Wind generating set variable pitch bearing state on-line monitoring and service safety early warning system

Abstract

The invention discloses a wind generating set variable pitch bearing state on-line monitoring and service safety early warning system, which realizes multi-dimensional data acquisition through a force sensor, a rotating speed sensor and a vibration sensor of an on-line monitoring module, and establishes a load-rotating speed-vibration characteristic mapping relation through a database module. A fault prediction model is constructed by using intelligent algorithms such as blind source separation of the performance prediction module, and state grading evaluation is realized in combination with a multi-level threshold value of the safety performance evaluation module. The hardware module completes precise sensor type selection and equipment cooperation, and the software module achieves the functions of data acquisition, analysis, display and the like. The system overcomes the problems of low monitoring precision, weak data processing, inaccurate early warning and the like in the prior art, can quickly and accurately predict and evaluate the bearing state, position faults and optimize maintenance decisions, reduces the operation and maintenance cost, and improves the operation safety and economy of a unit.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of safety warning of wind turbine generator, and particularly relates to a wind turbine generator variable pitch bearing state online monitoring and service safety warning system. BACKGROUND

[0002] With the increasing demand for clean energy worldwide, wind power, as an important part of green and sustainable energy, its industry continues to expand. The safe and stable operation of wind turbine generators directly affects the power generation efficiency and economic benefits, and the performance state of the variable pitch bearing, as a key component of the wind turbine generator, plays a decisive role in the overall operation of the unit. Once the variable pitch bearing fails, not only will it cause the unit to shut down, causing huge economic losses, but also may cause safety accidents. Therefore, online monitoring and service safety warning of the variable pitch bearing have become a key link to ensure the reliable operation of the wind turbine generator.

[0003] The prior art has many deficiencies in variable pitch bearing monitoring. In terms of monitoring accuracy, the traditional single sensor monitoring method is difficult to fully capture the complex operating state information of the variable pitch bearing, has weak early fault recognition ability, and cannot discover potential small damage of the bearing in time, leading to gradual accumulation of fault hazards. In terms of data processing, the previous data processing method is relatively simple, cannot fully mine effective information in the massive monitoring data, and cannot establish an accurate fault feature model, so that the evaluation of the operating state of the variable pitch bearing lacks scientificity and accuracy. In the safety warning link, the existing warning system is mostly based on fixed thresholds and cannot be dynamically adjusted according to different service environments and operating conditions, which is prone to false positives or false negatives, and cannot provide reliable decision basis for maintenance personnel. SUMMARY

[0004] In order to overcome the shortcomings and deficiencies of the prior art, the present application provides a wind turbine generator variable pitch bearing state online monitoring and service safety warning system.

[0005] The technical scheme adopted by the present application is that the wind turbine generator variable pitch bearing state online monitoring and service safety warning system comprises an online monitoring module, a database module, a performance prediction module, a safety performance evaluation module, a hardware module and a software module. The online monitoring module is used for collecting operating data based on daily operation and maintenance of the wind turbine generator variable pitch bearing, using multiple sensors to perform online monitoring on the load, rotating speed and vibration acceleration of the variable pitch bearing, and transmitting the monitoring data to the database module. The database module is used for establishing a mapping relationship between load-rotating speed-vibration characteristics based on measured data, and storing safety performance data under different service environments, operating conditions and corresponding safety performance data. The performance prediction module is used to quickly predict the service state of the variable pitch bearing by using an intelligent algorithm. The safety performance evaluation module is used to perform a graded evaluation of the variable pitch bearing state by using a multi-level threshold variable pitch bearing service safety performance evaluation method based on the mapping relationship between the load-speed-vibration characteristics. The hardware module includes sensors, data acquisition equipment, data processing units and communication modules that meet the selection requirements. The software module includes a data acquisition and monitoring module, a data analysis and decision module, a user interface module and a report generation module.

[0006] Further, in the online monitoring module, the force sensor uses a strain gauge or a load sensor, which is installed at the connection between the variable pitch bearing outer ring and the blade, for measuring the static and dynamic loads borne by the variable pitch bearing; the speed sensor uses an optical speed sensor or a capacitive speed sensor, which is arranged between the blade and the variable pitch bearing outer ring, for measuring the variable pitch speed; the vibration sensor uses an accelerometer or an acceleration vibration sensor, which is arranged at the bearing area and non-bearing area of the inner ring of the variable pitch bearing of the wind turbine generator, for capturing vibration signals during the operation of the variable pitch bearing.

[0007] Further, the mapping relationship established by the database module is obtained based on the recorded dynamic response data of the variable pitch bearing in healthy and different fault states under different variable pitch speeds and load states.

[0008] Further, the intelligent algorithm used by the performance prediction module includes a blind source separation and a manifold learning algorithm, which extracts the characteristics of the variable pitch bearing under different working conditions and faults by training the historical operation data and real-time sensor data, and establishes a variable pitch bearing fault prediction model.

