Offshore wind turbine tower health comprehensive monitoring method and system

Abstract

The present invention proposes a comprehensive monitoring method and system for the health of offshore wind turbine towers. Belonging to the field of offshore wind power technology, the method includes: collecting full-range data on the tower based on a multi-sensor network, and transmitting the collected full-range data to the edge computing node for preliminary data preprocessing; through machine vision algorithms, based on high-definition cameras and combined with drone inspections, real-time monitoring of the tower surface and surrounding environmental data is carried out, and convolutional neural networks and generative adversarial networks are used to intelligently analyze images, perform three-dimensional reconstruction, and assess the degree of damage. Through full-range data collection by a multi-sensor network and preliminary data preprocessing by edge computing nodes, the integrity and accuracy of the monitoring data can be ensured, noise and outliers can be eliminated in a timely manner, and the effectiveness of subsequent analysis can be improved.

Description

Technical Field

[0001] The present invention provides a comprehensive health monitoring method and system for offshore wind power towers, belonging to the technical field of offshore wind power. Background Art

[0002] With the growing global demand for renewable energy, offshore wind power, as a key component of clean energy, is becoming increasingly important in terms of safety and reliability in its construction and operation. Offshore wind turbine towers, as key structures supporting wind turbines, are constantly exposed to the complex and ever-changing marine environment, facing multiple challenges such as wind, wave, and current impacts, salt spray corrosion, and the effects of sea ice. These environmental factors can easily lead to structural damage and performance degradation, impacting the overall operational efficiency and safety of the wind farm.

[0003] Traditional offshore wind turbine tower monitoring methods rely primarily on manual inspections and periodic testing, which are limited by long inspection cycles, low efficiency, and susceptibility to human influence. Furthermore, due to the unique characteristics of the marine environment, traditional monitoring methods often struggle to provide a comprehensive and real-time understanding of the tower's health, making it difficult to promptly identify and prevent potential safety hazards. Summary of the Invention

[0004] The present invention provides a comprehensive offshore wind turbine tower health monitoring method and system to solve the problems mentioned in the above background technology:

[0005] The present invention proposes a comprehensive health monitoring method for offshore wind turbine towers, the method comprising:

[0006] S1. Collect full-range data from the tower based on a multi-sensor network and transmit the collected full-range data to the edge computing node for preliminary data preprocessing.

[0007] S2. Using machine vision algorithms, high-definition cameras and drone inspections, we monitor the tower surface and surrounding environment in real time. Using convolutional neural networks and generative adversarial networks, we perform intelligent image analysis, 3D reconstruction, and damage assessment.

[0008] S3. Based on finite element analysis and computational fluid dynamics, a complex physical model of the offshore wind turbine tower is constructed to simulate its dynamic response under the influence of wind, waves and currents.

[0009] S4. Build a big data fusion platform to deeply integrate sensor data, machine vision data, environmental parameter data, and physical model simulation data to form a comprehensive monitoring data set;

[0010] S5. Through machine learning algorithms, the comprehensive monitoring data set is analyzed in real time to automatically identify potential tower failures and damages. Based on the diagnosis results, multi-level warning thresholds are set and warning information is issued to operation and maintenance personnel through multiple channels.

[0011] Furthermore, the S1 includes:

[0012] S11. Adopt a layered deployment strategy, dividing sensors into basic monitoring layer, key monitoring layer and emergency monitoring layer, and automatically adjust the sampling rate and accuracy according to environmental conditions through adaptive sensor algorithms;

[0013] S12, setting up redundancy configuration and fault tolerance mechanism for the sensor network;

[0014] S13. Collect data from the entire tower, compress the collected data using a compression algorithm, and transmit the data to the edge computing node via a low-power wide area network.

[0015] S14. During the transmission process, the data transmission strategy is dynamically adjusted according to the data transmission priority and network load through the intelligent scheduling algorithm;

[0016] S15. The edge computing node uses a time series analysis algorithm to denoise, filter, and smooth the received data, evaluate the data quality, identify and process outliers and missing data, and extract key feature parameters.

[0017] Furthermore, the S14 includes:

[0018] Build a comprehensive evaluation system based on the urgency, importance, and real-time requirements of the data, and prioritize the data to be transmitted;

[0019] During data transmission, the system continuously monitors the network load of the LPWAN, performs real-time analysis and prediction based on big data, and evaluates the current network status and future trends.

[0020] According to the evaluation results, the data transmission strategy is dynamically adjusted based on the intelligent scheduling algorithm of multi-objective optimization;

[0021] Through the transmission strategy effectiveness evaluation mechanism, performance indicators are regularly evaluated. By comparing and analyzing the performance indicators under different transmission strategies, the transmission strategy is identified and optimized. At the same time, a feedback mechanism is established to feed the evaluation results back to the intelligent scheduling algorithm to adaptively optimize and continuously improve the transmission strategy.

[0022] Through cross-layer collaborative transmission optimization strategy, during the data transmission process, cross-layer information sharing and collaborative decision-making are carried out based on the transmission strategy of the current layer and combined with the transmission requirements and resource conditions of other layers.

[0023] Furthermore, the S2 includes:

[0024] S21 uses high-resolution infrared thermal imaging cameras and visible light cameras to perform all-weather, multispectral imaging. Based on AI image recognition algorithms, it processes images taken by drones in real time to automatically track and lock suspected damaged areas on the tower surface.

[0025] S22. Build a deep neural network model and combine it with CNN and GAN to perform fine image segmentation and feature extraction. Apply transfer learning algorithms and use pre-trained models to accelerate the training process of the deep neural network model.

[0026] S23, based on the multi-angle images taken by the UAV, SLAM is used to perform 3D reconstruction and process the reconstructed 3D model.

[0027] S24. Combined with finite element analysis and material mechanics theory, the damaged area in the three-dimensional model is quantitatively evaluated.

[0028] Furthermore, the S3 includes:

[0029] S31. Based on the ocean environment in which the tower is located, set the initial ocean environment parameters of the model, and set the boundary conditions based on the interaction between the tower and the surrounding structures;

[0030] S32. Conduct sensitivity tests on initial and boundary conditions through uncertainty analysis, and evaluate the impact of different parameter changes on model results. Build structural mechanics, fluid mechanics, and thermodynamics models, and simulate the tower's structural response, hydrodynamic characteristics, and heat conduction processes.

[0031] S33. Through coupling algorithms, data exchange and mutual influence between structural mechanics, fluid mechanics, and thermodynamics models are achieved. Nonlinear material models are used to evaluate the nonlinear behavior of materials in extreme environments. Models are validated and calibrated, and model parameters and algorithms are adjusted by comparing historical monitoring data, experimental data, and theoretical analysis results.

[0032] S34. Use a multi-operating-condition simulation scheme to simulate the dynamic response of the tower under different operating conditions. Analyze the dynamic response data based on the time series analysis method and extract the changing trends of key characteristic parameters.

[0033] S35. By calculating the tower's dynamic parameters, the tower's stability under different operating conditions is evaluated. Failure mode and effects analysis is introduced to predict possible failure modes and assess their impact on the tower's overall performance and safety.

[0034] S36. Construct extreme event simulation scenarios, using historical data and numerical simulation methods to generate input parameters under extreme environmental conditions, simulate dynamic responses under extreme events, and evaluate the load-bearing capacity and safety performance of the tower under extreme environments;

[0035] S37. Conduct risk assessments, combine the results of the failure mode and effects analysis, and assess the extent and probability of damage to the tower caused by extreme events. Based on the risk assessment results, formulate appropriate emergency response plans and recovery strategies.

[0036] S38. Collect the differences between the model simulation results and the actual monitoring data, analyze the causes of the differences, optimize and adjust the model based on the analysis results, and perform iterative simulation.

[0037] Furthermore, the S32 includes:

[0038] Identify the sources of uncertainty that affect tower performance evaluation, quantify the identified uncertainty parameters through statistical methods and probability distribution functions, and establish uncertainty ranges or probability distribution models;

[0039] Based on the sensitivity testing scheme, by changing the values ​​of single or multiple uncertainty parameters, the changes in model output are obtained; and by using sensitivity analysis tools, the contribution of different parameter changes to model results is quantified, and key sensitive parameters are identified;

[0040] Based on the sensitivity test results, the impact of uncertainty on tower performance evaluation was evaluated, and structural mechanics models, fluid mechanics models, and thermodynamics models were constructed respectively;

[0041] Collect historical monitoring data, experimental data and theoretical analysis results as benchmark data for model verification, compare and analyze the model simulation results with the benchmark data, evaluate the accuracy and reliability of the model, and adjust and optimize the model parameters and algorithms based on the evaluation results.

[0042] Furthermore, the S36 includes:

[0043] Obtain historical data on extreme climate events, and based on this data and combined with numerical simulations, predict extreme event scenarios; and construct diverse extreme event simulation scenarios;

[0044] Based on the constructed extreme event scenario, the corresponding environmental input parameters are generated using numerical simulation methods and the generated input parameters are verified;

[0045] The verified input parameters are fed into the multi-physics coupling model to simulate the dynamic response under extreme events. The changes in the tower's key performance parameters under extreme conditions are recorded, and the dynamic response characteristics and ultimate load-bearing capacity are analyzed.

[0046] Based on the dynamic response simulation results, the safety and reliability of the tower in extreme environments are evaluated. Combined with the failure mode and effect analysis, the failure modes of the tower under extreme events are identified, and the degree of impact on overall performance and safety is evaluated.

[0047] Furthermore, the S4 includes:

[0048] S41. Design the platform architecture and define unified data interface standards, conduct quality assessments on data sources, and filter or modify data sources that do not meet the requirements;

[0049] S42. Dynamically adjust data collection and transmission strategies based on data source priority, real-time requirements, and data volume through an intelligent scheduling mechanism;

[0050] S43, preprocessing the data, and re-evaluating the quality of the pre-processed data based on the data quality assessment model, and performing secondary filtering and correction on the data based on the assessment results;

[0051] S44. Define the level and granularity of data fusion, select fusion strategies based on specific application scenarios, implement time synchronization and spatial alignment of multi-source data, introduce graph neural networks, and process data relationships and network structures;

[0052] S45. Based on the fusion result verification mechanism, the fusion effect is evaluated by comparing the data quality before and after fusion and the analysis results, and the fusion strategy is iteratively optimized;

[0053] S46. Through distributed data storage strategies, massive data can be distributedly stored and efficiently accessed, and data indexing and partitioning strategies can be set up and implemented.

[0054] Furthermore, the S5 includes:

[0055] S51. Continuously monitor the integrated monitoring data set through a real-time monitoring engine and identify potential faults in real time based on machine learning models.

