Yaw anomaly detection method, device and equipment for wind driven generator and medium

Abstract

The invention provides a yaw anomaly detection method, device, equipment and medium for a wind driven generator, and the method comprises the steps: collecting sensor parameters of the wind driven generator in real time, and carrying out the preprocessing of the sensor parameters as operation data; building a digital twinborn model, namely building a physical model of a yaw system based on a mechanical structure, a dynamic principle and an electrical characteristic of the wind driven generator, and building a data driving model by adopting machine learning and deep learning algorithms based on operation data of the wind driven generator; using the trained machine learning model to classify and predict the operation data; inputting the operation data into the digital twinborn model to simulate the actual operation state of the wind turbine generator, obtaining a predicted value, comparing the predicted value with an actual measurement value, and when an error exceeds a preset threshold value, detecting that the wind turbine generator has yaw abnormity, the problem that the service life of equipment is damaged due to the fact that a wind turbine generator yaw control system cannot accurately obtain the real operation state of a yaw system of a wind turbine generator is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of wind turbine yaw detection, and in particular to a method, device, equipment and medium for detecting abnormal yaw of a wind turbine. Background Art

[0002] Wind power, as a clean, renewable energy source, has been widely adopted worldwide. The yaw system of a wind turbine is central to capturing wind energy. Wind direction fluctuates constantly, and the yaw system precisely adjusts the nacelle to ensure the rotor is always facing the wind, thereby maximizing wind energy capture efficiency. Research has shown that poor yaw system performance can reduce wind energy capture efficiency by 10% to 30%, significantly impacting the wind turbine's power generation performance. Furthermore, the yaw system operates under complex conditions over long periods of time, and its reliability is directly related to the overall lifespan and operational safety of the wind turbine.

[0003] Currently, yaw control of wind turbines primarily relies on traditional fixed-rule control strategies, such as simple feedback control based on wind speed and direction. Although existing technologies can meet the basic operating requirements of wind turbines to a certain extent, they still face the following challenges under complex and changing environmental conditions:

[0004] 1. Single-data dependency: Traditional detection methods often rely on data from a single sensor, such as using only a wind vane to determine whether the yaw direction is accurate. However, in actual operation, yaw anomalies can be caused by a variety of factors. A single data point cannot fully reflect the true operating status of the yaw system, which can easily lead to missed detections or false detections.

[0005] 2. Flaws in Threshold Judgment: Simply setting thresholds to identify anomalies is inadequate for the complex and changing operating conditions of wind turbines. The normal operating parameter ranges for the yaw system vary across wind farm environments and operating phases. Static thresholds are difficult to cover all scenarios, and threshold judgments often fail to detect potential anomalies that gradually develop. This results in ineffective early-stage fault resolution and ultimately leads to serious equipment damage.

[0006] 3. Lack of analysis of the system's overall operating mechanism: Traditional methods fail to fully consider the complex interactions between the yaw system's mechanical structure, electrical characteristics, and dynamics. When an anomaly occurs, in-depth analysis of the root cause is impossible, and only superficial troubleshooting can be performed, making it difficult to fundamentally resolve the problem. This impacts equipment maintenance efficiency and long-term operational stability.

[0007] Therefore, it is urgent to propose a method for detecting yaw anomaly of a wind turbine to solve the problem that the yaw control system of the wind turbine cannot accurately obtain the actual operating status of the yaw system of the wind turbine, resulting in damage to the equipment life. Summary of the Invention

[0008] To overcome the problems existing in the related art, the present disclosure provides a method, device, equipment and medium for detecting yaw anomaly of a wind turbine, so as to solve the technical problem in the related art that the true operating status of the yaw system of the wind turbine cannot be accurately obtained, resulting in a damage to the equipment life.

