Multi-source parameter linkage high-pile wharf pile foundation structure anomaly diagnosis system and method

Abstract

The application provides a multi-source parameter linkage high-pile wharf pile structure anomaly diagnosis system and method, and relates to the technical field of intelligent health monitoring. The core lies in constructing an intelligent diagnosis architecture based on the cooperation of a multi-source monitoring parameter event diagnosis unit and a trend evaluation unit. The event diagnosis unit is used for realizing accurate differentiation of different modes such as elastic overload, pile body damage and overall damage. The trend evaluation unit focuses on the slow degradation process of the structure performance, quantifies the long-term degradation trend and realizes predictive maintenance. A dynamic feedback mechanism is established between the two units, so that the short-term event conclusion can correct the long-term baseline, and the long-term trend can adjust the sensitivity of the short-term event diagnosis, thereby forming an intelligent diagnosis closed loop. The application solves the problems of high false alarm rate, weak damage mode recognition ability and disconnection between long and short period analysis of the traditional single parameter threshold alarm, and can provide intelligent decision support for the safe operation and accurate maintenance of the high-pile wharf.

Description

Technical Field

[0001] This invention relates to the field of intelligent health monitoring technology for port hydraulic structures, and in particular to a multi-source parameter linkage system and method for diagnosing abnormalities in the pile foundation structure of high-pile wharves. Background Technology

[0002] With the deepening of globalization and increasingly close international economic and trade cooperation, ports, as a key link in import and export trade, are crucial for realizing global trade and interconnectivity. High-pile quays, as one of the main structural forms of modern ports, are prone to cumulative damage due to the long-term cyclical loads from ship impacts and wave impacts on their foundation piles. The health of this structure directly affects the safety and efficiency of port operations.

[0003] However, existing high-pile wharf foundation pile structure health monitoring systems have the following shortcomings: 1) Limitations of single-parameter threshold alarms: They rely solely on whether a single parameter exceeds the limit, resulting in a high false alarm rate and an inability to distinguish between elastic overload and structural damage; 2) Lack of multi-source parameter collaborative analysis: They do not fully utilize the spatiotemporal correlation of multi-source data; 3) Insufficient damage mode recognition capability: They struggle to distinguish between different damage modes such as pile body damage, connection node failure, and foundation soil problems; 4) Disconnection between long-term trends and short-term events: They fail to organically combine the response to instantaneous impact events with long-term performance degradation trends. Summary of the Invention

[0004] The purpose of this invention is to provide a method for diagnosing abnormalities in foundation pile structures to solve the above-mentioned technical problems.

[0005] This invention further provides a method for diagnosing structural anomalies in high-pile wharf foundations using multi-source parameter linkage, establishing a dual-unit collaborative architecture for event diagnosis and trend assessment, including the following steps: Step 1: Multi-source data acquisition and preprocessing: A surface strain sensor array was deployed at the key cross-section of the foundation pile, and inclinometers, vibration acceleration sensors, and settlement monitoring points were deployed at the wharf deck or key frame locations to simultaneously collect data from each sensor. Simultaneously, ship AIS data and wave height monitoring data were acquired within a set time window before and after the event trigger. The sensor data were processed through filtering, noise reduction, and temperature compensation to obtain the pre-processed pile strain. Panel tilt angle ,settlement Vibration acceleration The bending moment of the pile was calculated. ; Step 2: Benchmark learning, feature extraction, and acquisition of specified parameters: Set no less than During the initial learning period of 14 days, the static health baseline values ​​of each monitoring parameter were determined using statistical methods. Standard deviation of characteristic fluctuation amplitude under typical operating conditions and the initial healthy stiffness of the foundation piles Simultaneously, the design bearing capacity of various monitoring parameters of the wharf structure is obtained. ; Step 3, Event Diagnosis Unit Operation Flow: When any sensor reading deviates from its baseline by more than a dynamic threshold, i.e. ,in For the magnitude of the sensor's deviation from the baseline, This is the dynamic threshold sensitivity coefficient, which triggers the event diagnosis process. If the threshold exceeds the limit, the number of sensor types... Furthermore, the over-limit sensor types include strain sensors, and the peak time standard deviation of each sensor is less than the set threshold. If the time interval is reached within seconds, a global synchronous diagnosis is triggered to calculate the anomaly isolation degree and perform event diagnosis and classification. The anomaly isolation degree is also used to construct a multi-parameter linkage anomaly feature map: the anomaly isolation degree, strain normalized residual amplitude, and each normalized residual amplitude are used to form a multi-dimensional anomaly feature vector, which is mapped to a preset two-dimensional anomaly feature space to form feature points. By tracking the trajectory drift trend of the feature points, sudden damage and cumulative fatigue damage are distinguished. The cause of the event is also determined based on the ship's AIS data and wave height monitoring data. If the threshold exceeds the limit and the number of sensor types is less than 2, or if the peak times are not synchronized, a local asynchronous diagnosis is triggered. This will immediately trigger a security alarm, among which The threshold coefficient for local anomaly safety alarm; otherwise, check the sensor data quality, correlate with the ship's AIS and wave height monitoring data, mark it as a suspicious event and observe whether it recurs under similar conditions in the future; During the diagnostic process, the bending moment of the foundation piles is compared in real time to indicate the triggering event. With respect to the design bending moment bearing capacity of the pile ,like This will immediately trigger a security alarm, among which This is the load-bearing capacity safety alarm threshold coefficient; Step 4: Trend Assessment Unit Operation Process: This process is executed periodically. After filtering out all transient event data, the long-term static baseline values ​​of each sensor are updated, the long-term damage index of the pile structure is calculated, and the long-term degradation trend of the structural performance is evaluated based on the threshold range of the index. Step 5: Integrated Linkage and Decision-Making between Two Units The residual damage values ​​confirmed in the event diagnosis are fed back to the trend assessment unit to update the long-term baseline; the sensitivity threshold coefficient of the event diagnosis unit is dynamically adjusted according to the degree of degradation shown by the trend assessment; and the combined outputs of the two units are used to generate a comprehensive diagnostic report that includes event causes, location prediction, risk level and maintenance recommendations.

