drift, the fault type is classified as either minor drift or severe drift; Set the sensor's health status level according to the fault type; Calculate the basic data reliability score based on the health status level; A Bayesian network algorithm is used to dynamically adjust the base score to obtain the final credibility weight coefficient.

[0039] Furthermore, this embodiment analyzes data change patterns under abnormal conditions to identify drift types, including linear and nonlinear drift, and determines fault type classification labels based on drift amplitude and duration. Using these fault type classification labels as input parameters, a health status assessment matrix is ​​established. If the fault type is a minor drift, the health status level is set to level two; if the fault type is a severe drift, the health status level is set to level three. Based on the health status level values, a basic data reliability score is calculated using a linear mapping function to obtain a quantitative indicator reflecting the quality of sensor data. A Bayesian network algorithm is used to fuse historical fault frequencies and current drift levels to dynamically adjust the basic data reliability score, resulting in the final reliability weight coefficient. This reliability weight coefficient is used to weight the raw sensor data. If the weight coefficient is lower than a preset threshold, the influence weight of that node's data in the fusion process is reduced, resulting in sensor data output with reliability correction.

[0040] Specifically, when the sensor node detects a sudden jump in temperature data from 25.2℃ to 45.8℃ within 5 minutes, the system immediately initiates a Kalman filter-based fault diagnosis algorithm to analyze the data drift pattern. The algorithm first calculates the data change rate to be 4.12℃ / minute, far exceeding the normal threshold of 0.5℃ / minute. Simultaneously, sliding window statistical analysis reveals that the data variance increased from 0.03 to 2.47, and the standard deviation jumped from 0.17 to 1.57, indicating a severe deterioration in data stability. The system further employs a support vector machine-based fault classification algorithm, inputting the extracted time-domain feature vectors into a trained classification model. By calculating Euclidean distance and similarity matching, the system identifies the anomaly as a sensor drift fault type with a confidence level of 92.3%. Based on the fault diagnosis results, the system uses a fuzzy logic inference engine to evaluate the sensor's health status. The scores for five dimensions—data reliability, response time, drift degree, etc.—are set to 0.23, 0.78, 0.15, 0.67, and 0.41 respectively. A weighted average is used to calculate a comprehensive health index of 0.448, corresponding to a level three warning state. Simultaneously, based on the fault severity and historical reliability records, the system dynamically adjusts the data credibility weight coefficient using Bayesian inference, reducing the sensor's weight from 0.85 to 0.32. This ensures that the faulty sensor's impact on the overall system judgment is minimized during data fusion and triggers a backup sensor activation and maintenance warning mechanism.

[0041] Furthermore, the process of obtaining the corrected load-bearing state and stress accumulation monitoring values ​​includes: Acquire real-time monitoring data from abnormal sensor nodes and determine the specific time period and degree of deviation of the abnormal data; Based on the time period, normal data of adjacent nodes within the same time period are obtained, and the theoretical monitoring value is calculated using a spatial interpolation algorithm. Extract reference correction values ​​for similar operating conditions from historical data of abnormal sensors; The theoretical monitoring value and the reference correction value are weighted and fused to obtain the data compensation value; Based on the data compensation value, the corrected bearing state and stress accumulation monitoring value are calculated.

[0042] Furthermore, this embodiment acquires real-time monitoring data from all sensors in the sensor network. Through comparative analysis of preset threshold ranges, it determines whether the data from each sensor exceeds the normal fluctuation range, thus obtaining fault identification results. If the sensor data exceeds the normal range, an anomaly detection mechanism is activated. A sliding window method is used to analyze the recent data change trend of the sensor, determining the specific time period and degree of deviation of the abnormal data. Based on the time period information of the abnormal data, normal data from adjacent node sensors within the same time period are acquired, and the theoretical monitoring value of the faulty sensor location is calculated using a spatial interpolation algorithm. A historical pattern matching method is used to extract data features under similar operating conditions from the historical normal operating data of the faulty sensor, obtaining a time-series-based reference correction value. Data fusion technology is used to weight and synthesize the interpolation results of adjacent nodes with the historical pattern reference value, and the final data compensation value is determined based on the credibility weight of each data source. The compensated sensor data is acquired, and combined with the overall sensor network's load-bearing state analysis model, the corrected structural load-bearing capacity distribution is calculated. Based on the corrected load-bearing state data, a stress accumulation analysis method is used to process the stress change information at each monitoring point, obtaining accurate stress accumulation monitoring values ​​for state assessment.