[0009] One or more embodiments of this specification provide a method for detecting yaw anomaly of a wind turbine, comprising the following steps:

[0010] Real-time collection of wind turbine sensor parameters, which are then pre-processed as operating data;

[0011] Build a digital twin model, including a physical model of the yaw system based on the mechanical structure, dynamic principles, and electrical characteristics of the wind turbine, and use machine learning and deep learning algorithms to build a data-driven model based on the wind turbine's operating data;

[0012] Using the trained machine learning model to classify and predict the operating data;

[0013] The operating data is input into the digital twin model to simulate the actual operating status of the wind turbine generator set to obtain a predicted value, which is compared with the actual measured value. When the error exceeds a preset threshold, a yaw abnormality of the wind turbine generator is detected.

[0014] Preferably, the method further comprises the following steps:

[0015] When abnormal yaw of the wind turbine is detected, the abnormal warning mechanism is activated and abnormal information is sent;

[0016] Abnormal yaw is diagnosed and located through big data analysis technology.

[0017] Preferably, the real-time collection of sensor parameters of the wind turbine and the pre-processing thereof as operating data specifically include the following steps:

[0018] Set the sensor data collection frequency based on the dynamic characteristics of the wind turbine and the need for anomaly detection;

[0019] The operating data is transmitted to a data processing center in real time and classified and stored using distributed storage technology.

[0020] Preferably, the construction of the digital twin model includes establishing a physical model of the yaw system based on the mechanical structure, dynamic principles and electrical characteristics of the wind turbine and building a data-driven model based on the operating data of the wind turbine, using machine learning and deep learning algorithms, and specifically includes the following steps:

[0021] Use multi-body dynamics simulation technology to establish a physical model of the yaw system, simulate the operating status of the yaw system under different working conditions, and verify the accuracy of the physical model;

[0022] Use long short-term memory networks to learn historical operating data of wind turbines, establish dynamic mapping relationships between parameters, and build data-driven models;

[0023] Fusing the physical model with the data-driven model, and using prior knowledge of the physical model to constrain the training of the data-driven model;

[0024] The physical model is modified and optimized through the data-driven model.

[0025] One or more embodiments of this specification provide a device for detecting yaw anomaly in a wind turbine, including an acquisition module, a model building module, a synchronization module, and a detection module;

[0026] The acquisition module is used to collect sensor parameters of the wind turbine in real time and use them as operating data after preprocessing;

[0027] The model building module is used to build a digital twin model, including establishing a physical model of the yaw system based on the mechanical structure, dynamic principles and electrical characteristics of the wind turbine and building a data-driven model based on the operating data of the wind turbine using machine learning and deep learning algorithms;

[0028] The synchronization module is used to classify and predict the operating data using the trained machine learning model;

[0029] The detection module is used to input the operating data into the digital twin model to simulate the actual operating status of the wind turbine, obtain a predicted value, compare the predicted value with the actual measured value, and when the error exceeds a preset threshold, detect that the wind turbine has a yaw abnormality.

[0030] Preferably, it also includes an early warning and diagnosis module, which is used to activate the abnormality early warning mechanism and send abnormal information when the wind turbine is detected to have abnormal yaw;

[0031] Abnormal yaw is diagnosed and located through big data analysis technology.

[0032] Preferably, the acquisition module includes a setting unit and a storage unit;

[0033] The setting unit is used to set the data acquisition frequency of the sensor according to the dynamic characteristics and abnormality detection requirements of the wind turbine;

[0034] The storage unit is used to transmit the operating data to the data processing center in real time and adopt distributed storage technology to perform classified storage.

[0035] Preferably, the model building module includes a physical model unit, a data-driven model unit, a fusion unit and an optimization unit;

[0036] The physical model unit is used to establish a physical model of the yaw system using multi-body dynamics simulation technology, simulate the operating state of the yaw system under different working conditions, and verify the accuracy of the physical model;

[0037] The data-driven model unit is used to use a long short-term memory network to learn the historical operating data of the wind turbine, establish a dynamic mapping relationship between parameters, and build a data-driven model;

[0038] The fusion unit is used to fuse the physical model with the data-driven model, and use the prior knowledge of the physical model to constrain the training of the data-driven model;

[0039] The optimization unit is used to modify and optimize the physical model through the data-driven model.