[0006] Furthermore, the formula for calculating the degree of abnormal isolation in step 3 is as follows: in, To normalize the residual amplitude of the strain, For each sensor except strain gauges that exceeds the threshold limit The normalized residual amplitude, The total number of sensors whose parameter values ​​exceed the limit; The anomaly isolation degree is also used to construct a multi-parameter linkage anomaly feature map. Specifically, the anomaly isolation degree calculated in each event diagnosis is combined with the corresponding strain normalized residual amplitude and the normalized residual amplitude of each non-strain sensor to form a multi-dimensional anomaly feature vector. This feature vector is then mapped to a preset two-dimensional anomaly feature space to form feature points of the anomaly event. By tracking the trajectory drift trend of this feature point in the feature space, single sudden damage and cumulative fatigue damage are distinguished: if the feature point trajectory shows a jump migration, it is determined to be a sudden damage event; if the feature point trajectory shows a gradual drift, it is determined to be cumulative fatigue damage.

[0007] Furthermore, the formula for calculating the long-term damage index of the foundation pile structure in step 4 is as follows: in, The slope of the linear regression. The weighting coefficient for the urgency of the stress trend. is the weighting coefficient for the structural stiffness degradation rate, and ; The baseline stress value varies over time. This represents the design value of the pile stress. This is the latest stress baseline value. The ratio of the structural response stiffness of the pile foundation is given. This represents the initial healthy stiffness.

[0008] Furthermore, the event diagnosis and classification method in step 3 is as follows: like This was determined to be an elastic overload event. like and It was determined to be a suspected overall injury event; like and This was determined to be a suspected pile damage incident; in, and These are the threshold values ​​for judging strain-normalized residual amplitude and anomalous isolation degree, respectively.

[0009] Furthermore, the method for assessing the long-term degradation trend of structural performance in step 4 is as follows: like If so, the person is considered to be in a stable health condition. like If so, it is judged as mild degradation; like If so, it is judged as significant degradation; in, and These are the first and second thresholds for trend assessment, respectively.

[0010] Furthermore, the method for calculating the normalized residual amplitude is as follows: Normalized event amplitude ; Normalized residual amplitude ; Normalized outlier ; in, The peak value of the sensor's response. This is the residual value of the sensor's response. This represents the standard deviation of the characteristic fluctuation amplitude of the sensor under typical operating conditions.

