Digital twinborn modeling system for wind power plant and use method

Abstract

The invention belongs to the technical field of wind power generation, and discloses a wind power plant digital twin modeling system and a use method, the system comprises an equipment module, the equipment module comprises a biological bionic adaptive fan blade system, a quantum sensing network, a laser radar system, a high-resolution multi-modal imaging system and an intelligent autonomous flying robot group system; the network module comprises a 5G private network, a quantum communication network and an edge computing network; the data module comprises a distributed database system, a big data processing platform and an AI training platform; the AI module comprises a wind regime prediction model, a power prediction model, an equipment health state monitoring model and a self-adaptive control algorithm model; the application module comprises a digital twinborn visualization platform, a holographic mixed reality operation and maintenance system and an intelligent scheduling system. Modeling and intelligent management of the wind power plant can be realized, the power generation efficiency can be improved, the operation and maintenance cost can be reduced, and the service life of equipment can be prolonged.

Description

Technical Field

[0001] The present invention belongs to the technical field of wind power generation, and in particular relates to a wind farm digital twin modeling system and a use method thereof. Background Art

[0002] The digital twin system of a wind farm is a system that uses digital technology to create a virtual replica of a wind farm, achieving real-time interaction and synchronization between physical entities and virtual models. This system provides a powerful tool for wind farm management, optimization and decision-making by integrating technologies such as the Internet of Things, big data analysis and artificial intelligence.

[0003] At present, wind farm digital twin systems face many challenges in practical applications. Traditional data acquisition methods have limited accuracy and are difficult to capture instantaneous changes in wind farms; the lack of real-time data transmission and processing affects the response speed of the system; the accuracy and adaptability of the model also need to be improved, and it is difficult to accurately predict the complex and changeable wind farm environment. In addition, the existing system's capabilities in equipment health monitoring, predictive maintenance, and adaptive control are still limited. Therefore, there is an urgent need for a wind farm digital twin modeling system that can achieve real-time and intelligent management.

[0004] The Chinese patent publication number is CN117614130A, and the name is a patent application for a digital twin platform that assists in the operation and management of new energy wind farms. The platform includes an application layer, a business layer, a data layer, and a hardware layer. The hardware layer includes a hardware server, a server storage device, a Gis-Server server, and security software. The data layer includes an image data module, a vector data module, a professional surface data module, and a wind and light storage data module. The data layer is connected to the hardware layer in communication, the business layer is connected to the data layer in communication, and the application layer is connected to the business layer in communication. This patent application cannot achieve fast data transmission and processing of wind farms and long-term maintenance of equipment. Summary of the invention

[0005] In order to overcome the problems existing in the above-mentioned prior art, the purpose of the present invention is to provide a wind farm digital twin modeling system and a method of use. By integrating equipment modules, network modules, data modules, AI modules and application modules, it is possible to realize modeling and intelligent management of wind farms, improve power generation efficiency, reduce operation and maintenance costs, and extend equipment life.

[0006] To achieve the above object, the technical solution adopted by the present invention is: a wind farm digital twin modeling system, comprising: Equipment modules, including a biomimetic adaptive wind turbine blade system, a quantum sensor network, a lidar system, a high-resolution multimodal imaging system, and an intelligent autonomous flying robot swarm system; A network module, which includes a 5G private network, a quantum communication network, and an edge computing network; A data module, which includes a distributed database system, a big data processing platform, and an AI training platform; An AI module, wherein the AI ​​module includes a wind condition prediction model, a power prediction model, an equipment health status monitoring model, and an adaptive control algorithm model; Application modules, including a digital twin visualization platform, a holographic mixed reality operation and maintenance system, and an intelligent scheduling system; Among them, the equipment module collects real-time data of the wind farm, the equipment module communicates with the data module through the network module, the data module is communicatively connected to the AI ​​module, the AI ​​module is used to perform analysis and prediction based on the data processed by the data module, and the application module is communicatively connected to the AI ​​module and the network module.

[0007] Optionally, the biomimetic adaptive fan blade system of the equipment module includes: an intelligent deformable skeleton structure, which is made of a shape memory alloy and has an integrated microfluidic network inside; the intelligent deformable skeleton structure is coated with a high-performance composite material outer layer structure, which is made of a carbon fiber reinforced polymer composite material.

