Offshore booster station structure health monitoring system based on Internet of Things

Abstract

The invention discloses a maritime booster station structure health monitoring system based on the Internet of Things, and belongs to the field of maritime booster station monitoring. The maritime booster station structure health monitoring system comprises a data acquisition module, a data processing module, a data analysis module and a fault diagnosis and prediction module; the data acquisition module is used for acquiring various health data of the offshore booster station structure, generating a parameter acquisition signal and sending the parameter acquisition signal to the data processing module; the data processing module is used for carrying out normalization processing on the received data to obtain a hidden worry coefficient; and the data analysis module sets a hidden worry coefficient threshold value, compares the hidden worry coefficient with the hidden worry coefficient threshold value, generates an emergency signal or a normal signal according to a comparison result, and sends the generated emergency signal or normal signal to the fault diagnosis prediction module. According to the method, the structural health condition of the offshore booster station can be comprehensively mastered, the risk of safety accidents caused by structural faults is reduced, and stable operation of an offshore wind power system is guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of offshore booster station monitoring, and more specifically, to an offshore booster station structural health monitoring system based on the Internet of Things. Background Art

[0002] With the rapid development of the offshore wind power industry, offshore substations, as key hubs in offshore wind power systems, shoulder the important functions of gathering electricity, increasing voltage levels, and achieving stable power transmission. However, due to their long-term exposure to complex and harsh marine environments, offshore substations are subject to a variety of factors, including wave impact, strong winds, and seawater corrosion, making their structures susceptible to damage and performance degradation. For example, in extreme weather conditions such as typhoons, the impact of waves can reach tens or even hundreds of tons. Frequent impacts can cause fatigue cracks in the steel structures of offshore substations. The high salinity of seawater increases the corrosion thickness of steel structures by millimeters each year, severely affecting structural strength and, in turn, the safe and stable operation of the entire offshore wind power system. Once a structural failure occurs in an offshore substation, it will not only cause the offshore wind farm to shut down, resulting in huge economic losses, but may also trigger safety accidents and threaten the lives of offshore workers.

[0003] Currently, monitoring methods for the structural health of offshore substations are relatively limited, with manual inspections mostly used. Manual inspections require specialized personnel to travel to the stations by ship or helicopter, costing tens of thousands of yuan per inspection. Furthermore, due to factors such as weather and sea conditions, the average inspection cycle can take several months. This approach is not only inefficient and costly, but also hinders real-time and comprehensive understanding of the structural health of offshore substations. For example, manual inspections struggle to detect microcracks within steel structures or structural changes that occur between inspections. Some traditional monitoring equipment suffers from incomplete data collection, untimely data transmission, and low intelligence. Traditional monitoring equipment often monitors only a single parameter, such as vibration or displacement, and cannot comprehensively assess structural health from multiple dimensions. Data transmission relies on wired networks or low-bandwidth wireless transmission, which is prone to transmission delays or even interruptions in the complex electromagnetic environment of the ocean. Furthermore, it lacks intelligent analysis capabilities, making it impossible to predict potential faults and thus fails to meet the requirements for efficient and accurate monitoring of the structural health of offshore substations.

[0004] In order to solve the above problems, a technical solution is now provided. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides an offshore booster station structural health monitoring system based on the Internet of Things to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] An offshore booster station structural health monitoring system based on the Internet of Things, comprising a data acquisition module, a data processing module, a data analysis module, and a fault diagnosis and prediction module;

[0008] The data acquisition module is used to collect various health data of the offshore booster station structure and generate parameter acquisition signals to be sent to the data processing module;

[0009] The data processing module is used to normalize the received data to obtain the hidden worry coefficient;

[0010] The data analysis module sets a hidden worry coefficient threshold, compares the hidden worry coefficient with the hidden worry coefficient threshold, generates an emergency signal or a normal signal according to the comparison result, and sends the generated emergency signal or normal signal to the fault diagnosis and prediction module;

[0011] The fault diagnosis and prediction module combines the historical data of the offshore substation structure to establish a fault diagnosis model and a prediction model to conduct real-time evaluation and prediction of the health status of the offshore substation structure.

