Fan blade clearance monitoring method and system

Abstract

The invention belongs to the technical field of fan operation and maintenance, and particularly discloses a fan blade clearance monitoring method and system, and the system comprises a composite sensing module, an attitude sensing module, a deformation detection module, a data processing module, a linkage control module, a data operation and maintenance module and a power supply module. The fan blade clearance monitoring system can overcome the defects that an existing fan blade clearance monitoring product is poor in extreme environment adaptability, lags in early warning, is low in data utilization rate, is insufficient in adaptability and is high in operation and maintenance cost, and can remarkably improve the adaptability of the fan blade clearance monitoring system to the extreme environment and the monitoring stability in the extreme environment. Dynamic risk pre-judgment is achieved, the problem of early warning lag is avoided, the system is high in multi-source data fusion efficiency, high in monitoring precision, high in adaptability and low in operation and maintenance cost, closed-loop operation and maintenance management is conveniently formed, the operation efficiency of the draught fan is improved, and safe operation of the draught fan is guaranteed.

Description

Technical Field

[0001] This invention belongs to the field of wind turbine operation and maintenance technology, specifically relating to a method and system for monitoring the air clearance of wind turbine blades. Background Technology

[0002] Wind turbine blade clearance monitoring is a core component in ensuring the safe and stable operation of wind turbine generators. Its purpose is to monitor the minimum distance between the blades and the tower in real time and accurately, providing timely warnings of collision risks and preventing "tower-sweeping" accidents, thereby reducing economic losses caused by blade damage, tower breakage, and downtime. Currently, wind turbine blade clearance monitoring technology and products have achieved a certain scale of application, but existing products still have the following shortcomings in practical applications: 1. Poor adaptability to harsh environments and unstable monitoring accuracy: Existing wind turbine blade clearance monitoring products mostly rely on lidar or visual cameras for distance measurement. In harsh environments such as sandstorms, rain, snow, strong winds, and high salt spray, they are easily blocked, interfered with by signal reflection, or corroded, which leads to increased measurement errors, and even data loss, false alarms, or missed alarms, resulting in low monitoring reliability.

[0003] 2. Lack of dynamic risk prediction capability and delayed early warning: Existing technologies mainly focus on "real-time distance measurement," which can only trigger early warnings when the clearance distance is close to the safety threshold. They cannot predict short-term clearance risks based on dynamic factors such as blade oscillation, tower deformation, and wind field changes. This lag often results in the control system or maintenance personnel not having enough time to adjust the wind turbine's operating status, thus leading to accidents.

[0004] 3. Poor multi-sensor fusion effect and low data utilization: Although some existing products have tried to use multi-sensor combinations, they are mostly simple data superposition applications, lacking an effective data fusion mechanism. They cannot dynamically adjust the weight of sensor data according to environmental changes, making it difficult to balance the accuracy and continuity of monitoring data.

[0005] 4. Poor system adaptability and high operation and maintenance costs: Most existing monitoring systems are designed with fixed specifications, making it difficult to flexibly adapt to different models of wind turbines. Installation and debugging are complicated, and some core sensors rely on imports, resulting in high overall costs and limiting large-scale promotion and application.

[0006] 5. Insufficient data integration hinders collaborative operation and maintenance decision-making: Airspace monitoring data is mostly stored independently and is not effectively integrated with the wind turbine main control system and operation and maintenance management platform. The monitoring system can only provide a single early warning function and cannot combine airspace monitoring data with pitch control, shutdown protection, operation and maintenance scheduling, etc., failing to form an effective closed-loop management system and making it difficult to fully realize the value of airspace monitoring data. Summary of the Invention

[0007] The purpose of this invention is to provide a method and system for monitoring the clearance of wind turbine blades, in order to solve the above-mentioned problems existing in the prior art.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: Firstly, a method for monitoring the clearance of wind turbine blades is provided, including: Collect multi-dimensional monitoring datasets of the target wind turbine, as well as real-time wind field data and predicted wind field data of the wind field where the target wind turbine is located. The multi-dimensional monitoring datasets include initial blade clearance data, blade tip polarized light image data, real-time environmental data, blade attitude data, and tower deformation data. Data preprocessing is performed on the multi-dimensional monitoring dataset to obtain the preprocessed multi-dimensional monitoring dataset; Multi-source data fusion processing is performed on the preprocessed multi-dimensional monitoring dataset to obtain the fused blade clearance distance; By integrating blade clearance distance, real-time wind field data, and predicted wind field data into a pre-trained dynamic coupling model for clearance prediction, the trend data of blade clearance distance change of the target wind turbine in the future time period is obtained. The predicted blade clearance distance at each future time point is extracted from the trend data of blade clearance distance change, and a graded early warning judgment is made for the predicted blade clearance distance at each future time point. When the corresponding predicted blade clearance distance meets the set graded early warning conditions, the corresponding graded early warning information is output to the wind turbine main control system and operation and maintenance management platform of the target wind turbine.

