Fan fault simulation and diagnosis training system and method

Abstract

The invention discloses a fan fault simulation and diagnosis training system and method. A plurality of modules are integrated to realize comprehensive fan fault simulation, diagnosis and training functions. The virtual reality simulation environment module is responsible for creating a fan three-dimensional model and an operation environment and collecting operation data of students; the multi-modal sensor fusion module is used for collecting multi-dimensional data of the running state of the fan; the fault generation and diagnosis module receives the data, generates a simulation fault scene, performs intelligent diagnosis, and feeds back a result to the virtual reality simulation environment module for student reference; the quantum enhancement prediction and optimization module carries out fault prediction according to historical and real-time data of fan operation and provides optimization suggestions; the self-adaptive learning and evaluation module dynamically adjusts the training content and difficulty according to the performance of the trainee; the multi-scale digital twin platform simulates a fan system in an omnibearing manner and shares data with each module; the system provides a vivid training environment for trainees, and is helpful for improving fan fault processing skills.

Description

Technical Field

[0001] The invention belongs to the technical field of wind power generation, and relates to a wind turbine fault simulation and diagnosis training system and method. Background Art

[0002] As the global demand for clean energy continues to grow, wind power generation, as a clean and renewable energy form, has become increasingly prominent in the global energy structure. With the continuous advancement of wind power technology and the continuous expansion of wind farms, the complexity and operation and maintenance difficulty of wind power generation systems have also increased. This has put forward higher challenges to the professional skills and fault diagnosis capabilities of wind turbine operators.

[0003] Traditional wind turbine fault simulation and diagnosis training systems, as an important tool to improve operator skills, have played a role in training and guidance to a certain extent, but their limitations are becoming increasingly apparent. These systems mainly rely on simple computer simulations and limited physical training, making it difficult to fully and realistically reproduce complex wind turbine operating conditions and fault scenarios.

[0004] Specifically, the existing training system has obvious deficiencies in the authenticity of the simulation environment. Due to the lack of in-depth simulation of the actual wind turbine operating environment, it is difficult for operators to gain an experience close to actual operation through the training system. This results in operators being unable to make quick and accurate judgments and decisions when faced with real wind turbine failures due to a lack of sufficient response experience.

[0005] In addition, the diversity and authenticity of fault scenarios are also a major shortcoming of existing training systems. Due to technical limitations and cost considerations, existing systems can usually only simulate limited fault scenarios and cannot fully cover the various complex situations that may be encountered in actual work. This limits operators' comprehensive understanding and mastery of fault types and response strategies.

[0006] In terms of training content, the existing training system also has the problem of insufficient personalization and adaptability. Different trainees have different learning needs and backgrounds, but the existing system often adopts unified teaching content and methods, which is difficult to meet the personalized learning needs of different trainees. This not only affects the training effect, but may also cause trainees to resist the training content.

[0007] The existing training system also has obvious deficiencies in terms of the accuracy and timeliness of fault diagnosis and prediction. Due to the limitations of the simulation environment and technical restrictions, the accuracy and timeliness of fault diagnosis and prediction in the existing system often cannot meet the requirements of actual work. This affects the efficiency and safety of wind turbine operation and maintenance, and may also lead to potential safety hazards.

[0008] Finally, the existing training system is also difficult to update and expand. With the rapid development of wind power technology and the expansion of wind farms, the existing system is often difficult to update and expand in time to adapt to new training needs and technical requirements. This not only limits the scope of use and application effect of the training system, but may also cause the training content to be out of touch with actual work. Summary of the invention

[0009] The purpose of the present invention is to solve the technical problem that the prior art mainly relies on simple computer simulation and limited physical training, and it is difficult to fully and realistically reproduce complex fan operating conditions and fault scenarios, and to provide a fan fault simulation and diagnosis training system and method.

[0010] In order to achieve the above object, the present invention adopts the following technical solutions: A first aspect of the present invention provides a fan fault simulation and diagnosis training system, comprising: Virtual reality simulation environment module, used to create a three-dimensional model of the wind turbine and the operating environment, and collect trainees' operating data on the virtual reality simulation environment module; Multimodal sensor fusion module, used to collect multi-dimensional data of wind turbine operation status; The fault generation and diagnosis module is used to receive data from the virtual reality simulation environment module and the multimodal sensor fusion module, generate a simulated fault scenario based on the received data, and perform intelligent diagnosis on the fault to generate a diagnosis result; the diagnosis result and the simulated fault scenario are transmitted to the virtual reality simulation environment module for reference by the trainees; The quantum enhanced prediction and optimization module is used to predict the faults of the wind turbine based on the historical data and real-time data of the wind turbine operation, and to make optimization suggestions based on the prediction results; the prediction results and optimization suggestions are transmitted to the virtual reality simulation environment module for the trainees’ reference; Adaptive learning and assessment module, used to dynamically adjust the training content and difficulty based on the trainees’ performance in the virtual reality simulation environment module; A multi-scale digital twin platform is used to simulate wind turbine systems from micro to macro. The multi-scale digital twin platform shares data and fault generation and diagnosis modules, quantum enhanced prediction and optimization modules, and virtual reality simulation environment modules.