[0009] Further, the multi-level threshold of the safety performance evaluation module includes a warning threshold, a maintenance threshold and a component replacement threshold; when the monitoring data approaches the warning threshold, the variable pitch bearing state is analyzed and judged; when the monitoring data exceeds the maintenance threshold, the system triggers an alarm and locates the fault position by using the prediction model; when the monitoring data reaches or exceeds the component replacement threshold, the system determines that there is a serious failure form inside the bearing and locates it and issues a safety warning.

[0010] Further, in the hardware module, the selection of the sensors is as follows: the force sensor is suitable for a range of 0.5-10000kN, the accelerometer has a range of 0-5g and a frequency range of 0-100Hz, and the optical speed sensor has a measurement range of 0-60rpm; the data acquisition equipment includes a data acquisition card supporting multiple input channels, the data processing unit uses a high-performance embedded computing platform or an industrial computer, and the communication module is equipped with a wireless transmission module.

[0011] Further, in the software module, the data acquisition and monitoring module is used to acquire sensor data in real time and perform preliminary processing and storage; the data analysis and decision module includes an algorithm library and an analysis engine, the algorithm library supports safety evaluation, early warning and decision algorithms, and the analysis engine is used to perform data analysis, model prediction and fault diagnosis; the user interface module provides an intuitive graphical interface to display real-time monitoring data, historical analysis results and warning information; the report generation module automatically generates safety evaluation reports and maintenance recommendations.

[0012] Further, the system architecture of the software module is divided into front end and back end, the front end supports multiple terminals, and the back end is responsible for data processing and storage; the module division includes data acquisition module, data analysis module, user management module and report generation module; the development technology front end uses WebStorm, the back end uses Python, the database uses SQLServer, and the machine learning model uses support vector machine, manifold learning or deep learning algorithm for training and prediction.

[0013] Further, it also includes a core algorithm integration module, which includes a running state rapid evaluation algorithm, an early warning algorithm and a maintenance decision algorithm; the running state rapid evaluation algorithm is based on real-time sensor data and uses a multi-level threshold method constructed by time domain characteristic values to evaluate the running state of the variable pitch bearing; the early warning algorithm automatically identifies potential faults based on machine learning models and fault prediction models; the maintenance decision algorithm considers the running state, performance degradation and maintenance history, and optimizes the maintenance decision process through risk assessment and cost-benefit analysis.

[0014] Further, in the core algorithm integration module, each core algorithm is developed as an independent module, data interaction and calling between modules are performed through a unified interface, and testing, verification and iterative optimization are performed.

[0015] Beneficial Effects: This invention proposes an online monitoring and service safety early warning system for the pitch bearing of wind turbine generators. In terms of monitoring, the system employs multiple sensors to perform multi-dimensional online monitoring of the load, speed, and vibration acceleration of the pitch bearing. Compared to traditional single-sensor systems, this system can more comprehensively capture bearing operating status information, greatly improving early fault identification capabilities and timely detection of potential minor damage. In data processing, the database module establishes a load-speed-vibration characteristic mapping relationship based on measured data. The performance prediction module utilizes intelligent algorithms such as blind source separation and manifold learning, combined with historical and real-time data training, to deeply mine the value of monitoring data and construct a precise fault characteristic model, making the operational status assessment more scientific and accurate. In the safety early warning stage, the safety performance evaluation module sets multi-level thresholds, dynamically adjusting them according to different service environments and operating conditions. Combined with the early warning algorithm in the core algorithm integration module, it automatically identifies potential faults based on machine learning and fault prediction models, effectively avoiding false alarms and missed alarms, and providing reliable decision-making basis for maintenance personnel. Furthermore, the reasonable selection of hardware modules and the scientific architecture design of software modules ensure the efficiency and stability of data acquisition, transmission, analysis, and display. This system comprehensively improves the accuracy, reliability, and intelligence of pitch bearing monitoring, reduces the risk of unit failure and operation and maintenance costs, effectively ensures the safe and stable operation of wind turbine generators, and improves overall power generation efficiency. Attached Figure Description

[0016] Figure 1 This is a system composition diagram of the present invention; Figure 2 This is a flowchart of the system operation of the present invention. Detailed Implementation

[0017] It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. The following describes the application in further detail with reference to the accompanying drawings and specific embodiments.

