Wind power plant operation and maintenance personnel virtual reality simulation training method, system, equipment and medium

Abstract

The invention provides a virtual reality simulation training method, system and equipment for wind power plant operation and maintenance personnel and a medium. The virtual reality simulation training method comprises the following steps: step 1, constructing a virtual wind power plant; 2, realizing man-machine interaction between the user and the virtual wind power plant; 3, generating a virtual coach with professional knowledge and training experience, and guiding the operation of the user to realize simulation training of the operation and maintenance personnel of the wind power plant; according to the method, the application potential of emerging technologies such as virtual reality and artificial intelligence in the field of wind power operation and maintenance training is fully exerted, and the method has great theoretical significance and practical value in the aspects of improving industry talent quality, reducing training cost, promoting green and low-carbon development and the like.

Description

Technical Field

[0001] The present invention belongs to the technical field of training operation and maintenance personnel in the wind power industry, and in particular, relates to a virtual reality simulation training method, system, equipment and medium for wind farm operation and maintenance personnel. Background Art

[0002] As a clean, renewable green energy, wind power plays an important role in addressing global climate change and promoting energy transformation. However, the efficient and stable operation of wind farms places high demands on the professional skills of operation and maintenance personnel. Operation and maintenance personnel need to master professional knowledge such as the structure and control principle of wind turbines, fault diagnosis and disposal methods, and have excellent practical skills to ensure the continuous safe production of wind farms. Therefore, the training of operation and maintenance personnel is crucial to the sustainable development of the wind power industry.

[0003] At present, the training of wind power operation and maintenance personnel mainly adopts the traditional model of physical operation. Although this model can provide real training scenarios, it also has many shortcomings, such as limited training resources, high practical operation risks, lack of team collaboration training, and general training results. At the same time, the traditional training model is difficult to achieve personalized guidance and intelligent adjustment, and lacks systematic recording and analysis and evaluation of the practical operation data of operation and maintenance personnel, and cannot form data closed-loop management and continuous optimization. Summary of the invention

[0004] The purpose of the present invention is to provide a virtual reality simulation training method, system, equipment and medium for wind farm operation and maintenance personnel, which solves many shortcomings of traditional wind power operation and maintenance personnel training, such as limited training resources, high practical operation risks and difficulties in team collaboration training.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is:

[0006] The present invention provides a virtual reality simulation training method for wind farm operation and maintenance personnel, comprising the following steps:

[0007] Step 1, construct a virtual wind farm;

[0008] Step 2, realizing human-computer interaction between the user and the virtual wind farm;

[0009] Step 3: Generate a virtual trainer with professional knowledge and training experience, and guide the user's operation to achieve simulation training for wind farm operation and maintenance personnel.

[0010] Preferably, in step 1, a virtual wind farm is constructed, and the specific method is:

[0011] Based on the geographic information data of the actual wind farm, the unit structure drawings and equipment parameters, a virtual wind farm is constructed by combining 3D modeling and physical rendering technology.

[0012] Preferably, in step 2, human-computer interaction between the user and the virtual wind farm is achieved, and the specific method is:

[0013] Wearable devices are combined with Kalman filtering algorithm to realize human-computer interaction between users and virtual wind farms.

[0014] Preferably, in step 3, a virtual trainer with professional knowledge and training experience is generated to guide the user's operation, and the specific method is:

[0015] Generate a virtual coach based on digital human technology;

[0016] Capture typical operation and maintenance actions of wind farms to obtain capture data;

[0017] Generate animation sequence from the captured data;

[0018] Bind the animation sequence to the virtual trainer;

[0019] Build a decision-making model for the virtual trainer and use the reinforcement learning algorithm to train the decision-making model so that the decision-making model can adjust the scene difficulty and teaching rhythm according to the real-time performance of the target user.

[0020] A virtual reality simulation training system for wind farm operation and maintenance personnel, comprising:

[0021] Virtual wind farm environment modeling module, used to build a virtual wind farm;

[0022] Multi-sensor fusion motion capture interaction module, used to realize human-computer interaction between users and virtual wind farms;

[0023] The virtual trainer and intelligent training module is used to generate a virtual trainer with professional knowledge and training experience, and to guide the user's operations, thereby realizing simulation training for wind farm operation and maintenance personnel.

[0024] Preferably, it also includes a training data management and blockchain module for real-time recording and analysis of the operation data of the operation and maintenance personnel and generating personalized training evaluation reports.

[0025] Preferably, it also includes cloud deployment and multi-person collaboration modules for simultaneous access by multiple users and remote online training.

[0026] A computer device comprising:

[0027] a processor suitable for executing a computer program;

[0028] A computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the method described is performed.

[0029] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described above is implemented.

[0030] A computer program product comprises a computer program, wherein the computer program implements the method when executed by a processor.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] The present invention provides a virtual reality simulation training method for wind farm operation and maintenance personnel, which highly restores the real wind farm environment through virtual reality technology, and integrates artificial intelligence technology to simulate various typical operation and maintenance scenarios and fault cases, providing an immersive training experience for operation and maintenance personnel. The intelligent virtual trainer can dynamically adjust the training difficulty, strategy and scenario according to the real-time performance of the operation and maintenance personnel, realize personalized intelligent training, thereby greatly improving the professionalism, pertinence and effectiveness of the training, and significantly enhancing the practical skills and emergency response capabilities of the operation and maintenance personnel.

[0033] Compared with traditional physical training, this invention does not need to occupy actual wind farm resources, avoiding the risk of equipment damage and personal injury. At the same time, through cloud deployment and multi-person remote collaboration, it reduces fixed cost investment such as teachers and venues, greatly saves training costs, and improves economy and resource utilization efficiency.

[0034] The present invention incorporates artificial intelligence technologies, such as reinforcement learning algorithms and digital human technology, and can dynamically adjust the difficulty, strategy and scenario of training according to the real-time performance of operation and maintenance personnel, and tailor a personalized training plan for each student, thus realizing the intelligence and personalization of training and greatly improving the accuracy and effectiveness of training.

[0035] The operation process of the present invention does not generate direct environmental pollution and energy consumption, which is conducive to achieving green and low-carbon development of wind power industry training, promoting the efficient use of renewable energy, and contributing to addressing climate change and achieving carbon neutrality.