[0008] Optionally, the quantum sensing network of the equipment module is arranged on a biomimetic adaptive wind turbine blade system, and the quantum sensing network includes: a quantum wind speed sensor, an optical quantum gyroscope wind direction sensor, a superconducting quantum interferometer temperature sensor, a graphene-based quantum capacitance humidity sensor, and an optical microcavity pressure sensor.

[0009] Optionally, the laser radar system of the equipment module adopts coherent Doppler laser radar.

[0010] Optionally, the intelligent autonomous flying robot swarm system includes multiple drones, each of which includes: a multi-rotor carbon fiber composite material frame, on which a multi-sensor fusion system, an edge computing unit, and a dual-mode communication system combining 5G and satellite communications are provided.

[0011] Optionally, the edge computing network of the network module includes: an edge computing node arranged on each wind turbine tower, the edge computing node is equipped with a multi-core processor, a high-performance GPU, a large-capacity high-speed memory, a high-speed solid-state storage and a high-speed network interface.

[0012] Optionally, the wind condition prediction model of the AI ​​module adopts a spatiotemporal graph neural network architecture, the input features include historical meteorological data, terrain data and satellite images, and the output results are wind speed and wind direction.

[0013] The method for using the wind farm digital twin modeling system comprises the following steps: Collect wind farm data through the equipment module and transmit the data to the data module through the network module; Receive and process data from the equipment module through the data module; Perform in-depth analysis and prediction based on the data processed by the data module through the AI ​​module; The analysis and prediction results of the AI ​​module are converted into visual information and control instructions through the application module.

[0014] Optionally, the blade morphology and stress data are collected through a biomimetic adaptive wind turbine blade system; wind speed, wind direction, temperature, humidity and air pressure data are collected through a quantum sensor network; three-dimensional wind field data are obtained through a laser radar system scan; image and video data of the appearance and operating status of the wind turbine are collected through a high-resolution camera system; inspection tasks are performed by an autonomous flying robot group to collect detailed images and thermal imaging data of the wind turbine; and data are transmitted to a data module through a network module; The distributed database system is used to store and manage structured data. The big data processing platform is used to clean, convert and pre-process data. The AI ​​training platform is used to prepare data sets for subsequent model training and optimization. Generate future wind forecasts through wind prediction models; predict future power generation through power prediction models; evaluate equipment conditions and predict potential failures through equipment health monitoring models; and generate optimal control strategies based on current conditions and prediction results through adaptive control algorithm models; The virtual model of the wind farm is updated in real time through the digital twin visualization platform; augmented reality guidance is provided to on-site staff through the holographic mixed reality operation and maintenance system; and the wind turbine parameters and operation strategies are automatically adjusted according to the optimization results through the intelligent scheduling system.

[0015] Optionally, the adaptive control algorithm model of the AI ​​module adopts a method that combines model predictive control and deep reinforcement learning. The state space includes wind speed, wind direction, blade angle and generator speed, the action space includes blade pitch angle, yaw angle and generator torque, and the reward function is the weighted sum of power generation minus load and wear; the intelligent scheduling system of the application module adopts a multi-objective optimization algorithm, including a genetic algorithm and a particle swarm optimization algorithm. The optimization goals are to maximize power generation, minimize grid fluctuations and minimize equipment wear, and the constraints include grid stability, equipment operating parameters and environmental protection requirements.

[0016] Compared with the prior art, the present invention has the following beneficial effects: The wind farm digital twin modeling system provided by the present invention significantly improves the overall performance and management level of the wind farm by integrating multiple modules. The system of the present invention can realize ultra-real-time monitoring of the wind farm environment, greatly improving the accuracy of wind condition prediction and power generation prediction. Through biomimetic adaptive wind turbine blades and intelligent control algorithms, the system can optimize wind energy capture efficiency under various wind conditions, while reducing equipment load and extending service life. Through health monitoring and predictive maintenance functions, the equipment failure rate and maintenance costs can be significantly reduced. The intelligent autonomous flying robot swarm system can realize all-weather, high-efficiency inspections, improving operation and maintenance safety.

[0017] The system of the present invention significantly improves the economic benefits and sustainability of wind farms by improving power generation efficiency, reducing operation and maintenance costs, and extending equipment life. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The drawings described herein are for explanation purposes only and are not intended to limit the scope of the present invention in any way. In addition, the shapes and proportional dimensions of the components in the drawings are only for illustration purposes to help understand the present invention, and are not intended to specifically limit the shapes and proportional dimensions of the components of the present invention. In the drawings: Figure 1 This is a schematic diagram of the overall architecture of a wind farm digital twin modeling system according to Embodiment 1 of the present invention; Figure 2 This is a workflow diagram of a wind farm digital twin modeling system according to Embodiment 2 of the present invention. DETAILED DESCRIPTION

[0019] In order to enable those skilled in the art to better understand the technical solutions in the present invention, 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 only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0020] 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.