[0056] S52. Diagnose the identified potential faults, analyze the fault causes, impact scope, and potential consequences, and generate intelligent maintenance decision recommendations based on the fault diagnosis results;

[0057] S53. Setting different levels of warning thresholds according to the severity of the fault, and dynamically adjusting the warning thresholds based on historical data and current monitoring results based on an adaptive warning mechanism;

[0058] S54, and send early warning information to operation and maintenance personnel through multiple channels.

[0059] The present invention proposes a comprehensive offshore wind turbine tower health monitoring system, comprising a memory, a processor, and a computer program stored in and runnable on the memory. The processor executes the program to implement any of the above-described comprehensive offshore wind turbine tower health monitoring methods.

[0060] The beneficial effects of the present invention are as follows: through full-range data collection of multi-sensor networks and preliminary data preprocessing of edge computing nodes, the integrity and accuracy of monitoring data can be ensured, noise and outliers can be eliminated in time, and the effectiveness of subsequent analysis can be improved; combined with machine vision and drone inspections, using convolutional neural networks (CNN) and generative adversarial networks (GAN), it is possible to automatically identify tower surface damage, perform three-dimensional reconstruction and damage degree assessment, thereby achieving accurate monitoring of the health status of the tower structure; through finite element analysis and computational fluid dynamics (CFD) models, the dynamic response of the tower under the action of wind, waves and currents is simulated, which helps to predict the long-term performance and potential risks of the tower; the constructed big data fusion platform can integrate multiple data sources to form The system generates comprehensive monitoring data sets, and machine learning algorithms can analyze these data in real time, automatically identify potential tower failures, and improve the accuracy and timeliness of fault diagnosis. The adaptive sensor algorithm can automatically adjust the sampling rate and accuracy according to environmental conditions, while the intelligent scheduling algorithm can dynamically adjust the data transmission strategy according to the data transmission priority and network load to ensure the efficiency and reliability of data transmission. The multi-level warning threshold and multi-channel warning information sending mechanism can promptly notify operation and maintenance personnel of potential problems, while the intelligent maintenance decision-making suggestions can guide fast and effective fault response measures. Whether it is the data transmission strategy, data fusion strategy or warning threshold, they all have the ability of self-optimization and adaptive adjustment to ensure the long-term stability and effectiveness of the monitoring system. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is a step diagram of the method of the present invention. DETAILED DESCRIPTION

[0062] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, in the absence of conflict, the embodiments of the present application and the features therein may be combined with each other.

[0063] The following description sets forth numerous specific details to facilitate a thorough understanding of the present invention. The embodiments described are merely a portion of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are intended to fall within the scope of protection of the present invention.

[0064] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

[0065] One embodiment of the present invention, as Figure 1 The method for comprehensive monitoring of offshore wind turbine tower health comprises:

[0066] S1. Collect full-range data from the tower based on a multi-sensor network and transmit the collected full-range data to the edge computing node for preliminary data preprocessing.

[0067] S2. Using machine vision algorithms, high-definition cameras and drone inspections, the tower surface and surrounding environment are monitored in real time. Using convolutional neural networks (CNNs) and generative adversarial networks (GANs), images are intelligently analyzed to identify subtle damage such as cracks, rust, and contamination, and perform 3D reconstruction and damage assessment.

[0068] S3. Based on finite element analysis (FEA) and computational fluid dynamics (CFD) technology, a complex physical model of the offshore wind turbine tower is constructed to simulate its dynamic response under the influence of wind, waves and currents.

[0069] S4. Build a big data fusion platform to deeply integrate sensor data, machine vision data, environmental parameter data, and physical model simulation data to form a multi-dimensional, high-dimensional comprehensive monitoring data set;

[0070] S5. Through machine learning algorithms, comprehensive monitoring data sets are analyzed in real time to automatically identify potential tower failures and damage, such as cracks, loose bolts, and weld cracks. Based on the diagnosis results, multi-level warning thresholds are set and warning information is issued to operation and maintenance personnel through multiple channels.

[0071] The working principle of the above technical solution is as follows: by deploying a variety of sensors (such as stress sensors, vibration sensors, temperature sensors, humidity sensors, wind speed and direction sensors, etc.) at key locations of offshore wind turbine towers (such as foundations, tower bodies, transition sections, cabin connections, etc.), the environmental parameters, structural status, and operating status data of the entire tower range can be collected; the collected data is transmitted to edge computing nodes in real time via wired or wireless means; the edge computing nodes perform preliminary processing on the received raw data, including data cleaning (removing noise, outliers, etc.), data compression (reducing transmission and storage burdens), and preliminary data aggregation and classification. Provide a basis for subsequent analysis; combine high-definition cameras fixed on the tower and regular or on-demand drone inspections to obtain detailed images and videos of the tower surface and surrounding environment; use convolutional neural networks (CNN) for image feature extraction and recognition, and generative adversarial networks (GAN) for image enhancement and repair to improve image quality, thereby achieving accurate identification of minor damage such as cracks, rust, and stains; based on the identified damage information, combine 3D reconstruction technology to build a 3D model of the tower surface and quantitatively evaluate the degree of damage; build an accurate finite element model of the offshore wind turbine tower, taking into account material properties, geometric shape, boundary conditions, etc. The mechanical behavior of the tower under static and dynamic loads is simulated by combining factors such as components; CFD technology is used to simulate the complex fluid effects of environmental factors such as wind, waves, and currents on the tower, and analyze their impact on the dynamic response of the tower; the results of FEA and CFD are combined to simulate the dynamic response of the tower under the combined action of wind, waves and currents, and evaluate its structural safety and stability; a big data fusion platform is built that integrates multiple data sources (sensor data, machine vision data, environmental parameter data, physical model simulation data, etc.); advanced data fusion algorithms are used to deeply fuse data from different sources, formats, and precisions to form a multi-dimensional and high-dimensional comprehensive monitoring system. The system further optimizes the fused data set, such as data dimensionality reduction and feature extraction, to improve data quality and analysis efficiency. It uses machine learning algorithms (such as deep learning, support vector machines, random forests, etc.) to conduct real-time analysis of the comprehensive monitoring data set to automatically identify potential tower failures and damage. It accurately identifies fault types and locations such as cracks, loose bolts, and weld cracks through pattern recognition, anomaly detection, and other technologies. It sets multi-level warning thresholds based on the diagnostic results. When the monitoring data exceeds the set threshold, it sends warning information to operation and maintenance personnel through multiple channels such as SMS, email, and APP push to ensure timely response and processing.

[0072] The above technical solution achieves the following results: Through a multi-sensor network, comprehensive data collection is achieved on all key parts of the tower and its surrounding environment, ensuring comprehensive monitoring of the tower's operating status and environmental conditions. Combined with machine vision algorithms and drone inspections, subtle changes in the tower surface and surrounding environment are captured in real time, ensuring timely detection of potential problems. Advanced artificial intelligence technologies such as convolutional neural networks (CNNs) and generative adversarial networks (GANs) are used for intelligent image analysis, accurately identifying subtle damage such as cracks, rust, and stains, improving detection accuracy and efficiency. Three-dimensional reconstruction technology can visually display and quantitatively assess damage to the tower surface, providing a scientific basis for maintenance decisions. A complex physical model constructed using finite element analysis (FEA) and computational fluid dynamics (CFD) techniques simulates the tower's dynamic response in complex environments such as wind, waves, and currents, predicting its long-term operating performance and potential risks. Real-time analysis of comprehensive monitoring data sets using machine learning algorithms automatically identifies potential faults and damage, and sets multi-level warning thresholds to provide early warning information to operation and maintenance personnel to prevent accidents. The constructed big data fusion platform can deeply integrate data from multiple sources and types to form a multi-dimensional and high-dimensional comprehensive monitoring data set, thereby improving the efficiency of data processing and analysis. Early warning information is sent to operation and maintenance personnel through multiple channels such as SMS, email, and APP push to ensure the timeliness and accuracy of information transmission and improve operation and maintenance efficiency. Through real-time monitoring and early warning, preventive maintenance of the tower can be achieved, and emergency repairs after a fault occurs can be avoided, thereby reducing maintenance costs and downtime. Based on monitoring data and early warning information, operation and maintenance personnel and resources can be reasonably arranged, operation and maintenance plans and resource allocation can be optimized, and overall operation and maintenance efficiency can be improved.

[0073] In one embodiment of the present invention, the S1 includes:

[0074] S11. Adopt a layered deployment strategy, dividing sensors into basic monitoring layer, key monitoring layer, and emergency monitoring layer. Adaptive sensor algorithms automatically adjust the sampling rate and accuracy according to environmental conditions (such as wind speed and temperature).

[0075] S12, setting up redundancy configuration and fault tolerance mechanism for the sensor network;

[0076] S13. Collect data from the entire tower, compress the data using a compression algorithm, and transmit the data to an edge computing node via a low-power wide area network (LPWAN), such as LoRa or NB-IoT.

[0077] S14. During the transmission process, the data transmission strategy is dynamically adjusted according to the data transmission priority and network load through the intelligent scheduling algorithm. The transmission strategy is calculated using the following formula:

[0078] T′ i (t) = T i (t)×(α×P i +β×(1-L t ))

[0079] Among them, T′ i (t) represents the transmission strategy of the i-th data packet after adjustment at time t, T i (t) represents the current transmission strategy of the i-th data packet, P i Indicates the priority of the i-th data packet, L t It represents the network load at time t and can be a value between 0 and 1, where 1 indicates a fully loaded network. α and β represent the weighting factors of priority and network load, and α + β = 1.

[0080] S15. The edge computing node uses a time series analysis algorithm to denoise, filter, and smooth the received data, evaluate the data quality, identify and process outliers and missing data, and extract key feature parameters.