[0036] In summary, this invention gives full play to the application potential of emerging technologies such as virtual reality and artificial intelligence in the field of wind power operation and maintenance training, and has great theoretical significance and practical value in improving the quality of industry talents, reducing training costs, and promoting green and low-carbon development. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram of a module of the present invention;

[0038] Figure 2 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0039] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0040] Example 1

[0041] like Figure 1 and Figure 2 As shown, this embodiment provides a virtual reality simulation training system for wind farm operation and maintenance personnel, including the following main modules:

[0042] Virtual wind farm environment modeling module, used to build a virtual wind farm;

[0043] Multi-sensor fusion motion capture interaction module, used to realize human-computer interaction between users and virtual wind farms;

[0044] Virtual trainer and intelligent training module, used to generate virtual trainers with professional knowledge and training experience, and guide users' operations;

[0045] Training data management and blockchain modules are used to record and analyze the operation data of operation and maintenance personnel in real time and generate personalized training evaluation reports;

[0046] Cloud deployment and multi-person collaboration modules are used for simultaneous access by multiple users and remote online training.

[0047] Example 2

[0048] Based on Example 1, this example provides a virtual reality simulation training system for wind farm operation and maintenance personnel. The virtual wind farm environment modeling module constructs a virtual wind farm based on the geographic information data, unit structure drawings and equipment parameters of the actual wind farm, using three-dimensional modeling and physically based rendering (PBR) technology.

[0049] In this embodiment, the detailed method of the virtual wind farm environment modeling module is as follows:

[0050] Step 1: Data collection and preprocessing

[0051] Obtain geographic information data of the actual wind farm, including topography, landforms and vegetation distribution; obtain structural drawings and parameters of each wind turbine model in the actual wind farm, including but not limited to 3D model parameters of components such as machine houses, blades, bearings, etc.; obtain design drawings and parameters of auxiliary facilities such as substations, high and low voltage lines, duty rooms, and maintenance workshops in the actual wind farm; organize, optimize and convert the above data into a format to prepare for subsequent 3D modeling.

[0052] Step 2: Construction of virtual wind farm scene

[0053] Using geographic information system (GIS) software, an overall topographic model of the wind farm is constructed; wind turbine models, line models and other auxiliary facility models are added to the overall topographic model of the wind farm according to actual layout, so as to obtain an initial virtual wind farm with the auxiliary facility models added;

[0054] According to the actual scenario, a road model and a vegetation model are added to the initial virtual wind farm with the auxiliary facilities model, and the weather effect is rendered to obtain a virtual wind farm scene.

[0055] Step 3: Modeling of fans and key equipment

[0056] Obtain detailed structural drawings and parameter specifications of the wind turbine, including the size and material information of major components such as the nacelle, blades, hub, tower, etc.;

[0057] Based on the structural drawings and parameter specifications of each wind turbine model, 3D computer-aided design (CAD) software is used to create accurate 3D models of the wind turbine components:

[0058] -First build the cabin model, including the internal structure such as the generator, gearbox, etc.

[0059] -Build a blade model, paying attention to the blade twist angle and airfoil profile changes.

[0060] -Create the hub model, including details of the blade connections.

[0061] -Build a tower model, taking into account its taper and internal ladder structures.

[0062] The use of parametric modeling technology ensures that the model can be quickly adjusted according to the specifications of different types of wind turbines.

[0063] Optimize the details of the model, such as adding small parts such as bolts and rivets to improve realism.

[0064] Substation Modeling

[0065] Based on the plan and elevation drawings of the substation, the infrastructure model is built using CAD software.

[0066] Detailed modeling of major equipment such as transformers, circuit breakers, disconnectors, etc.:

[0067] -For transformers, accurately model details such as heat sinks, high and low voltage bushings, etc.

[0068] -For circuit breakers, simulate their internal structure and operating mechanism.

[0069] Add models of auxiliary facilities such as control cabinets and cable trenches.

[0070] High and low voltage line modeling

[0071] According to the line design drawing, use CAD software to build a model of the telephone pole or tower.

[0072] Create a wire model, paying attention to simulating its curvature and sag.

[0073] Add models of auxiliary equipment such as insulators and lightning arresters.

[0074] Control system interface modeling

[0075] Get interface screenshots and operation logic of the actual wind turbine control system.

[0076] Rebuild the control interface using UI design software such as Adobe XD or Sketch.

[0077] Import interface designs into 3D modeling software to create interactive virtual control panels.

[0078] Refined Modeling

[0079] For parts that require frequent operation, such as control buttons, handles, fasteners, etc., ultra-high precision modeling is performed:

[0080] Use 3D scanning technology to obtain accurate three-dimensional data of physical objects.

[0081] Reconstruct high-precision models based on scan data in CAD software.

[0082] Add material information, such as metallic luster, plastic texture, etc.

[0083] Model Optimization

[0084] Simplify the model and remove internal structures that are not visible in the VR environment.

[0085] Use polygon reduction technology to reduce model complexity while ensuring visual effects.

[0086] Create models with different levels of accuracy (LOD) to suit different rendering needs.

[0087] Rigging and Animation

[0088] Assemble the various components into complete fan and equipment models in 3D software.

[0089] Add appropriate constraints and motion parameters for movable parts (such as blades, circuit breakers, etc.).

[0090] Create basic animation sequences, such as fan startup and shutdown.

[0091] Step 4: Physically Based Rendering (PBR)

[0092] Using PBR technology, the material, lighting, shadow, etc. of the virtual wind farm scene obtained in step 2 and the wind turbine and key equipment model obtained in step 3 are physically simulated and rendered. Specifically:

[0093] The following models are simulated for physical properties:

[0094] a) Fan model:

[0095] -Simulate the bending deformation of blades at different wind speeds

[0096] -Simulates the slight vibration of the cabin under strong wind

[0097] -Simulation of stress distribution at the connection between hub and blade

[0098] b) Substation equipment model:

[0099] -Simulate the flow and heat dissipation characteristics of transformer oil

[0100] -Simulate the action process of circuit breaker contacts

[0101] -Simulate the changes in insulation performance of insulators under different weather conditions

[0102] c) Cable and wire models:

[0103] -Simulate deformation under different temperatures and loads

[0104] -Simulate the impact of wind load on high voltage lines

[0105] d) Terrain and vegetation models:

[0106] -Simulate the effect of wind on vegetation, such as grass waving and leaves swaying

[0107] -Simulate the erosion effect of rain on terrain. Perform physically based rendering on all models, including but not limited to:

[0108] a) Fan model:

