Wind power intelligent operation and maintenance system of ''machine-group-domain'' architecture

Abstract

The invention discloses a wind power intelligent operation and maintenance system of a'machine-group-domain 'architecture. The wind power intelligent operation and maintenance system comprises a plurality of fan management modules, a plurality of field group management modules and a region management module, the fan management module is used for adjusting control parameters of the fan according to operation data and environmental parameters of the fan, and the field group management module is used for performing electric quantity prediction according to the operation data of the fan and weather prediction information to obtain field group power generation prediction data; the regional management module is used for scheduling a wind power plant group according to the field group power generation prediction data of each field group management module and the actual load of the power grid and generating field group scheduling information, and the field group management module is used for generating a fan control instruction according to the field group scheduling information, the operation data of the fan and the environmental parameters and sending the fan control instruction to the fan management module. According to the invention, management control of a single fan is realized, management control of a wind power plant group and a wind power area is also realized, unmanned and intelligent management of the whole area of the wind power plant is realized, and the wind power utilization efficiency and the power grid stability are improved.

Description

Technical Field

[0001] This invention relates to the field of wind power generation technology, and in particular to a wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture. Background Technology

[0002] As the global energy structure accelerates its transformation towards cleaner and lower-carbon energy, wind power, as an important renewable energy source, has experienced rapid development in recent years. With the continuous expansion of installed capacity, the operation and maintenance management of wind farms faces increasingly severe challenges.

[0003] Traditional operation and maintenance methods mainly rely on manual inspections and periodic maintenance, which suffer from low efficiency, high costs, and slow response times. This is especially problematic when dealing with widely distributed wind farms in complex geographical environments, making it difficult to achieve real-time monitoring of equipment operating status and fault early warning. Therefore, building an efficient and intelligent wind power operation and maintenance management system to improve the automation and intelligence level of wind farms has become an inevitable trend in the industry.

[0004] In recent years, with the development of artificial intelligence, the Internet of Things, edge computing, and sensor technologies, the wind power sector has gradually introduced intelligent operation and maintenance systems to improve operation and maintenance efficiency and equipment reliability. However, existing intelligent operation and maintenance management systems for wind power focus on the management of individual wind turbine units and lack a unified architecture design from wind turbines and wind farm clusters to wind power areas, making it difficult to achieve unmanned and intelligent management of the entire wind farm. Summary of the Invention

[0005] In view of this, the present invention provides a wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture, the main purpose of which is to solve the problem that the existing wind farm intelligent operation and maintenance management system is unable to achieve unmanned and intelligent management of the entire wind farm.

[0006] According to one aspect of this application, a wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture is provided. Multiple wind turbines form a wind farm cluster, and multiple wind farm clusters constitute a wind power region. The system includes: each wind turbine corresponding to a wind turbine management module, each wind farm cluster corresponding to a cluster management module, and a wind power region corresponding to a region management module.

[0007] Each wind turbine management module is connected to the corresponding wind turbine and the wind farm group management module where the wind turbine is located. Each wind farm group management module is also connected to the area management module and multiple wind turbine management modules within the corresponding wind farm group.

[0008] The wind turbine management module is used to acquire the wind turbine's operating data and environmental parameters of the environment where the wind turbine is located, and adjust the wind turbine's control parameters based on the wind turbine's operating data and environmental parameters. The wind farm cluster management module is used to acquire the wind turbine's operating data and weather forecast information, and perform power generation forecast based on the wind turbine's operating data and weather forecast information to obtain wind farm cluster power generation forecast data. The area management module is used to schedule the wind farm cluster based on the wind farm cluster power generation forecast data of each wind farm cluster management module and the actual load of the power grid, and generate wind farm cluster scheduling information. The wind farm cluster management module is used to generate wind turbine control commands based on the wind farm cluster scheduling information, the wind turbine's operating data and environmental parameters, and send them to the wind turbine management module.

[0009] Optionally, each wind turbine management module includes a wind turbine control unit, a wind turbine status detection unit, and an environmental parameter detection unit, wherein...

[0010] The wind turbine control unit is connected to the corresponding wind turbine and the wind farm group management module, wind turbine status detection unit and environmental parameter detection unit corresponding to the wind turbine group. The wind turbine control unit is a neuromorphic chip.