[0011] Furthermore, the virtual reality simulation environment module includes a high-performance graphics rendering engine, an adaptive level of detail system, an advanced physics engine, and a multi-point force feedback haptic device and a high-precision hand tracking system; The high-performance graphics rendering engine is used to dynamically adjust model details according to the user's gaze point and optimize rendering performance by cooperating with the adaptive detail level system; The adaptive level of detail system is used to dynamically adjust the level of detail of the model according to the user's gaze point in the virtual environment; The advanced physics engine is used for rigid and soft body dynamics simulation; The multi-point force feedback tactile device and the high-precision hand tracking system are used to provide tactile feedback and hand tracking.

[0012] Furthermore, it also includes a secure communication and data management center, which is used to ensure the secure transmission and storage of system data.

[0013] Furthermore, the multimodal sensor fusion module includes a micro-electromechanical system sensor network, a fiber Bragg grating strain sensing system, a millimeter wave radar array, and a multi-source heterogeneous data dynamic weight fusion algorithm based on a graph attention network; The micro-electromechanical system sensor network, fiber grating strain sensor system and millimeter wave radar array collect the operating data of the wind turbine equipment. The operating data of the wind turbine equipment is fused through a multi-source heterogeneous data dynamic weight fusion algorithm based on a graph attention network to obtain multi-dimensional data of the wind turbine operating status.

[0014] Further, the fault generation and diagnosis module includes a fault generation model, a diagnosis model, a structural causal model and a model compression module; The fault generation model generates fault data using a conditional generative adversarial network; The diagnostic model uses graph convolutional networks and long short-term memory networks to extract features from fault data to achieve fault type diagnosis; The structural causal model analyzes the causal relationship in the fault data to determine the root cause and propagation path of the fault; The model compression module is used to compress the model size.

[0015] Furthermore, the quantum enhanced prediction and optimization module includes a superconducting quantum processor, a quantum approximate optimization algorithm, a variational quantum eigensolver, and an adaptive quantum circuit design system; The quantum approximate optimization algorithm is used to optimize the layout of the wind farm; The variational quantum eigensolver is used for material fatigue analysis; The superconducting quantum processor is used to run a quantum approximate optimization algorithm and a variational quantum eigensolver; The adaptive quantum circuit design system dynamically adjusts the quantum gate sequence according to the execution of the quantum approximate optimization algorithm and the variational quantum eigensolver and the state of the superconducting quantum processor.

[0016] Furthermore, the adaptive learning and evaluation module includes a learning framework based on a model-independent meta-learning algorithm, an active learning module, a dynamic difficulty adjustment mechanism, and a knowledge graph module; The learning framework based on the model-independent meta-learning algorithm is used to process and analyze the learning data of students using the model-independent meta-learning algorithm; The active learning module is used to analyze the most valuable learning content for students; based on the results of the analysis of the most valuable learning content for students, appropriate learning resources are selected from the knowledge graph module; The dynamic difficulty adjustment mechanism is used to adjust the difficulty of the learning content according to the real-time performance of the students; The knowledge graph module uses a graph neural network to model a knowledge graph, which is used to generate a personalized learning path for students based on the association relationship of the knowledge graph.

[0017] Furthermore, the multi-scale digital twin platform includes a micro-scale model, a meso-scale model, a macro-scale model, and a scale bridging algorithm based on deep learning; The microscopic scale model uses molecular dynamics software to simulate the microscopic world; The mesoscale model uses finite element analysis software to construct the mesoworld; The macro-scale model uses a large nest simulation to simulate wind field flow; The deep learning-based scale bridging algorithm is used to achieve the transmission and integration of cross-scale information.

[0018] Furthermore, it also includes a system integration and deployment module, which adopts a distributed architecture that supports horizontal expansion.

[0019] A second aspect of the present invention provides a fan fault simulation and diagnosis training method, comprising the following steps: Collect data from trainees’ operations in the pre-established 3D wind turbine model and operating environment; Collect multi-dimensional data of fan operation status; Generate simulated fault scenarios based on the data of trainees' operations in the pre-established three-dimensional model of the wind turbine and the operating environment and the multi-dimensional data of the wind turbine's operating status; generate fault diagnosis results based on the simulated fault scenarios; transmit the simulated fault scenarios and fault diagnosis results to the three-dimensional model of the wind turbine and the operating environment for trainees' reference; Based on the multi-dimensional data of the wind turbine's operating status, the quantum enhanced prediction and optimization module is used to predict the wind turbine's faults and obtain the fault prediction results; based on the fault prediction results, suggestions for optimizing the wind turbine's operation are put forward; the fault prediction results and suggestions for optimizing the wind turbine's operation are fed back to the wind turbine's three-dimensional model and operating environment for students' reference; Based on the trainees' operation data of the wind turbine 3D model and operating environment, the training difficulty of the wind turbine 3D model and operating environment is adjusted through the adaptive learning and evaluation module.