[0011] A high-pile wharf foundation pile structure anomaly diagnosis system based on multi-source parameter linkage using the above method is characterized by including a multi-source data acquisition module, a data analysis and processing module, and a data visualization and interaction module. The multi-source data acquisition module is used to acquire real-time data collected by monitoring sensors for key parameters of the foundation piles at typical locations of the high-pile wharf. Specifically, this includes: monitoring strain data at key locations of the pile body through a strain sensor array, measuring overall vibration data of the wharf through a vibration acceleration sensor, measuring the inclination angle data of the wharf deck through an inclinometer, and acquiring settlement data of the wharf through settlement monitoring points; The data analysis and processing module, connected to the multi-source data acquisition module, is used to preprocess the received real-time monitoring data and execute a pile anomaly diagnosis and analysis process based on the dual-unit collaboration of event diagnosis and trend assessment. The diagnosis method includes: calculating the anomaly isolation degree and constructing a multi-parameter linkage anomaly feature map based on the anomaly isolation degree; distinguishing between sudden damage and cumulative fatigue damage by tracking the trajectory drift trend of feature points in the two-dimensional anomaly feature space; and also used to access and acquire real-time data from the Automatic Identification System (AIS) and wave height monitoring data of the wharf area. The data visualization and interaction module, connected to the data analysis and processing module, is used to integrate and display the real-time health status of the entire foundation piles of the wharf, multi-parameter linkage analysis views, event diagnosis results, long-term trend assessment reports, and operation and maintenance management information.

[0012] Furthermore, the strain sensor array in the multi-source data acquisition module is deployed on the key cross-section of the monitored pile. Specifically, four strain sensors are symmetrically arranged around the pile at 90° angles to each other, establishing a comprehensive sensing capability for the bending response of the cross-section and improving the robustness of the system. The vibration acceleration sensor, inclinometer, and settlement monitoring point are deployed at the front edge of the wharf panel or at the key frame position.

[0013] Furthermore, the data analysis and processing module is connected to the multi-source data acquisition module and is used to preprocess the data acquired by the multi-source data acquisition module, acquire external data, and perform pile anomaly diagnosis and analysis.

[0014] Furthermore, the data visualization and interaction module is connected to the data analysis and processing module to display a global situation overview, the health status of individual pile foundations, multi-parameter linkage analysis and decision support, task management views, etc., and to perform human-computer interaction.

[0015] This invention aims to overcome the shortcomings of existing technologies and provide a multi-source parameter linkage system and method for diagnosing structural anomalies in high-pile wharf foundations. The system achieves all-weather monitoring through integrated multi-source data acquisition, intelligent analysis, and visualization modules. The core of the method lies in constructing an intelligent diagnostic architecture based on the collaborative operation of two units: event diagnosis and trend assessment, using multi-source monitoring parameters. The event diagnosis unit focuses on capturing instantaneous anomalies caused by ship impacts, wave impacts, etc., and accurately distinguishes and classifies different modes such as elastic overload, pile damage, and overall damage by analyzing the synchronicity of multi-sensor responses, normalized residual amplitude, and anomaly isolation. The trend assessment unit focuses on the slow degradation process of structural performance, quantifying long-term damage indices and degradation trends by tracking stress baseline drift and stiffness decay, thus achieving predictive maintenance. A dynamic feedback mechanism is established between the two units, allowing short-term event conclusions to correct long-term baselines, and long-term trends to adjust the sensitivity of short-term event diagnosis, thereby forming a self-evolving intelligent diagnostic closed loop.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: Firstly, by using multi-parameter linkage analysis, the false alarm rate is significantly reduced, and the accuracy of damage identification is improved. Secondly, it can distinguish different types of damage modes, providing a basis for precise maintenance; Third, it organically combines instantaneous events with long-term trends to achieve a leap from early warning to prediction capabilities; Fourth, it can accurately correlate structural response anomalies with external conditions, forming a complete chain of evidence of 'load-response-state'.

[0017] Fifth, the visual interactive interface greatly improves the efficiency of interpreting monitoring data and the level of decision support. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the overall decision-making process of the method of this invention. Detailed Implementation

[0019] The specific embodiments of the present invention will now be described with reference to the accompanying drawings.

[0020] To address the technical problems of existing high-pile wharf foundation health monitoring methods, such as high false alarm rates for single-parameter threshold alarms, lack of linkage analysis of multi-source data, weak damage pattern recognition capabilities, and a disconnect between short-term event and long-term trend assessment, this invention proposes a multi-source parameter linkage method and system for diagnosing structural anomalies in high-pile wharf foundations. This system constructs an intelligent diagnostic architecture that coordinates event-driven and trend assessment units, deeply integrating multi-source monitoring data such as pile strain, overall vibration, panel tilt angle, and settlement. This enables comprehensive and intelligent diagnosis of the foundation structure, from identifying instantaneous abnormal impacts to assessing long-term performance degradation, significantly improving early warning accuracy and decision support levels.