[0011] The wind turbine status detection unit is used to detect the operating status of the wind turbine and obtain wind turbine operating data. The environmental parameter detection unit is used to detect the environmental parameters of the environment in which the wind turbine is located. The wind turbine control unit is used to calculate the real-time control parameters of the wind turbine based on the wind turbine operating data and the environmental parameters using an adaptive control algorithm. The real-time control parameters include real-time control parameters for pitch angle, real-time control parameters for generator speed, and real-time control parameters for yaw angle.

[0012] Optionally, the wind farm management module includes a wind farm control unit, which is connected to the area management module and each wind turbine management module within the corresponding wind farm area;

[0013] The field cluster control unit is used to acquire meteorological forecast data, input the meteorological forecast data and wind turbine operation data into a preset power prediction model to obtain field cluster power prediction data, and input the weather forecast data and real-time environmental information into a risk prediction model to obtain risk assessment information.

[0014] Optionally, the wind turbine control unit is further configured to input the wind turbine's operating data into a trained wind turbine state prediction model to obtain the difference between the wind turbine's current state and its normal state. When the difference is greater than a preset threshold, the wind turbine is determined to be abnormal. When the wind turbine is abnormal, the wind turbine's operating data is input into a preset fault analysis model to obtain the fault type that caused the wind turbine's abnormality, and alarm information is output according to the fault type.

[0015] Optionally, the wind farm control unit is also used to receive grid dispatch instructions, calculate the pitch angle and generator speed of the wind turbine based on the grid dispatch instructions and the wind turbine's operating data, generate control instructions based on the calculated pitch angle and generator speed, and send them to the wind turbine management module.

[0016] Optionally, the cluster management module further includes an inspection intelligent agent, which is connected to the cluster control unit. The inspection intelligent agent is used to collect image information and infrared thermal images of wind turbine components and send the image information and infrared thermal images to the cluster control unit.

[0017] Optionally, the field control unit is further configured to extract features from the image information to obtain wind turbine component features, determine the state of the wind turbine component based on the wind turbine component features, input the image information corresponding to the wind turbine component into the wind turbine component life prediction model when the wind turbine component is undamaged, obtain the predicted life of the wind turbine component, and output alarm information when the wind turbine component is damaged.

[0018] Optionally, the field control unit is also used to perform temperature field analysis on the infrared thermal image, identify overheated areas based on the analysis results, determine the status of the fan components based on the overheated areas, and output alarm information when the fan components are in a damaged state.

[0019] Optionally, the inspection intelligent agent includes an unmanned aerial vehicle and a robot, wherein the control unit of the aircraft and the control unit of the robot are both neuromorphic chips and integrated sensing and computing chips.

[0020] Optionally, the regional management module is used to obtain the actual load of the power grid, construct an objective function based on the power generation prediction data of the power group and the actual load of the power grid, solve the objective function with the ramp rate of the power group and the actual output of the power group as constraints, obtain the output of each power group, and generate power group scheduling information according to the output of each power group.

[0021] This application provides a wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture. The cluster management module predicts the power generation of each wind turbine based on weather forecast information and the actual operating data of each turbine, obtaining cluster power generation forecast data. The area management module combines the cluster power generation forecast data with the actual load of the power grid to coordinate control and optimize scheduling over a large area, generating cluster scheduling information. The cluster management module outputs wind turbine control commands based on the cluster scheduling information and the turbine operating data. The turbine management module adjusts the turbines based on the turbine control commands, turbine operating data, and environmental parameters. This system not only achieves management and control of individual wind turbines but also the management and control of wind farm clusters and wind power areas, realizing unmanned and intelligent management of the entire wind farm area, improving wind power utilization efficiency and power grid stability.

[0022] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0023] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0024] Figure 1 This paper illustrates a structural block diagram of a wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture, as provided in an embodiment of this application.

[0025] Figure 2 This illustration shows the structural connection diagram of a neuromorphic chip in a wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture, as provided in an embodiment of this application.

[0026] Figure 3 This illustration shows another structural block diagram of a wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture provided in an embodiment of this application.

[0027] Figure 4 This application illustrates a holographic information perception and digital twin framework for a wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture, as provided in an embodiment of this application.

[0028] In the diagram: 1-Wind turbine management module; 2-Farm cluster management module; 3-Area management module. Detailed Implementation

[0029] The present invention will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the embodiments of the present invention can be combined with each other.