[0020] Compared with the prior art, the present invention has the following beneficial effects: The invention discloses a fan fault simulation and diagnosis training system. The virtual reality simulation environment module provides trainees with a highly realistic three-dimensional model and operating environment of the fan, so that trainees can learn and practice in a close-to-real working scene, significantly improving the practicality and effectiveness of the training. The multimodal sensor fusion module can comprehensively collect multidimensional data of the fan operation status, providing a rich data source for the fault generation and diagnosis module, thereby improving the accuracy of fault simulation and the reliability of diagnosis. The fault generation and diagnosis module can generate simulated fault scenarios based on these data and perform intelligent diagnosis, providing trainees with valuable fault handling experience. The quantum enhanced prediction and optimization module uses advanced data analysis technology to predict the fault of the fan and put forward optimization suggestions, which helps trainees master the methods of fault prediction and optimization and improve the efficiency and quality of fault handling. The adaptive learning and evaluation module dynamically adjusts the training content and difficulty according to the performance of the trainees in the virtual reality simulation environment, realizes a personalized training experience, and helps improve the learning effect and satisfaction of the trainees; the multi-scale digital twin platform provides the system with a full range of fan system simulation capabilities, realizes data sharing and interaction, and further enhances the overall performance and practicality of the system.

[0021] The present invention provides a method for simulating and diagnosing wind turbine faults. By collecting the trainees' operation data in a pre-established three-dimensional model and operating environment of the wind turbine, as well as the multi-dimensional data of the wind turbine's operating status, we can accurately simulate various fault scenarios and generate accurate fault diagnosis results. This not only provides trainees with abundant practical opportunities, but also helps them better understand and master the diagnosis and treatment methods of wind turbine faults, significantly improving the training effect. Using the quantum enhanced prediction and optimization module to predict wind turbine faults can detect potential faults in advance and put forward targeted wind turbine operation optimization suggestions. This helps trainees understand the basic principles and methods of fault prediction, improves their ability to monitor and analyze the operating status of the wind turbine, and further enhances the practicality and foresight of the training. The adaptive learning and evaluation module can dynamically adjust the training difficulty according to the trainees' operation data in the three-dimensional model and operating environment of the wind turbine. This personalized training method can ensure that trainees obtain the best learning effect at a level suitable for them, while also improving the flexibility and pertinence of the training. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0023] Figure 1 This is a block diagram of the fan fault simulation and diagnosis training system of the present invention; Figure 2 A block diagram of the blower fault simulation and diagnosis training method of the present invention; Figure 3 This is a structural diagram of a fan fault simulation and diagnosis training system according to an embodiment of the present invention.

[0024] Among them, 101-virtual reality simulation environment module; 102-multimodal sensor fusion module; 103-fault generation and diagnosis module; 104-quantum enhanced prediction and optimization module; 105-adaptive learning and evaluation module; 106-multi-scale digital twin platform. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. The components of the embodiments of the present invention described and marked in the drawings here can be arranged and designed in various different configurations.

[0026] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0028] In the description of the embodiments of the present invention, it should be noted that if the terms "upper", "lower", "horizontal", "inner", etc. indicate an orientation or positional relationship based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the product of the invention is usually placed when in use, it is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In addition, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0029] In addition, if the term "horizontal" appears, it does not mean that the component must be absolutely horizontal, but can be slightly tilted. For example, "horizontal" only means that its direction is more horizontal than "vertical", which does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0030] In the description of the embodiments of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal connection of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0031] The present invention is further described in detail below in conjunction with the accompanying drawings: See also Figure 1 The present invention provides a fan fault simulation and diagnosis training system, comprising: A virtual reality simulation environment module 101 is used to create a three-dimensional model of a wind turbine and an operating environment, and collect the operating data of trainees on the virtual reality simulation environment module; A multimodal sensor fusion module 102, used to collect multidimensional data of the wind turbine operating status; The fault generation and diagnosis module 103 is used to receive data from the virtual reality simulation environment module and the multimodal sensor fusion module, generate a simulated fault scenario based on the received data, and perform intelligent diagnosis on the fault to generate a diagnosis result; the diagnosis result and the simulated fault scenario are transmitted to the virtual reality simulation environment module for reference by the trainees; The quantum enhanced prediction and optimization module 104 is used to predict the faults of the wind turbine according to the historical data and real-time data of the wind turbine operation, and to make optimization suggestions based on the prediction results; the prediction results and optimization suggestions are transmitted to the virtual reality simulation environment module for the trainees’ reference; An adaptive learning and assessment module 105, used to dynamically adjust the training content and difficulty according to the trainees' performance in the virtual reality simulation environment module; The multi-scale digital twin platform 106 is used to simulate the wind turbine system from micro to macro in all directions. The data and fault generation and diagnosis module, quantum enhanced prediction and optimization module and virtual reality simulation environment module of the multi-scale digital twin platform are shared.