[0009] In one possible design, the data preprocessing of the multi-dimensional monitoring dataset to obtain a preprocessed multi-dimensional monitoring dataset includes: The multi-dimensional monitoring dataset is synchronized, denoised, and outlier removed to obtain a preprocessed multi-dimensional monitoring dataset.

[0010] In one possible design, the multi-source data fusion processing of the preprocessed multi-dimensional monitoring dataset to obtain the fused blade clearance distance includes: Based on real-time environmental data, an adaptive weighted fusion algorithm is used to dynamically adjust the weights of the initial blade clearance data and the blade tip polarized light image data, and the initial blade clearance data and the blade tip polarized light image data are weighted and fused to obtain the real-time fused clearance distance. The real-time fusion clearance distance is corrected using blade attitude data and tower deformation data to obtain the fusion blade clearance distance.

[0011] In one possible design, when dynamically adjusting the weights of the initial blade clearance data and the blade tip polarized light image data using an adaptive weighted fusion algorithm based on real-time environmental data, the weight adjustment rules of the adaptive weighted fusion algorithm are as follows: when the real-time environmental data is within the set normal environmental data range, the weight of the initial blade clearance data accounts for 30%-40%, and the weight of the blade tip polarized light image data accounts for 60%-70%; when the real-time environmental data is within the set severe environmental data range, the weight of the initial blade clearance data accounts for 80%-90%, and the weight of the blade tip polarized light image data accounts for 10%-20%.

[0012] In one possible design, the step of correcting the real-time fused clearance distance using blade attitude data and tower deformation data to obtain the fused blade clearance distance includes: Based on the blade attitude data, the blade swing angle variable is determined, and the blade swing clearance correction amount is determined according to the blade swing angle variable. The real-time fused clearance distance is corrected using the blade swing clearance correction amount to obtain the initial corrected clearance distance. The tower deformation clearance correction amount is determined based on the tower deformation data, and the initial correction clearance distance is corrected using the tower deformation clearance correction amount to obtain the fusion blade clearance distance.

[0013] In one possible design, the step of performing graded early warning determination on the predicted blade clearance distance at each future time point, and outputting the corresponding graded early warning information to the wind turbine's main control system and operation and maintenance management platform when the corresponding predicted blade clearance distance meets the set graded early warning conditions, includes: The predicted blade clearance distance at each future time point is compared with the set tiered early warning rules. When the corresponding predicted blade clearance distance falls within the set tier 3 early warning distance range, the corresponding tier 3 early warning information is output to the wind turbine's main control system and operation and maintenance management platform. When the corresponding predicted blade clearance distance falls within the set tier 2 early warning distance range, the corresponding tier 2 early warning information is output to the wind turbine's main control system and operation and maintenance management platform. When the corresponding predicted blade clearance distance falls within the set tier 1 early warning distance range, the corresponding tier 1 early warning information is output to the wind turbine's main control system and operation and maintenance management platform.

[0014] In one possible design, the method further includes: Multi-dimensional monitoring datasets, real-time wind field data, predicted wind field data, fused blade clearance distance, blade clearance distance change trend data, and graded early warning information are uploaded to a cloud database for storage.

[0015] Secondly, a wind turbine blade clearance monitoring system is provided, comprising a composite sensing module, an attitude sensing module, a deformation detection module, a data processing module, and a linkage control module. The composite sensing module is installed on the tower of the target wind turbine and includes a millimeter-wave radar, a polarized light camera, and an environmental sensing unit. The millimeter-wave radar is used to collect initial blade clearance data of the target wind turbine. The polarized light camera is used to collect polarized light image data of the blade tips of the target wind turbine. The environmental sensing unit is used to collect real-time environmental data of the target wind turbine. The attitude sensing module is installed at the root of the blades of the target wind turbine and includes an IMU (Inertial Measurement Unit) for collecting blade attitude data of the target wind turbine. The deformation detection module is installed on the sidewall of the tower of the target wind turbine and includes distributed strain sensors for collecting tower deformation data of the target wind turbine. The data processing module is used to collect initial blade clearance data, blade tip polarized light image data, real-time environmental data, blade attitude data, and tower deformation data. Deformation data constitutes a multi-dimensional monitoring dataset for the target wind turbine. This dataset undergoes preprocessing to obtain a preprocessed multi-dimensional monitoring dataset. Multi-source data fusion processing is then performed on this preprocessed dataset to obtain a fused blade clearance distance. This fused blade clearance distance, real-time wind field data, and predicted wind field data are input into a pre-trained dynamic coupling model for clearance prediction, yielding trend data on the target wind turbine's blade clearance distance over future time periods. Predicted blade clearance distances for each future time point are extracted from this trend data, and tiered early warning judgments are made for these predicted blade clearance distances. When the corresponding predicted blade clearance distance meets the set tiered early warning conditions, the corresponding tiered early warning information is output. The linkage control module establishes a communication connection with the target wind turbine's main control system and operation and maintenance management platform to transmit the tiered early warning information to these systems.