[0012] In a preferred embodiment, the data analysis module operation specifically includes the following:

[0013] Acceleration sensors are placed on the tower of the offshore booster station to obtain vibration acceleration data of the offshore booster station under the action of waves and sea wind;

[0014] Strain sensors are arranged at the stress-bearing parts of the steel structure of the offshore booster station to collect strain data of the steel structure in real time;

[0015] Arrange displacement sensors at the connection between the foundation and the superstructure of the offshore booster station to collect displacement data of the connection between the foundation and the superstructure of the offshore booster station;

[0016] Corrosion monitoring sensors are arranged on the surface of the steel structure of the offshore booster station to monitor the corrosion data of the steel structure of the offshore booster station in real time.

[0017] In a preferred embodiment, the data processing module operation specifically includes the following:

[0018] The vibration acceleration data, strain data, displacement data and corrosion data are marked as D1, D2, D3 and D4 respectively. The vibration acceleration data, strain data, displacement data and corrosion data are normalized to obtain the hidden worry coefficient. The calculation formula is:

[0019]

[0020] Where DF is the hidden worry coefficient, f1, f2, f3, and f4 are the preset proportional coefficients of vibration acceleration data, strain data, displacement data, and corrosion data, respectively, and f1, f2, f3, and f4 are all greater than 0.

[0021] In a preferred embodiment, the data analysis module operation specifically includes the following:

[0022] The thresholds for the hidden danger coefficient are set based on the historical operating data of the offshore booster station, and are divided into the upper limit of the normal threshold, the lower limit of the emergency threshold, and the upper limit of the emergency threshold;

[0023] Under normal circumstances, the potential worry factor should be lower than the upper limit of the normal threshold. When the potential worry factor exceeds the lower limit of the emergency threshold and is lower than the upper limit of the emergency threshold, it indicates that there is a potential problem in the offshore booster station structure;

[0024] When the risk factor exceeds the upper limit of the emergency threshold, it means that there is a serious risk to the structure.

[0025] In a preferred embodiment, the data analysis module operation specifically further includes the following:

[0026] The calculated hidden worry coefficient is compared with the set threshold in real time. If the hidden worry coefficient is less than or equal to the upper limit of the normal threshold, a normal signal is generated and sent to the fault diagnosis and prediction module;

[0027] If the hidden worry coefficient is greater than the lower emergency threshold and less than or equal to the upper emergency threshold, a low-level emergency signal is generated;

[0028] If the hidden worry coefficient is greater than the upper limit of the emergency threshold, a high-level emergency signal is generated, and the corresponding signal is sent to the fault diagnosis and prediction module together with the specific hidden worry coefficient and data source information;

[0029] The data analysis module also regularly performs trend analysis on data collected by various types of sensors, predicts data change trends through the sliding average method, and assists in determining whether there are any abnormal developments.

[0030] In a preferred embodiment, the operation of the fault diagnosis and prediction module specifically includes the following:

[0031] Construct a fault diagnosis model, use the hidden worry coefficients under normal and different fault conditions in historical data, various sensor data and their combinations as training samples, train the fault diagnosis model, and determine the optimal classification hyperplane;

[0032] After receiving the signals and data sent by the data analysis module, the real-time data is input into the trained fault diagnosis model. By calculating the distance and position between the data point and the classification hyperplane, the current health status of the offshore substation structure is determined, including normal, minor fault and serious fault, and the fault type and corresponding fault location are output.

[0033] In a preferred embodiment, the operation of the fault diagnosis and prediction module specifically further includes the following:

[0034] A prediction model is established using a long-short-term memory network, which processes long-term dependencies in time series data and predicts the structural health status of offshore booster stations.

[0035] The historical worry coefficient sequence and each sensor data sequence are used as training data, and training is performed by adjusting the network parameters.

[0036] In a preferred embodiment, the operation of the fault diagnosis and prediction module specifically further includes the following:

[0037] After the training is completed, the prediction model is used to predict the trend of the potential risk factor of the offshore booster station structure and the occurrence of faults in the future based on the current and recent potential risk factors and sensor data. Early warning information is provided in advance to provide a basis for maintenance decision-making.

[0038] At the same time, the fault diagnosis and prediction module also regularly updates and optimizes the fault diagnosis model and prediction model, and retrains the model based on newly collected historical data.