[0021] In the description of the embodiments of the present invention, it should be noted that if the terms "upper", "lower", "horizontal", "inner", etc. appear, the orientation or position relationship indicated is based on the orientation or position relationship shown in the accompanying drawings, or is the orientation or position relationship in which the product of the invention is usually placed when used. 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.

[0022] When an element is referred to as being "disposed on" another element, it may be directly on the other element or there may also be a central element. When an element is considered to be "connected to" another element, it may be directly connected to the other element or there may also be a central element. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are for illustrative purposes only and are not intended to be the only embodiments. If the term "horizontal" appears, it does not mean that the component is required to be absolutely horizontal, but it 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 it can be slightly tilted.

[0023] It should be noted that similar reference numerals and letters represent similar items in the following figures, so once an item is defined in one figure, it does not need to be further defined and explained in the subsequent figures. In the description of the present invention, it is understood that the terms "include" and "comprise" indicate the presence of the described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or their collections.

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by those skilled in the art of the technical field of the present invention. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, the singular forms of "a", "an" and "the" are intended to include plural forms unless the context clearly indicates otherwise.

[0025] The present invention is described in detail below with reference to the accompanying drawings.

[0026] A wind farm digital twin modeling system of the present invention comprises: Equipment modules, including a biomimetic adaptive wind turbine blade system, a quantum sensor network, a lidar system, a high-resolution multimodal imaging system, and an intelligent autonomous flying robot swarm system; A network module, which includes a 5G private network, a quantum communication network, and an edge computing network; A data module, which includes a distributed database system, a big data processing platform, and an AI training platform; An AI module, wherein the AI ​​module includes a wind condition prediction model, a power prediction model, an equipment health status monitoring model, and an adaptive control algorithm model; Application modules, including a digital twin visualization platform, a holographic mixed reality operation and maintenance system, and an intelligent scheduling system; Among them, the equipment module collects real-time data of the wind farm, the equipment module communicates with the data module through the network module, the data module is communicatively connected to the AI ​​module, the AI ​​module is used to perform analysis and prediction based on the data processed by the data module, and the application module is communicatively connected to the AI ​​module and the network module.

[0027] The present invention can optimize wind energy capture efficiency under various wind conditions through biomimetic adaptive wind turbine blades and intelligent control, while reducing equipment load and extending service life. The health monitoring and predictive maintenance functions can significantly reduce equipment failure rate and maintenance costs. The intelligent autonomous flying robot swarm system can achieve all-weather and efficient inspections, improving operation and maintenance safety.

[0028] Example 1 like Figure 1 As shown, the wind farm digital twin modeling system of the present invention includes: equipment module, network module, data module, AI module and application module. The interaction relationship between the modules is as follows: Interaction between the equipment module and the network module: Various sensors, devices and autonomous robots in the equipment module transmit the collected real-time data to the upper module through the 5G private network and quantum communication network provided by the network module.

[0029] The network module provides a high-speed, low-latency, and secure data transmission channel for the device module, ensuring the real-time and integrity of the data.

[0030] Interaction between the network module and the data module: The network module transmits the massive data collected from the device module to the distributed database system and big data processing platform of the data module.

[0031] The data module sends control instructions and configuration updates to the device module through the network module.

[0032] Interaction between the data module and the AI ​​module: The data module provides the AI ​​module with preprocessed and structured data as input for AI model training and reasoning. The AI ​​module stores the trained model parameters and intermediate results in the data module for subsequent use and optimization.

[0033] Interaction between AI module and application module: AI module passes analysis results, prediction data and optimization suggestions to application module for visualization and decision support. Application module feeds back user operations and decisions to AI module for online model learning and strategy adjustment.

[0034] Interaction between application module and equipment module: The application module sends control instructions to the equipment module through the network module to achieve remote control of the fan and other equipment. The real-time status and operation results of the equipment module are fed back to the application module through the network module for real-time monitoring and decision verification.

[0035] Interaction between data module and application module: The data module provides historical data and statistical information to the application module for generating reports and long-term trend analysis. The application module sends user-defined data processing requirements and query requests to the data module.