[0109] - Rendering the microstructure of metal surfaces, such as the frosted effect of the cabin shell

[0110] - Rendering the glossiness and reflective properties of the leaf surface

[0111] - Rendering the coating effect on the tower surface

[0112] b) Substation equipment model:

[0113] -Rendering the metallic luster of the transformer heat sink

[0114] - Rendering the ceramic material properties of the insulator

[0115] - Rendering of the plastic and metal composite material of the control panel

[0116] c) Cable and wire models:

[0117] - Rendering the subtle texture and reflective properties of the wire surface

[0118] - Rendering semi-transparent effect of insulation layer

[0119] d) Terrain and vegetation models:

[0120] -Multiple layers of textures for rendering ground surfaces, such as dirt, rocks, and grass

[0121] - Rendering of translucent effects of vegetation, such as light transmission through leaves

[0122] e) Building and auxiliary facilities model:

[0123] -Rendering the roughness and weathering effects of concrete surfaces

[0124] - Rendering of rust and wear on metal structures

[0125] f) Sky and Atmospheric Effects:

[0126] - Render realistic sky color changes

[0127] - Rendering volumetric and light scattering effects of clouds Lighting and shadow processing:

[0128] a) Global Illumination:

[0129] -Simulate direct and indirect sunlight in the scene

[0130] -Simulate the diffuse reflection effect of sky light

[0131] b) Local illumination:

[0132] -Simulation of artificial lighting in substations

[0133] -Simulate the lighting effect of device indicator lights

[0134] c) Dynamic shadows:

[0135] - Rendering dynamic shadows produced by the rotation of wind turbine blades

[0136] - Rendering the light and shadow changes caused by the movement of clouds

[0137] d) Ambient Occlusion:

[0138] - Rendering subtle shadows in complex structures, such as shadows between parts inside a cabin

[0139] This detailed physical simulation and rendering process can greatly improve the realism and immersion of the virtual wind farm environment. Physical simulation can make the objects in the scene show more realistic dynamic effects, while physically based rendering can make the scene more realistic visually, thus providing a highly simulated training environment for wind farm operation and maintenance personnel. This highly simulated environment can not only improve the training effect, but also help operation and maintenance personnel better understand the performance of equipment under different environmental conditions.

[0140] Step 5: VR Integration and Optimization

[0141] Model Ensemble

[0142] - Import the 3D models created in steps 2 and 3 into your chosen VR engine (such as Unity or Unreal Engine).

[0143] - Arrange and organize these models in a virtual environment according to the actual wind farm layout.

[0144] -Add terrain, vegetation, sky and other environmental elements to build a complete virtual wind farm scene.

[0145] VR Device Support

[0146] -Configure stereo rendering settings for VR headsets to ensure correct depth perception.

[0147] -Implement input mapping for VR controllers to support intuitive gesture interactions.

[0148] -Optimize the rendering pipeline to meet the high frame rate requirements of VR devices (usually above 90fps).

[0149] Performance Optimization

[0150] a)LOD (Level of Detail) system implementation:

[0151] - Create multiple level of detail versions for complex models (such as wind turbines and substation equipment).

[0152] - Set LOD parameters in the VR engine to define the switching distances for different levels.

[0153] -Implement LOD groups to ensure coordinated switching of components of complex objects.

[0154] b) Frustum Culling:

[0155] - Calculate the view frustum based on the camera parameters.

[0156] -Set bounding volumes (such as bounding spheres, bounding boxes) for scene objects.

[0157] -Implement frustum testing and only render objects within the frustum.

[0158] c) Rendering optimization:

[0159] - Use GPU instancing to render recurring models.

[0160] -Implement occlusion culling to avoid rendering occluded objects.

[0161] -Develop and use optimized shaders to reduce the amount of calculations per frame.

[0162] Multiplayer collaboration function

[0163] - Select and implement an appropriate network architecture (such as client-server or P2P).

[0164] - Implement data compression and incremental updates to reduce network transmission volume.

[0165] -Develop prediction and interpolation algorithms to reduce the impact of network latency.

[0166] -Implement an interest management system to only synchronize the status of relevant areas.

[0167] Motion Capture Integration

[0168] -Integrate selected motion capture system SDK.

[0169] -Smooth motion data processing to reduce jitter.

[0170] -Develop a motion retargeting system to map the captured data to the virtual character.

[0171] Network Optimization

[0172] -Implement network message priority queue to ensure that critical information is transmitted first.

[0173] -Select appropriate transmission protocols (TCP / UDP) for different types of data.

[0174] -Develop frame synchronization or state synchronization mechanism to ensure consistency among multiple users.

[0175] Comprehensive testing and tuning

[0176] - Conduct comprehensive functional and performance testing.

[0177] -Use performance analysis tools to identify bottlenecks and perform targeted optimizations.

[0178] -Collect user feedback and iteratively improve the responsiveness and smoothness of the system.

[0179] Through these steps, not only the 3D model is integrated into the virtual reality engine to build a virtual wind farm, but also high-performance VR interaction, multi-person collaboration and motion capture functions are realized. These optimization measures ensure that the system can maintain high frame rate and low latency in a complex virtual environment, providing users with a smooth and immersive training experience.

[0180] Example 3

[0181] Based on Example 1, this example provides a virtual reality simulation training system for wind farm operation and maintenance personnel, in which the multi-sensor fusion motion capture interaction module mainly uses algorithms to achieve high-precision human-computer interaction. This module uses existing wearable devices on the market, but through data fusion and processing algorithms, it significantly improves the accuracy and naturalness of the interaction. The specific implementation is as follows:

[0182] 1. Hardware configuration:

[0183] -VR Helmet: Use a commercially available high-end VR helmet, such as Oculus Quest 2 or HTC Vive Pro.

[0184] -Hand tracking: Use the built-in hand tracking function of the VR headset or a standalone hand tracking device such as Leap Motion.

[0185] -Body tracking: using HTC Vive Trackers or similar trackers.

[0186] -Eye tracking: If the VR helmet does not have eye tracking, you can choose to use Tobii Eye Tracker.

[0187] -Haptic feedback: Use commercially available haptic feedback gloves, such as HaptX Gloves.

[0188] 2. Algorithm fusion:

[0189] a) Multi-sensor data fusion algorithm:

[0190] -Develop a multi-sensor fusion algorithm based on Kalman filtering to comprehensively process data from VR helmets, hand tracking devices and body trackers.