[0016] In one possible design, the system further includes a data operation and maintenance module and a power supply module. The data operation and maintenance module is used to upload multi-dimensional monitoring datasets, real-time wind field data, predicted wind field data, fused blade clearance distance, blade clearance distance change trend data, and graded early warning information to a cloud database for storage. The power supply module is used to supply power to the composite sensing module, attitude sensing module, deformation detection module, data processing module, linkage control module, and data operation and maintenance module.

[0017] Thirdly, a wind turbine blade clearance monitoring system is provided, including: Memory, used to store instructions; The processor is configured to read instructions stored in the memory and execute any one of the wind turbine blade clearance monitoring methods described in the first aspect above, according to the instructions.

[0018] Fourthly, a computer-readable storage medium is provided, on which instructions are stored, which, when executed on a computer, cause the computer to perform any one of the wind turbine blade clearance monitoring methods described in the first aspect. Simultaneously, a computer program product is also provided, which, when executed on a computer, performs any one of the wind turbine blade clearance monitoring methods described in the first aspect.

[0019] Beneficial effects: 1. This invention can significantly improve adaptability to extreme environments: through composite sensing design and adaptive weighted fusion mechanism, it can effectively resist interference from harsh environments and ensure the accuracy and continuity of wind turbine blade clearance monitoring data under different environments.

[0020] 2. This invention enables dynamic risk prediction and avoids delayed early warning: By integrating multi-dimensional dynamic data with predicted wind field data to perform blade clearance coupling prediction, it can accurately predict future clearance change trends, trigger graded early warnings in advance, achieve advanced early warning, reserve sufficient adjustment time for the wind turbine main control system, and effectively reduce the risk of "tower sweeping".

[0021] 3. This invention has high efficiency in multi-source data fusion and high monitoring accuracy: Through multi-dimensional data preprocessing and adaptive weighted fusion mechanism, the measurement errors caused by environmental interference, blade oscillation and tower deformation are effectively eliminated, which significantly improves the measurement accuracy of clearance distance and data utilization.

[0022] 4. This invention has strong adaptability and low operation and maintenance costs: Through modular design, it can be flexibly adapted to different models of wind turbines, reducing procurement costs. It can be paired with a remote operation and maintenance management platform, reducing the frequency of high-altitude operations and significantly reducing operation and maintenance costs and risks.

[0023] 5. This invention can form a closed-loop operation and maintenance management system to improve the operating efficiency of wind turbines: It realizes a closed-loop management of the entire process of "monitoring-prediction-control-operation and maintenance", which can effectively avoid wind turbine safety risks and improve wind turbine power generation efficiency while ensuring safety. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a flowchart illustrating the method in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the system installation in Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of the system configuration in Embodiment 3 of the present invention. Detailed Implementation

[0026] It should be noted that the descriptions of these embodiments are intended to aid in understanding the invention and do not constitute a limitation thereof. The specific structural and functional details disclosed herein are merely for describing exemplary embodiments of the invention. However, the invention may be embodied in many alternative forms and should not be construed as being limited to the embodiments described herein.

[0027] It should be understood that, unless otherwise explicitly specified and limited, the corresponding terms should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in the embodiments according to the specific circumstances.

[0028] Specific details are provided in the following description to provide a complete understanding of the exemplary embodiments. However, those skilled in the art will understand that the exemplary embodiments can be implemented without these specific details. For example, apparatus may be shown in block diagrams to avoid obscuring the examples with unnecessary details. In other embodiments, well-known processes, structures, and techniques may be omitted with non-essential details to avoid obscuring the embodiments.

[0029] Example 1: This embodiment provides a method for monitoring the clearance of wind turbine blades, such as... Figure 1 As shown, the method includes the following steps: S1. Collect multi-dimensional monitoring datasets of the target wind turbine, as well as real-time wind field data and predicted wind field data of the wind field where the target wind turbine is located. The multi-dimensional monitoring datasets include initial blade clearance data, blade tip polarized light image data, real-time environmental data, blade attitude data, and tower deformation data.

[0030] In practical implementation, real-time and predicted wind field data (such as predicted wind field data for the next 5-10 minutes, including wind direction and wind force) of the wind field where the target wind turbine is located can be obtained through the corresponding meteorological interface. A composite sensing module can be pre-deployed at a predetermined height on the target wind turbine tower (close to the optimal monitoring position near the blade rotation trajectory), an attitude sensing module can be deployed at the root of the target wind turbine blades, and a deformation detection module can be deployed on the sidewall of the target wind turbine tower. This allows for the synchronous acquisition of multi-dimensional monitoring data through various sensing / detection modules. The composite sensing module may include millimeter-wave radar, a polarized light camera, and an environmental sensing unit; the attitude sensing module may use an IMU (Inertial Measurement Unit); and the deformation detection module may use distributed strain sensors. Millimeter-wave radar is used to acquire initial blade clearance data between the blade tip and the tower; polarized light camera is used to acquire polarized light image data of the blade tip to assist in positioning; environmental sensing unit is used to acquire real-time environmental data (including rainfall, dust concentration, wind speed, salt spray concentration, etc.); IMU inertial measurement unit is used to acquire blade attitude data (including swing angle, angular velocity, acceleration); distributed strain sensor is used to acquire tower deformation data (including radial micro-deformation, axial strain).