[0043] Specifically, when sensor S001 detects an abnormal stress value of 15.8 MPa, the system first activates the fault identification algorithm. Using a Kalman filter, it calculates that the deviation rate between this value and the historical average of 12.3 MPa reaches 28.5%, exceeding the set threshold of 15%, and is therefore determined to be abnormal data. The adjacent sensor data acquisition mechanism is then activated, obtaining real-time stress values ​​of 12.1 MPa and 12.7 MPa from sensors S002 and S003, respectively. Simultaneously, historical data patterns from the same period over the past 30 days are retrieved, revealing that the normal stress range at this location is 11.8-13.2 MPa. The system uses a weighted average algorithm for data compensation, setting the weight of adjacent sensor data to 0.4 and the weight of historical data to 0.6, calculating a correction value of (12.1 × 0.2 + 12.7 × 0.2 + 12.3 × 0.6) = 12.46 MPa. Based on the corrected stress data, the load-bearing status assessment module uses a fuzzy logic algorithm, combined with the structural load-bearing capacity of 25 MPa, to calculate a safety factor of 2.01, determining the load-bearing status to be good. Stress buildup monitoring uses time series analysis to compare the corrected value with the data from the previous hour (12.2 MPa), showing a growth rate of 2.1%, which is lower than the warning threshold of 5%. The system automatically updates the monitoring database and generates a correction report to ensure the continuity and accuracy of structural safety monitoring.

[0044] Furthermore, based on the corrected stress accumulation monitoring values, a stress distribution model for the breakwater anchoring system is established, and the process of triggering a safety warning when the stress exceeds the fatigue limit includes: Based on the spatial coordinates and stress values ​​in the corrected stress accumulation monitoring values, a three-dimensional model of stress distribution in the breakwater anchoring system is constructed. The gradient calculation method is used to identify stress concentration regions and extract stress peak data from these regions. The peak stress data is compared with the material fatigue limit data. If the fatigue limit is exceeded, the structural safety early warning mechanism is triggered.

[0045] Furthermore, this embodiment constructs a three-dimensional model of the stress distribution of the anchoring system using a three-dimensional interpolation method based on the spatial coordinate information and corresponding stress values ​​in the corrected monitoring data, thereby obtaining a complete stress field distribution state within the anchoring system. The stress concentration region is identified using a gradient calculation method on the three-dimensional stress distribution model. If the rate of change of the stress gradient in a certain region exceeds a preset gradient threshold, that region is determined to be a stress concentration region. Stress peak data is extracted from the stress concentration regions, and a support vector machine algorithm is used to classify and identify the stress peaks, obtaining the stress peak distribution characteristics of different types of anchoring components. Based on the fatigue limit parameters pre-stored in the material property database of different anchoring components, a fatigue limit comparison analysis is performed on the stress peak distribution characteristics to determine whether the stress state of each component exceeds the fatigue limit. If the stress peak exceeds the corresponding material fatigue limit threshold, a structural safety early warning mechanism is triggered, generating early warning information data including the coordinates of the exceeding limit location, the degree of exceeding the limit, and the risk level. The early warning information data is transmitted to the monitoring center system in real time via a wireless communication module, while simultaneously updating the historical records and early warning status identifiers in the anchoring system stress monitoring database.