[0191] -Algorithm input: raw position and attitude data of each sensor.

[0192] -Algorithm output: high-precision, low-latency full-body motion data.

[0193] b) Gesture recognition enhancement algorithm:

[0194] -Gesture recognition algorithm based on deep learning to improve the accuracy and robustness of gesture recognition.

[0195] -Algorithm input: raw data from hand tracking device.

[0196] - Algorithm output: precise gesture type and parameters.

[0197] c) Eye movement intention prediction algorithm:

[0198] -Eye movement intention prediction algorithm based on LSTM (Long Short-Term Memory) network.

[0199] - Algorithm input: raw data from the eye tracker.

[0200] -Algorithm output: user gaze intention and point of interest prediction.

[0201] d) Adaptive tactile feedback algorithm:

[0202] - Adaptive tactile feedback algorithm based on physical simulation, dynamically adjusting tactile feedback according to the material and interaction mode of virtual objects.

[0203] -Algorithm input: virtual object attributes and interaction data.

[0204] - Algorithm output: precise control signal for haptic feedback device.

[0205] e) Motion prediction and smoothing algorithm:

[0206] -Action prediction algorithm based on machine learning to reduce system latency.

[0207] - Implemented an adaptive motion smoothing algorithm to remove sensor noise and jitter.

[0208] - Algorithm input: raw sensor data stream.

[0209] - Algorithm output: smooth, predicted motion trajectory.

[0210] 3. System Integration:

[0211] -Use middleware to achieve seamless integration of various hardware devices and virtual reality engines.

[0212] -Design a modular software architecture to facilitate future expansion of new interactive devices or algorithms.

[0213] Through these algorithms, the system can achieve high-precision full-body motion capture, natural gesture interaction, precise eye tracking, and realistic tactile feedback based on the use of existing hardware. This approach not only improves the performance of the system, but also maintains good compatibility and scalability, providing a more realistic and intuitive virtual reality training experience for wind farm operation and maintenance personnel.

[0214] In this embodiment, the detailed method of the multi-sensor fusion motion capture interaction module is as follows:

[0215] Step 1: Visual capture system

[0216] 1.1 Hardware Configuration

[0217] -Use the built-in binocular RGB camera system of VR helmets (such as Oculus Quest 2 or HTC Vive Pro Eye) for visual capture.

[0218] -For situations where more accurate hand tracking is required, a dedicated hand tracking device, such as the Leap Motion Controller, can be installed in front of the VR helmet.

[0219] 1.2 Acquisition of 3D Position Information

[0220] -Development of depth estimation algorithm based on stereo vision:

[0221] a) Apply a stereo matching algorithm, such as Semi-Global Matching (SGM) or a deep learning-based stereo matching network, to compute the disparity map.

[0222] b) Reconstruct the depth information of the scene through disparity information and camera parameters.

[0223] - Implement human and tool detection algorithms:

[0224] a) Detect and localize human parts and tools in images using optimized deep learning models such as YOLO or variants of Faster R-CNN.

[0225] b) Combine the 2D detection results with the depth information to infer the 3D spatial position.

[0226] 1.3 Skeleton Tracking Algorithm

[0227] -Integrate skeleton tracking algorithm into the main processing unit of VR helmet:

[0228] a) Develop a deep learning-based human pose estimation network, such as an improved version of HRNet or OpenPose.

[0229] b) Implement a specialized hand skeleton tracking algorithm, such as the improved HandNet or a custom hand keypoint detection network.

[0230] -Data fusion and optimization:

[0231] a) The output of the skeleton tracking algorithm is fused with the IMU data and an extended Kalman filter is used to improve the stability and accuracy of the pose estimation.

[0232] b) Implement a time series based posture smoothing algorithm to reduce jitter and instability.

[0233] 1.4 Tool Tracking

[0234] - Implement feature point based tool tracking algorithms, such as an improved version of ORB-SLAM.

[0235] - Enhanced tracking stability using depth information for precise 6DoF positioning of tools.

[0236] 1.5 System Optimization

[0237] -Develop an adaptive processing mechanism to dynamically adjust the algorithm precision according to the system load, achieving a balance between performance and accuracy.

[0238] -Implement the field of view optimization algorithm and expand the effective capture area through software methods to cover the user's main operating space.

[0239] Through this improved visual capture system, the hardware of the VR helmet can be fully utilized to achieve high-precision human motion and tool position tracking, providing accurate and real-time interactive experience for virtual reality training of wind farm operation and maintenance personnel. This method does not require additional external camera settings, improves the portability and ease of use of the system, and ensures the accuracy and stability of capture through advanced software algorithms.

[0240] Step 2, Inertial Sensor System

[0241] -Use a micro inertial measurement unit (IMU) that integrates sensors such as accelerometers and gyroscopes; -Integrate IMU sensors into wearable devices such as gloves, helmets, and belts of maintenance personnel; -Integrate algorithms such as Kalman filtering to accurately estimate the motion posture of human joints from IMU data.

[0242] 3. Hand data gloves

[0243] -Use flexible sensors or fiber optic sensors to make data gloves; -Accurately capture hand movements by measuring the bending deformation of fingers and palms; -Support the capture and tracking of subtle movements such as grasping and operating tools and spare parts.

[0244] 4. Multi-sensor fusion algorithm

[0245] - Integrate data from multiple sensors such as vision, inertia, and hand; - Optimize sensor data fusion based on algorithms such as Kalman filtering and particle filtering; - Obtain accurate, complete, and real-time trajectories of human motion and hand interaction.

[0246] 5. Multimodal Natural Interaction

[0247] -Support gesture interaction: realize the operation control of the system through the recognition of hand movement trajectory; -Support voice interaction: integrate the voice recognition module to complete the system interaction through voice commands; -Support eye movement interaction: use the eye tracking camera to capture eye movements for system interaction; -Support brain-computer interface: collect brain wave signals for control and interaction.

[0248] 6. Force feedback and physical effect simulation

[0249] -Use tactile feedback gloves to provide force feedback through piezoelectric ceramic actuators or pneumatic muscles; -Integrate vibration trackers to simulate vibration and impact effects in wind power operation and maintenance; -Use devices such as thermal resistors to simulate temperature changes during operation.

[0250] This technical solution achieves high-precision capture of operation and maintenance personnel's actions and human-computer interactions through the fusion of multiple sensors, supports natural and intuitive multimodal interaction methods, and simulates the physical effects in real operation and maintenance scenarios, greatly enhancing the immersion and realism of virtual reality.