[0018] like Figure 1 As shown, the online monitoring and service safety early warning system for the pitch bearing of a wind turbine generator includes an online monitoring module, a database module, a performance prediction module, a safety performance evaluation module, a hardware module, and a software module. The online monitoring module is used to monitor the load, speed, and vibration acceleration of the pitch bearing online based on the daily operation and maintenance data collected from the wind turbine generator pitch bearing, and transmits the monitoring data to the database module. Specifically, the online monitoring module serves as the basis for the system to perceive the operating state of the variable pitch bearing, and realizes real-time monitoring of key parameters through various sensors. Among them, the force sensor selects a strain gauge or a load sensor, which is installed at the connection between the variable pitch bearing outer ring and the blade. Its applicable range is 0.5-10000kN, which can accurately measure the static and dynamic load borne by the variable pitch bearing and obtain the bearing stress condition. The speed sensor adopts an optical speed sensor or a capacitive speed sensor, which is arranged between the blade and the variable pitch bearing outer ring, with a measurement range of 0-60rpm, which can monitor the variable pitch speed of the blade in real time. The vibration sensor adopts an accelerometer or an acceleration vibration sensor, which is distributed in the bearing load area and non-load area of the variable pitch bearing of the wind turbine generator set, with a range of 0-5g and a frequency range of 0-100Hz, which can effectively capture the vibration signals in the bearing operation process. The data collected by these sensors is transmitted to the database module for storage and analysis through the data acquisition equipment in the hardware module.

[0019] The implementation of this module provides the system with comprehensive and accurate raw data. The multi-dimensional monitoring method changes the limitations of traditional single sensor monitoring and can more timely and comprehensively reflect the operating state of the variable pitch bearing. Whether it is a small vibration anomaly or a sudden change in load, it can be effectively monitored, making early fault identification possible. Through continuous monitoring of these key parameters, the operation and maintenance personnel can master the operation trend of the variable pitch bearing and provide reliable data support for subsequent performance prediction and safety evaluation, thereby ensuring the stable operation of the wind turbine generator set.

[0020] The database module is used to establish a mapping relationship between load-speed-vibration characteristics based on measured data, and store safety performance data corresponding to different service environments and operating conditions; Specifically, the database module plays an important role in data storage and relationship modeling in the system. It establishes a mapping relationship between load-speed-vibration characteristics based on measured data. This relationship is established by recording the dynamic response data of the variable pitch bearing in healthy and different fault states under different variable pitch speeds and load states. The database not only stores real-time monitoring data, but also contains safety performance data corresponding to different service environments and operating conditions, forming a complete variable pitch bearing data resource library. These data provide rich historical references for the analysis and decision-making of the system. Through the integration and processing of a large amount of data, the operation law of the variable pitch bearing can be deeply mined.

[0021] In terms of implementation, the database module employs scientific data management methods, classifying, storing, and indexing data to ensure efficient data retrieval and querying. The established mapping relationships provide crucial theoretical support for the performance prediction and safety performance evaluation modules, enabling the system to analyze and predict the operating status of the pitch bearing based on data models. By continuously updating and improving the data in the database, the system can adapt to the monitoring needs of pitch bearings under different operating conditions and environments, providing a solid data foundation for the intelligent operation of the entire system and improving its accuracy and reliability.

[0022] The performance prediction module is used to quickly predict the service status of the pitch bearing using intelligent algorithms. Specifically, the performance prediction module is the core component of the system for predicting pitch bearing failures. It employs intelligent algorithms such as blind source separation and manifold learning to rapidly predict the service status of the pitch bearing. These algorithms, through deep training on historical operating data and real-time sensor data, are able to extract characteristics of the pitch bearing under different operating conditions and failures from complex data. By establishing a pitch bearing failure prediction model, this module can extrapolate the future operating status of the bearing and identify potential failure hazards in advance. The application of algorithms avoids the subjectivity and limitations of traditional experience-based judgments, and the data- and model-based prediction method is more scientific and accurate.

[0023] During implementation, the performance prediction module needs to work closely with the database module, acquiring a large amount of historical and real-time monitoring data from the database as training samples. As data accumulates and is updated, the algorithm can continuously optimize the model, improving prediction accuracy and reliability. By accurately predicting the operating status of the pitch bearing, maintenance personnel can develop maintenance plans in advance, avoiding failures, reducing downtime and maintenance costs, and ensuring the stable and efficient operation of wind turbine generators. This is of great significance for improving the economic benefits and safety of wind power generation.

[0024] The safety performance evaluation module is used to classify and evaluate the condition of the pitch bearing based on the mapping relationship between the load-speed-vibration characteristics and the multi-level threshold pitch bearing service safety performance evaluation method. Specifically, the safety performance evaluation module, based on the mapping relationship between load, speed, and vibration characteristics, employs a multi-level threshold method to classify and assess the condition of pitch bearings. These multi-level thresholds include warning thresholds, maintenance thresholds, and component replacement thresholds. When monitoring data approaches the warning threshold, the system analyzes the pitch bearing condition and alerts maintenance personnel to monitor its operation. When monitoring data exceeds the maintenance threshold, the system triggers an alarm and locates the fault using a predictive model. When monitoring data reaches or exceeds the component replacement threshold, the system determines that a severe failure mode has occurred within the bearing, locates it, and issues a safety warning. This tiered assessment method allows for different levels of response based on the severity of the bearing condition, improving the accuracy and relevance of warnings.