[0046] Specifically, based on the strain data collected by sensors, the system first uses a Kalman filter algorithm to correct the original monitoring data, setting the process noise variance to 0.01 and the observation noise variance to 0.05. Through iterative calculation of the state transition matrix and the observation matrix, the strain measurement value at the anchoring point is corrected from the original 1250 microstrain to 1180 microstrain. The corrected data is input into the finite element analysis module, and the breakwater anchoring system is meshed using eight-node hexahedral elements, generating a total of 45,000 element nodes. The material parameters are set as follows: elastic modulus 210 GPa, Poisson's ratio 0.3, and density 7850 kg / m³. By solving the stiffness matrix equation K×U=F, where K is the overall stiffness matrix, U is the displacement vector, and F is the load vector, the stress distribution of each node in the anchoring system is calculated. The system identifies stress concentration in the root region of the anchor bolts, with a peak stress reaching 420 MPa, while the fatigue limit of the Q345 steel used in this region is 380 MPa. When the peak stress exceeds the fatigue limit threshold, the early warning algorithm is immediately activated. The system automatically calculates the safety factor as 380 / 420=0.905, which is lower than the set safety threshold of 1.2, triggering a red early warning signal. At the same time, the early warning information is pushed to the monitoring center, recommending immediate structural repair or load adjustment to ensure the long-term safe operation of the breakwater anchoring system.

[0047] Furthermore, the process of determining the overall safety status level of the structure after secondary verification of the early warning information includes: Extract the warning information records to be verified from the warning information database, and retrieve the original monitoring data and historical warning records within the corresponding time period to construct a verification dataset; Cross-validation algorithm is used to perform consistency analysis on the validation dataset and filter out reliable early warning information; Obtain real-time operating status parameters of each component of the breakwater and construct a collaborative performance evaluation matrix; Identify key risk sources based on the collaborative work performance evaluation matrix; The overall safety status level of the structure is assessed based on key risk sources and collaborative performance indicators.

[0048] Furthermore, this embodiment extracts the early warning information records to be verified from the early warning information database, and simultaneously retrieves the original monitoring data and historical early warning records within the corresponding time period to construct a multi-source data verification dataset. A cross-validation algorithm is used to perform consistency analysis on the multi-source data verification dataset. If the deviation rate between the early warning information and the original monitoring data exceeds a preset deviation threshold, the early warning information is marked as a suspected false alarm. Based on the verification results, reliable early warning information is selected, and real-time operating status parameters of each component are obtained from the breakwater structure database, including the load distribution of the anchoring system, the deformation of the breakwater body, and foundation settlement data. The operating status parameters are normalized using a parameter fusion processing module to establish a collaborative performance evaluation matrix between components, and the mutual influence coefficients between each component are calculated. The analytic hierarchy process (AHP) is used to assign weights to the collaborative performance evaluation matrix. If the influence coefficient of a component exceeds a critical influence threshold, the component is identified as a key risk source. Based on the key risk source identification results and component collaborative performance indicators, the overall safety status of the structure is assessed according to a preset safety level classification standard, generating a status assessment report containing a safety level code and risk description. The status assessment report is transmitted to the decision support system via a data interface, while simultaneously updating the verification records and safety status history archives in the quality assurance database.

[0049] In one embodiment, an anchoring system stress over-limit warning record is extracted from the warning information database. This record shows that a red warning was triggered at 14:30 on March 15, 2024. The system simultaneously retrieves raw strain monitoring data from 30 minutes before and after this time period, including strain values, temperature compensation data, and historical load records for 12 anchoring points, to construct a three-dimensional verification dataset containing warning information, raw data, and historical warning records.

[0050] Specifically, the cross-validation algorithm checks the consistency between the calculated stress value of 420 MPa at the time of the warning trigger and the actual stress value extrapolated from the original strain sensor data during the same period. When the deviation rate between the warning stress value and the sensor-measured stress value of 385 MPa reaches 9.1%, exceeding the preset deviation threshold of 5%, the system automatically marks the warning information as a suspected false alarm to avoid erroneous warnings caused by data processing errors.

[0051] In one possible implementation, a verified and screened reliable early warning information triggering system extracts the current load distribution data of the anchoring system from the breakwater structure database. This data shows that anchor point 8 bears a tensile force of 850 kN, the horizontal displacement monitoring value at the top of the breakwater is 15.2 mm, and the cumulative foundation settlement reaches 8.7 mm. The parameter fusion processing module normalizes the operating status parameters of different dimensions, converting the load, displacement, and settlement values ​​into standardized values ​​between 0 and 1, and establishing a 3×3 collaborative performance evaluation matrix.