[0030] To further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, the specific embodiments, structures, features, and effects according to the present invention will be described in detail below with reference to the accompanying drawings and preferred embodiments. In the following description, different "an embodiment" or "an embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0031] To address the problem that existing intelligent operation and maintenance management systems for wind farms struggle to achieve unmanned and intelligent management across the entire wind farm area, this application provides a wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture. Multiple wind farms form a wind farm cluster, and multiple wind farm clusters constitute a wind power region, such as... Figure 1 As shown, this includes: a wind turbine management module 1 for each wind turbine, a wind farm group management module 2 for each wind farm group, and a region management module 3 for each wind power area.

[0032] Each wind turbine management module 1 is connected to the corresponding wind turbine and the wind farm group management module 2 corresponding to the wind farm group where the wind turbine is located. Each wind farm group management module 2 is connected to the area management module 3 and multiple wind turbine management modules 1 within the corresponding wind farm group.

[0033] The wind turbine management module 1 is used to acquire the wind turbine's operating data and environmental parameters of the environment where the wind turbine is located. Based on the wind turbine's operating data and environmental parameters, it adjusts the wind turbine's control parameters. The wind farm management module 2 is used to acquire the wind turbine's operating data and weather forecast information. Based on the wind turbine's operating data and weather forecast information, it performs power generation forecasting to obtain wind farm power generation forecast data. The area management module 3 is used to schedule the wind farm group based on the wind farm power generation forecast data of each wind farm management module and the actual load of the power grid, generating wind farm scheduling information. The wind farm management module is used to generate wind turbine control commands based on the wind farm scheduling information, wind turbine operating data, and environmental parameters, and send them to the wind turbine management module.

[0034] Specifically, a wind power region comprises multiple wind farm clusters, each wind farm cluster includes multiple wind turbines, each wind turbine is equipped with a turbine management module, each wind farm cluster is equipped with a cluster management module, and each wind power region is equipped with a region management module. The turbine management module and the cluster management module communicate with each other within the cluster via IoT technology, enabling the issuance and processing of control commands. Similarly, the region management module and the cluster management module communicate with each other within the region via IoT technology, enabling the issuance and processing of control commands.

[0035] Each wind turbine is equipped with numerous sensors to detect its operating status and environmental information. The turbine management module adjusts the turbine's control parameters based on the turbine's operating data and environmental parameters. The wind farm management module acquires weather forecast information and, based on the weather forecast and actual turbine operating data, predicts the power generation of each turbine, obtaining the wind farm power generation forecast data. This data is then sent to the regional management module. The regional management module, combining the actual load on the power grid and the power generation forecast data for each wind farm, schedules and controls the power generation of the wind farm, generating scheduling information which is sent to the wind farm management module. The wind farm management module then generates turbine control commands based on the scheduling information, controlling the turbines to operate as required.

[0036] This invention provides a wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture. Compared with existing technologies, the cluster management module predicts the power generation of each wind turbine based on weather forecast information and the actual operating data of each turbine, obtaining cluster power generation forecast data. The regional management module combines the cluster power generation forecast data with the actual load of the power grid to coordinate control and schedule optimization over a large area, generating cluster scheduling information. The cluster management module outputs wind turbine control commands based on the cluster scheduling information and the turbine operating data. The turbine management module adjusts the turbines by combining the turbine control commands, turbine operating data, and environmental parameters. This not only realizes the management and control of individual wind turbines but also the management and control of wind farm clusters and wind power areas, achieving unmanned and intelligent management of the entire wind farm, improving wind power utilization efficiency and power grid stability.

[0037] In one specific embodiment of the present invention, each wind turbine management module 1 includes a wind turbine control unit, a wind turbine status detection unit, and an environmental parameter detection unit, wherein,

[0038] The wind turbine control unit is connected to the corresponding wind turbine, as well as the wind farm management module, wind turbine status detection unit, and environmental parameter detection unit of the wind farm group where the wind turbine is located. The wind turbine control unit is a neuromorphic chip.

[0039] The wind turbine status detection unit is used to detect the operating status of the wind turbine and obtain wind turbine operating data. The environmental parameter detection unit is used to detect the environmental parameters of the environment in which the wind turbine is located. The wind turbine control unit is used to calculate the real-time control parameters of the wind turbine based on the wind turbine operating data and environmental parameters using an adaptive control algorithm. The real-time control parameters include real-time control parameters of pitch angle, real-time control parameters of generator speed, and real-time control parameters of yaw angle.