[0081] The working principle of the above technical solution is as follows: Based on the monitoring needs and importance of the tower, sensors are divided into a basic monitoring layer, a critical monitoring layer, and an emergency monitoring layer. The basic monitoring layer is responsible for collecting basic environmental and operational data, such as temperature and humidity. The critical monitoring layer focuses on monitoring key parameters for tower structural safety and operational status, such as vibration and stress. The emergency monitoring layer provides additional monitoring capabilities in special circumstances (such as extreme weather and emergencies). The sensors automatically adjust the sampling rate and accuracy based on changes in environmental conditions (such as wind speed and temperature). In harsh or volatile environments, the sampling rate and accuracy are increased to obtain more accurate data; in stable environments, the sampling rate and accuracy are reduced to save energy and extend sensor life. Redundant sensors are installed in the sensor network to ensure that the system can still obtain complete data even if some sensors fail. This configuration improves system reliability and stability. Through software algorithms and hardware design, the sensor network achieves fault tolerance. When a sensor failure is detected, the system automatically switches to a backup sensor or uses data from other sensors to compensate to maintain data continuity and accuracy. The collected raw data is compressed to reduce data transmission bandwidth and energy consumption. The choice of compression algorithm should be determined based on data characteristics and transmission requirements to ensure data quality is maintained while reducing transmission costs. Compressed data is transmitted to edge computing nodes using low-power wide area network technologies such as LoRa and NB-IoT. These technologies offer wide coverage, low power consumption, and low cost, making them suitable for long-range, large-scale data transmission scenarios such as offshore wind turbine towers. Data transmission strategies are dynamically adjusted based on data priority and network load. High-priority data (such as emergency warnings) is transmitted first, while low-priority data (such as routine monitoring data) is transmitted when network load is lower. This ensures the timely transmission of important data and the efficient use of network resources. After receiving the data, the edge computing node first performs denoising, filtering, and smoothing using a time series analysis algorithm. These processing steps eliminate noise and interference signals, improving data accuracy and reliability. The processed data is then quality assessed to identify and address outliers and missing data. Outliers may be caused by sensor failures, data transmission errors, and other reasons and need to be eliminated or corrected. Missing data needs to be filled through methods such as interpolation. Key characteristic parameters such as vibration frequency, stress peak, and temperature change rate are extracted from the processed data. These characteristic parameters can reflect the operating status and health of the tower, providing important basis for subsequent analysis and early warning.

[0082] The effects of the above technical solutions are as follows: through a hierarchical deployment strategy, sensor resources are allocated according to different levels of monitoring needs to ensure the accuracy and frequency of key data collection, while reducing unnecessary data redundancy and improving the overall efficiency of data collection; the adaptive sensor algorithm can automatically adjust the sampling rate and accuracy according to environmental conditions, avoiding unnecessary high sampling rates when the environment is stable, thereby saving energy and extending the service life of the sensor; the setting of redundant configuration and fault-tolerant mechanism enables the system to quickly switch to backup sensors or use other sensor data for compensation when some sensors fail, ensuring data continuity and integrity, and improving the overall reliability of the system; by reducing the system downtime caused by sensor failure, the continuous operation of offshore wind turbine towers is guaranteed and the economic losses caused by downtime are reduced; the compression algorithm is used to compress the collected data, reducing the bandwidth requirements for data transmission and reducing transmission costs; low-power wide area network (LPWAN) technology The application of technologies such as LoRa and NB-IoT reduces energy consumption during data transmission and helps extend the life of sensors and communication equipment. Intelligent scheduling algorithms can dynamically adjust transmission strategies based on data transmission priority and network load conditions to ensure timely transmission of high-priority data while avoiding network congestion and improving the overall efficiency of data transmission. In complex environments such as offshore wind turbine towers, network conditions may fluctuate, and intelligent scheduling algorithms can respond flexibly to ensure stable data transmission. Edge computing nodes can denoise, filter, and smooth received data in real time, reducing the delay in data transmission to the cloud or data center for processing and improving the response speed of the system. Data quality is evaluated through time series analysis algorithms to identify and process outliers and missing data, ensuring the accuracy of subsequent analysis and early warning. The extraction of key characteristic parameters provides an important basis for subsequent analysis and early warning, helping to promptly discover and deal with potential problems. The application of the above formula can make sensor network systems more intelligent and efficient. By dynamically adjusting transmission strategies, optimized data transmission under different environmental conditions is achieved, improving the overall performance of the system and its ability to cope with complex environments. T in the formula i ′(t) can be calculated based on the network load L at the current time t. t and the priority of the data packet P i Dynamically adjust transmission strategy T i (t). This means that the system can flexibly adjust transmission behavior according to the real-time network status and data priority to optimize the efficiency and reliability of data transmission; and balance the impact of priority and network load through weight factors α and β. Priority P i The higher the data packet, the greater the weight it will get in the transmission strategy, and the network load L tThis determines the practical feasibility of data transmission. This dynamic adjustment can effectively manage network load and avoid data loss or transmission delays caused by network congestion or full load. Through intelligent scheduling algorithms, the system can adjust the transmission strategy in real time according to the current network load, thereby maximizing data transmission efficiency. When the network load is low or the packet priority is high, the system will prioritize the transmission of critical data to ensure that important information can be delivered to the edge computing node in a timely manner. The real-time adjustment mechanism in the formula enables the system to quickly respond to changes in network status, thereby ensuring the real-time nature of data transmission. This is particularly important for application scenarios that require rapid decision-making and response, such as ensuring data timeliness and accuracy in monitoring and control systems. By dynamically adjusting the transmission strategy, network bandwidth and transmission resources can be effectively utilized, avoiding waste and inefficient use of resources. This is particularly important in resource-constrained environments such as low-power wide area networks (LPWANs), which can maximize device battery life and reduce transmission costs.

[0083] In one embodiment of the present invention, the step S11 includes:

[0084] Based on the specific operating environment of the wind farm (such as terrain, climate conditions, wind speed distribution, etc.) and operation and maintenance requirements, the monitoring needs are analyzed to obtain demand analysis results; the demand analysis results include clarifying which parameters are necessary for basic monitoring (such as basic wind speed and temperature), which are key performance indicators (such as blade vibration and generator bearing temperature), and which parameters only require high-precision monitoring in emergency situations (such as structural stress in extreme weather); based on the demand analysis results, a three-layer monitoring architecture is constructed; the basic monitoring layer is responsible for continuously monitoring basic environmental parameters and adopts low-power, high-reliability sensors; the key monitoring layer focuses on the operating status of core equipment and deploys high-precision, fast-response sensors; the emergency monitoring layer is equipped with sensors with ultra-high sensitivity and dynamic adjustment capabilities for specific emergency situations, such as typhoons and extreme temperature differences.

[0085] Through adaptive sensing algorithms, the changing trends of environmental parameters (wind speed, temperature, humidity, etc.) are analyzed in real time to predict the impact of environmental parameters on monitoring accuracy and frequency requirements. For example, when the wind speed suddenly increases, the sampling rate and accuracy of the blade vibration sensor are automatically increased to capture more subtle vibration changes.

[0086] Sensors dynamically adjust their operating modes (e.g., sleep, low-power monitoring, high-precision monitoring) based on current task requirements and remaining battery life. Through collaborative mechanisms across sensor networks, sensors at each layer share and collaborate on local information via low-power wireless communication protocols (e.g., Zigbee), making joint decisions on monitoring strategies and resource allocation. For example, when the key monitoring layer detects an abnormal trend, it can trigger the basic monitoring layer to increase monitoring density in related areas, creating a coordinated monitoring effect.

[0087] The working principle of the above technical solution is to conduct an in-depth needs analysis based on the specific operating environment of the wind farm (topography, climate conditions, wind speed distribution, etc.) and the actual operation and maintenance needs. This step aims to clarify which parameters are necessary for daily operation monitoring (such as basic wind speed and temperature), which parameters are key indicators for evaluating wind turbine equipment performance (such as blade vibration and generator bearing temperature), and which parameters require special attention and high-precision monitoring in emergency situations (such as structural stress in extreme weather). Based on the results of the needs analysis, a three-layer monitoring architecture is constructed: basic monitoring layer, key monitoring layer, and emergency monitoring layer. Each layer selects different types of sensors based on its specific monitoring needs and sets corresponding monitoring accuracy and frequency. Adaptive sensing algorithms integrated in sensors or edge computing nodes perform real-time analysis of changing trends in environmental parameters (such as wind speed, temperature, and humidity). These algorithms predict the specific impact of environmental parameter changes on monitoring accuracy and frequency and automatically adjust the sensor sampling rate and monitoring accuracy accordingly. For example, when wind speed suddenly increases, the algorithm triggers the blade vibration sensor to increase its sampling rate and accuracy to ensure it can capture more subtle vibration changes. The sensor dynamically adjusts its operating mode based on its current mission requirements and remaining battery power. This includes entering sleep or low-power monitoring mode to save energy when high-precision monitoring is not required, and switching to high-precision monitoring mode when high-precision monitoring is required. Through this mode adjustment, the sensor can maximize its service life while ensuring monitoring quality. Sensors at each layer share local information and collaborate via low-power wireless communication protocols (such as Zigbee). This collaborative mechanism enables sensors to perceive each other's status and monitoring results, enabling more effective monitoring decisions. When sensors in the key monitoring layer detect abnormal trends, they can trigger the base monitoring layer to increase monitoring density in related areas, creating a coordinated monitoring effect. This coordinated monitoring effect not only improves the comprehensiveness and accuracy of monitoring but also helps to promptly identify and address potential problems.

[0088] The above technical solution achieves the following benefits: Through in-depth analysis of the specific operating environment and O&M requirements of a wind farm, monitoring parameters can be precisely defined, avoiding ineffective monitoring and improving monitoring efficiency. The constructed three-layer monitoring architecture (basic monitoring layer, key monitoring layer, and emergency monitoring layer) enables customized monitoring using sensors of varying precision to meet different monitoring needs. It analyzes the changing trends of environmental parameters in real time and automatically adjusts the sensor sampling rate and accuracy accordingly, ensuring efficient monitoring capabilities in complex and changing environments. Sensors dynamically adjust their operating modes based on current mission requirements and remaining battery power, ensuring monitoring quality, extending sensor lifespan, and optimizing resource utilization. While S11 primarily describes the monitoring architecture and adaptive adjustment, considering the entire system, this layered deployment and adaptive adjustment strategy helps enhance the overall reliability of the system. If some sensors fail, other sensors can quickly fill in the gaps, ensuring data continuity. Ultra-high-sensitivity sensors configured for specific emergency situations can provide critical data support under extreme conditions, effectively supporting the wind farm's emergency response. Through a collaborative mechanism across the sensor network, sensors at each layer can share information and collaborate to make joint decisions on monitoring strategies and resource allocation. This linkage effect helps to improve the comprehensiveness and accuracy of monitoring data; preliminary processing of data at edge computing nodes (such as denoising, filtering, smoothing, etc.) can further improve data quality and provide a reliable basis for subsequent analysis and early warning; the basic monitoring layer uses low-power, high-reliability sensors to reduce energy consumption and operation and maintenance costs; through intelligent scheduling algorithms and dynamic adjustment strategies, it can optimize data transmission and resource allocation, further reducing operation and maintenance costs.

[0089] In one embodiment of the present invention, the S14 includes:

[0090] Build a comprehensive assessment system based on the urgency, importance, and real-time requirements of the data, and prioritize the data to be transmitted. For example, data from the critical monitoring layer (such as abnormal blade vibration and generator overheating) will be given the highest priority, while routine environmental data from the basic monitoring layer may be given a lower priority.