[0039] The technical effects and advantages of the IoT-based offshore booster station structural health monitoring system of the present invention are as follows:

[0040] 1. By deploying multiple sensors on the tower, stress-bearing areas of the steel structure, foundation connections, and the surface of the steel structure, real-time collection of multi-dimensional data such as vibration acceleration, strain, displacement, and corrosion is achieved, comprehensively covering key structural parts of the offshore booster station. The data processing module normalizes data of different dimensions into a potential risk factor. The data analysis module accurately determines the health status of the structure by setting thresholds and conducting real-time comparisons, combined with trend analysis. It can clearly identify everything from normal operation to potential problems and serious risks, enabling operation and maintenance personnel to fully understand the structural health of the offshore booster station in real time.

[0041] 2. Rapidly and accurately determine fault type and location, mine data temporal evolution patterns, and predict potential faults in advance, achieving full intelligence in the entire process from fault diagnosis to risk prediction. Compared with traditional monitoring methods, this significantly improves fault identification and early warning capabilities, effectively reducing the risk of safety accidents caused by structural failures and ensuring the stable operation of offshore wind power systems.

[0042] 3. It can automatically collect data in real time, reducing the frequency and cost of manual inspections. Through accurate fault diagnosis and early warning, it can avoid unplanned downtime and major failures, reduce economic losses caused by equipment damage and power outages, and accurately make maintenance decisions to optimize maintenance resource allocation, reduce maintenance costs, and improve the economic benefits and operational efficiency of offshore substations. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a structural schematic diagram of an offshore booster station structural health monitoring system based on the Internet of Things according to the present invention. DETAILED DESCRIPTION

[0044] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0045] Example 1

[0046] Figure 1 The present invention provides an offshore booster station structural health monitoring system based on the Internet of Things, including a data acquisition module, a data processing module, a data analysis module and a fault diagnosis and prediction module;

[0047] The data acquisition module is used to collect various health data of the offshore booster station structure and generate parameter acquisition signals to be sent to the data processing module;

[0048] The data processing module is used to normalize the received data to obtain the hidden worry coefficient;

[0049] The data analysis module sets a hidden worry coefficient threshold, compares the hidden worry coefficient with the hidden worry coefficient threshold, generates an emergency signal or a normal signal according to the comparison result, and sends the generated emergency signal or normal signal to the fault diagnosis and prediction module;

[0050] The fault diagnosis and prediction module combines the historical data of the offshore substation structure to establish a fault diagnosis model and a prediction model to conduct real-time evaluation and prediction of the health status of the offshore substation structure.

[0051] The data analysis module operation specifically includes the following:

[0052] Acceleration sensors are placed on the tower of the offshore booster station to obtain vibration acceleration data of the offshore booster station under the action of waves and sea wind;

[0053] Strain sensors are arranged at the stress-bearing parts of the steel structure of the offshore booster station to collect strain data of the steel structure in real time;

[0054] Arrange displacement sensors at the connection between the foundation and the superstructure of the offshore booster station to collect displacement data of the connection between the foundation and the superstructure of the offshore booster station;

[0055] Corrosion monitoring sensors are arranged on the surface of the steel structure of the offshore booster station to monitor the corrosion data of the steel structure of the offshore booster station in real time.

[0056] The tower, stress-bearing areas of the steel structure, the connection between the foundation and the superstructure, and the surface of the steel structure are all critical areas that affect the structural safety of an offshore substation. Accelerometers placed on the tower provide real-time monitoring of the tower's vibration under the influence of waves and wind. As the primary load-bearing and wind-receiving component, the tower's vibration directly reflects the stability of the structure. Strain sensors placed on stress-bearing areas of the steel structure accurately capture stress changes. The steel structure bears the entire load of the substation, and strain data helps determine whether the structure is within a safe stress range. Displacement sensors at the connection between the foundation and the superstructure monitor foundation settlement and relative displacement. The foundation is the foundation of the entire substation, and even small displacement changes can pose safety hazards. Corrosion monitoring sensors on the steel structure's surface can promptly detect corrosion problems caused by the high-salt and humid marine environment, preventing corrosion-related structural strength degradation. By collecting data from these key areas, comprehensive monitoring of the substation's structural health is achieved.

[0057] Promptly detect abnormal signs: The marine environment is complex and ever-changing. Factors such as waves, winds, and seawater corrosion constantly threaten the safety of offshore substations. Various sensors collect data in real time. If structural anomalies occur, such as abnormal tower vibration, excessive steel structure strain, sudden foundation displacement, or accelerated steel corrosion, the sensors quickly capture the data changes and transmit the signals to subsequent modules for analysis and processing. Compared to manual inspections, sensor monitoring offers greater real-time performance and sensitivity, enabling early detection of fault signs.