[0036] Indirect interaction between AI module and equipment module: AI module indirectly controls the equipment of equipment module through application module and network module, such as adjusting the angle of wind turbine blades or changing generator parameters. The sensor data of equipment module is processed by network module and data module and becomes the basis for training and optimization of AI module model.

[0037] The multi-module architecture design of the present invention realizes the intelligent management of the whole process of data collection, transmission, storage, analysis, decision-making and control. The close cooperation between the modules ensures the efficient operation and rapid response capability of the system. For example, when the equipment module detects abnormal wind conditions, the data can be quickly transmitted to the data module and AI module through the network module for analysis. The optimization strategy generated by the AI ​​module is converted into specific control instructions through the application module, and finally transmitted back to the equipment module for execution through the network module. The whole process can be completed within milliseconds.

[0038] In addition, the wind farm digital twin modeling system of the present invention also implements multiple feedback loops: Short-term feedback loop: The real-time control strategy of the AI ​​module achieves real-time optimized control of the wind turbine through rapid interaction between the application module and the equipment module.

[0039] Medium-term feedback loop: The operating data accumulated by the data module is used for regular updating and optimization of the AI ​​module model to improve the accuracy of prediction and control.

[0040] Long-term feedback loop: The data analysis results and user feedback of the application module are used for long-term optimization and upgrading of the system, including the update of device module equipment and the iteration of AI algorithm.

[0041] Through this multi-module, multi-timescale closed-loop design, the system can continuously optimize itself, adapt to different operating environments and conditions, and achieve continuous improvement and efficiency enhancement of wind farm operations.

[0042] Specifically, the equipment module is used to collect real-time data of the wind farm, and the equipment module includes the following subsystems: A biomimetic adaptive fan blade system, the biomimetic adaptive fan blade system comprising: Intelligent deformation skeleton structure: The intelligent deformation skeleton structure is made of shape memory alloy and has temperature responsive deformation capability. A microchannel network is integrated inside the skeleton for the circulation of hydraulic working medium.

[0043] The intelligent deformable skeleton structure is covered with a high-performance composite material outer layer structure. The high-performance composite material outer layer structure adopts a carbon fiber reinforced polymer (CFRP) composite material, which significantly reduces the overall weight while ensuring the structural strength and has excellent elastic properties.

[0044] The intelligent deformable skeleton structure is controlled by a precision distributed hydraulic control system, which consists of multiple independent control units, each of which is equipped with a micro hydraulic actuator and an electromagnetic control valve. The precision distributed hydraulic control system has sub-millimeter position control accuracy and millisecond dynamic response capability. By accurately adjusting the fluid pressure distribution inside the skeleton, dynamic optimization of the blade morphology is achieved.

[0045] High-density optical fiber strain monitoring network: Fiber Bragg grating (FBG) is used for sensing, and the high-density optical fiber strain monitoring network is evenly distributed along the axial direction of the blade. The high-density optical fiber strain monitoring network has a sampling frequency of the kilohertz level and a measurement accuracy of the micro-strain level, and is used to monitor the stress distribution and deformation state of the blade in real time.

[0046] Through the coordinated operation of the above subsystems, the adaptive blade system can dynamically adjust its aerodynamic shape according to real-time wind field conditions, thereby significantly improving wind energy capture efficiency while effectively reducing structural loads. This adaptive capability greatly expands the effective operating range of the wind turbine, improves energy conversion efficiency, and helps extend the service life of the equipment.

[0047] Quantum sensor networks can achieve super-measurement of environmental parameters. Quantum sensor networks include: Quantum wind speed sensor based on nitrogen vacancy (NV) center diamond: This sensor uses the high sensitivity of the spin state of the NV center to magnetic field changes to accurately measure wind speed by detecting tiny magnetic field changes caused by wind. Its accuracy can reach sub-centimeter per second level and the sampling rate can reach hundreds of hertz.

[0048] Optical quantum gyroscope wind direction sensor: The optical quantum gyroscope designed based on the Sagnac effect principle has excellent angular resolution. Its accuracy can reach one decimal place and the sampling rate can reach hundreds of hertz.

[0049] Superconducting quantum interference device (SQUID) temperature sensor: Utilizing the sensitivity of superconducting Josephson junction to temperature changes, its accuracy can reach one hundredth of a degree Celsius and the sampling rate is tens of hertz.