[0021] like Figure 1 As shown, the core of the method of this invention, namely the diagnostic process running in the data analysis and processing module, specifically includes the following steps: Step 1: Synchronous Acquisition and Preprocessing of Multi-Source Data: Raw data from various sensors are simultaneously acquired, and filtered, denoised, temperature compensated, and time-stamp aligned to obtain high-quality time-series data on pile strain, panel tilt angle, settlement, and vibration acceleration. Based on the strain data, the bending moment time-series data for key pile sections is calculated. Surface strain sensors are deployed at key pile sections, and tilt angle sensors, vibration acceleration sensors, and settlement monitoring points are deployed on the wharf panel and key frames to collect data synchronously. The data sequences from each sensor are processed with noise reduction and temperature compensation to obtain the pile strain data. Panel tilt angle ,settlement Vibration acceleration The bending moment of the pile was calculated. The formula for calculating the bending moment of the pile is as follows: in, The cross section to be measured Directional bending moment, The cross section to be measured Directional bending moment. Measured cross-section. The formula for calculating directional bending moment is as follows: in, The elastic modulus of concrete. For the converted moment of inertia of the pile section, for The difference in load strain between two monitoring points in the direction, This represents the straight-line distance between the two monitoring points. The measured cross-section. The calculation of directional bending moment is similar.

[0022] The formula for calculating the moment of inertia of the converted section of the pile is as follows: Step 2: Benchmark learning, feature extraction, and acquisition of specified parameters: Set no less than During the initial learning period of 14 days, the static health baseline values ​​of each monitoring parameter were determined using statistical methods. The standard deviation of the characteristic fluctuation amplitude under typical operating conditions and the initial healthy stiffness of the foundation piles Simultaneously, the design bearing capacity of various monitoring parameters of the wharf structure is obtained. The formula for calculating the degree of anomalous isolation is: in, To normalize the residual amplitude of the strain, For each sensor except strain gauges that exceeds the threshold limit The normalized residual amplitude, This represents the total number of sensors whose parameter values ​​exceed the limit.

[0023] Anomaly isolation degree is also used to construct a multi-parameter linkage anomaly feature map. Specifically, the anomaly isolation degree calculated in each event diagnosis is combined with the corresponding strain normalized residual amplitude and the normalized residual amplitude of each non-strain sensor to form a multi-dimensional anomaly feature vector. This feature vector is then mapped to a preset two-dimensional anomaly feature space to form feature points of the anomaly event. By tracking the trajectory drift trend of this feature point in the feature space, single sudden damage and cumulative fatigue damage are distinguished: if the feature point trajectory shows a jump migration, it is determined to be a sudden damage event; if the feature point trajectory shows a gradual drift, it is determined to be cumulative fatigue damage.

[0024] Step 3, Event Diagnosis Unit Operation Flow: When any sensor reading deviates from its baseline by more than a dynamic threshold ( If ), then it is determined to be an abnormal event triggered (where (For the sensitivity coefficient), perform the following steps: Step 3.1: If the threshold exceeds the limit for the number of sensor types Furthermore, the over-limit sensor types include strain sensors, and the peak time standard deviation of each sensor is less than the set threshold. If the event occurs within seconds, a global synchronous diagnostic is triggered. Simultaneously, associated AIS and wave height data are queried to confirm whether any vessels were berthing or experiencing extreme winds and waves at the time the event was triggered. The process is as follows: Step 3.1.1: Normalized residual amplitude calculation: For each sensor exceeding the threshold... Calculate the normalized event amplitude of the strain sensor array with the largest absolute reading. With normalized residual amplitude .

[0025] in, The peak value of the sensor's response. This represents the residual value of the sensor's response.

[0026] It can solve for each sensor that exceeds the threshold. Normalized outlier amplitude .in, The peak value of the sensor's response. This is the residual value of the sensor's response. This represents the standard deviation of the characteristic fluctuation amplitude of the sensor under typical operating conditions.

[0027] Step 3.1.2: Calculation of anomaly isolation degree: Let the strain normalized residual amplitude be... Calculate the degree of abnormal isolation .

[0028] The formula for calculating the degree of anomalous isolation is: .

[0029] Step 3.1.3: Event Diagnosis and Classification, if If it is determined to be an elastic overload event; and It was determined to be a suspected event of systemic injury; if and This was determined to be a suspected damage incident to the pile body.

[0030] in, and These are the threshold values ​​for judging strain-normalized residual amplitude and anomalous isolation degree, respectively.