[0052] For example, the influence coefficient of the anchoring system load change on the embankment deformation is calculated to be 0.73, and the influence coefficient of the foundation settlement on the redistribution of the anchoring load is 0.68. When a certain influence coefficient exceeds the critical influence threshold of 0.7, the system judges that the component as a key risk source. The analytic hierarchy process (AHP) is used to assign weights to the collaborative work performance evaluation matrix, with the anchoring system weight set at 0.45, the embankment structure weight at 0.35, and the foundation weight at 0.20. The overall structural safety status score is calculated to be 72 points.

[0053] It should be noted that, according to the preset safety level classification standard, a score of 90 or above is considered excellent, 70-90 is good, 50-70 is average, and below 50 is dangerous. Based on a score of 72, the system assesses the overall safety status of the breakwater as good, generating a status assessment report containing the safety level code B and a risk description of "local stress concentration exists in the anchoring system, requiring enhanced monitoring." This report is transmitted in real-time to the decision support system via a standardized data interface, simultaneously updating verification records and historical safety status archives in the quality assurance database, providing reliable data support for subsequent maintenance decisions and risk management.

[0054] Furthermore, the process of updating the load-bearing capacity and stability assessment indicators of breakwaters includes: Acquire data on changes in marine environmental conditions and breakwater attitude sensing parameters to construct a multi-dimensional environmental parameter dataset; The Kalman filter algorithm is used to perform time-series fusion of environmental parameter datasets to calculate the correlation coefficient between environmental changes and structural responses. Based on the correlation coefficients, an environment-structure association feature matrix is ​​constructed, and key environmental variables are identified; Based on the historical change sequences of key environmental variables, an environmental change trend prediction model is established and predictions are made to obtain trend prediction results. Based on the trend prediction results, update the parameters of the dynamic safety assessment model of the breakwater and calculate the dynamic bearing capacity assessment value. Based on the dynamic bearing capacity assessment values, the bearing capacity and stability assessment indicators of the breakwater are updated.

[0055] Furthermore, this embodiment acquires environmental parameters such as wave height, tidal changes, wind speed and direction, and ocean current speed from marine environmental monitoring stations, and simultaneously collects data on tilt angle, vibration frequency, and displacement amplitude output from breakwater attitude sensors to establish a multi-dimensional environmental parameter dataset. A Kalman filter algorithm is used to perform time-series fusion processing on the multi-dimensional environmental parameter dataset. If the rate of change of environmental parameters exceeds a preset fluctuation threshold, the structural response feature vector for that period is extracted, and the correlation coefficient between environmental changes and attitude response is calculated. An environment-structure correlation feature matrix is ​​constructed based on the correlation coefficient, and the dominant environmental factors are determined through feature weight analysis. If the weight value of a certain environmental factor exceeds the dominant influence threshold, it is marked as a key environmental variable. Historical change sequences of key environmental variables are obtained, and a time series analysis method is used to establish an environmental change trend prediction model to calculate the change trend and peak probability of environmental loads in future periods. The parameters of the breakwater dynamic safety assessment model are updated based on the change trend prediction results. The stress distribution and deformation response of the structure under predicted environmental conditions are calculated using the load-response transfer function to obtain the dynamic bearing capacity assessment value. A sliding window method is used to update the dynamic load-bearing capacity assessment value in real time. If the assessment value is lower than the safe load-bearing threshold, a stability index recalculation procedure is triggered to obtain the structural safety state parameters at the current moment. The overall stability level of the breakwater is determined through comparative analysis of the state parameters. Based on the stability level and the load-bearing capacity assessment results, a structural safety state update record is generated and stored in the breakwater safety monitoring database.

[0056] In one embodiment, the marine environmental monitoring station measured the current velocity as 1.2 m / s using an acoustic Doppler current profiler, recorded a significant wave height of 3.8 m using a wave height sensor, and detected a southeasterly wind speed of 15.6 m / s using an anemometer. Simultaneously, the breakwater attitude sensor output data showing a top tilt angle of 0.8 degrees, a dominant vibration frequency of 2.3 Hz, and a horizontal displacement amplitude of 12.5 mm. The system uniformly sampled these different types of environmental parameters and structural response data at a frequency of 10 Hz, constructing a time-series data matrix containing six dimensions of parameters.