[0050] Graphene-based quantum capacitance humidity sensor: Taking full advantage of the extremely sensitive property of graphene single-atom module structure to humidity changes, humidity measurement is achieved. Its accuracy can reach one decimal place of relative humidity, and the sampling rate is tens of hertz.

[0051] Optical microcavity pressure sensor: Based on the sensitive characteristics of the optical microcavity resonant frequency to pressure changes, pressure measurement is achieved. Its accuracy can reach 100% Pa level and the sampling rate is tens of Hz.

[0052] Together, these quantum sensors form a comprehensive environmental monitoring network that can provide multi-dimensional environmental parameter measurement data, providing a solid data foundation for accurate modeling and efficient operation of wind farms.

[0053] The LiDAR system uses coherent Doppler LiDAR to achieve three-dimensional scanning of the wind farm. The LiDAR system can achieve full 360-degree coverage on the horizontal plane and tens of degrees in the vertical direction, ensuring comprehensive monitoring of the entire wind farm. The spatial resolution is at the decimeter level, and the speed resolution is at the centimeter per second level, providing refined data support for wind farm analysis. The detection distance is up to kilometers, meeting the monitoring needs of large wind farms. The scanning frequency switches between panoramic and fixed-point modes according to demand, and the frequency range is from Hertz to tens of Hertz.

[0054] The LiDAR system provides high-resolution three-dimensional wind farm data, providing accurate input for wind prediction models and reliable decision-making basis for intelligent control strategies of wind turbines. The long-range detection capability of the LiDAR system significantly improves the overall operational efficiency and safety of wind farms.

[0055] The high-resolution multimodality imaging system integrates a variety of imaging devices, including: High dynamic range visible light camera: It adopts ultra-high-definition resolution, has the ability to collect dozens of frames per second, and integrates optical image stabilization to ensure clear and stable images in all weather conditions.

[0056] High-sensitivity thermal imaging camera: It has a resolution of hundreds of thousands of pixels, can achieve an acquisition rate of dozens of frames per second, and a temperature resolution of one hundredth of a degree Celsius, which can accurately capture tiny temperature changes of the equipment.

[0057] Hyperspectral imaging camera: covers the visible light to near-infrared band, has hundreds of independent spectral channels, and wavelength resolution reaches the nanometer level, providing rich spectral information for material analysis and fault diagnosis.

[0058] The high-resolution multi-modal imaging system works in tandem to achieve all-round, multi-dimensional monitoring of key components such as wind turbine blades and towers. The high-resolution multi-modal imaging system can not only capture visible structural changes, but also detect potential problems such as thermal anomalies and material aging, improving the accuracy and timeliness of preventive maintenance.

[0059] The intelligent autonomous flying robot swarm system consists of multiple highly intelligent drones, which are configured according to the scale of the wind farm, usually with several units for every tens of megawatts of installed capacity.

[0060] Each drone uses a multi-rotor carbon fiber composite frame with a diameter in the meter range and a weight controlled at several kilograms, achieving the optimal balance of high strength and lightweight.

[0061] The single flight time of each drone can meet the inspection needs of large wind farms.

[0062] Each drone is equipped with a multi-sensor fusion system that integrates ultra-high-definition visible light cameras, high-resolution thermal imagers, millimeter-wave radars and multi-line lidars, achieving all-round and multi-scale data collection capabilities.

[0063] Each drone has edge computing capabilities, and each drone is equipped with an embedded processor with powerful real-time AI reasoning capabilities, which can perform preliminary data analysis and anomaly detection during flight.

[0064] Each drone uses a dual-mode system that combines 5G and satellite communications to ensure high-bandwidth, low-latency real-time data transmission in complex geographical locations.

[0065] The intelligent autonomous flying robot swarm system achieves all-round and efficient inspection of wind farms through collaborative work. The system not only greatly improves the coverage and frequency of inspections, but also improves the accuracy and real-time performance of anomaly detection through edge computing and AI technology. It reduces the safety risks of manual inspections and improves the preventive and accurate nature of maintenance, which is of great significance to improving the overall operational efficiency and safety of wind farms.

[0066] Specifically, the network module includes the following components: 5G private network, the 5G private network adopts 3.5 GHz and 28 GHz dual frequency bands, and the bandwidth of the 5G private network is 100 MHz in the 3.5 GHz band and 800 MHz in the 28 GHz band.

[0067] Quantum communication network is a key component to ensure the security of data transmission. Quantum communication network includes: Quantum key distribution (QKD) system based on entanglement distribution: the key generation rate is greater than 10 megabits per second and the effective distance on the ground is 100 kilometers.

