Wind power plant monitoring system integrating historical data and AI technology

Abstract

The invention, which relates to the monitoring field of the wind power plant, discloses a historical data and AI technology integrated wind power plant monitoring system comprising a data acquisition module, a historical data integration module, a preprocessing module, an AI application module, a visualization module, and an intelligent decision module. According to the wind power plant monitoring system fusing the historical data and the AI technology, through an intelligent fan group cooperative control strategy driven by reinforcement learning, autonomous cooperation and optimized operation between fans are achieved, the wind energy utilization rate and the power generation efficiency are improved, meanwhile, mutual interference and abrasion between the fans are reduced, and the wind power plant monitoring system is suitable for being popularized and applied. According to the method, the service life of equipment is prolonged, the overall economic benefit and environmental benefit of the wind power plant are improved, the success experience and optimization model of the existing wind power plant can be quickly applied to the new wind power plant by utilizing the transfer learning technology, the performance improvement process of the new wind power plant is accelerated, and the operation risk and cost of the new wind power plant are reduced.

Description

Technical Field

[0001] The present invention relates to the field of wind farm monitoring, and in particular to a wind farm monitoring system integrating historical data and AI technology. Background Art

[0002] A wind farm is a place that uses wind energy to generate electricity. It is a collection of power generation facilities that install multiple wind turbines in an area and convert wind energy into electrical energy, which is then transmitted to the power grid for users to use.

[0003] With the rapid development of renewable energy, wind power, as a clean and renewable energy form, has an increasing installed capacity. However, wind farms are usually located in remote areas with complex and changeable environments, and traditional monitoring systems are difficult to meet the requirements of real-time and accuracy. In the existing technology, wind farm monitoring systems mainly rely on real-time data collection and analysis, but often ignore the value of historical data, resulting in the inability to fully utilize historical information for trend prediction and fault warning. In addition, existing data analysis methods mostly rely on manual experience, lack intelligent processing capabilities, and are difficult to cope with large-scale data processing needs.

[0004] Therefore, it is necessary to propose a wind farm monitoring system that integrates historical data and AI technology to solve the above problems. Summary of the invention

[0005] The main purpose of the present invention is to provide a wind farm monitoring system integrating historical data and AI technology, which can effectively solve the problems in the background technology.

[0006] To achieve the above object, the technical solution adopted by the present invention is: A wind farm monitoring system integrating historical data and AI technology, comprising a data acquisition module, a historical data integration module, a preprocessing module, an AI application module, a visualization module, and an intelligent decision-making module. The data acquisition module collects data from the wind farm based on intelligent sensors, image acquisition devices, and audio acquisition devices arranged at the wind farm. The historical data integration module centrally stores and manages the massive data accumulated by the wind farm over the years based on the constructed historical data warehouse; The preprocessing module is used to perform noise removal, gap filling, outlier processing and standardization conversion on historical data; The AI ​​application module generates highly realistic simulated fault data by learning from historical fault data and normal operation data; it regards the wind turbine group of the entire wind farm as a multi-agent system, trains it through a reinforcement learning algorithm, and autonomously decides on the optimal operating parameters; The visualization module constructs an immersive 3D visualization monitoring interface of the wind farm based on virtual reality and augmented reality technologies; The intelligent decision-making module develops an intelligent decision-making assistance platform based on the analysis results of the AI ​​application module to provide all-round decision-making support for wind farm managers.

[0007] Preferably, the data acquisition module includes a data fusion module, a data acquisition frequency adjustment module, and a data transmission protocol module, wherein the data fusion module performs data acquisition of wind speed, wind direction, temperature, humidity, air pressure, generator speed, torque, power output, blade pitch angle, and gearbox oil temperature based on intelligent sensors deployed at various key locations of the wind farm, and also includes high-resolution image acquisition equipment and audio acquisition equipment, the image acquisition equipment is used to regularly photograph the surface conditions of wind turbine blades and towers, and the audio acquisition equipment is used to capture abnormal sounds generated during the operation of wind farm equipment, including abnormal noise caused by bearing wear and whistling sounds caused by abnormal aerodynamic performance of blades; The data acquisition frequency adjustment module is dynamically adjusted based on the importance and change characteristics of the parameters. For key power and wind speed data, high-frequency acquisition of 10 times per second can be achieved. For some relatively stable environmental parameters, the acquisition frequency is appropriately reduced to 2 times per second to balance data transmission and storage pressure; The data transmission protocol module is based on establishing a unified data transmission protocol to transmit various types of collected data to the local data center of the wind farm in real time, and is used to automatically adjust the transmission strategy in the event of network instability or partial node failure to ensure data integrity and continuity.