[0251] Example 4

[0252] Based on Example 1, this example provides a virtual reality simulation training system for wind farm operation and maintenance personnel. The virtual trainer and intelligent training module generate a virtual trainer with professional knowledge and training experience based on digital human technology, and guide the operation and maintenance personnel through voice / text / action. At the same time, an intelligent algorithm based on reinforcement learning is introduced to dynamically adjust the training difficulty, scenario and teaching strategy according to the real-time performance of the operation and maintenance personnel to achieve adaptive personalized training. The module also presets various typical fault cases, and dynamically generates new fault scenarios in combination with artificial intelligence technology to improve the diversity and challenge of training.

[0253] In this embodiment, the detailed methods of the virtual coach and the intelligent training module are as follows:

[0254] Step 1: Generate a virtual trainer

[0255] -Use digital human technology to model a virtual coach with professional appearance and body movements;

[0256] - Based on acoustic modeling and speech synthesis technology, the virtual trainer is given natural voice interaction capabilities; - Natural language processing technology is introduced to enable the virtual trainer to understand and respond to questions from operation and maintenance personnel; - A professional knowledge base of wind power is implanted to enable the virtual trainer to have professional theoretical and practical guidance capabilities.

[0257] Voice Recognition:

[0258] -Integrate high-precision speech recognition engines, such as end-to-end speech recognition models based on deep learning.

[0259] - Optimize the model to accommodate the specialized terminology and possible background noise of the wind farm environment.

[0260] Natural Language Understanding:

[0261] - Implement deep learning based natural language processing models, such as BERT or improved versions of GPT.

[0262] - Train the model to understand professional questions and instructions related to wind farm operation and maintenance.

[0263] -Develop intent recognition algorithms to accurately capture the questions and request intentions of operation and maintenance personnel.

[0264] Dialogue Management:

[0265] -Design intelligent dialogue management systems to maintain dialogue context and achieve coherent interactions.

[0266] -Implement dialogue strategies based on reinforcement learning and optimize the interaction process.

[0267] Knowledge base integration:

[0268] -Build a comprehensive wind power operation and maintenance knowledge graph, including information on equipment operation, fault diagnosis, safety regulations, etc.

[0269] -Implement a reasoning mechanism based on knowledge graph to support answering complex questions.

[0270] Practical guide generation:

[0271] -Develop task decomposition algorithms to decompose complex O&M operations into a sequence of steps.

[0272] -Enable context-aware guidance generation to provide appropriate guidance based on the operator's current status and environment.

[0273] Speech Synthesis:

[0274] - Use advanced neural network speech synthesis technology such as WaveNet or Tacotron 2.

[0275] -Customized voice models to create professional, clear and personalized virtual trainer voices.

[0276] Emotion recognition and expression:

[0277] -Integrate speech emotion recognition algorithm to capture the emotional state of operation and maintenance personnel.

[0278] -Dynamically adjust the virtual trainer's tone and delivery based on the situation.

[0279] Multimodal interactive fusion:

[0280] - Combine visual information to achieve coordination between voice and action instructions, and enhance the intuitiveness of guidance.

[0281] -Develop gesture recognition algorithms to support compound commands combining voice and gestures.

[0282] Real-time feedback mechanism:

[0283] -Implement a real-time voice feedback system to provide immediate evaluation and suggestions to the operations of operation and maintenance personnel.

[0284] -Develop error detection algorithms to identify and correct erroneous operations in a timely manner.

[0285] Continuous learning and optimization:

[0286] -Design incremental learning mechanisms to enable the virtual trainer to continuously learn and improve from new interactions.

[0287] - Fine-tune the model based on user feedback to continuously improve the quality of interaction.

[0288] Through these comprehensive voice interaction technologies, the virtual trainer can not only understand and respond to the questions of the operation and maintenance personnel, but also provide accurate and personalized practical guidance, achieving natural, smooth and efficient training interaction. This comprehensive interactive capability will significantly improve the effect and experience of virtual reality training for wind farm operation and maintenance personnel.

[0289] Step 2: Motion capture and animation generation

[0290] - Invite senior wind power operation and maintenance experts to capture typical operation actions; - Based on the captured data, generate smooth animation sequences through action relocation and key frame interpolation technologies;

[0291] -Bind animation sequences to the virtual trainer model for refined action demonstration.

[0292] Step 3: Reinforcement Learning Intelligent Training

[0293] - Build a decision-making model, including state space (training scenarios), behavior space (teaching strategies) and reward function (training effect evaluation); - Train the decision-making model based on reinforcement learning algorithms (such as deep Q learning) so that the decision-making model can adjust the scenario difficulty and teaching rhythm according to the real-time performance of operation and maintenance personnel; - Set up a reasonable reward mechanism to improve training quality while taking into account training efficiency.

[0294] The decision model is trained based on reinforcement learning algorithms (such as deep Q-learning), specifically:

[0295] Training Data:

[0296] Training data comes mainly from two sources:

[0297] a) Simulated data: Data generated by simulating the performance of students of different skill levels in various scenarios.

[0298] b) Actual training data: trainee performance data collected from real VR training sessions.

[0299] Relationship with previously built models:

[0300] Previously built models (such as natural language processing, action recognition, etc.) provide the necessary input information for reinforcement learning. For example:

[0301] -The action recognition model provides students with operation accuracy data

[0302] -Speech recognition and natural language processing models provide information about the quality of students’ questions and answers

[0303] Reinforcement learning process:

[0304] State:

[0305] - The student's current skill level

[0306] -Completion of the current task

[0307] - Indicators of students’ fatigue, concentration, etc.

[0308] Action:

[0309] - Adjust the difficulty of the scene

[0310] -Change the teaching rhythm

[0311] - Provide additional guidance or challenges

[0312] Reward:

[0313] - The speed and accuracy with which students complete tasks

[0314] - The level of progress of the students

[0315] -Student engagement and satisfaction

[0316] Training process:

[0317] -The agent learns the optimal strategy through repeated interaction with the simulated environment or the actual training system

[0318] -Using algorithms such as Deep Q-Learning, the agent learns to choose the best actions in different states to maximize long-term rewards

[0319] Build the environment simulator:

[0320] -Create an environment that simulates the performance and reactions of different students

[0321] -Integrate previously developed models such as motion recognition and voice interaction to provide real status feedback

[0322] Define the state space and action space:

[0323] - Status: including student performance indicators, task progress, environmental parameters, etc.