[0068] Quantum random number generator: provides true random numbers for encryption processes.

[0069] Quantum entanglement distribution network: achieving global quantum communication coverage through satellite relay.

[0070] Furthermore, each wind turbine tower is equipped with an edge computing node. The edge computing node processor adopts a multi-core processor with no less than 64 cores, a main frequency of no less than 2.5GHz, and supports advanced vector extension instruction set.

[0071] The graphics processor of the edge computing node is equipped with a high-performance GPU with no less than 80GB of video memory.

[0072] The memory of the edge computing node is equipped with large-capacity DDR4 high-speed memory with a capacity of no less than 1TB and a frequency of no less than 3200MHz.

[0073] The storage of edge computing nodes uses NVMe solid-state drives with a capacity of no less than 20TB and a read and write speed of no less than 7000MB / s.

[0074] The network interface of the edge computing node is equipped with a high-speed Ethernet interface with a bandwidth of no less than 100Gbps.

[0075] Edge computing nodes can realize localized data processing, significantly reduce data transmission volume, and improve system response speed.

[0076] Specifically, the data module includes: a distributed database system, a big data processing platform and an AI training platform. The distributed database system includes: Main database: The main database adopts a high-performance distributed database to store structured data.

[0077] Time series database: The time series database is specifically used to process and store time series data.

[0078] Graph database: The graph database is used to store and process complex relational data.

[0079] Data warehouse: The data warehouse is used for large-scale data analysis and storage.

[0080] The big data processing platform adopts a distributed computing framework, is configured with a 10,000-core cluster, uses a real-time stream processing engine, is configured with a 5,000-core cluster, adopts a cloud-native data lake solution, and has a storage capacity of 100PB.

[0081] The AI ​​training platform has a GPU cluster, a CPU cluster, a distributed training framework and an automatic machine learning platform. The GPU cluster is configured with 100 high-performance GPU servers. The CPU cluster is configured with 1,000 multi-core CPU nodes. The distributed training framework adopts a framework that supports distributed deep learning.

[0082] Specifically, the AI ​​module includes a wind condition prediction model, a power prediction model, an equipment health status monitoring model and an adaptive control algorithm model.

[0083] The wind forecast model uses a spatiotemporal graph neural network (STGNN) architecture, with input features including historical meteorological data, terrain data, and satellite images. The output is the wind speed and direction forecast for a 1 km x 1 km grid in the next 72 hours.

[0084] The graph convolution of the wind prediction model is:

[0085] in, Indicates The node feature matrix of the module, is the adjacency matrix with self-loops added, yes The degree matrix of is a learnable weight matrix, is the activation function.

[0086] The power prediction model adopts an integrated learning method, with input features: wind condition forecast, historical power generation data, and equipment status. The output result is: power generation forecast for the next 48 hours at a 15-minute granularity.

[0087] The equipment health monitoring model uses a combination of deep learning networks and long short-term memory networks. The input features are vibration data, temperature data, oil analysis data, and electrical parameters. The output results are: equipment remaining useful life (RUL) prediction and anomaly detection probability. Among them, the RUL prediction error is less than 5%, and the anomaly detection F1 score is greater than 0.95.

[0088] The adaptive control algorithm model adopts a method that combines model predictive control (MPC) and deep reinforcement learning (DRL). The state space of the adaptive control algorithm model includes wind speed, wind direction, blade angle, and generator speed. The action space includes: blade pitch angle, yaw angle, and generator torque. The reward function is: ;in, is the reward value, is the power generation, is the load, For wear, is the weight coefficient. The training method adopts the SoftActorCritic algorithm. The goal of the control strategy is to maximize the long-term cumulative reward, which is expressed as:

[0089] in, is the policy function, It's the trajectory. is the discount factor, and are the state and action at time step $t$ respectively.

[0090] Specifically, the application modules include a digital twin visualization platform, a holographic mixed reality operation and maintenance system, and an intelligent scheduling system.

[0091] The digital twin visualization platform uses real-time ray tracing technology combined with deep learning super sampling technology to provide high-quality, high-frame-rate visual effects. The digital twin visualization platform uses a physical engine that can accurately simulate the dynamic behavior of wind turbine components. The terrain data of the digital twin visualization platform uses a digital elevation model (DEM) with a resolution of no less than 0.5 meters. The real-time data update frequency of the digital twin visualization platform is 10 Hz.