[0040] Specifically, wind turbine status monitoring units, such as sensors for speed, vibration, temperature, and pressure, are installed in key components of the wind turbine (e.g., blades, gearbox, generator, bearings) to collect real-time operating data. Environmental parameter monitoring units, such as wind speed, wind direction, temperature, and humidity sensors, are also installed around the wind turbine to obtain environmental parameters of the environment in which the wind turbine is located, including wind speed, wind direction, temperature, humidity, and air pressure. Each wind turbine is equipped with a wind turbine control unit, which automatically adjusts the controller parameters according to the wind turbine's operating status and environmental changes to maintain optimal control performance.

[0041] The wind turbine control unit is a neuromorphic chip, which is a heterogeneous hybrid core industrial control neuromorphic chip, such as... Figure 2 As shown, the system comprises three cores: CPU, NPU, and FPGA. The CPU handles general task processing, the NPU accelerates neural networks, and the FPGA handles low-latency control. Logically, the CPU manages the industrial control system's operation and task scheduling, the NPU runs a large-scale AI model to optimize control strategy parameters in real time, and then transmits the optimized parameters to the FPGA core. The FPGA core receives the control strategy parameters and controls the fan in real time. This neuromorphic control chip employs a heterogeneous design, enabling efficient data processing, neural network computation, and real-time control at the hardware level, significantly improving the system's computational efficiency and response speed.

[0042] In one specific embodiment of the present invention, the wind farm management module 2 includes a wind farm control unit, which is connected to the area management module and each wind turbine management module within the corresponding wind farm area.

[0043] The wind farm control unit is used to acquire meteorological forecast data, input the meteorological forecast data and wind turbine operation data into a preset power prediction model to obtain wind farm power prediction data, and input the weather forecast data and real-time environmental information into a risk prediction model to obtain risk assessment information.

[0044] Specifically, the power generation control unit uses a power prediction model to consider weather data and turbine status over a future period to predict power generation for the power generation cluster. This allows for proactive optimization control, smooth power output, and better response to grid demands. The control unit also uses environmental risk assessment algorithms, combined with weather forecast data and environmental information, to evaluate potential safety risks (such as lightning strikes and freezing) and take preventative measures in advance.

[0045] In one specific embodiment of the present invention, the wind turbine control unit is further configured to input the wind turbine's operating data into a trained wind turbine state prediction model to obtain the difference between the current state and the normal state of the wind turbine. When the difference is greater than a preset threshold, the wind turbine is determined to be abnormal. When the wind turbine is abnormal, the wind turbine's operating data is input into a preset fault analysis model to obtain the fault type that caused the wind turbine to be abnormal, and alarm information is output according to the fault type.

[0046] Specifically, the wind turbine control unit calculates the difference between the predicted normal state and the actual collected data using a wind turbine condition prediction model. When the difference exceeds a preset threshold (the threshold is usually set based on historical fault data and equipment safety standards), the system determines that the wind turbine has malfunctioned, enabling early detection of potential faults, such as slight vibration abnormalities in the early stages of bearing wear or a gradual decrease in generator efficiency. Once a wind turbine malfunction is determined, the type of fault causing the malfunction is further identified, and alarm information is output based on the fault type.

[0047] In one specific embodiment of the present invention, the wind farm control unit is also used to receive grid dispatch instructions, calculate the pitch angle of the wind turbine and the generator speed according to the grid dispatch instructions and the wind turbine's operating data, generate control instructions according to the calculated pitch angle and generator speed, and send them to the wind turbine management module.

[0048] Specifically, grid dispatch instructions are based on the overall operation of the power grid, such as power supply and demand balance and grid stability, and include requirements for wind farm power generation capacity and generation rhythm. The wind farm cluster control unit receives these instructions in real time to understand the grid's expectations and requirements for the wind farms. The control unit acquires wind turbine operating data, such as turbine output power, actual generator speed, and actual pitch angle, and, combined with the grid's expectations and requirements, outputs control parameters for the wind turbines to meet those expectations and requirements.

[0049] In one specific embodiment of the present invention, the field management module 2 further includes an inspection intelligent agent, which is connected to the field control unit. The inspection intelligent agent is used to collect image information and infrared thermal images of the wind turbine components and send the image information and infrared thermal images to the field control unit. The inspection intelligent agent includes an unmanned aerial vehicle and a robot. The control unit of the aircraft and the control unit of the robot are both neuromorphic chips and sensing and computing integrated chips.