[0031] For example, for a 1.5MW onshore wind turbine, a composite sensing module can be installed at a height of 35m above the ground on the turbine tower (the optimal monitoring position for the blade rotation trajectory). The composite sensing module includes a 24GHz millimeter-wave radar, a polarized light camera, and an environmental sensing unit. One IMU (Inertial Measurement Unit) (attitude sensing module) is installed at the root of each of the three blades. A distributed strain sensor (deformation detection module) is arranged every 5m along the axial direction on the sidewall of the tower, for a total of six sensors. The following data is collected simultaneously: millimeter-wave radar collects data from the blade tips. The data consists of the initial blade clearance data of the tower (acquisition frequency 10Hz), the polarized light image data of the blade tip acquired by the polarized light camera (acquisition frequency 5Hz), the real-time environmental data acquired by the environmental sensing unit (acquisition frequency 1Hz) including rainfall, dust concentration, wind speed, and salt spray concentration, the blade attitude data acquired by the IMU inertial measurement unit (acquisition frequency 10Hz) including blade swing angle variables, angular velocity, and acceleration, and the tower deformation data acquired by the distributed strain sensor (acquisition frequency 5Hz) including radial micro-deformation and axial strain of the tower.

[0032] S2. Perform data preprocessing on the multi-dimensional monitoring dataset to obtain the preprocessed multi-dimensional monitoring dataset.

[0033] In practice, the multi-dimensional monitoring dataset can be synchronized, denoised, and outlier removed to obtain a preprocessed multi-dimensional monitoring dataset. During data synchronization, all data can be synchronized based on timestamps to ensure data temporal consistency. For data denoising, a Kalman filter algorithm can be used to denoise distance, attitude, and deformation data, eliminating interference from environmental noise and equipment vibration. For outlier removal, an adaptive threshold algorithm can be used to remove abnormal data caused by environmental interference, ensuring data validity, such as removing data with rainfall ≥5mm / h or dust concentration ≥100μg / m³. 3 Abnormal polarized light image data at times, and abnormal clearance distance data when wind speed ≥ 25m / s are removed.

[0034] S3. Perform multi-source data fusion processing on the preprocessed multi-dimensional monitoring dataset to obtain the fused blade clearance distance.

[0035] In practice, an adaptive weighted fusion algorithm can be used to dynamically adjust the weights of the initial blade clearance data and the blade tip polarized light image data based on real-time environmental data. The initial blade clearance data and the blade tip polarized light image data are then weighted and fused to obtain the real-time fused clearance distance. When dynamically adjusting the weights of the initial blade clearance data and the blade tip polarized light image data based on real-time environmental data using the adaptive weighted fusion algorithm, the weight adjustment rules can be set as follows: when the real-time environmental data is within the set normal environmental data range, the weight of the initial blade clearance data is 30%-40%, and the weight of the blade tip polarized light image data is 60%-70%; when the real-time environmental data is within the set severe environmental data range, the weight of the initial blade clearance data is 80%-90%, and the weight of the blade tip polarized light image data is 10%-20%. When the wind speed is ≤12m / s and there is no sandstorm, rain or snow, the weight of the polarized light image data at the blade tip of the polarized light camera is set to 65%, and the weight of the initial blade clearance data of the millimeter-wave radar is set to 35%. The initial blade clearance data and the polarized light image data at the blade tip are weighted and fused to obtain the real-time fused clearance distance. When the wind speed is >12m / s or there is sandstorm, rain or snow, the weight of the initial blade clearance data of the millimeter-wave radar is set to 85%, and the weight of the polarized light image data at the blade tip of the polarized light camera is set to 15%. The initial blade clearance data and the polarized light image data at the blade tip are weighted and fused to obtain the real-time fused clearance distance.

[0036] Then, the real-time fused clearance distance is corrected using blade attitude data and tower deformation data to obtain the fused blade clearance distance. The correction process includes: determining the blade swing angle variable based on the blade attitude data, determining the blade swing clearance correction amount based on the blade swing angle variable, and using the blade swing clearance correction amount to correct the real-time fused clearance distance to obtain the initial corrected clearance distance. For example, when the blade swing angle increases by 5°, the real-time fused clearance distance correction decreases by 3-5mm to ensure measurement accuracy. The tower deformation clearance correction amount is determined based on the tower deformation data, and the initial corrected clearance distance is corrected using the tower deformation clearance correction amount to obtain the fused blade clearance distance.