[0092] The digital twin visualization platform can present the operating status of the wind farm in real time, including key information such as wind turbine operation, wind farm distribution and power generation, providing intuitive visualization support for operational decisions.

[0093] The holographic mixed reality operation and maintenance system includes: hardware, including lightweight AR glasses weighing no more than 50 grams, with a field of view of no less than 110 degrees, a resolution of 4K, and a battery life of no less than 8 hours. The holographic mixed reality operation and maintenance system also includes an operating system customized based on an open source mobile operating system, optimized for AR application scenarios. The holographic mixed reality operation and maintenance system also includes a network module that supports the latest generation of mobile communication technology and high-speed WiFi standards, as well as a remote collaboration system.

[0094] The holographic mixed reality operation and maintenance system can provide on-site staff with real-time equipment status information, maintenance guidance and remote expert support, greatly improving maintenance efficiency and accuracy.

[0095] The intelligent dispatching system adopts a multi-objective optimization algorithm combining genetic algorithm and particle swarm optimization. The optimization objectives are to maximize power generation, minimize grid fluctuations and minimize equipment wear. Constraints: grid stability, equipment operating parameters, environmental protection requirements. Optimization cycle: 5-minute rolling optimization, 24-hour forward-looking planning. Execution delay: less than 100 milliseconds.

[0096] The multi-objective optimization problem is expressed as:

[0097]

[0098]

[0099] in, is the objective function vector, It is The objective function, and are inequality and equality constraints respectively.

[0100] The intelligent dispatching system can automatically adjust the wind turbine operating parameters according to real-time data and prediction results to optimize the overall operation of the wind farm.

[0101] Example 2 A method for using a wind farm digital twin modeling system based on Example 1 is as follows: Figure 2 As shown, the following steps are included: The blade morphology and stress data are collected through the biomimetic adaptive wind turbine blade system; wind speed, wind direction, temperature, humidity and air pressure data are collected through the quantum sensor network; three-dimensional wind field data is obtained through laser radar system scanning; image and video data of the appearance and operating status of the wind turbine are collected through the high-resolution camera system; inspection tasks are performed by autonomous flying robot swarms to collect detailed images and thermal imaging data of the wind turbine; and data is transmitted to the data module through the network module; Receive and process data from the equipment module through the data module, store and manage structured data through the distributed database system, clean, convert and pre-process data through the big data processing platform, and prepare data sets for subsequent model training and optimization through the AI ​​training platform; The AI ​​module performs in-depth analysis and prediction based on the data processed by the data module, generates future wind forecasts through the wind prediction model; predicts future power generation through the power prediction model; evaluates equipment status and predicts potential failures through the equipment health status monitoring model; and generates the optimal control strategy based on the current status and prediction results through the adaptive control algorithm model; The analysis and prediction results of the AI ​​module are converted into visual information and control instructions through the application module. The virtual model of the wind farm is updated in real time through the digital twin visualization platform. The holographic mixed reality operation and maintenance system provides augmented reality guidance to on-site staff. The intelligent scheduling system automatically adjusts wind turbine parameters and operation strategies according to the optimization results.

[0102] Through a closed-loop workflow, the wind farm digital twin modeling system can realize digital twin modeling and intelligent management of wind farms, significantly improving the wind farm's power generation efficiency, operation and maintenance efficiency, and overall equipment effectiveness.

[0103] Unless otherwise specified, the device elements involved in the above embodiments are all conventional device elements, and the structural settings, working modes or control modes involved are all conventional settings, working modes or control modes in the art unless otherwise specified.

[0104] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Other modifications or equivalent substitutions made to the technical solution of the present invention by ordinary technicians in the field should be included in the scope of the claims of the present invention as long as they do not depart from the spirit and scope of the technical solution of the present invention.

Claims

1. A wind farm digital twin modeling system, characterized in that: include: Equipment modules, including a biomimetic adaptive wind turbine blade system, a quantum sensor network, a lidar system, a high-resolution multimodal imaging system, and an intelligent autonomous flying robot swarm system; A network module, which includes a 5G private network, a quantum communication network, and an edge computing network; A data module, which includes a distributed database system, a big data processing platform, and an AI training platform; An AI module, wherein the AI ​​module includes a wind condition prediction model, a power prediction model, an equipment health status monitoring model, and an adaptive control algorithm model; Application modules, including a digital twin visualization platform, a holographic mixed reality operation and maintenance system, and an intelligent scheduling system; Among them, the equipment module collects real-time data of the wind farm, the equipment module communicates with the data module through the network module, the data module is communicatively connected to the AI ​​module, the AI ​​module is used to perform analysis and prediction based on the data processed by the data module, and the application module is communicatively connected to the AI ​​module and the network module.