[0091] During data transmission, the network load of the Low Power Wide Area Network (LPWAN) is continuously monitored, including key indicators such as bandwidth utilization, latency, and packet loss rate. Real-time analysis and prediction based on big data are performed to assess the current network status and future trends.

[0092] According to the evaluation results, the data transmission strategy is dynamically adjusted based on the intelligent scheduling algorithm of multi-objective optimization. For example, during peak network load periods, the transmission frequency of low-priority data is automatically reduced or temporarily cached locally, and then transmitted again after the network load is reduced. For high-priority data, network resources are allocated first to ensure the real-time and accuracy of the data.

[0093] Through the transmission strategy effectiveness evaluation mechanism, performance indicators such as data transmission efficiency, data quality, and sensor energy consumption are regularly evaluated. By comparing and analyzing the performance indicators under different transmission strategies, the transmission strategy is identified and optimized. At the same time, a feedback mechanism is established to feed the evaluation results back to the intelligent scheduling algorithm to adaptively optimize and continuously improve the transmission strategy.

[0094] Through cross-layer collaborative transmission optimization strategies, cross-layer information sharing and collaborative decision-making are carried out during data transmission based on the transmission strategy of the current layer and the transmission requirements and resource status of other layers (such as the basic monitoring layer and the key monitoring layer). For example, when emergency monitoring layer data needs to be transmitted urgently, network resources of other layers can be temporarily borrowed or their transmission strategies can be adjusted to ensure the timely transmission of emergency data.

[0095] The working principle of the above technical solution is to build a comprehensive assessment system based on the data's urgency, importance, and real-time requirements. This system considers various factors, such as the data's impact on wind farm operational safety and the frequency of data updates required. Based on this assessment system, data to be transmitted is prioritized. Data from the critical monitoring layer, such as abnormal blade vibration and generator overheating, will be given the highest priority due to their direct impact on wind farm safety. Routine environmental data from the basic monitoring layer, such as temperature and humidity, while also important, typically does not immediately impact wind farm operations and may therefore be given a lower priority. During data transmission, the system continuously monitors the network load of the low-power wide area network (LPWAN), including key metrics such as bandwidth utilization, latency, and packet loss rate. Based on the collected big data, the system performs real-time analysis and forecasting to assess current network status and future trends. This helps the system anticipate potential network congestion or failures and prepare accordingly. Based on the network load assessment results, the intelligent scheduling algorithm uses a multi-objective optimization strategy to dynamically adjust the data transmission strategy. This includes adjusting data transmission priority, frequency, and rate. During peak network load periods, the algorithm automatically reduces the transmission frequency of low-priority data or temporarily caches it locally, retransmitting it again when the network load decreases. For high-priority data, network resources are prioritized to ensure real-time and accurate data. A transmission strategy effectiveness evaluation mechanism regularly assesses performance indicators, including data transmission efficiency, data quality, and sensor energy consumption. These indicators reflect the effectiveness and cost-effectiveness of the current transmission strategy. By comparing and analyzing performance indicators under different transmission strategies, the system can identify the optimal transmission strategy and optimize the existing strategy. The evaluation results are fed back to the intelligent scheduling algorithm as a basis for optimization and continuous improvement. This feedback mechanism enables the system to adaptively adjust transmission strategies to respond to changing network environments and monitoring needs. During data transmission, the system not only monitors the transmission strategy of the current layer but also shares information across layers, including the transmission requirements and resource availability of other layers (such as the basic monitoring layer and the key monitoring layer). Based on this cross-layer information sharing, the system can make collaborative decisions and optimize the overall transmission strategy. For example, when data from the emergency monitoring layer needs to be transmitted urgently, the system can temporarily borrow network resources from other layers or adjust their transmission strategies to ensure timely transmission of emergency data. This collaborative transmission optimization strategy helps improve the overall performance and response speed of the system.

[0096] The above technical solution achieves the following results: By establishing a comprehensive evaluation system and prioritizing data to be transmitted, critical and urgent data is ensured to be transmitted first, thereby improving the efficiency and real-time nature of data transmission. An intelligent scheduling algorithm based on multi-objective optimization dynamically adjusts data transmission strategies and rationally allocates network resources based on network load. This not only avoids network congestion but also optimizes resource utilization and reduces overall energy consumption. Continuous monitoring of network load and real-time analysis and prediction based on big data enable the system to anticipate network status changes and take appropriate preventive measures. This helps reduce packet loss and latency during data transmission, improving data quality and reliability. A cross-layer collaborative transmission optimization strategy enables resource complementarity and collaboration between different monitoring layers through information sharing and collaborative decision-making. This facilitates rapid response in emergency situations, ensuring the timely transmission of critical data and the stable operation of the wind farm. A transmission strategy effectiveness evaluation mechanism regularly assesses performance indicators and feeds the results back into the intelligent scheduling algorithm. This feedback mechanism enables the system to adaptively adjust transmission strategies to respond to changing network environments and monitoring needs. Through continuous optimization and improvement, the system maintains efficient and stable data transmission capabilities. The entire technical solution implements intelligent data transmission management, automating and intelligently processing every aspect, from data prioritization, network load monitoring, intelligent scheduling, to cross-layer collaborative transmission optimization. This not only reduces the burden of manual intervention but also improves system reliability and stability. Through intelligent scheduling and cross-layer collaborative transmission optimization strategies, the system efficiently utilizes network resources and avoids unnecessary waste. This helps reduce operation and maintenance costs and improves the economic benefits of wind farms. Through real-time monitoring and data analysis, the system can promptly identify potential network failures and data transmission issues and take appropriate preventive measures. This helps reduce the risk of failures and lowers the losses and repair costs caused by such failures.

[0097] In one embodiment of the present invention, the S2 includes:

[0098] S21 uses high-resolution infrared thermal imaging cameras and visible light cameras to perform all-weather, multispectral imaging. Based on AI image recognition algorithms, it processes images taken by drones in real time to automatically track and lock suspected damaged areas on the tower surface.

[0099] S22. Build a deep neural network model and combine it with CNN and GAN to perform fine image segmentation and feature extraction. Apply transfer learning algorithms and use pre-trained models to accelerate the training process of the deep neural network model.

[0100] S23: Based on multi-angle images taken by drones, 3D reconstruction is performed using SLAM (Simultaneous Localization and Mapping), and the reconstructed 3D model is refined, including surface smoothing and texture mapping.

[0101] S24. Combined with finite element analysis and material mechanics theory, the damaged area in the three-dimensional model is quantitatively evaluated.

[0102] In one embodiment of the present invention, S3 includes:

[0103] S31. Based on the ocean environment in which the tower is located, including current velocity, water depth, seabed topography, etc., set the initial ocean environment parameters of the model, and based on the interaction between the tower and surrounding structures (such as other towers, submarine cables, and offshore wind farm infrastructure), set boundary conditions; such as interaction forces between structures and distance restrictions.

[0104] S32. Conduct sensitivity tests on initial and boundary conditions through uncertainty analysis, and evaluate the impact of different parameter changes on model results. Build structural mechanics, fluid mechanics, and thermodynamics models, and simulate the tower's structural response, hydrodynamic characteristics, and heat conduction processes.

[0105] S33. Through coupling algorithms, data exchange and mutual influence between structural mechanics, fluid mechanics, and thermodynamics models are achieved. Nonlinear material models are introduced to evaluate the nonlinear behavior of materials in extreme environments (such as plastic deformation and fatigue cumulative damage). The models are verified and calibrated, and model parameters and algorithms are adjusted by comparing historical monitoring data, experimental data, and theoretical analysis results.

[0106] S34. Use a multi-operating-condition simulation scheme, including normal operating conditions, design operating conditions, and extreme operating conditions (such as typhoons and tsunamis), to simulate the dynamic response of the tower under different operating conditions. Analyze the dynamic response data based on the time series analysis method, and extract the changing trends of key characteristic parameters (such as displacement, stress, vibration frequency, etc.);

[0107] S35. By calculating the tower's dynamic parameters, including natural frequency and damping ratio, the tower's stability under different operating conditions is evaluated. Failure Mode and Effects Analysis (FMEA) is then introduced to predict possible failure modes and assess their impact on the tower's overall performance and safety.

[0108] S36. Construct extreme event simulation scenarios, such as extreme storms, tsunamis, and other natural disasters. Utilize historical data and numerical simulation methods to generate input parameters under extreme environmental conditions, simulate dynamic responses under extreme events, and evaluate the load-bearing capacity and safety performance of the tower under extreme conditions.

[0109] S37. Conduct risk assessments, combine the results of the failure mode and effects analysis, and assess the extent and probability of damage to the tower caused by extreme events. Based on the risk assessment results, formulate appropriate emergency response plans and recovery strategies.

[0110] S38. Collect the differences between the model simulation results and the actual monitoring data, analyze the causes of the differences, optimize and adjust the model based on the analysis results, and perform iterative simulation.

[0111] The working principle of this technical solution is as follows: High-resolution infrared thermal imaging cameras and visible light cameras perform all-weather, multispectral imaging to obtain detailed image information of the tower surface. Using AI image recognition algorithms, drone-captured images are processed in real time to automatically identify and track suspected damaged areas on the tower surface. A deep neural network model is constructed, combining CNN (convolutional neural network) and GAN (generative adversarial network) to perform fine image segmentation and extract the precise contours and features of damaged areas. A transfer learning algorithm is applied, leveraging pre-trained models to accelerate the training of the deep neural network model, improving recognition efficiency and accuracy. Based on multi-angle images captured by the drone, SLAM (Simultaneous Localization and Mapping) technology is used to perform 3D reconstruction to generate a 3D model of the tower. The reconstructed 3D model undergoes refined processing, including surface smoothing and texture mapping, to enhance model fidelity and analytical accuracy. Finite element analysis and material mechanics theory are combined to quantitatively assess the damaged areas in the 3D model, including damage size, depth, and impact area. Initial and boundary conditions were set based on the tower's marine environment (such as current velocity, water depth, and seabed topography) and the interaction with surrounding structures (such as other towers and submarine cables). Structural, fluid, and thermodynamic models were developed to simulate the tower's structural response, hydrodynamic characteristics, and heat conduction processes. Sensitivity tests were conducted on the initial and boundary conditions using uncertainty analysis to assess the impact of varying parameters on the model results. A coupling algorithm enabled data exchange and interaction between the structural, fluid, and thermodynamic models, ensuring dynamic consistency across multiple physical fields. A nonlinear material model was introduced to assess the nonlinear behavior of materials under extreme conditions (such as plastic deformation and fatigue damage accumulation). The model was validated and calibrated, and model parameters and algorithms were adjusted by comparing historical monitoring data, experimental data, and theoretical analysis results. A multi-condition simulation scheme (normal, design, and extreme conditions) was used to simulate the tower's dynamic response under different operating conditions. The dynamic response data was analyzed using time series analysis to extract trends in key characteristic parameters (such as displacement, stress, and vibration frequency). Calculate the dynamic parameters of the tower (such as natural frequency, damping ratio, etc.) and evaluate its stability under different working conditions; introduce failure mode and effects analysis (FMEA) to predict possible failure modes and their impact on the overall performance and safety of the tower. Construct extreme event simulation scenarios (such as extreme storms, tsunamis, etc.) and use historical data and numerical simulation methods to generate input parameters under extreme environmental conditions. Conduct dynamic response simulations under extreme events to evaluate the load-bearing capacity and safety performance of the tower in extreme environments. Implement risk assessment and, combined with the results of failure mode and effects analysis, evaluate the extent and probability of damage to the tower caused by extreme events. Develop corresponding emergency response plans and recovery strategies based on the risk assessment results.Collect discrepancies between model simulation results and actual monitoring data and analyze the causes of the discrepancies. Based on the analysis results, optimize the model (such as adjusting model parameters, improving coupling algorithms, introducing new physical models, etc.), and perform iterative simulations to improve the accuracy and reliability of the model.