[0324] -Actions: Define adjustable teaching parameters such as difficulty level, instruction frequency, etc.

[0325] Design reward function:

[0326] -Design a reasonable reward mechanism based on factors such as students’ progress, completion, and satisfaction

[0327] Implementing the Deep Q-Learning Algorithm:

[0328] -Build a deep neural network as an approximator of the Q function

[0329] - Implement techniques such as experience replay and target network to improve learning stability

[0330] Training process:

[0331] - Initial training using simulated data

[0332] - Fine-tuning and continuous learning in the actual training system

[0333] Integration and Deployment:

[0334] -Integrate the trained agent into the virtual coach system

[0335] - Implement real-time decision-making mechanism to dynamically adjust training parameters based on trainee performance

[0336] Through this process, an intelligent teaching system was created that can adaptively adjust the difficulty and rhythm of training, providing a personalized learning experience based on the real-time performance of each operation and maintenance personnel.

[0337] Personalized training strategy

[0338] -Record the operation data, error types and weak links of operation and maintenance personnel;

[0339] - Based on data analysis, tailor-made personalized training plans for each student; - Intelligently generate corresponding training cases and key explanations for common error types; - Through the guidance of virtual coaches, strengthen the training of weak links in a targeted manner.

[0340] Dynamic fault case generation

[0341] - Preset various common fault cases and their root causes for wind turbines; - Dynamically combine and generate new composite fault scenarios based on Bayesian networks or other artificial intelligence technologies; - Intelligently adjust the time nodes and locations of faults to increase the diversity and uncertainty of training;

[0342] -Dynamically adjust the difficulty of faults based on the actual performance of the operation and maintenance personnel to avoid situations that are too easy or difficult.

[0343] Intelligently adjust the time and location of fault occurrence:

[0344] a) Fault library construction:

[0345] - Create a comprehensive fault database covering all possible fault types, locations and severity levels.

[0346] -Assign attribute labels to each fault, such as complexity, criticality, frequency of occurrence, etc.

[0347] b) Time node selection algorithm:

[0348] -Implement a Markov chain-based fault timing model to simulate the temporal pattern of faults in the real world.

[0349] - Generate random but plausible failure occurrence times using Monte Carlo methods.

[0350] c) Site selection mechanism:

[0351] -Develop an algorithm based on weighted random selection to select the fault location according to the fault probability distribution of different parts.

[0352] -Implement association rule mining to ensure the logical correlation between multiple faults.

[0353] d) Context-aware adjustments:

[0354] -Design a rule-based expert system to adjust fault parameters according to the current training scenario and student status.

[0355] -Use reinforcement learning algorithms to optimize fault generation strategies and maximize training results.

[0356] e) Dynamic Difficulty Control:

[0357] - Implemented an adaptive difficulty adjustment algorithm to dynamically adjust the complexity of the faults based on the student's performance history.

[0358] -Use fuzzy logic controller to smoothly adjust fault parameters and avoid abrupt difficulty changes.

[0359] Dynamically adjust the difficulty of faults based on the actual performance of the operation and maintenance personnel:

[0360] a) Performance evaluation model:

[0361] -Develop a multi-dimensional performance evaluation model that comprehensively considers factors such as response time, diagnostic accuracy, and resolution efficiency.

[0362] -Use deep neural networks to extract high-level features from students’ operation sequences and assess their skill level.

[0363] b) Difficulty Quantification System:

[0364] -Design a multi-parameter difficulty quantification system to quantify various aspects of failure (such as complexity, time pressure, information uncertainty, etc.).

[0365] -Implement a difficulty assessment model based on fuzzy set theory to deal with the ambiguity of difficulty definition.

[0366] c) Dynamic Difficulty Adjustment Algorithm:

[0367] -Develop a difficulty adjuster based on PID (Proportional-Integral-Derivative) control to dynamically adjust parameters based on the difference between student performance and target difficulty.

[0368] - Implemented a parameter tuning strategy based on Bayesian optimization to find a balance between keeping the game challenging and achievable.

[0369] d) Personalized difficulty curve:

[0370] -Use online learning algorithms to build a personalized difficulty-performance curve for each student.

[0371] -Apply regression analysis techniques to predict students' likely performance at different difficulty levels.

[0372] e) Real-time adjustment mechanism:

[0373] -Implement a real-time difficulty adjustment system based on event triggering, which can instantly respond to changes in student performance during training.

[0374] -Develop a smooth transition algorithm to ensure that difficulty adjustments do not cause a sudden interruption in the training experience.

[0375] f) Multi-objective optimization:

[0376] - Use multi-objective optimization algorithms (such as NSGA-II) to balance multiple objectives such as training effect, student experience and training efficiency.

[0377] -Implement a dynamic weight adjustment mechanism to adjust the importance of different objectives according to the training stage and student characteristics.

[0378] Through these specific implementation methods, the system can intelligently adjust the time, location and difficulty of the fault, providing each wind farm operation and maintenance personnel with a highly personalized, challenging but not discouraging training experience. This dynamic and intelligent adjustment not only increases the authenticity and unpredictability of the training, but also ensures that each trainee is always in his or her "zone of proximal development", thereby maximizing the learning effect.

[0379] This technical solution uses digital human technology to generate professional virtual coaches, integrates voice interaction, action demonstration and knowledge base, and combines artificial intelligence algorithms such as reinforcement learning to achieve innovative functions such as intelligence, personalization, and dynamic scenario generation, greatly improving the quality and flexibility of training.

[0380] Example 5

[0381] Based on Example 1, this example provides a virtual reality simulation training system for wind farm operation and maintenance personnel, a training data management and blockchain module, which is used to record and analyze the operation data of the operation and maintenance personnel in real time, such as movement trajectory, time, accuracy, etc., and generate a personalized training evaluation report. Blockchain technology is used to build a distributed, tamper-proof training data management system to ensure data security and traceability. Blockchain-based data provides strong support for the skill level certification of operation and maintenance personnel.

[0382] In this embodiment, the detailed technical solutions of the training data management and blockchain module are as follows:

[0383] 1. Training data collection and analysis

[0384] - Real-time collection of various operation data of operation and maintenance personnel in the virtual environment, including action trajectory, time, accuracy, etc.;

[0385] -Establish data standardization specifications to clean, standardize and structure the collected raw data;

[0386] -Based on data mining and machine learning algorithms, conduct in-depth analysis of the operation data of operation and maintenance personnel to discover regular problems and weak links.