[0037] S4. Input the fused blade clearance distance, real-time wind field data, and predicted wind field data into the pre-trained clearance dynamic coupling model to predict the clearance distance and obtain the blade clearance distance change trend data of the target wind turbine in the future time period.

[0038] In practical implementation, a dynamic coupling model of airspace clearance can be pre-constructed, and a reinforcement learning algorithm can be used to train the dynamic coupling model of airspace clearance based on historical operational data samples (including airspace clearance and wind field data) to accurately predict the trend of airspace clearance changes over a future period. For example, the dynamic coupling model of airspace clearance can adopt LSTM (Long Short-Term Memory Network), BiLSTM (Bidirectional Long Short-Term Memory Network), Transformer time series prediction model, or GNN-LSTM hybrid model (time series prediction model based on graph neural network).

[0039] In application, the fused blade clearance distance, real-time wind field data, and predicted wind field data can be input into a pre-trained dynamic coupling model for clearance prediction, thereby obtaining the blade clearance distance change trend data of the target wind turbine in the future time period. The blade clearance distance change trend data includes the predicted blade clearance distance at each future time point.

[0040] S5. Extract the predicted blade clearance distance for each future time point from the blade clearance distance change trend data, and make a graded early warning judgment on the predicted blade clearance distance for each future time point. When the corresponding predicted blade clearance distance meets the set graded early warning conditions, output the corresponding graded early warning information to the wind turbine main control system and operation and maintenance management platform of the target wind turbine.

[0041] In practice, the predicted blade clearance distance for each future time point can be extracted from the blade clearance distance change trend data. Then, the predicted blade clearance distance for each future time point is compared with the set graded early warning rules. When the corresponding predicted blade clearance distance is within the set third-level early warning distance range (e.g., 1.3m < predicted blade clearance distance ≤ 1.5m), the corresponding third-level early warning information is output to the wind turbine main control system and operation and maintenance management platform of the target wind turbine. When the corresponding predicted blade clearance distance is within the set second-level early warning distance range (e.g., 1.2m < predicted blade clearance distance ≤ 1.3m), the corresponding second-level early warning information is output to the wind turbine main control system and operation and maintenance management platform of the target wind turbine. When the corresponding predicted blade clearance distance is within the set first-level early warning distance range (e.g., predicted blade clearance distance ≤ 1.2m), the corresponding first-level early warning information is output to the wind turbine main control system and operation and maintenance management platform of the target wind turbine. The wind turbine main control system can automatically adjust the pitch angle of the target wind turbine blades, reduce the turbine speed, or trigger shutdown protection based on the corresponding level of early warning information to avoid collision accidents. The operation and maintenance management platform can simultaneously push the corresponding early warning information to the operation and maintenance personnel to guide them in conducting targeted inspections. For example, in the case of a level 3 early warning, the operation and maintenance management platform pushes an early warning information to remind the operation and maintenance personnel to pay attention; in the case of a level 2 early warning, the wind turbine main control system adjusts the pitch angle of the blades by 5° and reduces the turbine speed to 80% of the rated speed; in the case of a level 1 early warning, the wind turbine main control system triggers shutdown protection to stop the wind turbine from running and avoid collision accidents.

[0042] S6. Upload the multi-dimensional monitoring dataset, real-time wind field data, predicted wind field data, fused blade clearance distance, blade clearance distance change trend data, and graded early warning information to the cloud database for storage.

[0043] In practice, multi-dimensional monitoring datasets, real-time wind field data, predicted wind field data, fused blade clearance distance, blade clearance distance change trend data, and tiered early warning information can be uploaded to a cloud database for storage. All monitoring data and early warning information can be stored in the cloud database to form historical data. Each quarter, based on historical data, transfer learning algorithms can be used to optimize the weight parameters of the adaptive weighted fusion algorithm and the prediction parameters of the clearance dynamic coupling model, continuously improving monitoring accuracy.

[0044] This method overcomes the shortcomings of existing technologies in wind turbine blade clearance monitoring, such as poor adaptability to extreme environments, delayed early warning, low data utilization, insufficient adaptability, and high operation and maintenance costs. It enables accurate and continuous monitoring of clearance distance, advance prediction of dynamic risks, improves monitoring stability in extreme environments, reduces operation and maintenance costs, forms closed-loop operation and maintenance management, and ensures the safe and efficient operation of wind turbines.