[0112] The effect of the above technical solution is: through high-resolution infrared thermal imaging cameras and visible light cameras, combined with AI image recognition algorithms, it is possible to automatically track and lock suspected damaged areas on the tower surface. This method not only improves the efficiency and accuracy of detection, but also enables all-weather, multi-spectral monitoring to ensure timely detection of potential problems. Furthermore, through the combination of deep neural network models, CNN and GAN, fine segmentation and feature extraction of images are achieved, making the assessment of damaged areas more accurate, including quantitative assessment of damage size, depth and impact range, providing strong support for subsequent maintenance decisions; this technical solution takes into account the complex marine environment in which the tower is located and its interaction with surrounding structures. By setting reasonable initial conditions and boundary conditions, structural mechanics, fluid mechanics and thermodynamics models are established, and data exchange and mutual influence between these models are realized. This multi-physics coupled simulation approach more accurately reflects the dynamic response and performance changes of towers in actual operation, particularly by assessing the nonlinear behavior of materials under extreme conditions, thereby improving the accuracy and comprehensiveness of the simulation. Through a multi-condition simulation scheme encompassing normal, design, and extreme conditions, the dynamic response of the tower under various conditions is comprehensively simulated. Trends in key characteristic parameters are extracted using time series analysis. This approach facilitates a deeper understanding of the tower's behavior under various conditions, providing a scientific basis for tower design and maintenance. Furthermore, by constructing extreme event simulation scenarios, the tower's load-bearing capacity and safety performance under extreme conditions are evaluated. Risk assessments are then conducted, and corresponding emergency response plans and recovery strategies are developed, enhancing the safety and reliability of offshore wind farms. This technical approach also emphasizes continuous model optimization and iteration. By collecting discrepancies between model simulation results and actual monitoring data, the causes of the discrepancies are analyzed, and the model is optimized and adjusted based on the results. This iterative simulation approach enables the model to continuously adapt to changing conditions, improving its accuracy and reliability.

[0113] In one embodiment of the present invention, the step S32 includes:

[0114] Identify sources of uncertainty that affect tower performance assessment, including ocean environmental parameters (such as current velocity, water depth, and variability in seabed topography), uncertainty in structural geometry and material properties, and boundary conditions (such as ambiguity in inter-structural interaction forces and distance constraints). Quantify identified uncertainty parameters using statistical methods and probability distribution functions, and establish uncertainty ranges or probability distribution models.

[0115] Based on the sensitivity testing scheme, by changing the values ​​of single or multiple uncertainty parameters, the changes in model output (such as structural response, hydrodynamic characteristics, and heat conduction process) are obtained. By using sensitivity analysis tools (such as the Sobol index and FAST method), the contribution of different parameter changes to the model results is quantified and key sensitive parameters are identified.

[0116] Based on the sensitivity test results, the impact of uncertainty on tower performance evaluation was evaluated, and structural mechanics models, fluid mechanics models, and thermodynamics models were constructed respectively;

[0117] Structural mechanics model: Considering the geometric nonlinearity, material nonlinearity and inter-structural interactions of the tower, the static and dynamic responses of the tower are simulated.

[0118] Fluid mechanics model: Computational fluid dynamics (CFD) methods are used to simulate the effects of ocean currents, waves and other fluids on the tower, including hydrodynamic loads and scouring effects.

[0119] Thermodynamic model: Considering factors such as temperature fluctuations in the marine environment and seawater corrosion, it simulates processes such as heat conduction and thermal stress in the tower.

[0120] Collect historical monitoring data, experimental data and theoretical analysis results as benchmark data for model verification, compare and analyze the model simulation results with the benchmark data, evaluate the accuracy and reliability of the model, and adjust and optimize the model parameters and algorithms based on the evaluation results.

[0121] The working principle of the above technical solution is as follows: First, a comprehensive identification of sources of uncertainty affecting tower performance assessment is conducted. These sources of uncertainty include, but are not limited to, variability in marine environmental parameters (such as current velocity, water depth, and changes in seabed topography), uncertainty in structural geometry and material properties (such as manufacturing tolerances and material aging), and ambiguity in boundary conditions (such as the difficulty in accurately quantifying inter-structural interaction forces and estimation errors in distance constraints). Next, the identified uncertainty parameters are quantified using statistical methods and probability distribution functions. This involves establishing uncertainty ranges or probability distribution models to describe the variation patterns and likelihood of these parameters in real-world environments. This step provides the foundational data for subsequent sensitivity testing, enabling a systematic analysis of the impact of uncertainty on model results. Based on the sensitivity testing scheme, by varying the values ​​of one or more uncertainty parameters, the changes in model outputs (such as structural response, hydrodynamic characteristics, and heat conduction processes) are observed and recorded. This step aims to reveal the contribution and impact of different parameter changes on the model results. To quantify this impact, sensitivity analysis tools (such as the Sobol index and the FAST method) are used to conduct in-depth analysis of the sensitivity test results. These tools calculate the global and local sensitivity of different parameters to model outputs, thereby identifying key sensitive parameters. These key sensitive parameters are important focus points for subsequent model optimization and risk assessment. Based on sensitivity test results and uncertainty analysis, structural mechanics models, fluid mechanics models, and thermodynamics models are constructed. These models consider various nonlinear factors (such as geometric nonlinearity and material nonlinearity) in actual tower operation, as well as the influence of environmental factors (such as ocean currents, waves, and temperature fluctuations).

[0122] Structural mechanics model: simulates the tower's response behavior under static and dynamic loads, including changes in parameters such as displacement, stress, and vibration.

[0123] Fluid dynamics model: Computational fluid dynamics (CFD) methods are used to simulate the forces acting on the tower due to fluids such as ocean currents and waves, and to evaluate the impact of hydrodynamic loads and scouring effects on tower performance.

[0124] Thermodynamic model: Considering the impact of factors such as temperature fluctuations and seawater corrosion in the marine environment on the performance of the tower material, the heat conduction and thermal stress process of the tower are simulated.

[0125] To verify the accuracy and reliability of these models, historical monitoring data, experimental data, and theoretical analysis results were collected as baseline data. Model simulation results were compared with the baseline data to evaluate the models' predictive capabilities and consistency. Based on these evaluation results, model parameters and algorithms were adjusted and optimized to improve the models' accuracy and applicability.

[0126] The entire workflow is an iterative optimization process. By continuously collecting new monitoring data and experimental results, the model is verified and calibrated, deficiencies and deviations are identified, and the model is refined and optimized accordingly. This continuous improvement mechanism helps ensure that the model can continuously adapt to changes in the actual environment and evolving trends in tower performance, providing strong support for health monitoring and performance evaluation of offshore wind turbine towers.

[0127] The above technical solution achieves the following: by identifying and quantifying the sources of uncertainty affecting tower performance evaluation, an uncertainty range or probability distribution model is established. This helps more accurately describe parameter variations in real-world environments and improves the accuracy of model evaluation. Through sensitivity testing and analysis, key sensitive parameters with the greatest impact on model results are identified. This helps focus on these key parameters during model construction and optimization, thereby improving the overall accuracy of the model. Separate structural mechanics, fluid mechanics, and thermodynamics models are constructed, and their interactions are considered. This multi-physics coupled modeling approach can more comprehensively reflect the complex behavior of towers in actual operation, enhancing the robustness and adaptability of the model. The impact of uncertainty factors is fully considered during model construction, ensuring that the model maintains good predictive power and stability in the face of uncertainties in real-world environments. The key sensitive parameters identified through sensitivity analysis provide important insights for tower design optimization and operation and maintenance decisions. For example, these key parameters can be more strictly controlled or monitored to ensure tower safety and reliability. The constructed model can accurately evaluate tower performance under different operating conditions, including structural response, hydrodynamic characteristics, and heat conduction processes. This helps identify potential problems in advance, develop appropriate countermeasures, and improve the overall operational efficiency of offshore wind farms. This technical solution combines statistical methods, probability distribution functions, sensitivity analysis tools, and other techniques to provide new insights and methods for evaluating the performance of offshore wind turbine towers. This technological innovation helps promote research and development in related fields. By collecting historical monitoring data, experimental data, and theoretical analysis results as benchmark data for model validation, data-driven model optimization and adjustment is achieved. This data-driven approach helps continuously improve the accuracy and reliability of the model, driving continuous technological advancement.

[0128] In one embodiment of the present invention, the step S33 includes:

[0129] According to the specific needs and computing resources of the tower simulation, the most appropriate coupling strategy (such as partitioned coupling, direct coupling, iterative coupling, etc.) is selected, and customized optimization is performed for the selected coupling algorithm;

[0130] Nonlinear material models, such as elastic-plastic models and damage accumulation models, are introduced to describe the complex mechanical behavior of tower materials in extreme environments, and the nonlinear material models are integrated into the multi-physics field coupling model to ensure that the mechanical properties of the material can be updated in real time during the simulation process to reflect the nonlinear changes of the material.

[0131] Define data interfaces and exchange protocols between multiple physical fields to ensure smooth data transfer and sharing between structural mechanics, fluid mechanics, and thermodynamics models. By coupling the structural mechanics model with the fluid mechanics model, simulate the effects of fluid on the tower structure; including the application of hydrodynamic loads and feedback of fluid-structure interactions.