[0025] The implementation of this module relies on accurate mapping relationships and reasonable threshold settings. Through the analysis and research of a large amount of data, combined with actual operating experience of pitch bearings, specific values ​​for each threshold level are determined. In actual operation, the safety performance evaluation module receives data transmitted from the online monitoring module in real time and compares it with the thresholds to quickly determine the bearing's safety status. This scientific evaluation method avoids the false alarms and missed alarms of traditional fixed threshold early warning methods, providing maintenance personnel with a reliable basis for decision-making, effectively ensuring the safe operation of wind turbine pitch bearings and reducing the risk of equipment failure.

[0026] The hardware module includes sensors, data acquisition devices, data processing units, and communication modules that meet the selection requirements; Specifically, the hardware module is the physical foundation for the system's data acquisition, transmission, and processing. It includes sensors that meet selection requirements, data acquisition equipment, a data processing unit, and a communication module. Sensor selection strictly adheres to the pitch bearing monitoring requirements; for example, the force sensor has an application range of 0.5-10000kN, the acceleration sensor has a range of 0-5g and a frequency range of 0-100Hz, and the photoelectric speed sensor has a measurement range of 0-60rpm, ensuring accurate acquisition of key parameters of the pitch bearing's operation. The data acquisition equipment uses a multi-input data acquisition card, capable of simultaneously receiving data from multiple sensors and performing preliminary signal conditioning and digitization. The data processing unit uses a high-performance embedded computing platform or industrial computer, possessing powerful data processing capabilities, enabling real-time analysis and calculation of the acquired data. The communication module is equipped with a wireless transmission module to achieve remote data transmission, ensuring timely data delivery to the server or maintenance terminal.

[0027] In practical applications, the various components of the hardware module work closely together to form a complete data acquisition and transmission system. Sensors convert physical signals into electrical signals. After the data acquisition device collects and processes the signals, they are transmitted to the data processing unit for further analysis. Finally, the processed data is transmitted to other modules of the system or external terminals via the communication module. The stable operation of the hardware modules is a prerequisite for the normal operation of the entire system. Their appropriate selection and configuration ensure the accuracy of data acquisition, the stability of transmission, and the efficiency of processing, providing a solid hardware guarantee for the reliable operation of the system.

[0028] The software modules include a data acquisition and monitoring module, a data analysis and decision-making module, a user interface module, and a report generation module.

[0029] Specifically, the software module, as the core control and application part of the system, encompasses a data acquisition and monitoring module, a data analysis and decision-making module, a user interface module, and a report generation module, employing a front-end and back-end architecture design. The data acquisition and monitoring module is responsible for acquiring sensor data in real time, performing preliminary processing and storage to ensure data integrity and accuracy. The data analysis and decision-making module includes an algorithm library and an analysis engine. The algorithm library supports safety assessment, early warning, and decision-making algorithms, while the analysis engine performs data analysis, model prediction, and fault diagnosis. By calling algorithms from the performance prediction and safety performance evaluation modules, it achieves in-depth analysis and decision support for the pitch bearing's operating status. The user interface module provides an intuitive graphical interface, allowing maintenance personnel to view real-time monitoring data, historical analysis results, and early warning information, enabling human-computer interaction. The report generation module automatically generates safety assessment reports and maintenance recommendations, providing written evidence for maintenance decisions.

[0030] The software modules were developed using appropriate technologies and algorithm models. The front-end was developed using WebStorm to ensure a visually appealing and interactive interface; the back-end used Python, leveraging its powerful data processing and algorithm implementation capabilities; SQL Server was used for the database to ensure efficient data storage and management; and machine learning models employed support vector machines, manifold learning, or deep learning algorithms for training and prediction. All modules interact and call functions through a unified interface. Rigorous testing, verification, and iterative optimization ensured the stability and reliability of the software modules. The software modules enabled comprehensive management and analysis of pitch bearing monitoring data, providing maintenance personnel with a convenient and efficient operating platform, and enhancing the system's intelligence and application value.

[0031] Preferably, in the online monitoring module, the force sensor is a strain gauge or load sensor, installed at the connection between the outer ring of the pitch bearing and the blade, used to measure the static and dynamic loads borne by the pitch bearing; the speed sensor is a photoelectric speed sensor or a capacitive speed sensor, arranged between the blade and the outer ring of the pitch bearing, used to measure the pitch speed; the vibration sensor is an accelerometer or an acceleration-vibration sensor, arranged in the load-bearing area and non-load-bearing area of ​​the inner ring of the wind turbine pitch bearing, used to capture vibration signals during the operation of the pitch bearing.

[0032] Specifically, the force sensors, employing strain gauges or load cells, are installed at the connection between the outer ring of the pitch bearing and the blade, with a measurement range of 0.5-10000kN, accurately acquiring the magnitude and changes of bearing force. The speed sensors, using photoelectric or capacitive speed sensors, are positioned between the blade and the outer ring of the pitch bearing, enabling real-time monitoring of the blade's pitch speed within a measurement range of 0-60rpm. The vibration sensors, employing accelerometers or acceleration-vibration sensors, are distributed in the load-bearing and non-load-bearing areas of the bearing's inner ring, with a measurement range of 0-5g and a frequency range of 0-100Hz, effectively capturing bearing vibration signals. This implementation method, through targeted sensor selection and precise placement, achieves comprehensive monitoring of multiple parameters of the pitch bearing, providing accurate and reliable raw data for the system and laying the foundation for subsequent condition analysis and fault diagnosis.