[0040] One or more embodiments of this specification provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the above-mentioned method for detecting yaw anomaly of a wind turbine is implemented.

[0041] One or more embodiments of this specification provide a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the steps of the above-mentioned method for detecting yaw anomaly of a wind turbine are implemented.

[0042] The present invention provides a method, device, equipment and medium for detecting yaw anomaly of a wind turbine, which has the advantages of collecting sensor parameters of the wind turbine in real time, pre-processing them as operating data, collecting and pre-processing data through multiple sensors, comprehensively obtaining operating information and ensuring data quality, and providing a rich and reliable data basis for subsequent analysis; building a digital twin model, including building a physical model of the yaw system based on the mechanical structure, dynamic principles and electrical characteristics of the wind turbine and the operating data of the wind turbine, using machine learning and deep learning algorithms to build a data-driven model, combining physical characteristics to build a physical model, using algorithms to build a data-driven model based on operating data, accurately simulating system operation, mining data patterns and Improve model generalization and interpretability; use the trained machine learning model to classify and predict the operating data, and with the help of the trained machine learning model, accurately classify and predict the yaw system operating data to provide an effective basis for anomaly detection; input the operating data into the digital twin model to simulate the actual operating status of the wind turbine, obtain the predicted value, and compare the predicted value with the actual measured value. When the error exceeds the preset threshold, it is detected that the wind turbine has yaw anomaly. Use the trained digital twin model to simulate the actual operation and compare the predicted value with the measured value. It can efficiently and accurately detect yaw anomaly, assist in fault diagnosis and analysis, greatly improve detection accuracy and timeliness, effectively ensure stable operation of equipment, and improve power generation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate one or more embodiments of this specification or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0044] Figure 1 A schematic flow chart of a method for detecting yaw anomaly of a wind turbine provided in one or more embodiments of this specification;

[0045] Figure 2 A schematic structural diagram of a device for detecting yaw anomaly in a wind turbine provided in one or more embodiments of this specification;

[0046] Figure 3 A schematic diagram of the structure of a computer device provided in one or more embodiments of this specification. DETAILED DESCRIPTION

[0047] In order to enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the technical solutions in one or more embodiments of this specification will be clearly and completely described below in conjunction with the drawings in one or more embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this invention document.

[0048] The present invention will be described in detail below with reference to specific implementation methods and the accompanying drawings.

[0049] Method Example

[0050] According to an embodiment of the present invention, a method for detecting yaw anomaly of a wind turbine is provided. Figure 1 FIG. 1 is a flow chart of a method for detecting yaw anomaly of a wind turbine provided by this embodiment. The method for detecting yaw anomaly of a wind turbine according to an embodiment of the present invention includes the following steps:

[0051] S110 , collecting sensor parameters of the wind turbine in real time, and using them as operating data after pre-processing.

[0052] Specifically, various types of sensors are installed in key areas of the yaw system, such as the yaw motor, yaw bearing, wind vane, and anemometer. For example, current sensors, temperature sensors, and speed sensors are installed on the yaw motor to monitor the motor's operating current, temperature, and speed; vibration sensors are placed on the yaw bearing to detect bearing vibration; and wind vanes and anemometers are used to obtain real-time wind direction and speed data. The sensor placement must adhere to the principles of comprehensive coverage and focused monitoring to ensure accurate collection of key data reflecting the operating status of the yaw system.

[0053] S120. Build a digital twin model, including using computer-aided design (CAD) and multi-body dynamics simulation software (such as ADAMS) to create a physical model of the yaw system based on the wind turbine's mechanical structure, dynamics, and electrical characteristics. This model details key parameters such as the yaw motor's electromagnetic torque, gear ratio, yaw bearing friction torque, and nacelle's moment of inertia, as well as their interactions. Build a data-driven model using machine learning and deep learning algorithms based on wind turbine operating data, such as Supervisory Control and Data Acquisition (SCADA) system data.