[0132] The thermodynamic model is incorporated into the coupling system to conduct a thermal-mechanical coupling analysis based on the influence of ocean ambient temperature changes on the tower material properties and structural response.

[0133] Through verification cases, including theoretical verification, experimental verification and field monitoring data verification, the accuracy and reliability of the multi-physics field coupling model are evaluated, and the model parameters, algorithms and coupling strategies are adjusted based on the evaluation results.

[0134] The working principle of the above technical solution is to select the most appropriate coupling strategy based on the specific needs of the tower simulation and the available computing resources. These strategies may include partitioned coupling (solving different physical fields separately and then exchanging data through interfaces), direct coupling (solving all physical fields simultaneously in the same solver), and iterative coupling (gradually approximating the interactions between multiple physical fields through iterations). After selecting the coupling algorithm, customized optimization is performed to ensure that it can efficiently and accurately handle the complex problems in tower simulation. To more accurately describe the complex mechanical behavior of tower materials under extreme environments, nonlinear material models such as elastic-plastic models and damage accumulation models are introduced. These models can account for changes in the material's stress-strain relationship with factors such as loading history and temperature. Integrating nonlinear material models into the multiphysics coupling model ensures that the material's mechanical properties can be updated in real time during the simulation to reflect the material's nonlinear changes. Clearly defined data interfaces and exchange protocols are key to ensuring smooth data transfer and sharing between structural mechanics, fluid mechanics, and thermodynamics models. By defining standardized data formats, exchange timing, and synchronization mechanisms, accurate and timely data transmission between different physical fields is ensured, thus achieving tight multi-physics coupling. By coupling the structural mechanics model with the fluid mechanics model, the effects of fluids (such as currents and waves) on the tower structure are simulated. This includes the application of hydrodynamic loads and feedback from fluid-structure interactions. Fluid impact forces and friction on the structure are transmitted to the structural mechanics model, affecting the structural response. Simultaneously, structural deformation and displacement are fed back into the fluid mechanics model, influencing the fluid flow state. A thermodynamic model is incorporated into the coupling system to account for the impact of ocean ambient temperature changes on the tower's material properties and structural response. Thermal-mechanical coupling analysis is used to assess the impact of temperature-induced material property changes (such as thermal expansion and thermal stress) on the overall performance of the tower structure. The accuracy and reliability of the multi-physics coupling model are comprehensively evaluated through validation cases, including theoretical verification (comparison with existing theoretical or analytical solutions), experimental verification (comparison with experimental data under laboratory conditions), and field monitoring data verification (comparison with actual offshore wind farm operating data). Based on the evaluation results, the model parameters, algorithms and coupling strategies are adjusted and optimized to improve the predictive ability and applicability of the model.

[0135] The effects of the above technical solution are: selecting and optimizing the most appropriate coupling strategy based on the specific needs and computing resources of the tower simulation, ensuring that the simulation process achieves optimal accuracy while effectively utilizing resources; introducing nonlinear material models, such as elastoplastic models and damage accumulation models, can more accurately describe the complex mechanical behavior of tower materials in extreme environments. This precise material model can significantly improve the accuracy of simulation results, especially when considering material nonlinearity and damage accumulation effects; through the definition of clearly defined data interfaces and exchange protocols, smooth coupling between structural mechanics, fluid mechanics, and thermodynamic models is achieved, which can simulate the complex multi-physics interactions of the tower in actual operation. This comprehensive coupling model enhances the model's adaptability and robustness to actual operating conditions; the mechanical properties of the material are updated in real time during the simulation to reflect the nonlinear changes of the material, allowing the model to more accurately predict the performance of the tower under different operating conditions; incorporating the thermodynamic model into the coupling system and conducting thermal-mechanical coupling analysis can reveal the profound impact of changes in the ocean environment temperature on the tower material properties and structural response. This helps to consider the impact of temperature factors during the design and operation and maintenance stages, thereby formulating more scientific and reasonable solutions. The accuracy and reliability of the multi-physics coupling model are comprehensively evaluated through various methods such as theoretical verification, experimental verification, and field monitoring data verification. Adjusting the model parameters, algorithms, and coupling strategies based on the evaluation results can continuously optimize the model performance and improve its practical application value in design and operation and maintenance. This technical solution combines multiple advanced technical means such as multi-physics coupling, nonlinear material models, and real-time data exchange to provide new ideas and methods for the simulation and analysis of offshore wind turbine towers. This technological innovation helps to promote research and development in related fields. The optimized multi-physics coupling model is not only suitable for the simulation and analysis of offshore wind turbine towers, but can also be extended to the simulation of other similar complex structures. Its broad application prospects will provide strong support for technological progress and industrial upgrading in related industries.

[0136] In one embodiment of the present invention, the step S36 includes:

[0137] Obtain historical data on extreme climate events, such as extreme storms and tsunamis, analyze their frequency, intensity, duration, and spatial distribution characteristics, and predict extreme event scenarios based on historical data and combined with numerical simulations (such as climate models and ocean models); including but not limited to extreme values ​​of parameters such as wind speed, wave height, water depth, and seabed topography changes; consider the possibility of different extreme event combinations, and construct a variety of extreme event simulation scenarios; in order to comprehensively evaluate the performance of the tower in complex extreme environments.

[0138] Based on the constructed extreme event scenarios, numerical simulation methods are used to generate corresponding environmental input parameters, such as wind speed time series, wave spectrum, ocean current velocity field, etc.; the generated input parameters are verified to ensure that they conform to the physical characteristics and statistical laws of extreme events, while considering the rationality of the parameters and computational feasibility.

[0139] The verified input parameters are fed into the multi-physics coupling model to simulate the dynamic response under extreme events. The changes in key performance parameters of the tower under extreme conditions, such as maximum stress, displacement, vibration frequency, and fatigue cumulative damage, are recorded to analyze the dynamic response characteristics and ultimate load-bearing capacity.

[0140] Based on the dynamic response simulation results, the safety and reliability of the tower in extreme environments are evaluated, including structural integrity, stability, durability and other aspects. Combined with Failure Mode and Effects Analysis (FMEA), the failure modes of the tower under extreme events are identified, and the degree of impact on overall performance and safety is evaluated.

[0141] The working principle of the above technical solution is as follows: First, historical data on extreme climate events (such as extreme storms and tsunamis) is collected, including their frequency, intensity, duration, and spatial distribution. This data is an important foundation for understanding and predicting future extreme events. Based on this historical data, numerical simulation tools such as climate models and ocean models are combined to predict extreme event scenarios. These predictions cover the extreme values ​​of key parameters such as wind speed, wave height, water depth, and seabed topography, as well as their temporal and spatial variations. Diverse extreme event simulation scenarios are constructed, taking into account the potential for different extreme event combinations. These scenarios are designed to comprehensively cover the complex extreme environments that may be encountered, allowing for a comprehensive performance evaluation of the tower. Based on these constructed extreme event scenarios, numerical simulation methods are used to generate corresponding environmental input parameters, such as wind speed time series, wave spectra, and ocean current velocity fields. These parameters form the basis for simulating the dynamic response of the tower under extreme conditions. The generated input parameters are then verified to ensure that they conform to the physical characteristics and statistical laws of extreme events. Furthermore, the rationality and computational feasibility of the parameters must be considered to ensure the accuracy and reliability of the simulation results. The verified input parameters are then input into the multi-physics coupling model. This model integrates multiple physics fields, including structural mechanics, fluid mechanics, and thermodynamics, to simulate the dynamic response of a tower in complex environments. It also simulates the tower's dynamic response under extreme event scenarios. It records changes in key tower performance parameters, such as maximum stress, displacement, vibration frequency, and fatigue damage accumulation, under extreme conditions. These changes reflect the tower's operating state and performance under extreme conditions. Dynamic response characteristics and ultimate load-bearing capacity are analyzed. The tower's stability and durability under extreme conditions are assessed, and whether it meets design requirements. Based on the dynamic response simulation results, the tower's safety under extreme conditions is evaluated. This includes assessments of structural integrity (e.g., cracks and fractures) and stability (e.g., overturning and slippage). Failure Mode and Effects Analysis (FMEA) is used to identify potential tower failure modes under extreme events. The impact of these failure modes on overall performance and safety, as well as the potential consequences, is assessed. Based on the evaluation results, optimization recommendations for tower design, operation, and maintenance are proposed. This improves the tower's safety and reliability under extreme conditions, ensuring the safe and stable operation of offshore wind farms.

[0142] The above technical solution achieves the following: By considering the possible combinations of different extreme events, a diverse range of extreme event simulation scenarios are constructed, fully covering the complex extreme environments that may be encountered. This helps to more accurately assess the performance of towers under extreme conditions, providing a comprehensive risk assessment basis for wind farm design, construction, and operation and maintenance. Predictions based on historical data and numerical simulations can identify potential extreme event risks in advance, enabling the development of effective preventive and response measures to reduce the damage caused by extreme events to wind farms. Numerical simulation methods are used to generate environmental input parameters that conform to the physical characteristics and statistical laws of extreme events, ensuring the accuracy and reliability of simulation results. These parameters are then validated and input into a multi-physics coupling model, which more realistically reflects the tower's operating conditions under extreme environments. The dynamic response simulation using the multi-physics coupling model comprehensively considers the interactions between multiple physical fields, such as structural mechanics, fluid mechanics, and thermodynamics, improving the accuracy and comprehensiveness of the simulation. By recording the changes in key tower performance parameters under extreme conditions, such as maximum stress, displacement, vibration frequency, and fatigue cumulative damage, an in-depth analysis of the tower's dynamic response characteristics and ultimate load-bearing capacity is possible. This helps to discover potential design flaws or performance bottlenecks, providing strong support for design optimization. Combined with Failure Mode and Effects Analysis (FMEA), it can identify possible tower failure modes under extreme events and assess their impact on overall performance and safety. This helps to formulate targeted improvement measures to improve the reliability and durability of the tower. This technical solution integrates a variety of technical means such as extreme climate event analysis, numerical simulation, and multi-physics field coupling models, promoting technological innovation and development in related fields. Through comprehensive performance evaluation and optimization suggestions, it can provide important reference for the formulation of industry standards and promote the standardization and standardization of offshore wind power technology. By preventing and reducing the damage to wind farms caused by extreme events in advance, operation and maintenance costs can be reduced and the economic benefits of wind farms can be improved. Through comprehensive performance evaluation and safety and reliability analysis, the public's trust and support for offshore wind power technology can be enhanced, promoting its wider application and development.