[0033] Preferably, the mapping relationship established by the database module is obtained by recording the dynamic response data of the pitch bearing under healthy and different fault conditions at different pitch speeds and loads.

[0034] Specifically, the database module establishes the load-speed-vibration characteristic mapping relationship based on the dynamic response data of the pitch bearing under different pitch speeds and load conditions, including both healthy and fault states. This data is then analyzed and processed by the system to form the mapping relationship. This relationship integrates bearing operating data under multiple conditions and states, storing not only real-time monitoring data but also safety performance data corresponding to different service environments and operating conditions, forming a complete data resource library. This mapping relationship provides core theoretical support for the performance prediction and safety performance evaluation modules, enabling the system to scientifically analyze and predict bearing operating conditions based on data models, thus enhancing the accuracy and reliability of system decisions.

[0035] Preferably, the intelligent algorithm used in the performance prediction module includes blind source separation and manifold learning algorithms. It is trained using historical operating data and real-time sensor data to extract the characteristics of the pitch bearing under different operating conditions and faults, and to establish a pitch bearing fault prediction model.

[0036] Specifically, the performance prediction module employs intelligent algorithms and their working principles. This module utilizes blind source separation and manifold learning algorithms, using historical operating data and real-time sensor data as training samples to deeply mine data features. Through data training and analysis, it extracts key features of the pitch bearing under different operating conditions and faults, constructs a fault prediction model, and achieves rapid prediction of the bearing's service status. Compared to traditional experience-based judgment, the prediction method based on algorithms and data models eliminates subjectivity and limitations. As data accumulates and the algorithm continuously optimizes the model, prediction accuracy can be improved, helping maintenance personnel to identify potential faults in advance, rationally formulate maintenance plans, and reduce unit failure risks and maintenance costs.

[0037] Preferably, the multi-level thresholds of the safety performance evaluation module include a warning threshold, a maintenance threshold, and a component replacement threshold; when the monitoring data approaches the warning threshold, the state of the pitch bearing is assessed; when the monitoring data exceeds the maintenance threshold, the system triggers an alarm and locates the fault location through a prediction model; when the monitoring data reaches or exceeds the component replacement threshold, the system determines that a serious failure mode has occurred inside the bearing, locates it, and issues a safety warning.

[0038] Specifically, the safety performance evaluation module features multi-level threshold settings and functions. This module sets three levels of thresholds: early warning, maintenance, and component replacement. When monitored data approaches the early warning threshold, the system activates an analysis mechanism, prompting maintenance personnel to monitor the bearing status. When data exceeds the maintenance threshold, an alarm is triggered, and a predictive model is used to locate the fault. When data reaches or exceeds the component replacement threshold, the system determines that the bearing has suffered a severe failure, accurately locates the fault, and issues a safety warning. This tiered evaluation model provides differentiated responses based on the severity of the bearing condition, overcoming the shortcomings of traditional fixed-threshold early warning systems, avoiding false alarms and missed alarms, providing maintenance personnel with reliable decision-making support, ensuring the safe operation of the pitch bearing, and reducing unit downtime losses due to bearing failure.

[0039] Preferably, in the hardware module, the sensors are selected as follows: the force sensor has an applicable range of 0.5-10000kN, the acceleration sensor has a measurement range of 0-5g and a frequency range of 0-100Hz, and the photoelectric speed sensor has a measurement range of 0-60rpm; the data acquisition device includes a data acquisition card that supports multiple inputs, the data processing unit adopts a high-performance embedded computing platform or an industrial computer, and the communication module is equipped with a wireless transmission module.

[0040] Specifically, the selection and function of each component in the hardware module are carefully considered. Sensor selection is strictly tailored to the pitch bearing monitoring requirements: force sensors with a range of 0.5-10000kN, accelerometers with a range of 0-5g and a frequency range of 0-100Hz, and photoelectric speed sensors with a measurement range of 0-60rpm, ensuring accurate data acquisition. The data acquisition equipment uses a multi-input data acquisition card to achieve synchronous acquisition and preliminary processing of data from multiple sensors. The data processing unit uses a high-performance embedded computing platform or industrial computer, possessing powerful real-time data processing capabilities. The communication module is equipped with a wireless transmission module to ensure rapid remote data transmission. All hardware components work together to form a stable data acquisition, transmission, and processing system, providing a solid physical foundation for system operation.