[0058] The operation of the data processing module specifically includes the following:

[0059] The vibration acceleration data, strain data, displacement data and corrosion data are marked as D1, D2, D3 and D4 respectively. The vibration acceleration data, strain data, displacement data and corrosion data are normalized to obtain the hidden worry coefficient. The calculation formula is:

[0060]

[0061] Where DF is the hidden worry coefficient, f1, f2, f3, and f4 are the preset proportional coefficients of vibration acceleration data, strain data, displacement data, and corrosion data, respectively, and f1, f2, f3, and f4 are all greater than 0.

[0062] The vibration acceleration data, strain data, displacement data, and corrosion data collected by offshore substations vary greatly in their dimensions and numerical ranges, making direct comparison and analysis impossible. Through normalization, these data are mapped to a unified scale, eliminating dimensional effects and making different types of data comparable. The hidden worry coefficient calculation formula further consolidates the processed data into a single value, using a single indicator to comprehensively reflect the health status of various aspects of the offshore substation structure. This allows operations and maintenance personnel to quickly and intuitively grasp the overall health of the structure without having to simultaneously monitor multiple data dimensions, greatly improving the efficiency of data interpretation and analysis.

[0063] The data analysis module operation specifically includes the following:

[0064] The thresholds for the hidden danger coefficient are set based on the historical operating data of the offshore booster station, and are divided into the upper limit of the normal threshold, the lower limit of the emergency threshold, and the upper limit of the emergency threshold;

[0065] Under normal circumstances, the potential worry factor should be lower than the upper limit of the normal threshold. When the potential worry factor exceeds the lower limit of the emergency threshold and is lower than the upper limit of the emergency threshold, it indicates that there is a potential problem in the offshore booster station structure;

[0066] When the risk factor exceeds the upper limit of the emergency threshold, it means that there is a serious risk to the structure.

[0067] The operation of the data analysis module specifically includes the following:

[0068] The calculated hidden worry coefficient is compared with the set threshold in real time. If the hidden worry coefficient is less than or equal to the upper limit of the normal threshold, a normal signal is generated and sent to the fault diagnosis and prediction module;

[0069] If the hidden worry coefficient is greater than the lower emergency threshold and less than or equal to the upper emergency threshold, a low-level emergency signal is generated;

[0070] If the hidden worry coefficient is greater than the upper limit of the emergency threshold, a high-level emergency signal is generated, and the corresponding signal is sent to the fault diagnosis and prediction module together with the specific hidden worry coefficient and data source information;

[0071] The data analysis module also regularly performs trend analysis on data collected by various types of sensors, predicts data change trends through the sliding average method, and assists in determining whether there are any abnormal developments.

[0072] Based on historical operational data from offshore substations, the system sets thresholds for the potential risk factor (PFC) and categorizes them into the upper normal threshold, the lower emergency threshold, and the upper emergency threshold. This grading approach aligns with actual operating patterns. The upper normal threshold effectively distinguishes normal operating fluctuations from potential risks, preventing misjudgments. The dual emergency threshold clearly delineates potential issues from serious risks. By comparing the PFC with the thresholds in real time, generating normal, low-level emergency, and high-level emergency signals, O&M personnel can quickly and precisely identify the structural risk level based on the signal level, providing clear guidance for subsequent action and significantly improving the targeted and efficient nature of risk management. Real-time comparison of the PFC with the thresholds allows immediate detection of changes in structural health. Once the PFC exceeds the threshold, the system immediately generates a corresponding signal, significantly reducing risk detection time compared to manual inspections or periodic monitoring. For example, if a sudden stress concentration in a steel structure causes abnormal strain data, real-time monitoring can quickly detect it and issue an emergency signal, enabling O&M personnel to take timely action, preventing structural damage or even accidents caused by undetected hazards and effectively ensuring the stable operation of the offshore substation.

[0073] Regularly analyzing sensor data using a sliding average method for trend analysis can uncover potential patterns of change and predict structural health trends. For example, by analyzing vibration acceleration data trends and observing a gradual increase in values, even if the current potential risk factor remains within the normal range, it can provide a warning of potential future vibration exacerbations. This allows operations and maintenance personnel to plan maintenance schedules in advance and intervene before failures occur, avoiding unplanned downtime, reducing maintenance costs, and extending the service life of offshore substation structures.