[0387] 2. Personalized training evaluation report generation

[0388] - Based on the analysis results, tailor a personalized training evaluation report for each operation and maintenance personnel;

[0389] -The report includes training process data, operation quality score, weak link analysis, improvement suggestions, etc.;

[0390] -Present reports to operation and maintenance personnel in the form of intuitive charts and text via the web or mobile terminal.

[0391] 3. Blockchain distributed data management

[0392] -Build a dedicated private chain for training data management based on mature blockchain platforms such as Ethereum;

[0393] -Design smart contracts to write the identity information, training data, evaluation reports and other data of operation and maintenance personnel into the blockchain;

[0394] - Use encryption technology to ensure the privacy of personal data and prevent data from being stolen or tampered with.

[0395] 4. Data traceability and skill certification based on blockchain

[0396] -Use the immutability and traceability of blockchain to provide reliable evidence for the historical training data of operation and maintenance personnel;

[0397] -Establish a skill level system and certification mechanism for operation and maintenance personnel based on blockchain data;

[0398] -Issuing encrypted learning certificates to ensure the authenticity of certificates and promote talent training and mobility.

[0399] 5. Cross-institutional data sharing and collaboration

[0400] - Allow different wind power companies and training institutions to join the blockchain alliance to achieve data interoperability and mutual recognition;

[0401] - Set different data access permissions through smart contracts to support cross-institutional training data sharing;

[0402] - Provide supervision nodes for industry regulatory authorities to strengthen the supervision and evaluation of training quality.

[0403] 6. Product application of training data

[0404] - Based on blockchain data, develop relevant data products, such as training quality analysis reports, talent pool, etc.;

[0405] -Open application interfaces in the form of APIs to support calls by third-party data consumers or regulatory authorities;

[0406] - Develop the data economy and achieve efficient use of training data resources.

[0407] This technical solution achieves reliable storage and traceability of training data by building a blockchain distributed data management system, and supports innovative functions such as skill certification, cross-institutional collaboration, and data product application based on blockchain data, providing technical guarantee for the full utilization of training data resources.

[0408] Example 6

[0409] Based on Example 1, this embodiment provides a virtual reality simulation training system for wind farm operation and maintenance personnel, cloud deployment and multi-person collaboration module, which is used to adopt a cloud deployment mode, support multi-user synchronous access and remote online training, and realize resource sharing and efficient utilization. It supports real-time collaboration among multiple people, simulates teamwork operation and maintenance scenes, and cultivates communication and coordination skills. Through 5G and edge computing technology, low-latency, high-bandwidth data transmission is achieved to ensure a smooth virtual reality experience.

[0410] In this embodiment, the detailed technical solutions of cloud deployment and multi-person collaboration module are as follows:

[0411] 1. Cloud deployment architecture

[0412] -Use cloud deployment mode to deploy the virtual reality simulation system in a public cloud or private cloud environment;

[0413] -Based on the microservice architecture, the system functional modules are decoupled into independent service components to improve the flexibility and scalability of the system;

[0414] -Deploy load balancing clusters and auto-scaling components to achieve high availability and resource elasticity of the system;

[0415] -Introduce containerization technology (such as Docker) to achieve unified packaging, release and management of applications.

[0416] 2. Real-time multi-person collaboration engine

[0417] -Develop a real-time communication engine based on WebRTC, supporting multiple communication modes such as audio, video, and data channels;

[0418] -Introducing network adaptive bit rate control technology to dynamically adjust the bit rate according to network conditions to ensure communication quality;

[0419] -Use a decentralized P2P topology to reduce the bandwidth and computing pressure on the central node;

[0420] -Supports multiple people to enter the same virtual scene in real time, presenting their own avatars and perspectives.

[0421] 3. Virtual Reality Multiplayer Collaboration

[0422] - Based on the real-time communication engine, synchronize the avatar position, action, interaction status and other data of each participant;

[0423] - Design a reasonable regional authority control mechanism to prevent participants from maliciously interfering with each other;

[0424] -Support multi-person voice / text chat to achieve efficient team collaboration and communication;

[0425] -Introduce role-based operation permission management, such as lecturer mode, experience mode, etc.

[0426] 4.5G+Edge Computing Optimization

[0427] -Pre-deploy hotspot simulation scenes on edge nodes and use edge computing resources to perform computationally intensive tasks such as rendering;

[0428] - Realize real-time interaction between users and edge nodes through the low latency and high bandwidth characteristics of 5G networks;

[0429] - For unpopular scenarios, migrate computing tasks back to the cloud to take advantage of cloud computing power;

[0430] -Use dynamic optimization algorithm to intelligently select the optimal cloud-edge collaboration mode.

[0431] 5. Ultra-large-scale concurrent support

[0432] -Introducing sharding instance technology to split a single virtual scene instance into multiple sharding instances;

[0433] - Based on the principle of spatial locality, adjacent participants are assigned to the same shard instance;

[0434] -Using asynchronous parallel instance technology, it supports running massive concurrent shard instances at the same time;

[0435] -For large-scale concurrent scenarios, deploy a distributed streaming processing platform to achieve real-time data processing.

[0436] This technical solution uses advanced technologies such as cloud deployment, real-time communication, and 5G+edge computing to enable large-scale multi-person remote synchronous access to a virtual reality environment and efficient collaborative training. It also has capabilities such as high-concurrency processing, instance sharding, and dynamic resource scheduling, ensuring the high performance and scalability of the system.

[0437] In summary, the virtual reality simulation training system for wind farm operation and maintenance personnel integrates multiple technologies and has significant beneficial effects, which are mainly reflected in the following aspects:

[0438] 1. Improve training quality and practical ability

[0439] The system can highly restore the real wind farm environment and simulate various typical fault scenarios, providing an immersive training experience for operation and maintenance personnel. Through the guidance and demonstration of virtual coaches and intelligent personalized training, the professionalism and pertinence of the training are effectively improved, and the practical skills and emergency response capabilities of operation and maintenance personnel are greatly enhanced.

[0440] 2. Reduce training risks and costs

[0441] Compared with traditional physical training, this virtual simulation system does not need to occupy actual wind farm resources, avoiding the risk of equipment damage and personal injury. At the same time, through cloud deployment and multi-person collaboration, it reduces the investment in fixed costs such as teachers and venues, significantly improving the economy and resource utilization efficiency of training.