[0045] Example 2: This embodiment provides a wind turbine blade clearance monitoring system, such as Figure 2 As shown, the system includes a composite sensing module, an attitude sensing module, a deformation detection module, a data processing module, and a linkage control module. The composite sensing module, installed on the tower of the target wind turbine, includes a millimeter-wave radar, a polarized light camera, and an environmental sensing unit. The millimeter-wave radar is used to collect initial blade clearance data of the target wind turbine. The polarized light camera is used to collect polarized light image data of the blade tips of the target wind turbine. The environmental sensing unit is used to collect real-time environmental data of the target wind turbine. The attitude sensing module, installed at the root of the blades of the target wind turbine, includes an IMU (Inertial Measurement Unit) for collecting blade attitude data. The deformation detection module, installed on the sidewall of the tower of the target wind turbine, includes distributed strain sensors for collecting tower deformation data. The data processing module collects initial blade clearance data, blade tip polarized light image data, real-time environmental data, blade attitude data, and tower deformation data to form the target wind turbine's data. The system employs a multi-dimensional monitoring dataset for wind turbines. This dataset undergoes preprocessing to obtain a preprocessed multi-dimensional monitoring dataset. Multi-source data fusion processing is then performed on the preprocessed dataset to obtain a fused blade clearance distance. This fused blade clearance distance, real-time wind field data, and predicted wind field data are input into a pre-trained dynamic coupling model for clearance prediction. This yields trend data on the blade clearance distance changes of the target wind turbine over future time periods. Predicted blade clearance distances for each future time point are extracted from this trend data, and tiered early warning judgments are made for the predicted blade clearance distances at each future time point. When the corresponding predicted blade clearance distance meets the set tiered early warning conditions, the corresponding tiered early warning information is output. The linkage control module establishes a communication connection with the target wind turbine's main control system and operation and maintenance management platform to transmit the tiered early warning information to these systems.

[0046] Furthermore, the system also includes a data operation and maintenance module and a power supply module. The data operation and maintenance module is used to upload multi-dimensional monitoring datasets, real-time wind field data, predicted wind field data, fused blade clearance distance data, blade clearance distance change trend data, and graded early warning information to a cloud database for storage. The power supply module is used to power the composite sensing module, attitude sensing module, deformation detection module, data processing module, linkage control module, and data operation and maintenance module. The power supply module can adopt a dual power supply mode of solar power + backup lithium battery power, which is suitable for outdoor, offshore, and other scenarios without stable power supply, ensuring continuous and stable operation of the system. At the same time, the power supply module has a low battery warning function to promptly remind maintenance personnel to replace the battery.

[0047] For example, for onshore wind turbines, a composite sensing module can be installed 35m above the ground on the turbine tower (the optimal monitoring position for the blade rotation trajectory), using a waterproof and corrosion-resistant housing (suitable for outdoor environments). The composite sensing module includes a 24GHz millimeter-wave radar (measurement range 0.5-50m, measurement accuracy ±3mm), a polarized light camera (resolution 1920×1080, with automatic exposure and reflective filtering functions), and environmental sensing units (rainfall sensor measurement range 0-50mm / h, dust concentration sensor measurement range 0-1000μg / m³, wind speed sensor measurement range 0-50m / s, salt spray concentration sensor measurement range 0-2000ppm); and an IMU inertial measurement unit (measurement range: angle ±180°, angular velocity ±2000°) is installed at the root of each of the three blades. ° / s, acceleration ±16g, with vibration resistance and waterproof function); one distributed strain sensor (measurement range ±2000με, measurement accuracy ±1με) is arranged every 5m along the axial direction on the side wall of the tower, for a total of 6; the following data are collected synchronously: millimeter-wave radar collects the initial blade clearance data between the blade tip and the tower (collection frequency 10Hz), polarized light camera collects polarized light image data of the blade tip (collection frequency 5Hz), environmental sensing unit collects rainfall, dust concentration, wind speed, and salt spray concentration (collection frequency 1Hz) to form real-time environmental data, IMU inertial measurement unit collects blade swing angle variable, angular velocity, and acceleration (collection frequency 10Hz) to form blade attitude data, and distributed strain sensors collect tower radial micro-deformation and axial strain (collection frequency 5Hz) to form tower deformation data. The data processing module uses an industrial-grade STM32 microcontroller with a main frequency of 180MHz, possessing high-speed data processing capabilities. It can perform data preprocessing, multi-source data fusion, and dynamic risk prediction, with a response time of ≤100ms. It supports wired and wireless connections with various sensor modules (wireless communication uses the LoRa protocol, with a transmission distance of ≥100m). The linkage control module uses a PLC controller, electrically connected to the data processing module and the wind turbine main control system. It can receive early warning information and supports both manual and automatic control modes. The data maintenance module interfaces with a cloud database, supporting massive data storage (storage time ≥3 years). The maintenance management platform adopts a web + mobile terminal design, supporting functions such as data query, historical curve analysis, early warning information push, sensor calibration reminders, and maintenance work order management. The power supply module adopts a dual power supply mode of a 100W solar panel + a 24V / 100Ah lithium battery. The solar panel can meet the daily power supply needs, while the lithium battery serves as a backup power source, ensuring continuous operation of the system for ≥72 hours in cloudy or rainy weather. It also features a low battery warning function (a warning is pushed when the remaining battery is ≤20%).The system configuration has been field-tested and can maintain an airspace distance measurement accuracy within ±5mm even in severe environments such as sandstorms (sandstorm concentration 80μg / m³), moderate rain (rainfall 3mm / h), and strong winds (wind speed 15m / s). The dynamic risk prediction accuracy is ≥98%, and the early warning response time is ≤100ms. It can effectively avoid "tower sweeping" accidents, reduce operation and maintenance costs by more than 30%, and improve wind turbine power generation efficiency by 5%-8%, fully meeting the airspace monitoring requirements of onshore wind turbines.