[0050] Specifically, the robot uses its own sensors to collect features of the target to be detected, including but not limited to audio noise detection, detection of images, videos, infrared images, etc., and performs some preprocessing and classification on the collected signals.

[0051] The robots include wheeled, tracked, humanoid, quadrupedal, and drone-based intelligent inspection equipment that integrates high-definition cameras. The drones include rotary-wing drones equipped with high-definition cameras, infrared cameras, radar, and other sensors.

[0052] Drones and robots utilize path planning algorithms, combined with sensor data, for real-time obstacle avoidance, enabling safe operation in complex environments. Path planning algorithms, such as variations of the Traveling Salesman Problem (TSP) and overlay path planning, optimize the inspection paths of drones and robots, improving efficiency.

[0053] By using intelligent inspection agents, unmanned intelligent inspections of the entire plant can be carried out in different time periods, areas, around the clock, and with full coverage. This ensures that the inspection results are comprehensive, accurate, and reliable, saves labor costs, and promotes the realization of unmanned or minimally staffed operations.

[0054] The integrated sensing and computing smart chip integrates sensing and computing functions, enabling it to perceive the status of devices in real time and process data, providing strong hardware support for intelligent operation and maintenance.

[0055] In traditional photoelectric detection architectures, detection, storage, and computing units are separated, resulting in high latency and high power consumption. To address this bottleneck, a "sensing-storage-computing integrated architecture" has been designed, integrating detection, storage, and computing functions. For example, when a robot tightens a screw, it needs to use a camera to photograph the target, align the robot's robotic arm with the target, and then tighten the screw little by little. After equipping it with a sensing-computing integrated chip, this action no longer requires the robot's main control chip to process the visual signal, then process the signal, and then give control instructions to the robotic arm. Instead, the entire action is completed directly through the robotic arm and the sensing-computing integrated chip.

[0056] In one specific embodiment of the present invention, the field control unit is further configured to extract features from image information to obtain features of wind turbine components, determine the state of wind turbine components based on the features of wind turbine components, input the image information corresponding to the wind turbine components into the wind turbine component life prediction model when the state of the wind turbine components is undamaged, obtain the predicted life of the wind turbine components, and output alarm information when the state of the wind turbine components is damaged.

[0057] Specifically, drones or robots collect images or videos of key components of the wind turbine and send these images or videos to the wind farm control unit. These images cover the appearance details of important components such as wind turbine blades, bearings, and gearboxes. Subsequently, the wind farm control unit uses digital image processing technology and pattern recognition algorithms to extract features from the acquired image information. For example, edge detection algorithms are used to extract the contour edge features of the blades, texture analysis algorithms are used to obtain the texture features of the component surface, and color analysis algorithms are used to identify color changes on the component surface. These features reflect the surface condition and structural morphology of the wind turbine components and are important criteria for judging the component status. The extracted wind turbine component features are compared and analyzed with a pre-established standard feature library. The standard feature library stores various feature data of wind turbine components under normal operating conditions and different degrees of damage. By comparison, the wind farm control unit judges the status of the wind turbine components. If the extracted features highly match the features of the undamaged state in the standard feature library, it indicates that the wind turbine component is in normal operating condition; if there are significant differences in the features, such as cracks on the blade surface or wear marks on the bearing, the wind farm control unit determines that the wind turbine component is in a damaged state and outputs an alarm message. When the condition of a wind turbine component is determined to be undamaged, the farm control unit inputs the image information corresponding to the wind turbine component into the wind turbine component life prediction model. This model integrates machine learning algorithms, materials science knowledge, and a large amount of historical data. It can comprehensively analyze the current condition, operating conditions, environmental factors, and other information of the component. Through complex calculations and analysis, it predicts the remaining service life of the wind turbine component in its current condition.

[0058] In one embodiment, the wind turbine control unit analyzes the image information corresponding to the wind turbine components through time series analysis and prediction algorithms to predict the remaining lifespan of key components.

[0059] In one specific embodiment of the present invention, the field group control unit is also used to perform temperature field analysis on the infrared thermal image, identify overheated areas based on the analysis results, determine the status of the fan components based on the overheated areas, and output alarm information when the fan components are in a damaged state.