[0031] Step 3.2: If the number of sensor types exceeding the threshold is less than 2 or the peak times are not synchronized, a local asynchronous diagnosis is triggered, as follows: Step 3.2.1: If This will immediately trigger a security alarm, among which This is the threshold coefficient for local anomaly safety alarms.

[0032] Step 3.2.2: Otherwise, check the quality of the sensor data, correlate it with the ship's AIS and wave height monitoring data, mark it as a suspicious event, and observe whether it is reproduced under subsequent similar operating conditions.

[0033] Step 3.3: Real-time comparison of the bending moment of the foundation piles triggered by the event. With respect to the design bending moment bearing capacity of the pile ,like This will immediately trigger a security alarm, among which This is the load-bearing capacity safety alarm threshold coefficient.

[0034] The event diagnosis and classification methods are as follows: like This was determined to be an elastic overload event. like and It was determined to be a suspected overall injury event; like and This was determined to be a suspected pile damage incident; in, and These are the threshold values ​​for judging strain-normalized residual amplitude and anomalous isolation degree, respectively.

[0035] Step 4: Trend Assessment Unit Operation Process: This process is performed regularly.

[0036] Step 4.1: Baseline Update: After filtering out all transient event data, update the long-term static baseline values ​​of stress for each pile. .

[0037] Step 4.2: Calculate the stress trend urgency separately. With structural stiffness degradation rate .

[0038] Step 4.2.1: Calculate the stress trend urgency For each monitored pile, the stress trend urgency is calculated based on the most unfavorable point on the key monitoring section of that pile. In each trend assessment cycle, the absolute stress value at each point is calculated, and the sensor with the largest absolute value is selected as the representative point for this trend calculation of that pile.

[0039] Extract all valid stress baseline values ​​from the historical data of this representative point to construct a time series. And perform linear regression on it to obtain the stress change rate. .

[0040] Stress trend urgency The calculation formula is: in, The baseline stress value varies over time. This represents the design value of the pile stress. This is the latest stress baseline value.

[0041] The formula for calculating pile stress is: in, This refers to the elastic modulus of concrete. For sensors exist The change in strain at any given moment.

[0042] Step 4.2.2: Calculate the structural stiffness degradation rate The calculation formula is: in, The ratio of the structural response stiffness of the pile foundation is given. This represents the initial healthy stiffness.

[0043] The formula for calculating the response stiffness ratio of a pile structure is: Step 4.3: Calculate the long-term damage index of the pile structure. The calculation formula is: in, The weighting coefficient for the urgency of the stress trend. This is the weighting coefficient for the structural stiffness degradation rate. .

[0044] Step 4.4: Trend assessment, if If so, the health status is considered stable; if If it is, it is judged as mild degradation; if If so, it is judged as significant degradation.

[0045] in, and These are the first and second thresholds for trend assessment, respectively.

[0046] The formula for calculating the long-term damage index of pile structures is as follows: in, The slope of the linear regression. The weighting coefficient for the urgency of the stress trend. is the weighting coefficient for the structural stiffness degradation rate, and ; The baseline stress value varies over time. This represents the design value of the pile stress. This is the latest stress baseline value. The ratio of the structural response stiffness of the pile foundation is given. This represents the initial healthy stiffness.

[0047] Step 5: Integrated linkage and decision-making between the two units.

[0048] Step 5.1: If the event diagnosis unit determines that a damage event has occurred, the resulting residual value will directly update the baseline of the trend assessment unit.

[0049] Step 5.2: If the trend assessment unit shows accelerated degradation, the trigger sensitivity and the judgment threshold of elastic overload events in the event diagnosis process will be dynamically adjusted to make the system more sensitive to subsequent events.

[0050] Step 5.3: Based on the event diagnosis conclusion and trend assessment conclusion, the system outputs a comprehensive diagnostic report including damage location prediction, risk level, and recommended measures, which is displayed in real time in the data visualization and interaction module.

[0051] Positive feedback: The residual values ​​resulting from damage events confirmed by the event diagnosis unit will be used to update the long-term baseline in the trend assessment unit, making the trend assessment more in line with the latest status.

[0052] Parameter reverse adjustment: When the trend assessment shows mild degradation or higher risk, the system dynamically adjusts the trigger sensitivity coefficient of the event diagnosis unit in that area and the judgment threshold of elastic overload event, thereby improving the system's monitoring sensitivity during the performance degradation period.

[0053] Comprehensive decision output: The system integrates the conclusions of both units to generate a comprehensive diagnostic report that includes event type, risk level, degradation trend and maintenance recommendations, and pushes it to the visualization and interaction module.