[0041] Preferably, in the software modules, the data acquisition and monitoring module is used to acquire sensor data in real time, and perform preliminary processing and storage; the data analysis and decision-making module includes an algorithm library and an analysis engine, the algorithm library supports safety assessment, early warning and decision-making algorithms, and the analysis engine is used to perform data analysis, model prediction and fault diagnosis; the user interface module provides an intuitive graphical interface to display real-time monitoring data, historical analysis results and early warning information; and the report generation module automatically generates safety assessment reports and maintenance suggestions.

[0042] Specifically, the functions and roles of each submodule in the software are as follows: the data acquisition and monitoring module is responsible for acquiring and initially processing stored sensor data in real time, ensuring data integrity; the data analysis and decision-making module includes an algorithm library and an analysis engine. The algorithm library integrates safety assessment, early warning, and decision-making algorithms, while the analysis engine calls algorithms from the performance prediction and safety performance evaluation modules to achieve in-depth analysis and decision support for bearing operating status; the user interface module displays monitoring data, analysis results, and early warning information in a graphical interface, facilitating human-computer interaction; and the report generation module automatically generates safety assessment reports and maintenance recommendations, providing written references for operation and maintenance decisions. These submodules work collaboratively to achieve full-process management and application of monitoring data, improving the system's intelligence level and operation and maintenance efficiency.

[0043] Preferably, the system architecture of the software module is divided into a front-end and a back-end. The front-end supports multiple terminals, and the back-end is responsible for data processing and storage. The module division includes a data acquisition module, a data analysis module, a user management module, and a report generation module. The development technology uses WebStorm for the front-end, Python for the back-end, SQL Server for the database, and support vector machines, manifold learning, or deep learning algorithms for training and prediction for the machine learning model.

[0044] Specifically, regarding the software module architecture and development technology, the system adopts a front-end and back-end architecture. The front-end is developed based on WebStorm to ensure a beautiful interface, smooth interaction, and support for multi-terminal access. The back-end uses Python, leveraging its powerful data processing and algorithm implementation capabilities. SQL Server is selected as the database to ensure efficient data storage and management. Machine learning models are trained and predicted using support vector machines, manifold learning, or deep learning algorithms. The modules are divided into data acquisition, analysis, user management, and report generation modules. Data interaction and function calls are realized through a unified interface. Rigorous testing and optimization ensure the stability and reliability of the software modules, providing technical support for the implementation of system functions and improving user experience and system application value.

[0045] Preferably, it also includes a core algorithm integration module, which includes a rapid operating status assessment algorithm, an early warning algorithm, and a maintenance decision algorithm. The rapid operating status assessment algorithm evaluates the operating status of the pitch bearing based on real-time sensor data and uses a multi-level threshold method constructed from time-domain feature values. The early warning algorithm automatically identifies potential faults based on machine learning models and fault prediction models. The maintenance decision algorithm comprehensively considers operating status, performance degradation, and maintenance history, and optimizes the maintenance decision-making process through risk assessment and cost-benefit analysis.

[0046] Specifically, the core algorithm integration module comprises three algorithms: a rapid operating status assessment algorithm, an early warning algorithm, and a maintenance decision-making algorithm. The rapid operating status assessment algorithm uses real-time sensor data and a multi-level threshold method constructed from time-domain feature values ​​to evaluate the bearing's operating status. The early warning algorithm automatically identifies potential faults using machine learning and fault prediction models. The maintenance decision-making algorithm comprehensively considers the bearing's operating status, performance degradation, and maintenance history, optimizing maintenance decisions through risk assessment and cost-benefit analysis. Each algorithm is developed independently yet works collaboratively, interacting through a unified interface. Through testing, verification, and iterative optimization, the system can quickly and accurately assess bearing status, provide timely fault warnings, and formulate scientific maintenance decisions, thereby improving the system's operation and maintenance management level and economic efficiency.

[0047] Preferably, in the core algorithm integration module, each core algorithm is developed as an independent module, and data interaction and invocation between modules are carried out through a unified interface, and testing, verification and iterative optimization are performed.

[0048] Specifically, the core algorithms (rapid operational status assessment algorithm, early warning algorithm, and maintenance decision algorithm) in the core algorithm integration module adopt an independent modular development model. A unified data interface standard enables efficient data interaction and function calls between modules, ensuring collaborative operation between algorithms. During implementation, each algorithm undergoes rigorous unit testing, integration testing, and system testing after development to verify its functional accuracy and stability. Iterative optimization is performed based on actual operational data feedback to continuously improve algorithm performance and prediction accuracy. This implementation method avoids chaotic coupling between algorithms, facilitates maintenance and upgrades, and enables the system to continuously, quickly, and accurately assess the operating status of the pitch bearing, provide timely and precise early warnings of potential faults, and formulate scientific maintenance decisions based on a comprehensive consideration of risk and cost. This effectively improves the intelligence and efficiency of system operation and maintenance management, and reduces the operation and maintenance costs and failure risks of wind turbine generators.