[0032] See also Figure 3 The present invention provides a wind turbine fault simulation and diagnosis training system, which aims to solve many problems existing in traditional wind turbine training methods, such as insufficient training authenticity, low diagnostic accuracy, and poor learning efficiency. The system integrates multiple cutting-edge technologies, including virtual reality, artificial intelligence, quantum computing, and digital twins, to provide a highly realistic and intelligent training platform for wind farm operation and maintenance personnel.

[0033] First, the overall architecture of this system consists of seven core modules, which are interconnected through a central data bus to achieve efficient data transmission and collaborative work.

[0034] The system consists of seven core functional modules and a central data bus, forming a highly integrated and collaborative training ecosystem. The details are as follows: 1. Virtual reality simulation environment module: This module provides students with a highly realistic virtual wind turbine operating environment and is the main interactive interface of the system.

[0035] 2. Multimodal sensor fusion module: responsible for collecting and integrating wind turbine operation data from various sensors to provide real-time and comprehensive status information for the system.

[0036] 3. Fault generation and diagnosis module: Use artificial intelligence technology to generate various fault scenarios and perform intelligent diagnosis of the fan operating status.

[0037] 4. Quantum enhanced prediction and optimization module: Apply quantum computing technology to improve the system's prediction ability and optimization efficiency.

[0038] 5. Adaptive learning and assessment module: Dynamically adjust the training content and difficulty based on the students’ performance to provide a personalized learning experience.

[0039] 6. Multi-scale digital twin platform: Create a comprehensive digital model of the wind turbine system and realize all-round simulation from micro to macro.

[0040] 7. Secure communication and data management center: ensure the secure transmission, storage and management of system data.

[0041] 8. Data bus: Located in the center of the diagram, it connects all other modules. As the nerve center of the system, the data bus is responsible for coordinating data exchange and information flow between modules.

[0042] The modular design of the present invention makes the system of the present invention highly flexible and scalable; each module can be upgraded or replaced independently without affecting the operation of the overall system. At the same time, this architecture is also convenient for expanding new functional modules according to needs in the future.

[0043] The virtual reality simulation environment module is one of the cores of this system, providing students with a highly realistic wind turbine operation environment. The virtual reality simulation environment module includes a high-performance graphics rendering engine, an adaptive level of detail system, an advanced physics engine, a multi-point force feedback tactile device, and a high-precision hand tracking system.

[0044] The high-performance graphics rendering engine supports high-resolution display and high refresh rate, ensuring the ultimate realism of the visual experience; specifically, the system can support at least 2K resolution per eye and a refresh rate of no less than 90Hz, and these parameters can be adjusted according to the specific hardware configuration and application requirements. In order to further improve rendering efficiency and visual quality, the present invention has developed an adaptive level of detail (LOD) system; the adaptive level of detail system can adjust the detail level of objects in the scene in real time according to the user's gaze point, which not only ensures the visual effect but also optimizes the system performance.

[0045] The working principle of the adaptive level of detail system is described as follows:

[0046] In the formula, Indicates i The level of detail of an object, as an integer value. It is the highest detail level supported by the system, also an integer. is a positive real number used to adjust LOD The rate of change. is the distance from the i-th object to the user's gaze point, and is a reference distance. Through this formula, the system can dynamically adjust the detail level of the object according to its distance from the gaze point, which not only ensures the fine performance of nearby objects, but also reduces the rendering burden of distant objects.

[0047] The virtual reality simulation environment module integrates an advanced physics engine to achieve accurate rigid body and soft body dynamics simulation. Especially in the deformation simulation of wind turbine blades, the system uses advanced numerical calculation methods for high-precision calculations, and the deformation accuracy can reach the millimeter level. This high-precision physical simulation provides a reliable data basis for fault diagnosis and prediction.

[0048] The virtual reality simulation environment module integrates a multi-point force feedback tactile device and a high-precision hand tracking system; it uses a multi-point force feedback tactile device to provide accurate force feedback. At the same time, it also integrates high-precision hand tracking technology with a tracking accuracy of sub-millimeter level. The combination of these technologies enables trainees to touch and feel the wind turbine components in a virtual environment, improving the immersion and practicality of the training.