2. A wind farm digital twin modeling system according to claim 1, characterized in that: The biomimetic adaptive fan blade system of the equipment module includes: an intelligent deformable skeleton structure, which is made of shape memory alloy and has an integrated microchannel network inside; the intelligent deformable skeleton structure is coated with a high-performance composite material outer layer structure, which is made of carbon fiber reinforced polymer composite material.

3. A wind farm digital twin modeling system according to claim 1, characterized in that: The quantum sensing network of the equipment module is arranged on the biomimetic adaptive wind turbine blade system, and the quantum sensing network includes: a quantum wind speed sensor, an optical quantum gyroscope wind direction sensor, a superconducting quantum interferometer temperature sensor, a graphene-based quantum capacitance humidity sensor and an optical microcavity pressure sensor.

4. A wind farm digital twin modeling system according to claim 1, characterized in that: The laser radar system of the equipment module adopts coherent Doppler laser radar.

5. A wind farm digital twin modeling system according to claim 1, characterized in that: The intelligent autonomous flying robot swarm system includes multiple drones, each of which includes: a multi-rotor carbon fiber composite material frame, on which a multi-sensor fusion system, an edge computing unit, and a dual-mode communication system combining 5G and satellite communications are provided.

6. A wind farm digital twin modeling system according to claim 1, characterized in that: The edge computing network of the network module includes: an edge computing node arranged on each wind turbine tower, and the edge computing node is equipped with a multi-core processor, a high-performance GPU, a large-capacity high-speed memory, a high-speed solid-state storage and a high-speed network interface.

7. A wind farm digital twin modeling system according to claim 1, characterized in that: The wind condition prediction model of the AI ​​module adopts a spatiotemporal graph neural network architecture. The input features include historical meteorological data, terrain data and satellite images, and the output results are wind speed and direction.

8. A method for using a wind farm digital twin modeling system according to any one of claims 1 to 7, characterized in that: The following steps are involved: Collect wind farm data through the equipment module and transmit the data to the data module through the network module; Receive and process data from the equipment module through the data module; Perform in-depth analysis and prediction based on the data processed by the data module through the AI ​​module; The analysis and prediction results of the AI ​​module are converted into visual information and control instructions through the application module.

9. The method for using a wind farm digital twin modeling system according to claim 8, characterized in that: The blade morphology and stress data are collected through the biomimetic adaptive wind turbine blade system; wind speed, wind direction, temperature, humidity and air pressure data are collected through the quantum sensor network; three-dimensional wind field data is obtained through laser radar system scanning; image and video data of the appearance and operating status of the wind turbine are collected through the high-resolution camera system; inspection tasks are performed by autonomous flying robot swarms to collect detailed images and thermal imaging data of the wind turbine; and data is transmitted to the data module through the network module; The distributed database system is used to store and manage structured data. The big data processing platform is used to clean, convert and pre-process data. The AI ​​training platform is used to prepare data sets for subsequent model training and optimization. Generate future wind forecasts through wind forecast models; predict future power generation through power forecast models; Evaluate equipment status and predict potential failures through equipment health monitoring models; generate optimal control strategies based on current status and prediction results through adaptive control algorithm models; The virtual model of the wind farm is updated in real time through the digital twin visualization platform; augmented reality guidance is provided to on-site staff through the holographic mixed reality operation and maintenance system; and the wind turbine parameters and operation strategies are automatically adjusted according to the optimization results through the intelligent scheduling system.

10. The method for using a wind farm digital twin modeling system according to claim 9, characterized in that: The adaptive control algorithm model of the AI ​​module adopts a method that combines model predictive control and deep reinforcement learning. The state space includes wind speed, wind direction, blade angle and generator speed. The action space includes blade pitch angle, yaw angle and generator torque. The reward function is the weighted sum of power generation minus load and wear. The intelligent scheduling system of the application module adopts a multi-objective optimization algorithm, including a genetic algorithm and a particle swarm optimization algorithm. The optimization goals are to maximize power generation, minimize grid fluctuations and minimize equipment wear. The constraints include grid stability, equipment operating parameters and environmental protection requirements.

Images