[0054] In this invention, the linkage analysis of multi-source parameters is reflected at multiple levels: Spatiotemporal linkage: Step 3.2 determines whether the anomaly is caused by external excitation within the same spatiotemporal range by checking the synchronization of the peak times of multiple sensors.

[0055] Physical mechanism linkage: Step 3.3 distinguishes between local damage and overall damage to the pile body by calculating the degree of abnormal isolation. The mechanism is that different damage modes have different degrees and ranges of influence on various sensors.

[0056] Linking long-term trends with instantaneous events: Step five enables mutual verification and parameter adjustment of event and trend diagnostic conclusions, forming an adaptive and self-evolving intelligent diagnostic closed loop.

[0057] The multi-source parametric linkage high-pile wharf foundation pile structure anomaly diagnosis system proposed in this invention specifically includes the following modules: The multi-source data acquisition module is used to acquire real-time monitoring data of the foundation piles and superstructure at typical locations of the high-pile wharf. Specifically, it includes: a pile mechanical response monitoring unit, which measures the bending strain data of the pile through an array of strain sensors deployed at key sections of the pile; and a structural overall response monitoring unit, which measures the overall vibration data of the wharf deck through a vibration acceleration sensor, measures the rotational angle data of the wharf deck through an inclinometer, and measures the differential settlement data of the wharf frame through settlement monitoring points.

[0058] The data analysis and processing module is connected to the multi-source data acquisition module and is used to preprocess the acquired real-time monitoring data and execute the core pile anomaly diagnosis algorithm of this invention.

[0059] The data visualization and interaction module, connected to the data analysis and processing module, is used to intuitively display the real-time health status, historical maintenance records, multi-parameter linkage analysis professional view, diagnostic event flow, and decision support and operation and maintenance management interface of all foundation piles in the entire wharf area.

[0060] The aforementioned system modules are designed to address the core needs of health monitoring of high-pile wharf foundation pile structures, aiming to solve the challenges of fusion analysis and intelligent diagnosis of multi-source heterogeneous data.

[0061] Multi-source data acquisition module It is responsible for synchronously collecting raw monitoring data from various sensors deployed at the dock site. Sensor selection and deployment follow these principles: Strain sensor array: used to directly sense the stress state of the pile body. It is symmetrically arranged around the pile and on the key stress section of the monitored pile. Usually, at least 4 sensors are arranged at 90-degree angles to each other to calculate the bending moment of the section.

[0062] Reason for installation: The critical stress section of the pile body is the area most prone to bending failure. Symmetrical layout can eliminate the influence of axial force and accurately calculate the bidirectional bending moment.

[0063] Vibration acceleration sensors, inclinometers, and settlement monitoring points: used to sense the overall response of the superstructure. Deployed at the leading edge of the dock deck or key scaffolding locations.

[0064] Reason for installation: The front edge of the panel is most sensitive to loads such as ship impacts and waves. Its vibration, rotation and settlement information can effectively reflect the integrity of the load transfer path and the collaborative working status of the pile-superstructure.

[0065] Data Analysis and Processing Module This module is the intelligent core of the invention, and it runs a dual-core algorithm for event diagnosis and trend assessment. This module is also used to access and acquire real-time data from the Automatic Identification System (AIS) and wave height monitoring data in the dock area.

[0066] Data visualization and interaction module This module transforms complex multi-source data and diagnostic results into intuitive graphics, charts, and alarm information, integrating views such as a global situation overview, individual pile foundation health records, and multi-parameter spatiotemporal linkage analysis tools, and supports human-computer interaction.

[0067] The multi-source data acquisition module includes a strain sensor array deployed at key sections of the monitored piles. Specifically, four strain sensors are symmetrically arranged at 90° intervals around the pile, establishing a comprehensive sensing capability for the bending response of the section and improving the system's robustness. Vibration acceleration sensors, inclinometers, and settlement monitoring points are deployed at the leading edge of the wharf panel or key frame locations. The data analysis and processing module, connected to the multi-source data acquisition module, is used for preprocessing the data collected, acquiring external data, and performing pile anomaly diagnosis and analysis. The data visualization and interaction module, connected to the data analysis and processing module, displays a global overview, the individual health status of the pile foundation, multi-parameter linkage analysis and decision support, task management views, and facilitates human-computer interaction.

[0068] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the technical solutions of the embodiments of the present invention.