[0057] Specifically, this embodiment uses the Kalman filter algorithm to perform noise filtering and state estimation on the multi-dimensional dataset. When the system detects that the wave height rises sharply from 2.1 meters to 4.2 meters within 30 minutes, with a change rate of 100%, far exceeding the preset 30% fluctuation threshold, the algorithm automatically extracts feature vectors such as vibration acceleration, displacement change, and stress response of the breakwater during that period. Pearson correlation analysis shows that the correlation coefficient between wave height and structural vibration amplitude is 0.87, the correlation coefficient between ocean current velocity and foundation shear stress is 0.72, and the correlation coefficient between wind speed and breakwater crest displacement is 0.65.

[0058] In one possible implementation, the system constructs a 4×4 environment-structure correlation feature matrix based on correlation coefficients and uses principal component analysis to calculate the feature weights of each environmental factor. The analysis results show that the wave load weight is 0.42, the current impact weight is 0.28, the wind pressure weight is 0.18, and the tidal variation weight is 0.12. Since the wave load weight exceeds the dominant influence threshold of 0.35, the system marks it as a key environmental variable and initiates a targeted depth monitoring program.

[0059] For example, the time series analysis module extracts the historical wave height variation sequence over the past 72 hours and uses an autoregressive moving average model to build a prediction model. The model calculations show that the wave height will continue to increase over the next 6 hours, with a predicted peak of 5.1 meters and a probability of 78%. Based on this prediction, the dynamic safety assessment model updates the load factor from 1.2 to 1.6, and recalculates using the load-response transfer function, determining that the maximum bending moment of the breakwater under extreme wave conditions will reach 2850 kNm, and the maximum horizontal displacement of the breakwater is expected to be 28.3 mm.

[0060] It should be noted that the sliding window method updates the dynamic bearing capacity assessment value in real time with a 15-minute time window. The current calculated bearing capacity assessment value is 0.73, which is lower than the safe bearing capacity threshold of 0.80. The system immediately triggers a recalculation of the stability index. By comprehensively considering the structural stress state, deformation degree, and cumulative material fatigue damage, the overall stability coefficient of the breakwater is obtained as 1.15, corresponding to a stability level of C-level warning. The system automatically generates a status update record including timestamps, environmental conditions, structural response, and safety level, and stores it in real time in the breakwater safety monitoring database, providing a reliable basis for emergency response decisions.

[0061] Furthermore, the process of dynamically adjusting the monitoring frequency and data acquisition strategy of the sensor network based on the changing trends of stability assessment indicators includes: Acquire real-time monitoring data from each sensor node and calculate a sequence of performance evaluation metrics; Based on the performance evaluation index sequence, analyze the changing trend parameters to determine the stability of the network operation status; If the trend parameter exceeds the threshold, the monitoring frequency adjustment coefficient is determined based on the magnitude of the change. The acquisition strategy of the sensor nodes is updated based on the adjustment coefficient according to the monitoring frequency. Monitor the communication quality between nodes. If there is an abnormal node, trigger the network self-healing function and use the neighboring node data fusion method to compensate for the missing data. Based on the feedback from the system's operational status, the network topology is adjusted to monitor the status of the floating breakwater in real time and obtain monitoring results.

[0062] Furthermore, this embodiment acquires real-time monitoring data from each node in the distributed sensor network, processes the monitoring data using time series analysis, and obtains a performance evaluation index sequence for each sensor node. Based on the performance evaluation index sequence, trend parameters are calculated, and a sliding window statistical method is used to analyze the fluctuation amplitude and rate of change of the indicators to determine the stability of the current network operation. If the trend parameters exceed a preset threshold range, a dynamic adjustment mechanism is activated, and a monitoring frequency adjustment coefficient is determined based on the index change amplitude to obtain a new data acquisition frequency configuration. The acquisition strategy of each sensor node is updated using the new data acquisition frequency configuration, while simultaneously monitoring inter-node communication latency and data transmission success rate to obtain a network connection quality assessment result. If the network connection quality assessment result indicates the presence of abnormal nodes, the network self-healing function is triggered, using a neighboring node data fusion method to compensate for the missing data of abnormal nodes, and the overall system operation status is determined. Based on the feedback information from the overall system operation status, the working mode and data transmission path of the sensor nodes are adjusted, and network resources are reallocated using a load balancing strategy to determine the optimized network topology. The optimized network topology is used to update the distributed monitoring system configuration, continuously monitoring the status changes of each node and data quality indicators to obtain a guarantee mechanism for the continuous and stable operation of the system.