[0060] Specifically, infrared thermal imaging technology uses the infrared radiation emitted by an object to create an image. Objects at different temperatures emit infrared radiation of varying intensities, resulting in different grayscale or color differences in the image. The field control unit uses specialized image processing algorithms and temperature analysis models to analyze the temperature field of the infrared thermal image. It maps each pixel in the image to an actual temperature value, and through calculation and processing, constructs a temperature distribution field on the surface of the fan components. Based on pre-set temperature thresholds and normal temperature ranges, it identifies areas in the image where the temperature is abnormally high, i.e., overheated areas. Different fan components have their reasonable temperature ranges during normal operation. When the temperature of a certain area exceeds this range, it is determined to be overheated. For example, the temperature of a bearing is relatively stable during normal operation. If the infrared thermal image shows that the temperature of the bearing area is significantly higher than normal, the field control unit can quickly identify it as an overheated area. Once an overheated area is identified, the field control unit combines information such as the location of the component in that area, the severity of the overheating, and the temperature change trend to comprehensively determine the state of the fan component. If the overheating is mild and the temperature change is relatively stable, the component may be judged to be in a slightly abnormal state, requiring enhanced monitoring; however, when the temperature in the overheated area is too high and continues to rise, exceeding the safe range that the component can withstand, the wind turbine control unit will determine that the wind turbine component is damaged.

[0061] In one specific embodiment of the present invention, the regional management module 3 is used to obtain the actual load of the power grid, construct an objective function based on the power generation prediction data of the power group and the actual load of the power grid, solve the objective function with the ramp rate of the power group and the actual output of the power group as constraints, obtain the output of each power group, and generate power group scheduling information according to the output of each power group.

[0062] Specifically, an objective function is constructed based on the power generation forecast data of each wind farm cluster and the actual load on the power grid. For example, with maximizing energy utilization as the objective function, factors such as the power generation forecast data of the wind farm cluster and the actual load on the power grid are incorporated into the function. The ramp rate and actual output of the wind farm cluster are used as constraints. Through mathematical calculations, the degree of objective achievement under different wind farm cluster power generation strategies is quantified, the output of each wind farm cluster is obtained, and wind farm cluster scheduling information is generated. This allows the wind farm cluster management module to control the wind turbines within the wind farm cluster based on the wind farm cluster scheduling information.

[0063] In one embodiment, the cluster management module is also used to take information from meteorological sensors and wind turbine operating parameters output from lidar installation locations as inputs to establish a deep neural network learning model and build an accurate virtual lidar model. A multi-scale, multi-zone wind condition simulation framework for complex terrain sites is established within the framework of region-wind field-wind turbine, and a near-surface wind field analysis method based on this multi-scale wind condition simulation framework is developed.

[0064] In one embodiment, Figure 3 As shown, the wind turbine management module includes multiple wind turbines equipped with sensors and heterogeneous neuromorphic chips and integrated sensing and computing chips, performing holographic perception and digital twin simulation based on various devices; the wind farm cluster management module includes a wind turbine intelligent control module, multiple UAV intelligent inspection bodies and robot intelligent inspection bodies equipped with heterogeneous neuromorphic chips and integrated sensing and computing chips, performing holographic perception and digital twin simulation based on the wind farm cluster; the regional management module includes regional wind farm information perception, regional wind farm power generation prediction, and regional wind farm coordinated control, performing holographic perception and digital twin simulation based on the regional distributed wind farm cluster.

[0065] The "machine-cluster-domain" holographic perception system and digital twin simulation cover a three-tiered scope: the turbine side, the site side, and the regional side. Based on the three-tiered digital twin technology of the "machine cluster domain," and building upon the holographic perception of the "machine cluster domain," the functions achieved include real-time simulation of the turbine layer, real-time simulation of the wind field layer, and long-term scale simulation of the wind field domain, such as... Figure 4 As shown.

[0066] The above embodiments are merely exemplary embodiments of this application and are not intended to limit this application. The scope of protection of this application is defined by the claims. Those skilled in the art can make various modifications or equivalent substitutions to this application within its substance and scope of protection, and such modifications or equivalent substitutions should also be considered to fall within the scope of protection of this application.