[0054] S130. Classify and predict the operating data using a trained machine learning model, such as a support vector machine (SVM) or random forest (RF). First, extract features from the historical data, such as time domain features such as mean, variance, standard deviation, and crest factor, as well as frequency domain features such as frequency and amplitude. Then, use these features to train the machine learning model so that it can distinguish between normal operating conditions and abnormal operating conditions. During the actual detection process, the real-time collected data is input into the trained model, and the output of the model is used to determine whether there is a yaw anomaly.

[0055] S140: Input the operating data into the digital twin model to simulate the actual operating state of the wind turbine, obtain a predicted value, compare the predicted value with the actual measured value, and calculate the error between the two. When the error exceeds a preset threshold, a yaw anomaly is detected in the wind turbine. By using the digital twin model for comparative testing, the model's understanding of the operating mechanism of the yaw system can be fully utilized to more accurately identify anomalies and conduct a preliminary analysis of the cause of the anomaly.

[0056] Specifically, appropriate thresholds are set for each monitoring parameter based on the normal operating range of the yaw system. For example, the yaw motor's normal operating current range, upper temperature limit, and speed fluctuation range are set, as well as the yaw bearing's vibration amplitude threshold. When the sensor data exceeds the preset threshold, an abnormality alarm is triggered. This simple and intuitive method is suitable for quickly detecting obvious abnormalities, but it may not be able to detect potential, gradually developing abnormalities in a timely manner.

[0057] The method provided in this embodiment collects sensor parameters of wind turbines in real time, pre-processes them as operating data, collects and pre-processes data through multiple sensors, comprehensively obtains operating information and ensures data quality, and provides a rich and reliable data foundation for subsequent analysis; constructs a digital twin model, including a physical model of the yaw system based on the mechanical structure, dynamic principles and electrical characteristics of the wind turbine and the operating data of the wind turbine, adopts machine learning and deep learning algorithms to build a data-driven model, combines physical characteristics to build a physical model, uses algorithms to build a data-driven model based on operating data, accurately simulates system operation, mines data patterns and improves model generalization and interpretability; utilizes The trained machine learning model classifies and predicts the operating data. With the help of the trained machine learning model, the yaw system operating data is accurately classified and predicted, providing an effective basis for anomaly detection; the operating data is input into the digital twin model to simulate the actual operating status of the wind turbine group, and the predicted value is obtained. The predicted value is compared with the actual measured value. When the error exceeds the preset threshold, the wind turbine is detected to have a yaw anomaly. The trained digital twin model is used to simulate the actual operation, and the predicted value is compared with the measured value. This can efficiently and accurately detect yaw anomalies, assist in fault diagnosis and analysis, greatly improve detection accuracy and timeliness, effectively ensure stable operation of equipment, and improve power generation efficiency.

[0058] In one embodiment, the following steps are further included:

[0059] When abnormal yaw of the wind turbine is detected, the abnormal warning mechanism is activated and abnormal information is sent.

[0060] Abnormal yaw is diagnosed and located through big data analysis technology.

[0061] The method provided in this embodiment can, with the help of big data analysis technology, conduct in-depth analysis of abnormal yaw, comprehensively integrate multi-source data, accurately locate the root cause of the abnormality, provide clear direction for maintenance personnel, greatly shorten troubleshooting time, effectively improve equipment repair efficiency, and ensure that wind turbines resume normal operation as soon as possible.

[0062] In one embodiment, the real-time collection of sensor parameters of the wind turbine and the pre-processing thereof as operating data specifically include the following steps:

[0063] Set the data collection frequency of the sensor according to the dynamic characteristics of the wind turbine and the requirements of anomaly detection.

[0064] The operating data is transmitted to a data processing center in real time and classified and stored using distributed storage technology.

[0065] The method provided in this embodiment can obtain parameters of different change rates in a targeted manner, without missing critical instant data and reasonably controlling the data volume. It also uses distributed storage technology for classified storage to ensure that data is timely aggregated and stored in an orderly manner. It provides comprehensive, accurate, and easy-to-call data support for subsequent digital twin model construction, machine learning analysis, and anomaly detection, effectively ensuring the efficient implementation of the entire yaw anomaly detection process.