[0048] For offshore wind turbines, considering the environmental characteristics of high salt spray, strong corrosion, and strong winds at sea, the composite sensing module, attitude sensing module, and deformation detection module can all adopt a corrosion-resistant and waterproof IP68-rated shell with an anti-salt spray coating to extend the service life of the equipment. The millimeter-wave radar uses the 77GHz band to improve penetration and effectively resist interference from dense fog and heavy rainfall at sea. The power supply module uses a 200W solar panel + 24V / 200Ah lithium battery, combined with the lighting conditions at sea, to ensure continuous and stable operation of the system. The system data transmission adopts the 5G communication protocol to replace the LoRa protocol, improving the data transmission rate and stability and adapting to the complex communication environment at sea. The system configuration has been tested on-site at offshore wind turbines and can still operate stably under salt spray concentrations of 1500ppm, wind speeds of 20m / s, and dense fog conditions, with an air clearance monitoring accuracy of ±4mm and a dynamic risk prediction accuracy of ≥97%, making it suitable for the operation and maintenance needs of offshore wind turbines.

[0049] Example 3: This embodiment provides a wind turbine blade clearance monitoring system, such as Figure 3 As shown, at the hardware level, it includes: The data interface is used to establish data communication between the processor and external data terminals; Memory, used to store instructions; The processor is used to read the instructions stored in the memory and execute the wind turbine blade clearance monitoring method in Embodiment 1 according to the instructions.

[0050] Optionally, the system also includes an internal bus, through which the processor, memory, and data interface can be interconnected. This internal bus can be a PCIe (Peripheral Component Interconnect Eexpress) bus, which can be divided into an address bus, a data bus, a control bus, etc. The memory can include, but is not limited to, Random Access Memory (RAM), Read Only Memory (ROM), Flash Memory, First Input First Output (FIFO), and / or First In Last Out (FILO). The processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0051] Example 4: This embodiment provides a computer-readable storage medium storing instructions. When these instructions are executed on a computer, the computer performs the wind turbine blade clearance monitoring method described in Embodiment 1. The computer-readable storage medium refers to a data storage medium, which may include, but is not limited to, floppy disks, optical disks, hard disks, flash memory, USB flash drives, and / or Memory Sticks. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0052] This embodiment also provides a computer program product that, when run on a computer, executes the wind turbine blade clearance monitoring method of Embodiment 1. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device.

[0053] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for monitoring the clearance of wind turbine blades, characterized in that, include: Collect multi-dimensional monitoring datasets of the target wind turbine, as well as real-time wind field data and predicted wind field data of the wind field where the target wind turbine is located. The multi-dimensional monitoring datasets include initial blade clearance data, blade tip polarized light image data, real-time environmental data, blade attitude data, and tower deformation data. Data preprocessing is performed on the multi-dimensional monitoring dataset to obtain the preprocessed multi-dimensional monitoring dataset; Multi-source data fusion processing is performed on the preprocessed multi-dimensional monitoring dataset to obtain the fused blade clearance distance; By integrating blade clearance distance, real-time wind field data, and predicted wind field data into a pre-trained dynamic coupling model for clearance prediction, the trend data of blade clearance distance change of the target wind turbine in the future time period is obtained. The predicted blade clearance distance at each future time point is extracted from the trend data of blade clearance distance change, and a graded early warning judgment is made for the predicted blade clearance distance at each future time point. When the corresponding predicted blade clearance distance meets the set graded early warning conditions, the corresponding graded early warning information is output to the wind turbine main control system and operation and maintenance management platform of the target wind turbine.

2. The method for monitoring the clearance of wind turbine blades according to claim 1, characterized in that, The process of preprocessing the multi-dimensional monitoring dataset to obtain a preprocessed multi-dimensional monitoring dataset includes: The multi-dimensional monitoring dataset is synchronized, denoised, and outlier removed to obtain a preprocessed multi-dimensional monitoring dataset.

3. The method for monitoring the clearance of wind turbine blades according to claim 1, characterized in that, The process of fusing preprocessed multi-dimensional monitoring datasets to obtain fused blade clearance distance includes: Based on real-time environmental data, an adaptive weighted fusion algorithm is used to dynamically adjust the weights of the initial blade clearance data and the blade tip polarized light image data, and the initial blade clearance data and the blade tip polarized light image data are weighted and fused to obtain the real-time fused clearance distance. The real-time fusion clearance distance is corrected using blade attitude data and tower deformation data to obtain the fusion blade clearance distance.