[0049] The online monitoring and service safety early warning system for the pitch bearing condition of wind turbine generator sets operates through the above steps, including: S1. Real-time acquisition of multi-dimensional data: Through the force sensor (installed at the connection between the outer ring of the pitch bearing and the blade), the speed sensor (arranged between the blade and the outer ring), and the vibration sensor (distributed in the load-bearing area and non-load-bearing area of ​​the inner ring of the bearing) of the online monitoring module, the load, speed, and vibration acceleration data of the pitch bearing are acquired in real time and transmitted to the database module through the data acquisition equipment of the hardware module.

[0050] S2. Data Modeling and Storage: The database module establishes a load-speed-vibration characteristic mapping relationship based on the dynamic response data of the health and fault status of the pitch bearing under different working conditions (pitch speed, load status), and stores safety performance data for multiple scenarios to form a standardized data model.

[0051] S3. Intelligent Prediction of Service Status: The performance prediction module uses intelligent algorithms such as blind source separation and manifold learning to train on historical operating data and real-time sensor data in the database, extract the characteristics of the pitch bearing under different operating conditions and faults, build a fault prediction model, and realize rapid prediction of service status.

[0052] S4. Safety Performance Grading Assessment: The safety performance evaluation module uses the load-speed-vibration characteristic mapping relationship and adopts a three-level threshold standard of early warning, maintenance, and component replacement to grade and assess the condition of the pitch bearing and determine whether the data has reached different threshold ranges.

[0053] S5. Fault Warning and Decision Generation: When the monitoring data triggers the threshold, the warning algorithm in the core algorithm integration module automatically identifies potential faults and, combined with the operating status rapid assessment algorithm, locates the fault location; the maintenance decision algorithm integrates the operating status, performance degradation, and maintenance history, and generates maintenance decision suggestions through risk assessment and cost-benefit analysis.

[0054] S6. Information Display and Report Output: The user interface module of the software module displays real-time monitoring data, status assessment results and early warning information in a graphical interface. The report generation module automatically generates security assessment reports and maintenance suggestions, which are transmitted to the operation and maintenance terminal through the communication module.

[0055] The system has been optimized and upgraded in terms of monitoring methods, data processing, and early warning decision-making, providing a reliable guarantee for the safe operation of the pitch bearing of wind turbine generators.

[0056] This system's online monitoring module employs multiple sensors working in tandem. Force sensors are installed at the connection between the pitch bearing's outer ring and the blade to accurately measure static and dynamic loads; speed sensors are positioned between the blade and the pitch bearing's outer ring to acquire real-time pitch speed; vibration sensors are distributed throughout the bearing's inner ring's load-bearing and non-load-bearing areas, capturing vibration signals from all directions. This multi-dimensional data acquisition significantly enhances the ability to identify early-stage pitch bearing faults, enabling timely detection of potential minor damage. It overcomes the low accuracy of traditional monitoring methods, providing a rich and accurate data foundation for bearing condition assessment.

[0057] The system's database module establishes a load-speed-vibration characteristic mapping relationship based on the dynamic response data of pitch bearing health and fault states under different operating conditions. The performance prediction module employs intelligent algorithms such as blind source separation and manifold learning, trained with historical and real-time data, to extract features under different operating conditions and faults, constructing an accurate fault prediction model. This approach deeply mines the effective information in the monitoring data, making the assessment of pitch bearing operating conditions more scientific and accurate, changing the previous situation of simple data processing and lack of basis for assessment.

[0058] The system's safety performance evaluation module sets multi-level thresholds for early warning, maintenance, and component replacement, which can be dynamically adjusted according to different service environments and operating conditions. When monitoring data triggers different thresholds, the system uses the core algorithm integration module's rapid operational status assessment algorithm, early warning algorithm, and maintenance decision algorithm to accurately assess the pitch bearing's condition, locate faults, and optimize maintenance decisions. This mechanism not only effectively avoids false alarms and missed alarms but also provides maintenance personnel with a scientific and reasonable basis for decision-making. Furthermore, through risk assessment and cost-benefit analysis, it reduces maintenance costs and improves the safety and economy of wind turbine operation. If you would like to summarize other aspects of the system or have modification needs, please feel free to contact me.

[0059] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," "link," and "fix" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0060] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A wind turbine generator pitch bearing condition online monitoring and service safety early warning system, characterized in that, It includes an online monitoring module, a database module, a performance prediction module, a security performance evaluation module, a hardware module, and a software module; The online monitoring module is used to monitor the load, speed, and vibration acceleration of the pitch bearing online using multiple sensors based on the operational data collected during the daily operation and maintenance of the wind turbine pitch bearing, and transmits the monitoring data to the database module. The database module is used to establish a mapping relationship between load, speed, and vibration characteristics based on the measured data, and to store the corresponding safety performance data under different service environments and operating conditions. The performance prediction module is used to quickly predict the service status of the pitch bearing using intelligent algorithms. The safety performance evaluation module is used to classify and evaluate the condition of the pitch bearing based on the mapping relationship between the load-speed-vibration characteristics and the multi-level threshold pitch bearing service safety performance evaluation method. The hardware module includes sensors, data acquisition equipment, data processing unit and communication module that meet the selection requirements. The software module includes data acquisition and monitoring module, data analysis and decision-making module, user interface module and report generation module.