[0066] In one embodiment, the construction of the digital twin model includes establishing a physical model of the yaw system based on the mechanical structure, dynamic principles, and electrical characteristics of the wind turbine, and building a data-driven model based on the operating data of the wind turbine using machine learning and deep learning algorithms. Specifically, the following steps are included:

[0067] The physical model of the yaw system is established using multi-body dynamics simulation technology to simulate the operating status of the yaw system under different working conditions and verify the accuracy of the physical model.

[0068] Long short-term memory networks are used to learn the historical operating data of wind turbines, establish dynamic mapping relationships between parameters, and build data-driven models.

[0069] The physical model is integrated with the data-driven model, and the prior knowledge of the physical model is used to constrain the training of the data-driven model, thereby improving the generalization ability and interpretability of the model.

[0070] The physical model is modified and optimized through the data-driven model.

[0071] The method provided in this embodiment uses multibody dynamics simulation technology to build a physical model of the yaw system. This method accurately simulates the system's operating state under different operating conditions, verifies the model's accuracy, and provides a foundation for understanding the operating principles of the yaw system. A long-short-term memory network is used to deeply learn historical wind turbine operating data, construct a data-driven model, establish complex dynamic mapping relationships between parameters, and explore potential patterns in the data. This effectively improves the model's generalization capabilities, making it more stable in different scenarios and enhancing its interpretability, making the overall digital twin model more consistent with the actual operating conditions of the yaw system.

[0072] Device embodiment

[0073] According to an embodiment of the present invention, a device for detecting abnormal yaw of a wind turbine is provided. Figure 2 As shown, it is a structural diagram of the wind turbine yaw anomaly detection device provided in this embodiment. According to the wind turbine yaw anomaly detection device of the embodiment of the present invention, it includes an acquisition module 21, a model building module 22, a synchronization module 23 and a detection module 24.

[0074] The acquisition module 21 is used to collect sensor parameters of the wind turbine in real time and use them as operating data after pre-processing.

[0075] The model building module 22 is used to build a digital twin model, including establishing a physical model of the yaw system based on the mechanical structure, dynamic principles and electrical characteristics of the wind turbine and the operating data of the wind turbine, and using machine learning and deep learning algorithms to build a data-driven model.

[0076] The synchronization module 23 is used to classify and predict the operating data using the trained machine learning model.

[0077] The detection module 24 is used to input the operating data into the digital twin model to simulate the actual operating status of the wind turbine, obtain a predicted value, compare the predicted value with the actual measured value, and when the error exceeds a preset threshold, detect that the wind turbine has a yaw abnormality.

[0078] The device provided in this embodiment is used to collect sensor parameters of wind turbines in real time through the acquisition module 21, and use them as operating data after preprocessing. The data is collected and preprocessed by multiple sensors to comprehensively obtain operating information and ensure data quality, providing a rich and reliable data basis for subsequent analysis; the model construction module 22 constructs a digital twin model, including a physical model of the yaw system based on the mechanical structure, dynamic principles and electrical characteristics of the wind turbine and the operating data of the wind turbine, and uses machine learning and deep learning algorithms to build a data-driven model. The physical model is established in combination with physical characteristics, and the data-driven model is established based on the operating data using algorithms, so as to accurately simulate the system operation, mine data patterns and improve the generalization and interpretability of the model; synchronous Module 23 is used to classify and predict the operating data using a trained machine learning model. With the help of the trained machine learning model, the yaw system operating data is accurately classified and predicted, providing an effective basis for anomaly detection; the detection module 24 inputs the operating data into the digital twin model to simulate the actual operating status of the wind turbine, obtains a predicted value, and compares the predicted value with the actual measured value. When the error exceeds the preset threshold, it is detected that the wind turbine has a yaw anomaly. The trained digital twin model is used to simulate the actual operation, and the predicted value is compared with the measured value. It can efficiently and accurately detect yaw anomalies, assist in fault diagnosis and analysis, greatly improve the detection accuracy and timeliness, and effectively ensure the stable operation of the equipment and improve the power generation efficiency.