4. The method for monitoring the clearance of wind turbine blades according to claim 3, characterized in that, When dynamically adjusting the weights of initial blade clearance data and blade tip polarized light image data using an adaptive weighted fusion algorithm based on real-time environmental data, the weight adjustment rules of the adaptive weighted fusion algorithm are as follows: when the real-time environmental data is within the set normal environmental data range, the weight of the initial blade clearance data accounts for 30%-40%, and the weight of the blade tip polarized light image data accounts for 60%-70%; when the real-time environmental data is within the set severe environmental data range, the weight of the initial blade clearance data accounts for 80%-90%, and the weight of the blade tip polarized light image data accounts for 10%-20%.

5. The method for monitoring the clearance of wind turbine blades according to claim 3, characterized in that, The process of correcting the real-time fused clearance distance using blade attitude data and tower deformation data to obtain the fused blade clearance distance includes: Based on the blade attitude data, the blade swing angle variable is determined, and the blade swing clearance correction amount is determined according to the blade swing angle variable. The real-time fused clearance distance is corrected using the blade swing clearance correction amount to obtain the initial corrected clearance distance. The tower deformation clearance correction amount is determined based on the tower deformation data, and the initial correction clearance distance is corrected using the tower deformation clearance correction amount to obtain the fusion blade clearance distance.

6. The method for monitoring the clearance of wind turbine blades according to claim 1, characterized in that, The process involves classifying and issuing early warnings based on the predicted blade clearance distance at each future time point. When the predicted blade clearance distance meets the set classification and early warning conditions, the corresponding classification and early warning information is output to the wind turbine's main control system and operation and maintenance management platform, including: The predicted blade clearance distance at each future time point is compared with the set tiered early warning rules. When the corresponding predicted blade clearance distance falls within the set tier 3 early warning distance range, the corresponding tier 3 early warning information is output to the wind turbine's main control system and operation and maintenance management platform. When the corresponding predicted blade clearance distance falls within the set tier 2 early warning distance range, the corresponding tier 2 early warning information is output to the wind turbine's main control system and operation and maintenance management platform. When the corresponding predicted blade clearance distance falls within the set tier 1 early warning distance range, the corresponding tier 1 early warning information is output to the wind turbine's main control system and operation and maintenance management platform.

7. The method for monitoring the clearance of wind turbine blades according to claim 1, characterized in that, The method further includes: Multi-dimensional monitoring datasets, real-time wind field data, predicted wind field data, fused blade clearance distance, blade clearance distance change trend data, and graded early warning information are uploaded to a cloud database for storage.

8. A wind turbine blade clearance monitoring system, characterized in that, The system includes a composite sensing module, an attitude sensing module, a deformation detection module, a data processing module, and a linkage control module. The composite sensing module, installed on the tower of the target wind turbine, includes a millimeter-wave radar, a polarized light camera, and an environmental sensing unit. The millimeter-wave radar is used to collect initial blade clearance data of the target wind turbine. The polarized light camera is used to collect polarized light image data of the blade tips of the target wind turbine. The environmental sensing unit is used to collect real-time environmental data of the target wind turbine. The attitude sensing module, installed at the root of the blades of the target wind turbine, includes an IMU (Inertial Measurement Unit) for collecting blade attitude data. The deformation detection module, installed on the sidewall of the tower of the target wind turbine, includes distributed strain sensors for collecting tower deformation data. The data processing module collects initial blade clearance data, blade tip polarized light image data, real-time environmental data, blade attitude data, and tower deformation data to form the target wind turbine data. The system uses a multi-dimensional monitoring dataset. Data preprocessing is performed on the multi-dimensional monitoring dataset to obtain a preprocessed multi-dimensional monitoring dataset. Multi-source data fusion processing is then performed on the preprocessed multi-dimensional monitoring dataset to obtain a fused blade clearance distance. This fused blade clearance distance, real-time wind field data, and predicted wind field data are input into a pre-trained dynamic coupling model for clearance prediction. This yields the blade clearance distance trend data for the target wind turbine over future time periods. Predicted blade clearance distances for each future time point are extracted from this trend data, and a tiered early warning judgment is made for each predicted blade clearance distance. When the corresponding predicted blade clearance distance meets the set tiered early warning conditions, the corresponding tiered early warning information is output. The linkage control module establishes a communication connection with the target wind turbine's main control system and operation and maintenance management platform to transmit the tiered early warning information to these systems.

9. A wind turbine blade clearance monitoring system according to claim 8, characterized in that, The system also includes a data operation and maintenance module and a power supply module. The data operation and maintenance module is used to upload multi-dimensional monitoring datasets, real-time wind field data, predicted wind field data, fused blade clearance distance, blade clearance distance change trend data, and graded early warning information to a cloud database for storage. The power supply module is used to supply power to the composite sensing module, attitude sensing module, deformation detection module, data processing module, linkage control module, and data operation and maintenance module.

10. A wind turbine blade clearance monitoring system, characterized in that, include: Memory, used to store instructions; A processor is configured to read instructions stored in the memory and execute the wind turbine blade clearance monitoring method according to any one of claims 1-7.

Images