[0008] Preferably, the historical data integration module centrally stores and manages the massive data accumulated in the wind farm in the past based on the constructed historical data warehouse. The historical data includes design parameters and equipment selection data from the initial construction of the wind farm, to operation records, maintenance logs, fault reports over the years, as well as detailed meteorological data and power generation data in different seasons and weather conditions.

[0009] Preferably, the preprocessing module performs denoising, gap filling, outlier processing and standardization conversion on the historical data based on data cleaning and preprocessing technology, identifies and corrects abnormal data points caused by sensor failure and severe weather conditions through data mining algorithms, and reasonably interpolates and fills missing data using time series analysis methods, so that the historical data has higher quality and availability.

[0010] Preferably, the AI ​​application module includes a fault simulation and diagnosis enhancement module based on a generative adversarial network, an intelligent wind turbine group collaborative control module driven by reinforcement learning, and a cross-wind farm performance optimization module based on transfer learning, wherein the fault simulation and diagnosis enhancement module based on a generative adversarial network is based on GAN technology, and generates highly realistic simulated fault data through learning from historical fault data and normal operating data, including various types of wind turbine faults and equipment operating status data under complex meteorological conditions. This data is used to expand the training data set of the fault diagnosis model and enhance the model's ability to recognize rare faults and complex working conditions; a fault diagnosis model based on a combination of a convolutional neural network and a long short-term memory network is constructed, and the convolutional neural network is used to extract features from the image data of the wind turbine equipment and automatically identify potential physical damage and structural defects; the long short-term memory network models the operating parameter data of the time series to capture the dynamic change trend of the equipment operating status, thereby achieving early warning and accurate diagnosis of faults.

[0011] Preferably, the reinforcement learning-driven intelligent wind turbine group collaborative control module is used to regard the wind turbine group of the entire wind farm as a multi-agent system, and each wind turbine is trained as an independent agent through a reinforcement learning algorithm to autonomously decide on the optimal operating parameters based on local wind speed, wind direction information and the operating status of adjacent wind turbines, including but not limited to pitch angle adjustment and speed control, so as to maximize the overall power generation efficiency and load balancing of the wind turbine group. A distributed collaborative reinforcement learning framework is introduced to enable information exchange and collaboration between wind turbines to avoid wake interference between each other, optimize the wind flow field distribution in the wind farm, and improve wind energy utilization.

[0012] Preferably, the cross-wind farm performance optimization module based on transfer learning is used to target the similarities and differences between different wind farms, and uses transfer learning technology to quickly migrate the performance optimization model trained in a mature wind farm to a newly built wind farm with similar operating conditions, so as to reduce the time and amount of data required for model training of new wind farms; at the same time, a meta-model for wind farm performance evaluation is established, and through a comprehensive analysis of historical data of multiple wind farms, common performance influencing factors and key indicators are extracted, including the influence of topography on wind speed distribution and the adaptability of different wind turbine models in different climatic zones, so as to provide decision support for the planning, construction and operation of new wind farms based on meta-knowledge.

[0013] Preferably, the visualization module constructs an immersive 3D visualization monitoring interface of the wind farm based on virtual reality and augmented reality technologies. Operation and maintenance personnel can view the real-time operating status of the wind farm by wearing VR helmets and using AR smart glasses, including but not limited to the rotation of wind turbines, changes in meteorological conditions, and key parameter information of equipment.

[0014] Preferably, the intelligent decision-making module develops an intelligent decision-making assistance platform based on the results of the AAI application module, which is used to generate multiple feasible decision-making plans based on different decision-making scenarios, including wind turbine maintenance plan formulation, power generation scheduling optimization, and equipment upgrade and transformation decisions, using machine learning algorithms, and quantitatively evaluates and compares the potential risks, economic benefits, and implementation difficulties of each plan.

[0015] Compared with the prior art, the present invention provides a wind farm monitoring system integrating historical data and AI technology, which has the following beneficial effects: 1. The wind farm monitoring system that integrates historical data and AI technology integrates all-round data of the wind farm, including traditional numerical data, image data, audio data, historical data and real-time data, and uses a variety of advanced AI technologies for in-depth mining and analysis. It can more comprehensively and accurately understand the operating status and equipment health of the wind farm, discover potential problems and optimization opportunities hidden behind the data, and provide strong support for the refined management of the wind farm. Based on the rich fault simulation data generated by GAN technology and advanced deep learning models, the system can achieve high-precision diagnosis and early prediction of wind turbine faults, discover potential fault hazards in advance, and leave enough time for operation and maintenance personnel to troubleshoot and repair faults, effectively reduce equipment failure rate and maintenance costs, and improve the reliability and availability of wind farms.