[0063] In one embodiment, the breakwater's distributed sensor network comprises 48 monitoring nodes deployed at different locations along the breakwater for structural health monitoring. The system processes real-time data from each node using time-series analysis methods. Each sensor node collects parameters such as acceleration, strain, and displacement every second, forming a continuous sequence of performance evaluation indicators.

[0064] Specifically, the mean vibration acceleration recorded at node A01 over the past 24 hours was 0.15 m / s², with a standard deviation of 0.032, while the mean strain measurement at node B12 was 85 microstrains, with a coefficient of variation of 0.18.

[0065] For example, the system uses a 60-minute sliding window statistical method to analyze the fluctuations of indicators at each node. When the displacement rate of node C05 is detected to increase from 0.02 mm / min to 0.08 mm / min within 30 minutes, a change of 300%, far exceeding the preset 50% fluctuation threshold, the system immediately activates a dynamic adjustment mechanism. Based on the magnitude of the indicator change, a monitoring frequency adjustment coefficient of 2.5 is calculated, increasing the data acquisition frequency of this node from 1 Hz to 2.5 Hz to ensure accurate capture of abnormal states.

[0066] In one possible implementation, the system continuously monitors inter-node communication latency and data transmission success rate. When it detects that node D18's data transmission success rate drops from the normal 98.5% to 76.2%, and its communication latency increases from an average of 15 milliseconds to 45 milliseconds, the network connection quality assessment module marks it as an abnormal node. At this point, the system triggers the network self-healing function, using weighted fusion compensation based on data from neighboring nodes D17 and D19, with weights set to 0.6 and 0.4 respectively, to estimate missing data based on the principle of spatial correlation.

[0067] It should be noted that the system dynamically adjusts the network topology based on feedback from the overall operational status. When it detects that the eight nodes in the eastern monitoring area are overloaded and the data processing latency exceeds 200 milliseconds, the load balancing strategy automatically reroutes some data streams to the relatively idle node group on the western side. By reallocating network resources, the system adjusts the data transmission path from the original single backbone to a dual-path redundant transmission mode, effectively reducing the risk of network congestion.

[0068] For example, the optimized network topology adopts a hierarchical star architecture, dividing the 48 sensor nodes into 6 subnetworks, each containing 8 nodes and 1 aggregation node. The system continuously monitors the battery level, signal strength, and data quality indicators of each node. When the battery level of node F23 drops below 20%, it automatically switches to a low-power operating mode, adjusting the data acquisition frequency to 0.5 Hz to extend the node's lifespan. Through this dynamic configuration mechanism, the entire distributed monitoring system achieves continuous and stable operation, maintaining data integrity above 99.2%, providing reliable technical support for breakwater safety monitoring.

[0069] like Figure 2 As shown, based on the same inventive concept, this embodiment also provides a floating breakwater status monitoring system based on a sensing system, including: The data acquisition module is used to acquire raw monitoring data from various sensors in the distributed sensor network of the floating breakwater; The data fusion processing module is used to fuse and process the raw monitoring data to obtain a dynamic evaluation parameter matrix of the overall structure of the breakwater; The anomaly detection module is used to identify abnormal sensor nodes based on the dynamic evaluation parameter matrix; The fault diagnosis module is used to determine the health status level and data reliability weight coefficient of abnormal sensor nodes; The data compensation module is used to compensate the data of abnormal sensor nodes according to the health status level and data reliability weight coefficient, so as to obtain the corrected bearing status and stress accumulation monitoring values. The model building module is used to establish a stress distribution model of the breakwater anchoring system based on the corrected stress accumulation monitoring values, and to trigger a safety warning when the stress exceeds the fatigue limit; after secondary verification of the warning information, the overall safety status level of the structure is determined. The dynamic adjustment module is used to update the bearing capacity and stability assessment indicators of the breakwater based on the safety status level and combined with data on changes in marine environmental conditions; and to dynamically adjust the monitoring frequency and data acquisition strategy of the sensor network according to the changing trend of the stability assessment indicators, so as to monitor the status of the floating breakwater in real time and obtain monitoring results.