[0049] When collecting data from the wind turbine, the present invention uses a multimodal sensor fusion module for data collection. The multimodal sensor fusion module includes a micro-electromechanical system sensor network, a fiber grating strain sensor system, a millimeter wave radar array, and a multi-source heterogeneous data dynamic weight fusion algorithm based on a graph attention network; The MEMS sensor network, fiber Bragg grating strain sensing system and millimeter wave radar array collect the operating data of the wind turbine equipment, realizing all-round and multi-dimensional monitoring of the wind turbine operating status. The MEMS sensor network has achieved high-density deployment with its small size (1.5mm * 1.5mm *0.5mm) and low power consumption (<1mW), and 100 sensors can be arranged per square meter. The fiber Bragg grating strain sensing system provides reliable data for strain monitoring of wind turbine blades with its high precision (strain resolution 0.1με) and wide measurement range (±10000με). The millimeter wave radar array operates in the 77-81GHz frequency band and has extremely high spatial resolution (1mm at a distance of 10m), and is mainly used for non-contact detection of surface defects of wind turbine blades. The operating data of the wind turbine equipment is fused through a multi-source heterogeneous data dynamic weight fusion algorithm based on a graph attention network to obtain multi-dimensional data of the wind turbine operating status. The data fusion algorithm based on the graph attention network (GAT) is expressed by the following formula:

[0050] In the formula, It is the output feature vector, representing the fused data. It is the activation function, ReLU is usually used. Representation Node i The neighborhood set of is the attention coefficient, which reflects the importance of different data sources. is a learnable weight matrix, and is the input feature vector. Through this algorithm, the system can adaptively adjust the weights of different data sources to achieve the best data fusion effect.

[0051] In terms of fault diagnosis, the present invention designs a fault generation and diagnosis module. The fault generation and diagnosis module includes a fault generation model, a diagnosis model, a structural causal model and a model compression module; The fault generation module is based on the conditional generative adversarial network (CGAN) architecture and integrates physical constraints to ensure that the generated fault data is both diverse and conforms to physical laws.

[0052] The diagnostic model combines a graph convolutional network (GCN) and a long short-term memory network (LSTM) to extract features from fault data and diagnose the fault type. It can not only accurately identify faults (accuracy > 99.5%), but also accurately locate the fault (accuracy < 5 cm).

[0053] In order to improve the interpretability of the diagnosis results, the fault generation and diagnosis module introduces the structural causal model (SCM). The principle of SCM is:

[0054] here, represents the i-th variable, It is a decision The function of yes The direct cause (parent node set), and Represents an unobserved exogenous variable. Through this model, not only can the existence of a fault be pointed out, but also the cause chain of the fault can be explained, providing students with more valuable information.

[0055] The quantum enhanced prediction and optimization module of the present invention includes a superconducting quantum processor, a quantum approximate optimization algorithm, a variational quantum eigensolver, and an adaptive quantum circuit design system; the quantum enhanced prediction and optimization module uses a superconducting quantum processor with more than 100 quantum bits and a coherence time of more than 100μs, and a quantum annealer with more than 5,000 quantum bits. These quantum hardware provide powerful computing power for the system. At the algorithm level, the present invention uses a quantum approximate optimization algorithm (QAOA) to optimize the wind farm layout and a variational quantum eigensolver (VQE) to perform material fatigue analysis. These quantum algorithms have shown significant performance advantages on specific problems, and can achieve a speed increase of 100-1,000 times compared to classical algorithms.

[0056] In order to fully tap the potential of quantum hardware, the quantum enhanced prediction and optimization module of the present invention integrates an adaptive quantum circuit design system. The adaptive quantum circuit design system uses a reinforcement learning method to dynamically adjust the quantum gate sequence to adapt to different problem characteristics and the noise characteristics of quantum hardware. This adaptive method significantly improves the robustness and efficiency of quantum computing, making quantum optimization feasible in actual wind farm management.

[0057] The adaptive learning and evaluation module of the present invention includes a learning framework based on a model-independent meta-learning algorithm, an active learning module, a dynamic difficulty adjustment mechanism, and a knowledge graph module; the core idea of ​​the adaptive learning framework based on model-independent meta-learning (MAML) is to find an initial model parameter that can quickly adapt to new tasks. Its optimization goal can be expressed as:

[0058] here, are the initial parameters of the model, is a task sampled from the task distribution, is the loss function of task $\tau_i$, It's on mission A model that has been updated with one or more gradients. This method enables the system to quickly adapt to new fault types or operating scenarios, greatly improving the efficiency and flexibility of training. The active learning module is used to analyze the most valuable learning content for trainees; based on the results of the analysis of the most valuable learning content for trainees, appropriate learning resources are selected from the knowledge graph module; the dynamic difficulty adjustment mechanism adjusts the complexity of the fault scenario based on the real-time performance of the trainees to ensure that the learning process is challenging but not too difficult. The knowledge graph module uses graph neural network modeling knowledge graphs to generate personalized learning routes and generate the optimal learning sequence for each trainee, while taking into account cognitive science theory to optimize the arrangement of knowledge points and review intervals.