Claims

1. A wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture, wherein multiple wind farms form a wind farm cluster, and multiple wind farm clusters constitute a wind power region, characterized in that, include: Each wind turbine corresponds to a wind turbine management module, each wind farm cluster corresponds to a cluster management module, and each wind power area corresponds to an area management module. Each wind turbine management module is connected to the corresponding wind turbine and the wind farm group management module where the wind turbine is located. Each wind farm group management module is also connected to the area management module and multiple wind turbine management modules within the corresponding wind farm group. The wind turbine management module is used to acquire the wind turbine's operating data and environmental parameters of the environment where the wind turbine is located, and adjust the wind turbine's control parameters based on the wind turbine's operating data and environmental parameters. The wind farm cluster management module is used to acquire the wind turbine's operating data and weather forecast information, and perform power generation forecast based on the wind turbine's operating data and weather forecast information to obtain wind farm cluster power generation forecast data. The area management module is used to schedule the wind farm cluster based on the wind farm cluster power generation forecast data of each wind farm cluster management module and the actual load of the power grid, and generate wind farm cluster scheduling information. The wind farm cluster management module is used to generate wind turbine control commands based on the wind farm cluster scheduling information, the wind turbine's operating data and environmental parameters, and send them to the wind turbine management module.

2. The wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture as described in claim 1, characterized in that, Each wind turbine management module includes a wind turbine control unit, a wind turbine status detection unit, and an environmental parameter detection unit. The wind turbine control unit is connected to the corresponding wind turbine and the wind farm group management module, wind turbine status detection unit and environmental parameter detection unit corresponding to the wind turbine group. The wind turbine control unit is a neuromorphic chip. The wind turbine status detection unit is used to detect the operating status of the wind turbine and obtain wind turbine operating data. The environmental parameter detection unit is used to detect the environmental parameters of the environment in which the wind turbine is located. The wind turbine control unit is used to calculate the real-time control parameters of the wind turbine based on the wind turbine operating data and the environmental parameters using an adaptive control algorithm. The real-time control parameters include real-time control parameters for pitch angle, real-time control parameters for generator speed, and real-time control parameters for yaw angle.

3. The wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture as described in claim 1, characterized in that, The wind farm cluster management module includes a wind farm cluster control unit, which is connected to the area management module and each wind turbine management module within the corresponding wind farm cluster. The field cluster control unit is used to acquire meteorological forecast data, input the meteorological forecast data and wind turbine operation data into a preset power prediction model to obtain field cluster power prediction data, and input the weather forecast data and real-time environmental information into a risk prediction model to obtain risk assessment information.

4. The wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture as described in claim 3, characterized in that, The wind turbine control unit is also used to input the wind turbine's operating data into a trained wind turbine state prediction model to obtain the difference between the wind turbine's current state and its normal state. When the difference is greater than a preset threshold, the wind turbine is determined to be abnormal. When the wind turbine is abnormal, the wind turbine's operating data is input into a preset fault analysis model to obtain the fault type that caused the wind turbine to be abnormal, and alarm information is output according to the fault type.

5. The wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture as described in claim 3, characterized in that, The wind farm control unit is also used to receive grid dispatch instructions, calculate the pitch angle and generator speed of the wind turbine based on the grid dispatch instructions and the wind turbine's operating data, generate control instructions based on the calculated pitch angle and generator speed, and send them to the wind turbine management module.

6. The wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture as described in claim 3, characterized in that, The cluster management module also includes an inspection intelligent agent, which is connected to the cluster control unit. The inspection intelligent agent is used to collect image information and infrared thermal images of wind turbine components and send the image information and infrared thermal images to the cluster control unit.

7. The wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture as described in claim 6, characterized in that, The field control unit is also used to extract features from the image information to obtain the features of the wind turbine components, determine the state of the wind turbine components based on the features of the wind turbine components, and when the state of the wind turbine components is undamaged, input the image information corresponding to the wind turbine components into the wind turbine component life prediction model to obtain the predicted life of the wind turbine components. When the state of the wind turbine components is damaged, output alarm information.

8. The wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture as described in claim 6, characterized in that, The field control unit is also used to perform temperature field analysis on the infrared thermal image, identify overheated areas based on the analysis results, determine the status of the fan components based on the overheated areas, and output alarm information when the fan components are damaged.

9. The wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture as described in claim 6, characterized in that, The inspection intelligent agent includes unmanned aerial vehicles and robots. The control units of the aircraft and the robot are both neuromorphic chips and integrated sensing and computing chips.

10. The wind power intelligent operation and maintenance system with a "machine-cluster-domain" architecture as described in claim 1, characterized in that, The regional management module is used to obtain the actual load of the power grid, construct an objective function based on the power generation prediction data of the power group and the actual load of the power grid, solve the objective function with the ramp rate and actual output of the power group as constraints, obtain the output of each power group, and generate power group scheduling information based on the output of each power group.

Images