[0070] The floating breakwater condition monitoring system based on a sensing system provided in this embodiment has all the advantages of the floating breakwater condition monitoring method based on a sensing system provided in Embodiment 1.

[0071] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for monitoring the condition of a floating breakwater based on a sensing system, characterized in that, include: The raw monitoring data of various sensors in the distributed sensor network of the floating breakwater are acquired, and the raw monitoring data are fused to obtain the dynamic evaluation parameter matrix of the overall structure of the breakwater. Based on the dynamic evaluation parameter matrix, identify abnormal sensor nodes; Determine the health status level and data reliability weight coefficient of the abnormal sensor node; Based on the health status level and data reliability weight coefficient, the data of abnormal sensor nodes are compensated to obtain the corrected load-bearing status and stress accumulation monitoring values. Based on the corrected stress accumulation monitoring values, a stress distribution model of the breakwater anchoring system is established, and a safety warning is triggered when the stress exceeds the fatigue limit. After verifying the early warning information a second time, the overall safety status level of the structure is determined; Based on the aforementioned safety status level and combined with data on changes in marine environmental conditions, the load-bearing capacity and stability assessment indicators of the breakwater are updated. Based on the changing trends of the stability assessment indicators, the monitoring frequency and data acquisition strategy of the sensor network are dynamically adjusted to monitor the status of the floating breakwater in real time and obtain monitoring results.

2. The method according to claim 1, characterized in that, The process of obtaining the dynamic evaluation parameter matrix of the overall breakwater structure includes: Acquire raw monitoring data from pressure sensors, strain sensors, accelerometers, and gyroscopes; The raw monitoring data is filtered and standardized to obtain a sensor dataset in a unified format; The Kalman filter algorithm was used to perform time-series fusion on the sensor dataset to obtain the load-bearing status assessment results of each part of the breakwater; Based on the correlation between the load-bearing state assessment results and acceleration and gyroscope data, an attitude perception model of the breakwater is established. Based on the attitude perception model and the load-bearing state data, a dynamic response feature vector of the breakwater structure is constructed. Principal component analysis was performed on the dynamic response feature vector of the breakwater structure to extract key evaluation parameters and obtain the dynamic evaluation parameter matrix.

3. The method according to claim 1, characterized in that, The process of identifying abnormal sensor nodes based on the dynamic evaluation parameter matrix includes: Based on the stress accumulation characteristic values ​​and attitude change amplitude in the dynamic evaluation parameter matrix; The stress accumulation characteristic value and the attitude change amplitude are compared with preset thresholds respectively. If either exceeds the threshold, the anomaly detection process is triggered. The support vector machine algorithm is used to analyze all sensor data at the current moment to obtain the anomaly probability distribution matrix; Based on the anomaly probability distribution matrix, cross-validation is performed on adjacent sensor data to obtain a validation result vector; If the data deviation of a certain sensor in the verification result vector exceeds a preset value and the correlation coefficient is lower than a preset value, then the corresponding sensor will be marked as an abnormal state.

4. The method according to claim 1, characterized in that, The process of determining the health status level and data reliability weighting coefficient of the abnormal sensor node includes: Analyze the data drift patterns of sensor nodes marked as abnormal to identify the drift type; Based on the drift type, the fault type is classified as minor drift or severe drift; Based on the fault type, the sensor's health status level is set; Calculate the basic data reliability score based on the health status level; The basic scores are dynamically adjusted using a Bayesian network algorithm to obtain the final credibility weight coefficient.

5. The method according to claim 1, characterized in that, The process of obtaining the corrected load-bearing state and stress accumulation monitoring values ​​includes: Acquire real-time monitoring data from abnormal sensor nodes and determine the specific time period and degree of deviation of the abnormal data; Based on the time period, normal data of adjacent nodes within the same time period are obtained, and the theoretical monitoring value is calculated using a spatial interpolation algorithm. Extract reference correction values ​​for similar operating conditions from historical data of abnormal sensors; The theoretical monitoring value and the reference correction value are weighted and fused to obtain the data compensation value; Based on the data compensation value, the corrected bearing state and stress accumulation monitoring value are calculated.