[0016] 2. The wind farm monitoring system that integrates historical data and AI technology realizes autonomous collaboration and optimized operation between wind turbines through a collaborative control strategy of intelligent wind turbine groups driven by reinforcement learning, thereby improving wind energy utilization and power generation efficiency. At the same time, it reduces mutual interference and wear between wind turbines, extends the service life of equipment, and improves the overall economic and environmental benefits of wind farms. By using transfer learning technology, the successful experience and optimization models of existing wind farms can be quickly applied to new wind farms, accelerating the performance improvement process of new wind farms and reducing the operating risks and costs of newly built wind farms. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a system block diagram of the present invention. DETAILED DESCRIPTION

[0018] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0019] like Figure 1As shown, a wind farm monitoring system integrating historical data and AI technology includes a data acquisition module, a historical data integration module, a preprocessing module, an AI application module, a visualization module, and an intelligent decision-making module. The data acquisition module collects data from the wind farm based on intelligent sensors, image acquisition devices, and audio acquisition devices arranged at the wind farm. The historical data integration module centrally stores and manages the massive data accumulated by the wind farm over the years based on the constructed historical data warehouse; The preprocessing module is used to perform noise removal, gap filling, outlier processing and standardization conversion on historical data; The AI ​​application module generates highly realistic simulated fault data by learning from historical fault data and normal operation data. It regards the wind turbine group of the entire wind farm as a multi-agent system, trains it through reinforcement learning algorithms, and autonomously decides on the optimal operating parameters. The visualization module builds an immersive 3D visualization monitoring interface for wind farms based on virtual reality and augmented reality technologies; The intelligent decision-making module develops an intelligent decision-making assistance platform based on the analysis results of the AI ​​application module to provide comprehensive decision-making support for wind farm managers.

[0020] The data acquisition module includes a data fusion module, a data acquisition frequency adjustment module, and a data transmission protocol module. The data fusion module is based on intelligent sensors deployed at various key locations in the wind farm to collect data on wind speed, wind direction, temperature, humidity, air pressure, generator speed, torque, power output, blade pitch angle, and gearbox oil temperature. It also includes high-resolution image acquisition equipment and audio acquisition equipment. The image acquisition equipment is used to regularly photograph the surface conditions of wind turbine blades and towers, and the audio acquisition equipment is used to capture abnormal sounds generated during the operation of wind farm equipment, including abnormal noise caused by bearing wear and whistling sounds caused by abnormal aerodynamic performance of blades. The data acquisition frequency adjustment module dynamically adjusts the parameters based on their importance and change characteristics. For key power and wind speed data, it can achieve high-frequency acquisition of 10 times per second. For some relatively stable environmental parameters, the acquisition frequency is appropriately reduced to 2 times per second to balance data transmission and storage pressure. The data transmission protocol module is based on the establishment of a unified data transmission protocol to transmit various types of collected data to the local data center of the wind farm in real time. It is used to automatically adjust the transmission strategy in the event of network instability or partial node failure to ensure data integrity and continuity.

[0021] The historical data integration module is based on the constructed historical data warehouse to centrally store and manage the massive data accumulated by the wind farm in the past. The historical data includes the design parameters and equipment selection data from the initial construction of the wind farm, as well as the operation records, maintenance logs, fault reports over the years, as well as detailed meteorological data and power generation data in different seasons and weather conditions.

[0022] The preprocessing module removes noise, fills in gaps, processes outliers, and standardizes historical data based on data cleaning and preprocessing techniques. It uses data mining algorithms to identify and correct abnormal data points caused by sensor failures and severe weather conditions, and uses time series analysis methods to reasonably interpolate and fill in missing data, making historical data of higher quality and availability.

[0023] AI application modules include a fault simulation and diagnosis enhancement module based on generative adversarial networks, a smart wind turbine group collaborative control module driven by reinforcement learning, and a cross-wind farm performance optimization module based on transfer learning. The fault simulation and diagnosis enhancement module based on generative adversarial networks is based on GAN technology. By learning from historical fault data and normal operating data, it generates highly realistic simulated fault data, including various types of wind turbine faults and equipment operating status data under complex meteorological conditions. This data is used to expand the training data set of the fault diagnosis model and enhance the model's ability to identify rare faults and complex working conditions. A fault diagnosis model based on a combination of convolutional neural networks and long short-term memory networks is constructed. The convolutional neural network is used to extract features from the image data of wind turbine equipment and automatically identify potential physical damage and structural defects. The long short-term memory network models the operating parameter data of the time series to capture the dynamic change trend of the equipment's operating status, thereby achieving early warning and accurate diagnosis of faults.