[0143] In one embodiment of the present invention, the S4 includes:

[0144] S41. Design the platform architecture, including a distributed architecture, using a microservices architecture model, with each microservice responsible for specific data processing tasks, such as data collection, preprocessing, storage, and analysis. Design a multi-level data security strategy, including data encryption, access control, and audit logs, to ensure data security during transmission, storage, and processing. Define unified data interface standards, such as RESTful APIs and message queues, conduct quality assessments on data sources, and filter or modify those that do not meet requirements.

[0145] S42. Dynamically adjust data collection and transmission strategies based on data source priority, real-time requirements, and data volume through an intelligent scheduling mechanism;

[0146] S43, preprocessing the data, and re-evaluating the quality of the pre-processed data based on the data quality assessment model, and performing secondary filtering and correction on the data based on the assessment results;

[0147] S44. Define the level and granularity of data fusion, including feature-level fusion and decision-level fusion. Select fusion strategies based on specific application scenarios. Implement temporal synchronization and spatial alignment of multi-source data. Introduce graph neural networks to process data relationships and network structures.

[0148] S45. Based on the fusion result verification mechanism, the fusion effect is evaluated by comparing the data quality before and after fusion and the analysis results, and the fusion strategy is iteratively optimized;

[0149] S46. Through distributed data storage strategies, adopt big data processing frameworks such as Hadoop and Spark to distribute and efficiently access massive data, and set up and implement data indexing and partitioning strategies.

[0150] The working principle of the above technical solution is as follows: By designing a distributed architecture and adopting a microservices architecture model, the system is divided into multiple independent service units, each dedicated to completing a specific data processing task (such as data collection, preprocessing, storage, and analysis). This design approach improves the system's maintainability, scalability, and high availability, as each service can be independently deployed, updated, and expanded without affecting the entire system. The introduction of containerization technologies such as Docker and container orchestration tools such as Kubernetes enables automated service deployment, management, and expansion. Containerization packages applications and their dependencies into independent units, simplifying application deployment and maintenance. Kubernetes provides container orchestration capabilities, automatically managing the container lifecycle, including deployment, scheduling, scaling, and fault recovery. A multi-level data security strategy, including data encryption, access control, and audit logging, ensures data security during transmission, storage, and processing. Furthermore, unified data interface standards (such as RESTful APIs and message queues) are defined to facilitate data exchange and integration between different services. In addition, data source quality is assessed to ensure data integrity, accuracy, and timeliness, and non-compliant data sources are filtered or corrected. An intelligent scheduling mechanism dynamically adjusts data collection and transmission strategies based on data source priority, real-time requirements, and data volume. This helps optimize resource utilization and ensures that critical data is collected and transmitted to the processing system in a timely and accurate manner. The collected data is preprocessed, including steps such as data cleaning and format conversion, to improve data quality. The preprocessed data is then re-evaluated based on a data quality assessment model. Based on the assessment results, the data is filtered and corrected again to further improve its accuracy and reliability. The data fusion hierarchy and granularity (such as feature-level fusion and decision-level fusion) are defined, and appropriate fusion strategies are selected based on specific application scenarios. Multi-source data is synchronized temporally and spatially aligned to ensure data consistency and comparability. Technologies such as graph neural networks are introduced to process data relationships and network structures to uncover deep connections and underlying patterns between data. A fusion result verification mechanism evaluates the fusion effectiveness by comparing the data quality before and after fusion and analyzing the results. Based on the evaluation results, the fusion strategy is iteratively optimized to improve fusion effectiveness and data application value. A distributed data storage strategy is implemented, employing big data processing frameworks such as Hadoop and Spark for distributed storage and efficient access to massive amounts of data. Data indexing and partitioning strategies are implemented to improve data retrieval and processing efficiency. This design approach supports large-scale data processing needs and ensures efficient data storage and fast access.

[0151] The above technical solution significantly improves system availability and scalability by splitting the system into multiple microservices, each of which can be independently deployed and scaled. If a service fails, other services continue to operate without impacting the stability of the entire system. Furthermore, as business volume grows, new service instances can be easily added to expand system processing capacity. The introduction of containerization technologies such as Docker and orchestration tools such as Kubernetes enables automated service deployment, management, and scaling. This significantly reduces operational costs, improves deployment efficiency, and reduces human error. Containerization also provides environmental consistency, ensuring service stability and reliability across diverse environments. A multi-layered data security strategy, including data encryption, access control, and audit logs, ensures data security during transmission, storage, and processing. This helps prevent data leaks and unauthorized access, protecting user privacy and enterprise data security. Data source quality assessment, including checks for completeness, accuracy, and timeliness, ensures the quality of input data. This helps reduce errors and biases caused by data quality issues and improves the accuracy and reliability of data analysis. An intelligent scheduling mechanism dynamically adjusts data collection and transmission strategies based on data source priority, real-time requirements, and data volume. This helps optimize resource utilization, improve the efficiency of data collection and transmission, and ensure timely processing of critical data. Data is preprocessed and re-evaluated for quality, and secondary filtering and correction are performed based on the evaluation results. This helps improve the accuracy and reliability of data processing and reduce errors and biases in subsequent analysis. The hierarchy and granularity of data fusion, including feature-level fusion and decision-level fusion, are defined, and appropriate fusion strategies are selected based on specific application scenarios. This helps fully explore the potential connections and patterns between multi-source data, enhancing the depth and breadth of data analysis. Technologies such as graph neural networks are introduced to process data relationships and network structures, further improving the effectiveness and accuracy of data fusion. Big data processing frameworks such as Hadoop and Spark are used for distributed storage and efficient access of massive amounts of data. This helps address the challenges of massive data storage and processing, improving data processing efficiency and performance. Implementing data indexing and partitioning strategies further improves the efficiency of data retrieval and processing. This helps accelerate data analysis and meet the needs of application scenarios with high real-time requirements.

[0152] In one embodiment of the present invention, the S5 includes:

[0153] S51. Continuously monitor the integrated monitoring data set through a real-time monitoring engine and identify potential faults in real time based on machine learning models.

[0154] S52. Diagnose identified potential faults, analyze the fault causes, impact scope, and potential consequences, and generate intelligent maintenance decision recommendations based on the fault diagnosis results, including maintenance plan formulation, maintenance resource scheduling, and maintenance time window arrangement;

[0155] S53. Setting different levels of warning thresholds based on the severity of the fault, such as yellow warning, orange warning, and red warning, and dynamically adjusting the warning thresholds based on historical data and current monitoring results based on an adaptive warning mechanism;

[0156] S54, and issue early warning information to operation and maintenance personnel through multiple channels, including sound and light alarms, SMS notifications, email reminders and remote monitoring system push.

[0157] The working principle of the above technical solution is as follows: the engine is responsible for continuously monitoring the integrated monitoring data set after fusion, which contains monitoring information from multiple data sources that has been processed and analyzed. Through real-time data stream processing technology, it ensures that the data can be processed and analyzed quickly; based on pre-trained machine learning models (such as classifiers, regression models or anomaly detection algorithms), the monitoring data is analyzed in real time to identify potential failure modes or abnormal behaviors. These models are trained with historical failure data and normal operation data, and can automatically learn and identify abnormal features in the data; when a potential fault is identified, the system further diagnoses the fault. This includes analyzing the cause of the fault (such as equipment aging, improper operation, external factors, etc.), the scope of impact (affected equipment, systems or business processes), and potential consequences (such as performance degradation, increased downtime, safety hazards, etc.); based on the results of the fault diagnosis, the system generates intelligent maintenance decision recommendations. These recommendations may include developing specific maintenance plans (such as replacing parts, adjusting parameters, and upgrading software), scheduling maintenance resources (such as maintenance personnel, spare parts, and tools), and arranging maintenance time windows (taking into account factors such as production plans, resource availability, and the urgency of the fault). Depending on the severity of the fault, the system presets different levels of warning thresholds (such as yellow, orange, and red). These thresholds represent different levels of fault risk and are used to trigger corresponding warning responses. Based on the adaptive warning mechanism, the system dynamically adjusts the warning thresholds based on historical data and current monitoring results. This means that as the system operating conditions change (such as equipment aging, environmental changes, changes in production load, etc.), the warning threshold will also be adjusted accordingly to maintain the sensitivity and accuracy of the warning system; the system issues warning information to operation and maintenance personnel through multiple channels, including sound and light alarms (emitting sound and light signals at the monitoring center or equipment site), SMS notifications (sending warning information to the operation and maintenance personnel's mobile phones), email reminders (sending warning information to the operation and maintenance personnel's email address) and remote monitoring system push (displaying warning information on the remote monitoring platform and sending notifications); warning information usually contains key information such as fault type, location, severity, scope of impact, recommended response measures and emergency contact information, so that operation and maintenance personnel can quickly understand the fault situation and take appropriate measures.

[0158] The effects of the above technical solutions are: continuous monitoring of the integrated monitoring data set after integration through the real-time monitoring engine to ensure that the system can respond quickly to any potential problems; using machine learning models to analyze monitoring data in real time, it can automatically learn and identify abnormal patterns, so as to issue early warnings before failures occur, greatly improving the ability to prevent failures; in-depth fault diagnosis of identified potential faults, clarifying the cause of the fault, the scope of impact and potential consequences, helps operation and maintenance personnel better understand the problem and develop more effective solutions; based on the fault diagnosis results, automatically generate intelligent maintenance decision suggestions, including maintenance plans, resource scheduling and time window arrangements, etc., reducing the subjectivity and uncertainty of human decision-making and improving maintenance efficiency and quality; setting different levels of warning thresholds (such as yellow, orange, and red) according to the severity of the fault, so that operation and maintenance personnel can Staff can quickly judge the urgency of the fault according to the warning level; through the adaptive warning mechanism, the warning threshold is dynamically adjusted according to historical data and current monitoring results to ensure the accuracy and sensitivity of the warning system and avoid false alarms and missed alarms; warning information is released through multiple channels such as sound and light alarms, SMS notifications, email reminders and remote monitoring system push, ensuring that operation and maintenance personnel can receive warning information in a timely manner no matter where they are; the combined use of multiple channels improves the reach and timeliness of warning information, which helps operation and maintenance personnel respond quickly and take corresponding measures; through real-time fault identification, accurate fault diagnosis and intelligent decision support, operation and maintenance personnel can locate problems more quickly and take effective measures, reducing fault handling time; early warning and timely response help reduce the impact of faults on production operations, reduce downtime and economic losses, and improve system reliability and stability.

[0159] One embodiment of the present invention provides an offshore wind turbine tower health comprehensive monitoring system, comprising a memory, a processor, and a computer program stored in and executable on the memory, wherein the processor executes the program to implement any of the above-described offshore wind turbine tower health comprehensive monitoring methods.