Claims

1. A method for diagnosing structural anomalies in high-pile wharf foundations using multi-source parametric linkage, characterized in that: Establish a dual-unit collaborative architecture for event diagnosis and trend assessment, including the following steps: Step 1: Multi-source data acquisition and preprocessing: A surface strain sensor array was deployed at the key cross-section of the foundation pile, and inclinometers, vibration acceleration sensors, and settlement monitoring points were deployed at the wharf deck or key frame locations to simultaneously collect data from each sensor. Simultaneously, ship AIS data and wave height monitoring data were acquired within a set time window before and after the event trigger. The sensor data were processed through filtering, noise reduction, and temperature compensation to obtain the pre-processed pile strain. Panel tilt angle ,settlement Vibration acceleration The bending moment of the pile was calculated. ; Step 2: Benchmark learning, feature extraction, and acquisition of specified parameters: Set no less than During the initial learning period of 14 days, the static health baseline values ​​of each monitoring parameter were determined using statistical methods. Standard deviation of characteristic fluctuation amplitude under typical operating conditions and the initial healthy stiffness of the foundation piles Simultaneously, the design bearing capacity of various monitoring parameters of the wharf structure is obtained. ; Step 3, Event Diagnosis Unit Operation Flow: When any sensor reading deviates from its baseline by more than a dynamic threshold, i.e. ,in For the magnitude of the sensor's deviation from the baseline, This is the dynamic threshold sensitivity coefficient, which triggers the event diagnosis process. If the threshold exceeds the limit, the number of sensor types... Furthermore, the over-limit sensor types include strain sensors, and the peak time standard deviation of each sensor is less than the set threshold. If the time interval is reached within seconds, a global synchronous diagnosis is triggered to calculate the anomaly isolation degree and perform event diagnosis and classification. The anomaly isolation degree is also used to construct a multi-parameter linkage anomaly feature map: the anomaly isolation degree, strain normalized residual amplitude, and each normalized residual amplitude are used to form a multi-dimensional anomaly feature vector, which is mapped to a preset two-dimensional anomaly feature space to form feature points. By tracking the trajectory drift trend of the feature points, sudden damage and cumulative fatigue damage are distinguished. The cause of the event is also determined based on the ship's AIS data and wave height monitoring data. If the threshold exceeds the limit and the number of sensor types is less than 2, or if the peak times are not synchronized, a local asynchronous diagnosis is triggered. This will immediately trigger a security alarm, among which The threshold coefficient for local anomaly safety alarm; otherwise, check the sensor data quality, correlate with the ship's AIS and wave height monitoring data, mark it as a suspicious event and observe whether it recurs under similar conditions in the future; During the diagnostic process, the bending moment of the foundation piles is compared in real time to indicate the triggering event. With respect to the design bending moment bearing capacity of the pile ,like This will immediately trigger a security alarm, among which This is the load-bearing capacity safety alarm threshold coefficient; Step 4: Trend Assessment Unit Operation Process: This process is executed periodically. After filtering out all transient event data, the long-term static baseline values ​​of each sensor are updated, the long-term damage index of the pile structure is calculated, and the long-term degradation trend of the structural performance is evaluated based on the threshold range of the index. Step 5: Integrated Linkage and Decision-Making between Two Units The residual damage values ​​confirmed in the event diagnosis are fed back to the trend assessment unit to update the long-term baseline; the sensitivity threshold coefficient of the event diagnosis unit is dynamically adjusted according to the degree of degradation shown by the trend assessment; and the combined outputs of the two units are used to generate a comprehensive diagnostic report that includes event causes, location prediction, risk level and maintenance recommendations.

2. The method for diagnosing structural anomalies in high-pile wharf foundations using multi-source parametric linkage as described in claim 1, characterized in that, The formula for calculating the degree of abnormal isolation in step 3 is as follows: in, To normalize the residual amplitude of the strain, For each sensor except strain gauges that exceeds the threshold limit The normalized residual amplitude, The total number of sensors whose parameter values ​​exceed the limit; The anomaly isolation degree is also used to construct a multi-parameter linkage anomaly feature map. Specifically, the anomaly isolation degree calculated in each event diagnosis, the corresponding strain normalized anomaly amplitude, and the normalized anomaly amplitude of each non-strain sensor are used to form a multi-dimensional anomaly feature vector. This feature vector is mapped to a preset two-dimensional anomaly feature space to form feature points of the anomaly event. By tracking the trajectory drift trend of the feature point in the feature space, single sudden damage and cumulative fatigue damage are distinguished: if the feature point trajectory shows a jump migration, it is determined to be a sudden damage event; if the feature point trajectory shows a gradual drift, it is determined to be cumulative fatigue damage.