[0074] The operation of the fault diagnosis and prediction module specifically includes the following:

[0075] Construct a fault diagnosis model, use the hidden worry coefficients under normal and different fault conditions in historical data, various sensor data and their combinations as training samples, train the fault diagnosis model, and determine the optimal classification hyperplane;

[0076] After receiving the signals and data sent by the data analysis module, the real-time data is input into the trained fault diagnosis model. By calculating the distance and position between the data point and the classification hyperplane, the current health status of the offshore substation structure is determined, including normal, minor fault and serious fault, and the fault type and corresponding fault location are output.

[0077] The operation of the fault diagnosis and prediction module specifically includes the following:

[0078] A prediction model is established using a long-short-term memory network, which processes long-term dependencies in time series data and predicts the structural health status of offshore booster stations.

[0079] The historical worry coefficient sequence and each sensor data sequence are used as training data, and training is performed by adjusting the network parameters.

[0080] The operation of the fault diagnosis and prediction module specifically includes the following:

[0081] After the training is completed, the prediction model is used to predict the trend of the potential risk factor of the offshore booster station structure and the occurrence of faults in the future based on the current and recent potential risk factors and sensor data. Early warning information is provided in advance to provide a basis for maintenance decision-making.

[0082] At the same time, the fault diagnosis and prediction module also regularly updates and optimizes the fault diagnosis model and prediction model, and retrains the model based on newly collected historical data.

[0083] By training the fault diagnosis model based on historical data, using the hidden danger coefficients under normal and different fault conditions, various sensor data, and their combinations as training samples, it can fully learn the data characteristics of the offshore booster station in different health states. After determining the optimal classification hyperplane, when real-time data is input, by calculating the distance and position between the data point and the classification hyperplane, the current health state of the structure can be quickly and accurately determined, and the fault type and corresponding fault location can be clearly output. This eliminates the need for operation and maintenance personnel to investigate each fault individually and can directly identify the root cause, significantly reducing troubleshooting time, improving maintenance efficiency, avoiding repeated repairs due to unclear fault location, and reducing manpower and time costs.

[0084] By using a long-short-term memory network (LSTM) to establish a predictive model, leveraging its ability to process long-term dependencies in time series data, it is possible to deeply explore the temporal evolution of offshore substation structural health data. Using historical risk factor sequences and sensor data sequences as training data, the model can predict future trends in the structural risk factor and potential failure scenarios based on current and recent data. For example, it can predict potential risks such as accelerated corrosion of steel structures and increased foundation settlement in advance, providing early warning information. This information allows operations and maintenance personnel to plan maintenance plans and take preventive measures to prevent safety incidents caused by worsening faults, thereby ensuring the stable operation of offshore substations and minimizing economic losses caused by such accidents.

[0085] Regularly retraining fault diagnosis and prediction models based on newly collected historical data enables them to promptly adapt to changes in offshore substation structure operation, such as changes in data characteristics caused by equipment aging and environmental changes. Over time and as data accumulates, the models are continuously optimized and updated, maintaining a high level of accuracy and reliability, providing effective support for offshore substation structural health monitoring. This prevents misdiagnosis or missed faults due to model lags, ensuring the long-term stable and accurate operation of the monitoring system.

[0086] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0087] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An offshore booster station structural health monitoring system based on the Internet of Things, characterized in that: It includes data acquisition module, data processing module, data analysis module and fault diagnosis and prediction module; The data acquisition module is used to collect various health data of the offshore booster station structure and generate parameter acquisition signals to be sent to the data processing module; The data processing module is used to normalize the received data to obtain the hidden worry coefficient; The data analysis module sets a hidden worry coefficient threshold, compares the hidden worry coefficient with the hidden worry coefficient threshold, generates an emergency signal or a normal signal according to the comparison result, and sends the generated emergency signal or normal signal to the fault diagnosis and prediction module; The fault diagnosis and prediction module combines the historical data of the offshore substation structure to establish a fault diagnosis model and a prediction model to conduct real-time evaluation and prediction of the health status of the offshore substation structure.