[0059] In order to achieve all-round wind turbine system simulation from micro to macro, the present invention integrates a multi-scale digital twin platform. At the micro scale, molecular dynamics simulation software is used for simulation, and the number of simulated atoms exceeds , used to study the microstructure and properties of materials. At the mesoscopic scale, finite element analysis software is used for analysis, and the number of grids in a single fan model exceeds , and aeroelastic simulation is performed simultaneously with a time step of less than 0.1ms. At the macro scale, large eddy simulation (LES) is used to simulate wind field flow with a spatial resolution of less than 1m, and power system simulation based on power system simulation software is integrated. This multi-scale digital twin platform can provide comprehensive simulation from material performance to the entire wind farm operation, providing a solid foundation for fault diagnosis and prediction. In order to achieve effective transmission of cross-scale information, the multi-scale digital twin platform integrates a scale bridging algorithm based on deep learning; this algorithm can transfer micro-scale material performance information to macro-scale wind turbine performance prediction, and can also feed back macro-scale load information to micro-scale material fatigue analysis. This two-way information flow ensures that the system can fully consider the impact of each scale, improving the accuracy of the simulation and the reliability of the prediction.

[0060] Considering the importance of data security, the present invention integrates a highly secure communication and data management system. The highly secure communication and data management system uses quantum key distribution (QKD) technology for encrypted communication, with a key generation rate exceeding 10Mbps@50km and a safe distance exceeding 200km (using quantum repeaters). This quantum encryption technology can effectively prevent man-in-the-middle attacks and ensure the absolute security of data transmission.

[0061] In terms of data storage, blockchain technology based on the Proof of Stake (PoS) consensus mechanism is used, with a throughput of more than 10,000 TPS (Transactions Per Second) and the ability to store 10PB of data per year. The application of blockchain technology not only ensures the immutability of data, but also improves data traceability, which is of great significance for the traceability analysis of wind turbine failures.

[0062] A wind turbine fault simulation and diagnosis training system according to an embodiment of the present invention also adopts a cloud-edge collaborative computing architecture. The edge node adopts a 64-core, 256GB high-performance ARM architecture processor, which can realize preliminary analysis of the data; the central cloud adopts a large-scale cluster containing more than 1,000 high-performance GPUs to process complex AI training tasks and large-scale simulation calculations.

[0063] In order to ensure the reliability and scalability of the system, the system of the present invention adopts a modular design based on a microservice architecture. Each core function is encapsulated as an independent microservice and deployed and managed through container technology. This design not only improves the flexibility and maintainability of the system, but also enables the system to be elastically expanded according to actual needs.

[0064] In terms of system monitoring, the system integrates a real-time monitoring dashboard and uses advanced monitoring tools to track various performance indicators of the system. At the same time, the system also uses a distributed log analysis system to quickly locate and solve potential problems.

[0065] An embodiment of the present invention further discloses that the system of the present invention further includes: Performance evaluation module, used to monitor and evaluate various system indicators in real time; Data analysis module, used to process and analyze large amounts of data collected from various subsystems; Prediction model optimization module, used to continuously optimize the system's prediction model based on real-time data and historical data; A resource scheduling module for dynamically allocating computing resources based on system load and user demand; and Interface adapter module, used for data interaction and function integration with existing wind farm management systems.

[0066] The wind turbine fault simulation and diagnosis training system of the present invention significantly improves the effect and efficiency of wind turbine operation training by integrating advanced virtual reality, artificial intelligence, quantum computing and digital twin technologies. The system provides a highly realistic simulation environment and a rich variety of fault scenarios, which greatly improves the learning interest and participation of trainees. The application of adaptive learning algorithms and personalized learning paths makes the training process more efficient, and the learning curve is accelerated by more than 50%. The combination of fault generation and diagnosis module and quantum optimization algorithm greatly improves the accuracy of fault prediction, can provide 24 hours advance warning, and the fault prediction accuracy rate exceeds 95%, thereby reducing the maintenance cost of wind turbines, significantly improving the operational safety of wind farms, and reducing the major accident rate by more than 80%. The intelligent optimization function of the system also helps to improve the overall power generation efficiency of wind farms by 5-10%. It not only solves the technical problems in traditional wind turbine training that mainly rely on simple computer simulation and limited physical training, and it is difficult to fully and realistically reproduce complex wind turbine operating conditions and fault scenarios, but also provides strong technical support for the intelligent operation and maintenance of wind farms.

[0067] The present invention integrates various modules through a data bus to achieve highly realistic wind turbine operating environment simulation, accurate fault diagnosis and prediction, personalized training program formulation, and all-round wind turbine system optimization. The present invention significantly improves the wind turbine operation training effect, enhances the fault prediction capability, and effectively improves the operational efficiency and safety of wind farms; with the continuous advancement of technology, the system still has broad development space, and more advanced technologies may be integrated in the future, such as brain-computer interfaces, programmable materials, etc., to further improve the system's performance and user experience.