[0160] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A comprehensive monitoring method for offshore wind turbine tower health, characterized in that: The method comprises: S1. Collect full-range data from the tower based on a multi-sensor network and transmit the collected full-range data to the edge computing node for preliminary data preprocessing. S2. Using machine vision algorithms, high-definition cameras and drone inspections, we monitor the tower surface and surrounding environment in real time. Using convolutional neural networks and generative adversarial networks, we perform intelligent image analysis, 3D reconstruction, and damage assessment. S3. Based on finite element analysis and computational fluid dynamics, a complex physical model of the offshore wind turbine tower is constructed to simulate its dynamic response under the influence of wind, waves and currents. S4. Build a big data fusion platform to deeply integrate sensor data, machine vision data, environmental parameter data, and physical model simulation data to form a comprehensive monitoring data set; S5. Use machine learning algorithms to conduct real-time analysis of comprehensive monitoring data sets, automatically identify potential tower failures and damage, set multi-level warning thresholds based on the diagnosis results, and issue warning information to operation and maintenance personnel through multiple channels; Said S1 comprises: S11, layering the sensors and adjusting the sampling rate and accuracy; S12, setting up redundancy configuration and fault tolerance mechanism for the sensor network; S13, transmitting data to the edge computing node; S14. Dynamically adjust the data transmission strategy based on data transmission priority and network load. The transmission strategy is calculated using the following formula: in, represents the transmission strategy of the i-th data packet after adjustment at time t, Indicates the current transmission strategy of the i-th data packet, represents the priority of the i-th data packet, Indicates the network load at time t; α and β represent the weight factors of priority and network load, and α + β = 1; S15. Process the received data, evaluate the data quality, and extract key feature parameters; Said S11 comprises: Analyze monitoring requirements based on the specific operating environment and operation and maintenance requirements of the wind farm and obtain demand analysis results; based on the demand analysis results, build a three-layer monitoring architecture; Through adaptive sensing algorithms, the changing trends of environmental parameters are analyzed in real time to predict the impact of environmental parameters on monitoring accuracy and frequency. Sensors dynamically adjust their working modes based on current task requirements and remaining power. Through the collaborative mechanism between sensor networks, sensors at each layer share and collaborate locally through low-power wireless communication protocols, making joint decisions on monitoring strategies and resource allocation.

2. The offshore wind turbine tower health comprehensive monitoring method according to claim 1, characterized in that: Said S1 comprises: S11. Adopt a layered deployment strategy, dividing sensors into basic monitoring layer, key monitoring layer and emergency monitoring layer, and automatically adjust the sampling rate and accuracy according to environmental conditions through adaptive sensor algorithms; S12, setting up redundancy configuration and fault tolerance mechanism for the sensor network; S13. Collect data from the entire tower, compress the collected data using a compression algorithm, and transmit the data to the edge computing node via a low-power wide area network. S14. During the transmission process, the data transmission strategy is dynamically adjusted according to the data transmission priority and network load through the intelligent scheduling algorithm; S15. The edge computing node uses a time series analysis algorithm to denoise, filter, and smooth the received data, evaluate the data quality, identify and process outliers and missing data, and extract key feature parameters.

3. The offshore wind turbine tower health comprehensive monitoring method according to claim 2, characterized in that: Said S14 comprises: Build a comprehensive evaluation system based on the urgency, importance, and real-time requirements of the data, and prioritize the data to be transmitted; During data transmission, the system continuously monitors the network load of the LPWAN, performs real-time analysis and prediction based on big data, and evaluates the current network status and future trends. According to the evaluation results, the data transmission strategy is dynamically adjusted based on the intelligent scheduling algorithm of multi-objective optimization; Through the transmission strategy effectiveness evaluation mechanism, performance indicators are regularly evaluated. By comparing and analyzing the performance indicators under different transmission strategies, the transmission strategy is identified and optimized. At the same time, a feedback mechanism is established to feed the evaluation results back to the intelligent scheduling algorithm to adaptively optimize and continuously improve the transmission strategy. Through cross-layer collaborative transmission optimization strategy, during the data transmission process, cross-layer information sharing and collaborative decision-making are carried out based on the transmission strategy of the current layer and combined with the transmission requirements and resource conditions of other layers.

4. The offshore wind turbine tower health comprehensive monitoring method according to claim 1, characterized in that: Said S2 comprises: S21 uses high-resolution infrared thermal imaging cameras and visible light cameras to perform all-weather, multispectral imaging. Based on AI image recognition algorithms, it processes images taken by drones in real time to automatically track and lock suspected damaged areas on the tower surface. S22. Build a deep neural network model and combine it with CNN and GAN to perform fine image segmentation and feature extraction. Apply transfer learning algorithms and use pre-trained models to accelerate the training process of the deep neural network model. S23, based on the multi-angle images taken by the UAV, SLAM is used to perform 3D reconstruction and process the reconstructed 3D model. S24. Combined with finite element analysis and material mechanics theory, the damaged area in the three-dimensional model is quantitatively evaluated.

5. The offshore wind turbine tower health comprehensive monitoring method according to claim 1, characterized in that: The S3 includes: S31. Based on the ocean environment in which the tower is located, set the initial ocean environment parameters of the model, and set the boundary conditions based on the interaction between the tower and the surrounding structures; S32. Conduct sensitivity tests on initial and boundary conditions through uncertainty analysis, and evaluate the impact of different parameter changes on model results. Build structural mechanics, fluid mechanics, and thermodynamics models, and simulate the tower's structural response, hydrodynamic characteristics, and heat conduction processes. S33. Through coupling algorithms, data exchange and mutual influence between structural mechanics, fluid mechanics, and thermodynamics models are achieved. Nonlinear material models are used to evaluate the nonlinear behavior of materials in extreme environments. Models are validated and calibrated, and model parameters and algorithms are adjusted by comparing historical monitoring data, experimental data, and theoretical analysis results. S34. Use a multi-operating-condition simulation scheme to simulate the dynamic response of the tower under different operating conditions. Analyze the dynamic response data based on the time series analysis method and extract the changing trends of key characteristic parameters. S35. By calculating the tower's dynamic parameters, the tower's stability under different operating conditions is evaluated. Failure mode and effects analysis is introduced to predict possible failure modes and assess their impact on the tower's overall performance and safety. S36. Construct extreme event simulation scenarios, using historical data and numerical simulation methods to generate input parameters under extreme environmental conditions, simulate dynamic responses under extreme events, and evaluate the load-bearing capacity and safety performance of the tower under extreme environments; S37. Conduct risk assessments, combine the results of the failure mode and effects analysis, and assess the extent and probability of damage to the tower caused by extreme events. Based on the risk assessment results, formulate appropriate emergency response plans and recovery strategies. S38. Collect the differences between the model simulation results and the actual monitoring data, analyze the causes of the differences, optimize and adjust the model based on the analysis results, and perform iterative simulation.

6. The method for comprehensive monitoring of offshore wind turbine tower health according to claim 5, characterized in that: The S32 includes: Identify the sources of uncertainty that affect tower performance evaluation, quantify the identified uncertainty parameters through statistical methods and probability distribution functions, and establish uncertainty ranges or probability distribution models; Based on the sensitivity testing scheme, by changing the values ​​of single or multiple uncertainty parameters, the changes in model output are obtained; and by using sensitivity analysis tools, the contribution of different parameter changes to model results is quantified, and key sensitive parameters are identified; Based on the sensitivity test results, the impact of uncertainty on tower performance evaluation was evaluated, and structural mechanics models, fluid mechanics models, and thermodynamics models were constructed respectively; Collect historical monitoring data, experimental data and theoretical analysis results as benchmark data for model verification, compare and analyze the model simulation results with the benchmark data, evaluate the accuracy and reliability of the model, and adjust and optimize the model parameters and algorithms based on the evaluation results.

7. The method for comprehensive monitoring of offshore wind turbine tower health according to claim 5, characterized in that: The S36 includes: Obtain historical data on extreme climate events, and based on this data and combined with numerical simulations, predict extreme event scenarios; and construct diverse extreme event simulation scenarios; Based on the constructed extreme event scenario, the corresponding environmental input parameters are generated using numerical simulation methods and the generated input parameters are verified; The verified input parameters are fed into the multi-physics coupling model to simulate the dynamic response under extreme events. The changes in the tower's key performance parameters under extreme conditions are recorded, and the dynamic response characteristics and ultimate load-bearing capacity are analyzed. Based on the dynamic response simulation results, the safety and reliability of the tower in extreme environments are evaluated. Combined with the failure mode and effect analysis, the failure modes of the tower under extreme events are identified, and the degree of impact on overall performance and safety is evaluated.

8. The offshore wind turbine tower health comprehensive monitoring method according to claim 1, characterized in that: Said S4 comprises: S41. Design the platform architecture and define unified data interface standards, conduct quality assessments on data sources, and filter or modify data sources that do not meet the requirements; S42. Dynamically adjust data collection and transmission strategies based on data source priority, real-time requirements, and data volume through an intelligent scheduling mechanism; S43, preprocessing the data, and re-evaluating the quality of the pre-processed data based on the data quality assessment model, and performing secondary filtering and correction on the data based on the assessment results; S44. Define the level and granularity of data fusion, select fusion strategies based on specific application scenarios, implement time synchronization and spatial alignment of multi-source data, introduce graph neural networks, and process data relationships and network structures; S45. Based on the fusion result verification mechanism, the fusion effect is evaluated by comparing the data quality before and after fusion and the analysis results, and the fusion strategy is iteratively optimized; S46. Through distributed data storage strategies, massive data can be distributedly stored and efficiently accessed, and data indexing and partitioning strategies can be set up and implemented.

9. The offshore wind turbine tower health comprehensive monitoring method according to claim 1, characterized in that: Said S5 comprises: S51. Continuously monitor the integrated monitoring data set through a real-time monitoring engine and identify potential faults in real time based on machine learning models. S52. Diagnose the identified potential faults, analyze the fault causes, impact scope, and potential consequences, and generate intelligent maintenance decision recommendations based on the fault diagnosis results; S53. Setting different levels of warning thresholds according to the severity of the fault, and dynamically adjusting the warning thresholds based on historical data and current monitoring results based on an adaptive warning mechanism; S54, and send early warning information to operation and maintenance personnel through multiple channels.

10. An offshore wind turbine tower health comprehensive monitoring system, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory and executable on the memory, wherein the processor executes the program to implement the comprehensive monitoring method for offshore wind turbine tower health as described in any one of claims 1 to 9.

Images