[0079] In one embodiment, it also includes an early warning and diagnosis module, which is used to activate the abnormal early warning mechanism and send abnormal information when abnormal yaw of the wind turbine is detected; and diagnose and locate the abnormal yaw through big data analysis technology.

[0080] The device provided in this embodiment can use big data analysis technology to conduct in-depth analysis of abnormal yaw, comprehensively integrate multi-source data, accurately locate the root cause of the abnormality, provide clear direction for maintenance personnel, greatly shorten troubleshooting time, effectively improve equipment repair efficiency, and ensure that the wind turbine resumes normal operation as soon as possible.

[0081] In one embodiment, the acquisition module 21 includes a setting unit and a storage unit.

[0082] The setting unit is used to set the data acquisition frequency of the sensor according to the dynamic characteristics and abnormality detection requirements of the wind turbine. The storage unit is used to transmit the operating data to the data processing center in real time and use distributed storage technology for classified storage.

[0083] The device provided in this embodiment can obtain parameters of different change rates in a targeted manner, without missing critical instant data and reasonably controlling the data volume. It also uses distributed storage technology for classified storage to ensure that data is timely aggregated and stored in an orderly manner. It provides comprehensive, accurate, and easy-to-call data support for subsequent digital twin model construction, machine learning analysis, and anomaly detection, effectively ensuring the efficient implementation of the entire yaw anomaly detection process.

[0084] In one embodiment, the model building module 22 includes a physical model unit, a data-driven model unit, a fusion unit, and an optimization unit.

[0085] The physical model unit is used to establish a physical model of the yaw system using multi-body dynamics simulation technology, simulate the operating state of the yaw system under different working conditions, and verify the accuracy of the physical model.

[0086] The data-driven model unit is used to use a long short-term memory network to learn the historical operating data of the wind turbine, establish a dynamic mapping relationship between parameters, and build a data-driven model.

[0087] The fusion unit is used to fuse the physical model with the data-driven model, and utilize the prior knowledge of the physical model to constrain the training of the data-driven model, thereby improving the generalization ability and interpretability of the model.

[0088] The optimization unit is used to modify and optimize the physical model through the data-driven model.

[0089] The device provided in this embodiment uses multi-body dynamics simulation technology to build a physical model of the yaw system. This can accurately simulate the system's operating state under different operating conditions, verify the model's accuracy, and provide a foundation for understanding the operating principles of the yaw system. Long-short-term memory networks are used to deeply learn historical wind turbine operating data, construct a data-driven model, establish complex dynamic mapping relationships between parameters, and explore potential patterns in the data. This effectively improves the model's generalization capabilities, making it more stable in different scenarios and enhancing its interpretability, making the overall digital twin model more consistent with the actual operating conditions of the yaw system.

[0090] The embodiment of the present invention is an apparatus embodiment corresponding to the above-mentioned method embodiment. The specific operations of the processing steps of each module can be understood by referring to the description of the method embodiment, and will not be repeated here.

[0091] like Figure 3 As shown, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which implements the wind turbine yaw anomaly detection method in the above embodiment when the computer program is executed by a processor, or implements the wind turbine yaw anomaly detection method in the above embodiment when the computer program is executed by a processor.

[0092] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0093] Each embodiment in this specification is described in a progressive manner. The same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the device or system embodiments, since they are basically similar to the method embodiments, the description is relatively simple. For the relevant parts, refer to the partial description of the method embodiments. The device and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of this embodiment. A person of ordinary skill in the art can understand and implement it without making any creative efforts.

[0094] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention, and the contents not described in detail in the specification of the present invention belong to the common knowledge of those skilled in the art.

Claims

1. A method for detecting yaw anomaly of a wind turbine, characterized in that: The following steps are involved: Real-time collection of wind turbine sensor parameters, which are then pre-processed as operating data; Build a digital twin model, including a physical model of the yaw system based on the mechanical structure, dynamic principles, and electrical characteristics of the wind turbine, and use machine learning and deep learning algorithms to build a data-driven model based on the wind turbine's operating data; Using the trained machine learning model to classify and predict the operating data; The operating data is input into the digital twin model to simulate the actual operating status of the wind turbine generator set to obtain a predicted value, which is compared with the actual measured value. When the error exceeds a preset threshold, a yaw abnormality of the wind turbine generator is detected.