[0068] See also Figure 2 The present invention discloses a fan fault simulation and diagnosis training method, comprising the following steps: S1, collect data of students' operations in the pre-established 3D model and operating environment of the fan; first, use advanced modeling software (such as SolidWorks, AutoCAD, etc.) to build a 3D model of the fan according to the actual structure and operating environment of the fan. At the same time, build a virtual operating environment similar to the actual fan operating interface to ensure that students can perform simulated operations in a close to real working environment. When students perform simulated operations in the 3D model and operating environment of the fan, the system will collect students' operation data in real time, including but not limited to operation sequence, operation time, operation accuracy, etc. These data will be used for subsequent analysis and evaluation.

[0069] S2, collects multi-dimensional data of the fan operation status; the system will collect multi-dimensional data of the fan during actual operation, such as vibration signals, temperature data, pressure data, etc. These data will be used to generate simulated fault scenarios and perform fault prediction.

[0070] S3, based on the data of the trainees' operation in the pre-established three-dimensional model of the wind turbine and the operating environment and the multi-dimensional data of the wind turbine's operating status, generate a simulated fault scenario; based on the simulated fault scenario, generate a fault diagnosis result; transmit the simulated fault scenario and the fault diagnosis result to the three-dimensional model of the wind turbine and the operating environment for the trainees' reference; Based on the trainees’ operation data and the multi-dimensional data of the wind turbine’s operating status, the system generates simulated fault scenarios through algorithms. Then, the simulated fault scenarios are analyzed using the fault diagnosis algorithm to generate fault diagnosis results. These results will be transmitted to the wind turbine’s three-dimensional model and operating environment in real time for trainees’ reference and learning.

[0071] S4, based on the multi-dimensional data of the wind turbine operation status, the fault prediction of the wind turbine is carried out through the quantum enhanced prediction and optimization module to obtain the fault prediction results; based on the fault prediction results, the optimization suggestions for the wind turbine operation are put forward; the fault prediction results and the optimization suggestions for the wind turbine operation are fed back to the three-dimensional model of the wind turbine and the operating environment for the trainees’ reference; The system predicts wind turbine failures through the quantum-enhanced prediction and optimization module. This module uses the efficiency and accuracy of quantum computing to conduct in-depth analysis of the multi-dimensional data of wind turbine operating status and predict the type and time of possible failures. Based on the prediction results, the system proposes optimization suggestions for wind turbine operation, such as adjusting operating parameters and replacing worn parts. These suggestions will also be fed back to the wind turbine 3D model and operating environment for students' reference.

[0072] S5, based on the trainee operation data of the wind turbine 3D model and the operating environment, the training difficulty of the wind turbine 3D model and the operating environment is adjusted through the adaptive learning and evaluation module; the trainee's operation data is analyzed through the adaptive learning and evaluation module to evaluate the trainee's learning progress and ability level. According to the evaluation results, the system will automatically adjust the training difficulty of the wind turbine 3D model and the operating environment to ensure that the trainees can learn at a difficulty level that suits them and improve their learning efficiency.

[0073] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A fan fault simulation and diagnosis training system, characterized in that: include: Virtual reality simulation environment module, used to create a three-dimensional model of the wind turbine and the operating environment, and collect trainees' operating data on the virtual reality simulation environment module; Multimodal sensor fusion module, used to collect multi-dimensional data of wind turbine operating status; The fault generation and diagnosis module is used to receive data from the virtual reality simulation environment module and the multimodal sensor fusion module, generate a simulated fault scenario based on the received data, and perform intelligent diagnosis on the fault to generate a diagnosis result; the diagnosis result and the simulated fault scenario are transmitted to the virtual reality simulation environment module for reference by the trainees; The quantum enhanced prediction and optimization module is used to predict the faults of the wind turbine based on the historical data and real-time data of the wind turbine operation, and to make optimization suggestions based on the prediction results; the prediction results and optimization suggestions are transmitted to the virtual reality simulation environment module for the trainees’ reference; Adaptive learning and assessment module, used to dynamically adjust the training content and difficulty based on the trainees’ performance in the virtual reality simulation environment module; A multi-scale digital twin platform is used to simulate wind turbine systems from micro to macro. The multi-scale digital twin platform shares data and fault generation and diagnosis modules, quantum enhanced prediction and optimization modules, and virtual reality simulation environment modules.

2. The fan fault simulation and diagnosis training system according to claim 1 is characterized in that: The virtual reality simulation environment module includes a high-performance graphics rendering engine, an adaptive level of detail system, an advanced physics engine, a multi-point force feedback tactile device and a high-precision hand tracking system; The high-performance graphics rendering engine is used to dynamically adjust model details according to the user's gaze point and optimize rendering performance by cooperating with the adaptive detail level system; The adaptive level of detail system is used to dynamically adjust the level of detail of the model according to the user's gaze point in the virtual environment; The advanced physics engine is used for rigid and soft body dynamics simulation; The multi-point force feedback tactile device and the high-precision hand tracking system are used to provide tactile feedback and hand tracking.