2. The IoT-based offshore booster station structural health monitoring system according to claim 1, characterized in that: The data analysis module operation specifically includes the following: Acceleration sensors are placed on the tower of the offshore booster station to obtain vibration acceleration data of the offshore booster station under the action of waves and sea wind; Strain sensors are arranged at the stress-bearing parts of the steel structure of the offshore booster station to collect strain data of the steel structure in real time; Arrange displacement sensors at the connection between the foundation and the superstructure of the offshore booster station to collect displacement data of the connection between the foundation and the superstructure of the offshore booster station; Corrosion monitoring sensors are arranged on the surface of the steel structure of the offshore booster station to monitor the corrosion data of the steel structure of the offshore booster station in real time.

3. The IoT-based offshore booster station structural health monitoring system according to claim 2, characterized in that: The operation of the data processing module specifically includes the following: The vibration acceleration data, strain data, displacement data and corrosion data are marked as D1, D2, D3 and D4 respectively. The vibration acceleration data, strain data, displacement data and corrosion data are normalized to obtain the hidden worry coefficient. The calculation formula is: Where DF is the hidden worry coefficient, f1, f2, f3, and f4 are the preset proportional coefficients of vibration acceleration data, strain data, displacement data, and corrosion data, respectively, and f1, f2, f3, and f4 are all greater than 0.

4. The IoT-based offshore booster station structural health monitoring system according to claim 3, characterized in that: The data analysis module operation specifically includes the following: The thresholds for the hidden danger coefficient are set based on the historical operating data of the offshore booster station, and are divided into the upper limit of the normal threshold, the lower limit of the emergency threshold, and the upper limit of the emergency threshold; Under normal circumstances, the potential worry factor should be lower than the upper limit of the normal threshold. When the potential worry factor exceeds the lower limit of the emergency threshold and is lower than the upper limit of the emergency threshold, it indicates that there is a potential problem in the offshore booster station structure; When the risk factor exceeds the upper limit of the emergency threshold, it means that there is a serious risk to the structure.

5. The IoT-based offshore booster station structural health monitoring system according to claim 4, characterized in that: The operation of the data analysis module specifically includes the following: The calculated hidden worry coefficient is compared with the set threshold in real time. If the hidden worry coefficient is less than or equal to the upper limit of the normal threshold, a normal signal is generated and sent to the fault diagnosis and prediction module; If the hidden worry coefficient is greater than the lower emergency threshold and less than or equal to the upper emergency threshold, a low-level emergency signal is generated; If the hidden worry coefficient is greater than the upper limit of the emergency threshold, a high-level emergency signal is generated, and the corresponding signal is sent to the fault diagnosis and prediction module together with the specific hidden worry coefficient and data source information; The data analysis module also regularly performs trend analysis on data collected by various types of sensors, predicts data change trends through the sliding average method, and assists in determining whether there are any abnormal developments.

6. The IoT-based offshore booster station structural health monitoring system according to claim 5, characterized in that: The operation of the fault diagnosis and prediction module specifically includes the following: Construct a fault diagnosis model, use the hidden worry coefficients under normal and different fault conditions in historical data, various sensor data and their combinations as training samples, train the fault diagnosis model, and determine the optimal classification hyperplane; After receiving the signals and data sent by the data analysis module, the real-time data is input into the trained fault diagnosis model. By calculating the distance and position between the data point and the classification hyperplane, the current health status of the offshore substation structure is determined, including normal, minor fault and serious fault, and the fault type and corresponding fault location are output.

7. The IoT-based offshore booster station structural health monitoring system according to claim 6, characterized in that: The operation of the fault diagnosis and prediction module specifically includes the following: A prediction model is established using a long-short-term memory network, which processes long-term dependencies in time series data and predicts the structural health status of offshore booster stations. The historical worry coefficient sequence and each sensor data sequence are used as training data, and training is performed by adjusting the network parameters.

8. The IoT-based offshore booster station structural health monitoring system according to claim 7, characterized in that: The operation of the fault diagnosis and prediction module specifically includes the following: After the training is completed, the prediction model is used to predict the trend of the potential risk factor of the offshore booster station structure and the occurrence of faults in the future based on the current and recent potential risk factors and sensor data. Early warning information is provided in advance to provide a basis for maintenance decision-making. At the same time, the fault diagnosis and prediction module also regularly updates and optimizes the fault diagnosis model and prediction model, and retrains the model based on newly collected historical data.

Images