6. The method according to claim 1, characterized in that, Based on the corrected stress accumulation monitoring values, the process of establishing a stress distribution model for the breakwater anchoring system and triggering a safety warning when the stress exceeds the fatigue limit includes: Based on the spatial coordinates and stress values ​​in the corrected stress accumulation monitoring values, a three-dimensional model of stress distribution in the breakwater anchoring system is constructed. The gradient calculation method is used to identify stress concentration regions and extract stress peak data from these regions. The peak stress data is compared with the material fatigue limit data. If the fatigue limit is exceeded, a structural safety early warning mechanism is triggered.

7. The method according to claim 1, characterized in that, The process of determining the overall safety status level of the structure after secondary verification of the early warning information includes: Extract the warning information records to be verified from the warning information database, and retrieve the original monitoring data and historical warning records within the corresponding time period to construct a verification dataset; The cross-validation algorithm is used to perform consistency analysis on the validation dataset to filter out reliable early warning information. Obtain real-time operating status parameters of each component of the breakwater and construct a collaborative performance evaluation matrix; Based on the collaborative work performance evaluation matrix, identify key risk sources; Based on the key risk sources and collaborative performance indicators, the overall safety status level of the structure is assessed.

8. The method according to claim 1, characterized in that, The process of updating the load-bearing capacity and stability assessment indicators of breakwaters includes: Acquire data on changes in marine environmental conditions and breakwater attitude sensing parameters to construct a multi-dimensional environmental parameter dataset; The Kalman filter algorithm is used to perform time-series fusion on the environmental parameter dataset to calculate the correlation coefficient between environmental changes and structural response. Based on the correlation coefficients, an environment-structure correlation feature matrix is ​​constructed, and key environmental variables are identified; Based on the historical change sequence of the key environmental variables, an environmental change trend prediction model is established and predictions are made to obtain trend prediction results. Based on the trend prediction results, update the parameters of the breakwater's dynamic safety assessment model and calculate the dynamic bearing capacity assessment value. Based on the dynamic bearing capacity assessment values, update the bearing capacity and stability assessment indicators of the breakwater.

9. The method according to claim 1, characterized in that, The process of dynamically adjusting the monitoring frequency and data acquisition strategy of the sensor network based on the changing trend of the stability evaluation index includes: Acquire real-time monitoring data from each sensor node and calculate a sequence of performance evaluation metrics; Based on the performance evaluation index sequence, analyze the changing trend parameters to determine the stability of the network operation status; If the change trend parameter exceeds the threshold, the monitoring frequency adjustment coefficient is determined based on the change magnitude; The acquisition strategy of the sensor nodes is updated according to the monitoring frequency adjustment coefficient. Monitor the communication quality between nodes. If there is an abnormal node, trigger the network self-healing function and use the neighboring node data fusion method to compensate for the missing data. Based on the feedback from the system's operational status, the network topology is adjusted to monitor the status of the floating breakwater in real time and obtain monitoring results.

10. A floating breakwater condition monitoring system based on a sensing system, characterized in that, include: The data acquisition module is used to acquire raw monitoring data from various sensors in the distributed sensor network of the floating breakwater; The data fusion processing module is used to fuse the original monitoring data to obtain a dynamic evaluation parameter matrix of the overall structure of the breakwater; An anomaly detection module is used to identify abnormal sensor nodes based on the dynamic evaluation parameter matrix; The fault diagnosis module is used to determine the health status level and data reliability weight coefficient of the abnormal sensor node; The data compensation module is used to compensate the data of abnormal sensor nodes according to the health status level and data reliability weight coefficient to obtain the corrected bearing status and stress accumulation monitoring values. The model building module is used to establish a stress distribution model of the breakwater anchoring system based on the corrected stress accumulation monitoring values, and to trigger a safety warning when the stress exceeds the fatigue limit. After verifying the early warning information a second time, the overall safety status level of the structure is determined; The dynamic adjustment module is used to update the bearing capacity and stability assessment indicators of the breakwater based on the safety status level and combined with the data on changes in marine environmental conditions; and to dynamically adjust the monitoring frequency and data acquisition strategy of the sensor network according to the changing trend of the stability assessment indicators, so as to monitor the status of the floating breakwater in real time and obtain monitoring results.

Images