[0442] 3. Make training intelligent and personalized

[0443] The system incorporates artificial intelligence technologies, such as reinforcement learning algorithms and digital human technology. It can dynamically adjust the training difficulty, strategies and scenarios based on the real-time performance of operation and maintenance personnel, and tailor personalized training plans for each student, greatly improving the accuracy and effectiveness of training.

[0444] 4. Promote training data and release value

[0445] With the help of blockchain technology, the system has built a distributed training data management system to ensure the security and traceability of data, laying the foundation for skill certification of operation and maintenance personnel and cross-institutional collaboration. At the same time, training data can be used to develop related data products, release data value, and promote industry talent training.

[0446] 5. Support large-scale remote collaborative training

[0447] Based on 5G, cloud computing and other technologies, the system supports a large number of operation and maintenance personnel to remotely and synchronously enter the virtual environment and carry out efficient team collaboration training, which is conducive to cultivating comprehensive qualities of communication, coordination, and efficient collaboration, and meeting the industry's demand for compound talents.

[0448] 6. Achieve low-carbon, environmentally friendly and sustainable development

[0449] The operation of this system does not produce direct environmental pollution and energy consumption, which is conducive to the green and low-carbon development of wind power industry operation and maintenance personnel training, promotes the efficient use of renewable energy, and contributes to the realization of carbon neutrality.

[0450] In short, this innovative system has fully tapped the huge potential of emerging technologies such as virtual reality and artificial intelligence in the field of wind power operation and maintenance training, and has important theoretical significance and application value in improving the quality of industry talents, reducing training costs, and promoting green development.

[0451] In summary, the system integrates multiple innovative technologies to provide immersive training, which can greatly improve the practical ability and training efficiency of wind power operation and maintenance personnel and solve many shortcomings in existing training.

[0452] Example 7

[0453] This embodiment 7 provides a computer device, including: a memory for storing a computer program; and a processor for implementing the steps of a computer method when executing the computer program.

[0454] When the processor executes the computer program, the steps of the above computer method are implemented.

[0455] Alternatively, when the processor executes the computer program, the functions of each module in the above system are realized.

[0456] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The computer device may include, but is not limited to, a processor and a memory. Those skilled in the art will appreciate that the above are examples of computer devices and do not constitute a limitation on computer devices. The computer device may include more components than the above, or a combination of certain components, or different components. For example, the computer device may also include input and output devices, network access devices, buses, etc.

[0457] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc. The processor is the control center of the computer device, and uses various interfaces and lines to connect various parts of the entire computer device.

[0458] The memory may be used to store the computer programs and / or modules, and the processor implements various functions of the computer device by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory.

[0459] The memory may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (SmartMed i aCard, SMC), a secure digital (SecureDi gita l, SD) card, a flash card (FlashCard), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0460] Example 8

[0461] This embodiment 8 also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the method described are implemented.

[0462] If the module / unit integrated in the computer system is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium.

[0463] Based on such understanding, the present invention implements all or part of the processes in the above method, and can also be completed by instructing related hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and when the computer program is executed by a processor, the steps of the above computer method can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or preset intermediate form, etc.

[0464] The computer-readable storage medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0465] It should be noted that the content contained in the computer-readable storage medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable storage media do not include electrical carrier signals and telecommunication signals.

[0466] Example 9

[0467] This embodiment 9 provides a computer product, which includes a computer program, and the computer program is stored in a computer-readable storage medium; the processor of the computer device reads the computer program from the computer-readable storage medium, and the processor executes the computer program, so that the computer device can execute the method in embodiment 1, which will not be repeated here.

[0468] It should be noted that a person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing related hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods.

[0469] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A virtual reality simulation training method for wind farm operation and maintenance personnel, characterized in that: The following steps are involved: Step 1, construct a virtual wind farm; Step 2, realizing human-computer interaction between the user and the virtual wind farm; Step 3: Generate a virtual trainer with professional knowledge and training experience, and guide the user's operation to achieve simulation training for wind farm operation and maintenance personnel.

2. A virtual reality simulation training method for wind farm operation and maintenance personnel according to claim 1, characterized in that: In step 1, a virtual wind farm is constructed. The specific method is: Based on the geographic information data of the actual wind farm, the unit structure drawings and equipment parameters, a virtual wind farm is constructed by combining 3D modeling and physical rendering technology.

3. A virtual reality simulation training method for wind farm operation and maintenance personnel according to claim 1, characterized in that: In step 2, human-computer interaction between the user and the virtual wind farm is realized. The specific method is: Wearable devices are combined with Kalman filtering algorithm to realize human-computer interaction between users and virtual wind farms.

4. A virtual reality simulation training method for wind farm operation and maintenance personnel according to claim 1, characterized in that: In step 3, a virtual trainer with professional knowledge and training experience is generated to guide the user's operation. The specific method is: Generate a virtual coach based on digital human technology; Capture typical operation and maintenance actions of wind farms to obtain capture data; Generate animation sequence from the captured data; Bind the animation sequence to the virtual trainer; Build a decision-making model for the virtual trainer and use the reinforcement learning algorithm to train the decision-making model so that the decision-making model can adjust the scene difficulty and teaching rhythm according to the real-time performance of the target user.

5. A virtual reality simulation training system for wind farm operation and maintenance personnel, characterized in that: include: Virtual wind farm environment modeling module, used to build a virtual wind farm; Multi-sensor fusion motion capture interaction module, used to realize human-computer interaction between users and virtual wind farms; The virtual trainer and intelligent training module is used to generate a virtual trainer with professional knowledge and training experience, and to guide the user's operations, thereby realizing simulation training for wind farm operation and maintenance personnel.

6. A virtual reality simulation training system for wind farm operation and maintenance personnel according to claim 5, characterized in that: It also includes training data management and blockchain modules, which are used to record and analyze the operation data of operation and maintenance personnel in real time and generate personalized training evaluation reports.

7. A virtual reality simulation training system for wind farm operation and maintenance personnel according to claim 5, characterized in that: It also includes cloud deployment and multi-person collaboration modules for simultaneous access by multiple users and remote online training.

8. A computer device, characterized in that: include: a processor suitable for executing a computer program; A computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by the processor, the method according to any one of claims 1 to 4 is performed.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 4 is implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 4 is implemented.

Images