3. The method for diagnosing structural anomalies in high-pile wharf foundations using multi-source parametric linkage as described in claim 1, characterized in that, The formula for calculating the long-term damage index of the foundation pile structure in step 4 is as follows: in, The slope of the linear regression. The weighting coefficient for the urgency of the stress trend. is the weighting coefficient for the structural stiffness degradation rate, and ; The baseline stress value varies over time. This represents the design value of the pile stress. This is the latest stress baseline value. The ratio of the structural response stiffness of the pile foundation is given. This represents the initial healthy stiffness.

4. The method for diagnosing structural anomalies in high-pile wharf foundations using multi-source parametric linkage as described in claim 1, characterized in that, The event diagnosis and classification method in step 3 is as follows: like This was determined to be an elastic overload event. like and It was determined to be a suspected overall injury event; like and This was determined to be a suspected pile damage incident; in, and These are the threshold values ​​for judging strain-normalized residual amplitude and anomalous isolation degree, respectively.

5. The method for diagnosing structural anomalies in high-pile wharf foundations using multi-source parametric linkage as described in claim 1, characterized in that, The method for assessing the long-term degradation trend of structural performance in step 4 is as follows: like If so, the person is considered to be in a stable health condition. like If so, it is judged as mild degradation; like If so, it is judged as significant degradation; in, and These are the first and second thresholds for trend assessment, respectively.

6. The method for diagnosing structural anomalies in high-pile wharf foundations using multi-source parametric linkage as described in claim 1, characterized in that, The method for calculating the normalized residual amplitude is as follows: Normalized event amplitude ; Normalized residual amplitude ; Normalized outlier ; in, The peak value of the sensor's response. This is the residual value of the sensor's response. This represents the standard deviation of the characteristic fluctuation amplitude of the sensor under typical operating conditions.

7. A multi-source parametric linkage high-pile wharf foundation pile structure anomaly diagnosis system using the method described in claim 1, characterized in that, It includes a multi-source data acquisition module, a data analysis and processing module, and a data visualization and interaction module; The multi-source data acquisition module is used to acquire real-time data collected by monitoring sensors for key parameters of the foundation piles at typical locations of the high-pile wharf. It monitors strain data at key locations of the pile body through a strain sensor array, measures overall vibration data of the wharf through a vibration acceleration sensor, measures the inclination angle of the wharf panels through an inclinometer, and acquires settlement data of the wharf through settlement monitoring points. The data analysis and processing module, connected to the multi-source data acquisition module, is used to preprocess the received real-time monitoring data and execute a pile anomaly diagnosis and analysis process based on the dual-unit collaboration of event diagnosis and trend assessment. The diagnosis method includes: calculating the anomaly isolation degree and constructing a multi-parameter linkage anomaly feature map based on the anomaly isolation degree; distinguishing between sudden damage and cumulative fatigue damage by tracking the trajectory drift trend of feature points in the two-dimensional anomaly feature space; and also used to access and acquire real-time data from the Automatic Identification System (AIS) and wave height monitoring data of the wharf area. The data visualization and interaction module, connected to the data analysis and processing module, is used to integrate and display the real-time health status of the entire foundation piles of the wharf, multi-parameter linkage analysis views, event diagnosis results, long-term trend assessment reports, and operation and maintenance management information.

8. The multi-source parametric linkage high-pile wharf foundation pile structure anomaly diagnosis system according to claim 7, characterized in that, The strain sensor array in the multi-source data acquisition module is deployed on the key sections of the monitored pile. Specifically, four strain sensors are symmetrically arranged around the pile at 90° angles to each other, establishing a comprehensive sensing capability for the bending response of the section and improving the robustness of the system. The vibration acceleration sensor, inclinometer, and settlement monitoring point are deployed at the front edge of the wharf panel or at key frame positions.

9. The multi-source parametric linkage high-pile wharf foundation pile structure anomaly diagnosis system according to claim 7, characterized in that, The data analysis and processing module is connected to the multi-source data acquisition module and is used to preprocess the data acquired by the multi-source data acquisition module, acquire external data, and perform abnormal pile diagnosis and analysis.

10. The multi-source parametric linkage high-pile wharf foundation pile structure anomaly diagnosis system according to claim 7, characterized in that, The data visualization and interaction module is connected to the data analysis and processing module and is used to display the overall situation overview, the health status of individual pile foundations, multi-parameter linkage analysis and decision support, task management view, and to perform human-computer interaction.

Images