2. The method for detecting yaw anomaly of a wind turbine according to claim 1, wherein: The following steps are also included: When abnormal yaw of the wind turbine is detected, the abnormal warning mechanism is activated and abnormal information is sent; Abnormal yaw is diagnosed and located through big data analysis technology.

3. The method for detecting yaw anomaly of a wind turbine according to claim 1, wherein: The real-time collection of sensor parameters of the wind turbine and the pre-processing thereof as operating data specifically include the following steps: Set the sensor data collection frequency based on the dynamic characteristics of the wind turbine and the need for anomaly detection; The operating data is transmitted to a data processing center in real time and classified and stored using distributed storage technology.

4. The method for detecting yaw anomaly of a wind turbine according to claim 1, wherein: The digital twin model is constructed by establishing a physical model of the yaw system based on the mechanical structure, dynamic principles, and electrical characteristics of the wind turbine, and building a data-driven model using machine learning and deep learning algorithms based on the wind turbine's operating data. The specific steps include: Use multi-body dynamics simulation technology to establish a physical model of the yaw system, simulate the operating status of the yaw system under different working conditions, and verify the accuracy of the physical model; Use long short-term memory networks to learn historical operating data of wind turbines, establish dynamic mapping relationships between parameters, and build data-driven models; Fusing the physical model with the data-driven model, and using prior knowledge of the physical model to constrain the training of the data-driven model; The physical model is modified and optimized through the data-driven model.

5. A wind turbine yaw anomaly detection device, characterized in that: It includes acquisition module, model building module, synchronization module and detection module; The acquisition module is used to collect sensor parameters of the wind turbine in real time and use them as operating data after preprocessing; The model building module is used to build a digital twin model, including establishing a physical model of the yaw system based on the mechanical structure, dynamic principles and electrical characteristics of the wind turbine and building a data-driven model based on the operating data of the wind turbine using machine learning and deep learning algorithms; The synchronization module is used to classify and predict the operating data using the trained machine learning model; The detection module is used to input the operating data into the digital twin model to simulate the actual operating status of the wind turbine, obtain a predicted value, compare the predicted value with the actual measured value, and when the error exceeds a preset threshold, detect that the wind turbine has a yaw abnormality.

6. The device for detecting yaw anomaly of a wind turbine according to claim 5, wherein: It also includes an early warning and diagnosis module for activating an abnormality early warning mechanism and sending abnormality information when abnormal yaw of the wind turbine is detected; Abnormal yaw is diagnosed and located through big data analysis technology.

7. The device for detecting yaw anomaly of a wind turbine according to claim 5, wherein: The acquisition module includes a setting unit and a storage unit; The setting unit is used to set the data acquisition frequency of the sensor according to the dynamic characteristics and abnormality detection requirements of the wind turbine; The storage unit is used to transmit the operating data to the data processing center in real time and adopt distributed storage technology to perform classified storage.

8. The device for detecting yaw anomaly of a wind turbine according to claim 5, wherein: The model building module includes a physical model unit, a data-driven model unit, a fusion unit and an optimization unit; The physical model unit is used to establish a physical model of the yaw system using multi-body dynamics simulation technology, simulate the operating state of the yaw system under different working conditions, and verify the accuracy of the physical model; The data-driven model unit is used to use a long short-term memory network to learn the historical operating data of the wind turbine, establish a dynamic mapping relationship between parameters, and build a data-driven model; The fusion unit is used to fuse the physical model with the data-driven model, and use the prior knowledge of the physical model to constrain the training of the data-driven model; The optimization unit is used to modify and optimize the physical model through the data-driven model.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method for detecting yaw anomaly of a wind turbine according to any one of claims 1 to 4 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method for detecting yaw anomaly of a wind turbine as claimed in any one of claims 1 to 4 are implemented.

Images