2. The online monitoring and service safety early warning system for the pitch bearing condition of a wind turbine generator set according to claim 1, characterized in that, In the online monitoring module, the force sensor is a strain gauge or load sensor, installed at the connection between the outer ring of the pitch bearing and the blade, used to measure the static and dynamic loads borne by the pitch bearing; the speed sensor is a photoelectric speed sensor or a capacitive speed sensor, arranged between the blade and the outer ring of the pitch bearing, used to measure the pitch speed; the vibration sensor is an accelerometer or an acceleration-vibration sensor, arranged in the load-bearing area and non-load-bearing area of ​​the inner ring of the wind turbine pitch bearing, used to capture vibration signals during the operation of the pitch bearing.

3. The online monitoring and service safety early warning system for the pitch bearing condition of a wind turbine generator set according to claim 1, characterized in that, The mapping relationship established by the database module is based on the dynamic response data of the pitch bearing under healthy and different fault conditions, recorded under different pitch speeds and load conditions.

4. The online monitoring and service safety early warning system for the pitch bearing condition of a wind turbine generator set according to claim 1, characterized in that, The performance prediction module employs intelligent algorithms including blind source separation and manifold learning algorithms. It is trained using historical operating data and real-time sensor data to extract the characteristics of the pitch bearing under different operating conditions and faults, and establishes a pitch bearing fault prediction model.

5. The online monitoring and service safety early warning system for the pitch bearing condition of a wind turbine generator set according to claim 1, characterized in that, The multi-level thresholds of the safety performance evaluation module include early warning thresholds, maintenance thresholds, and component replacement thresholds. When the monitoring data approaches the early warning threshold, the condition of the pitch bearing is assessed. When the monitoring data exceeds the maintenance threshold, the system triggers an alarm and locates the fault location using a prediction model. When the monitoring data reaches or exceeds the component replacement threshold, the system determines that a serious failure mode has occurred inside the bearing, locates it, and issues a safety warning.

6. The online monitoring and service safety early warning system for the pitch bearing condition of a wind turbine generator set according to claim 1, characterized in that, The hardware module includes the following sensor selections: a force sensor with an application range of 0.5-10000kN, an acceleration sensor with a measurement range of 0-5g and a frequency range of 0-100Hz, and a photoelectric speed sensor with a measurement range of 0-60rpm; the data acquisition device includes a data acquisition card that supports multiple inputs, the data processing unit adopts a high-performance embedded computing platform or an industrial computer, and the communication module is equipped with a wireless transmission module.

7. The online monitoring and service safety early warning system for the pitch bearing condition of a wind turbine generator set according to claim 1, characterized in that, In the software module, the data acquisition and monitoring module is used to acquire sensor data in real time and perform preliminary processing and storage; the data analysis and decision-making module includes an algorithm library and an analysis engine. The algorithm library supports safety assessment, early warning and decision-making algorithms, and the analysis engine is used to perform data analysis, model prediction and fault diagnosis. The user interface module provides an intuitive graphical interface that displays real-time monitoring data, historical analysis results, and early warning information; the report generation module automatically generates security assessment reports and maintenance recommendations.

8. The online monitoring and service safety early warning system for the pitch bearing condition of a wind turbine generator set according to claim 1, characterized in that, The system architecture of the software module is divided into a front-end and a back-end. The front-end supports multiple terminals, while the back-end is responsible for data processing and storage. The module division includes a data acquisition module, a data analysis module, a user management module, and a report generation module. The development technology uses WebStorm for the front-end and Python for the back-end. The database uses SQL Server, and the machine learning model uses support vector machines, manifold learning, or deep learning algorithms for training and prediction.

9. The online monitoring and service safety early warning system for the pitch bearing condition of a wind turbine generator set according to claim 1, characterized in that, It also includes a core algorithm integration module, which includes a rapid operation status assessment algorithm, an early warning algorithm, and a maintenance decision algorithm. The rapid operation status assessment algorithm evaluates the operation status of the pitch bearing based on real-time sensor data and uses a multi-level threshold method constructed from time-domain feature values. The early warning algorithm automatically identifies potential faults based on machine learning models and fault prediction models. The maintenance decision algorithm comprehensively considers the operation status, performance degradation, and maintenance history, and optimizes the maintenance decision process through risk assessment and cost-benefit analysis.

10. The online monitoring and service safety early warning system for the pitch bearing condition of a wind turbine generator set according to claim 9, characterized in that, In the core algorithm integration module, each core algorithm is developed as an independent module. Data interaction and invocation between modules are carried out through a unified interface, and testing, verification and iterative optimization are performed.

Images