3. The fan fault simulation and diagnosis training system according to claim 1 is characterized in that: It also includes a secure communication and data management center, which is used to ensure the secure transmission and storage of system data.

4. The fan fault simulation and diagnosis training system according to claim 1 is characterized in that: The multimodal sensor fusion module includes a micro-electromechanical system sensor network, a fiber Bragg grating strain sensing system, a millimeter wave radar array, and a multi-source heterogeneous data dynamic weight fusion algorithm based on a graph attention network; The micro-electromechanical system sensor network, fiber grating strain sensor system and millimeter wave radar array collect the operating data of the wind turbine equipment. The operating data of the wind turbine equipment is fused through a multi-source heterogeneous data dynamic weight fusion algorithm based on a graph attention network to obtain multi-dimensional data of the wind turbine operating status.

5. The fan fault simulation and diagnosis training system according to claim 1 is characterized in that: The fault generation and diagnosis module includes a fault generation model, a diagnosis model, a structural causal model and a model compression module; The fault generation model generates fault data using a conditional generative adversarial network; The diagnostic model uses graph convolutional networks and long short-term memory networks to extract features from fault data to achieve fault type diagnosis; The structural causal model analyzes the causal relationship in the fault data to determine the root cause and propagation path of the fault; The model compression module is used to compress the model size.

6. The fan fault simulation and diagnosis training system according to claim 1, characterized in that: The quantum enhanced prediction and optimization module includes a superconducting quantum processor, a quantum approximate optimization algorithm, a variational quantum eigensolver and an adaptive quantum circuit design system; The quantum approximate optimization algorithm is used to optimize the layout of the wind farm; The variational quantum eigensolver is used for material fatigue analysis; The superconducting quantum processor is used to run a quantum approximate optimization algorithm and a variational quantum eigensolver; The adaptive quantum circuit design system dynamically adjusts the quantum gate sequence according to the execution of the quantum approximate optimization algorithm and the variational quantum eigensolver and the state of the superconducting quantum processor.

7. The fan fault simulation and diagnosis training system according to claim 1, characterized in that: The adaptive learning and evaluation module includes a learning framework based on a model-independent meta-learning algorithm, an active learning module, a dynamic difficulty adjustment mechanism, and a knowledge graph module; The learning framework based on the model-independent meta-learning algorithm is used to process and analyze the learning data of students using the model-independent meta-learning algorithm; The active learning module is used to analyze the most valuable learning content for students; based on the results of the analysis of the most valuable learning content for students, appropriate learning resources are selected from the knowledge graph module; The dynamic difficulty adjustment mechanism is used to adjust the difficulty of the learning content according to the real-time performance of the students; The knowledge graph module uses a graph neural network to model a knowledge graph, which is used to generate a personalized learning path for students based on the association relationship of the knowledge graph.

8. The fan fault simulation and diagnosis training system according to claim 1 is characterized in that: The multi-scale digital twin platform includes a micro-scale model, a meso-scale model, a macro-scale model, and a scale bridging algorithm based on deep learning; The microscopic scale model uses molecular dynamics software to simulate the microscopic world; The mesoscale model uses finite element analysis software to construct the mesoworld; The macro-scale model uses a large nest simulation to simulate wind field flow; The deep learning-based scale bridging algorithm is used to achieve the transmission and integration of cross-scale information.

9. The fan fault simulation and diagnosis training system according to claim 1, characterized in that: It also includes a system integration and deployment module, which adopts a distributed architecture that supports horizontal expansion.

10. A fan fault simulation and diagnosis training method, based on the fan fault simulation and diagnosis training system according to claim 1, characterized in that: The following steps are involved: Collect data from trainees’ operations in the pre-established 3D wind turbine model and operating environment; Collect multi-dimensional data of fan operation status; Generate simulated fault scenarios based on the data of students' operations in the pre-established three-dimensional model of the wind turbine and the operating environment and the multi-dimensional data of the wind turbine's operating status; Generate fault diagnosis results based on the simulated fault scenarios; Transmit simulated fault scenarios and fault diagnosis results to the wind turbine 3D model and operating environment for trainees’ reference; Based on the multi-dimensional data of the wind turbine operating status, the fault prediction of the wind turbine is carried out through the quantum enhanced prediction and optimization module to obtain the fault prediction results; Propose optimization suggestions for wind turbine operation based on fault prediction results; Feedback fault prediction results and wind turbine operation optimization suggestions to the wind turbine 3D model and operating environment for trainees’ reference; Based on the trainees' operation data of the wind turbine 3D model and operating environment, the training difficulty of the wind turbine 3D model and operating environment is adjusted through the adaptive learning and evaluation module.

Images