[0024] The reinforcement learning-driven intelligent wind turbine group collaborative control module is used to regard the wind turbine group of the entire wind farm as a multi-agent system. Each wind turbine is an independent agent and is trained through a reinforcement learning algorithm. It is used to autonomously decide the optimal operating parameters based on local wind speed and wind direction information and the operating status of adjacent wind turbines, including but not limited to pitch angle adjustment and speed control, to maximize the overall power generation efficiency and load balance of the wind turbine group. A distributed collaborative reinforcement learning framework is introduced to enable information exchange and collaboration between wind turbines to avoid wake interference between each other, optimize the wind flow field distribution in the wind farm, and improve wind energy utilization.

[0025] The cross-wind farm performance optimization module based on transfer learning is used to target the similarities and differences between different wind farms. It uses transfer learning technology to quickly migrate the performance optimization model trained in a mature wind farm to a newly built wind farm with similar operating conditions, thereby reducing the time and data volume required for model training in new wind farms. At the same time, a meta-model for wind farm performance evaluation is established. Through a comprehensive analysis of historical data from multiple wind farms, common performance influencing factors and key indicators are extracted, including the influence of topography on wind speed distribution and the adaptability of different wind turbine models in different climatic zones. Based on meta-knowledge, decision support is provided for the planning, construction and operation of new wind farms.

[0026] The visualization module builds an immersive 3D visualization monitoring interface for wind farms based on virtual reality and augmented reality technologies. Operation and maintenance personnel can view the real-time operating status of wind farms by wearing VR helmets and using AR smart glasses, including but not limited to the rotation of wind turbines, changes in meteorological conditions, and key parameter information of equipment.

[0027] Based on the results of the AAI application module, the intelligent decision-making module develops an intelligent decision-making assistance platform to generate multiple feasible decision-making plans based on different decision-making scenarios, including wind turbine maintenance plan formulation, power generation scheduling optimization, and equipment upgrade and transformation decisions. It also uses machine learning algorithms to quantify and compare the potential risks, economic benefits, and implementation difficulties of each plan.

[0028] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A wind farm monitoring system integrating historical data and AI technology, including a data acquisition module, a historical data integration module, a preprocessing module, an AI application module, a visualization module, and an intelligent decision-making module, characterized in that: The data acquisition module collects data from the wind farm based on intelligent sensors, image acquisition devices and audio acquisition devices arranged at the wind farm; The historical data integration module centrally stores and manages the massive data accumulated by the wind farm over the years based on the constructed historical data warehouse; The preprocessing module is used to perform noise removal, gap filling, outlier processing and standardization conversion on historical data; The AI ​​application module generates highly realistic simulated fault data by learning from historical fault data and normal operation data; it regards the wind turbine group of the entire wind farm as a multi-agent system, trains it through a reinforcement learning algorithm, and autonomously decides on the optimal operating parameters; The visualization module constructs an immersive 3D visualization monitoring interface of the wind farm based on virtual reality and augmented reality technologies; The intelligent decision-making module develops an intelligent decision-making assistance platform based on the analysis results of the AI ​​application module to provide all-round decision-making support for wind farm managers.

2. A wind farm monitoring system integrating historical data and AI technology according to claim 1, characterized in that: The data acquisition module includes a data fusion module, a data acquisition frequency adjustment module, and a data transmission protocol module, wherein the data fusion module is based on intelligent sensors deployed at various key locations of the wind farm to collect data on wind speed, wind direction, temperature, humidity, air pressure, generator speed, torque, power output, blade pitch angle, and gearbox oil temperature, and also includes high-resolution image acquisition equipment and audio acquisition equipment. The image acquisition equipment is used to regularly photograph the surface conditions of wind turbine blades and towers, and the audio acquisition equipment is used to capture abnormal sounds generated during the operation of wind farm equipment, including abnormal noise caused by bearing wear and whistling sounds caused by abnormal aerodynamic performance of blades; The data acquisition frequency adjustment module is dynamically adjusted based on the importance and change characteristics of the parameters. For key power and wind speed data, high-frequency acquisition of 10 times per second can be achieved. For some relatively stable environmental parameters, the acquisition frequency is appropriately reduced to 2 times per second to balance data transmission and storage pressure; The data transmission protocol module is based on establishing a unified data transmission protocol to transmit various types of collected data to the local data center of the wind farm in real time, and is used to automatically adjust the transmission strategy in the event of network instability or partial node failure to ensure data integrity and continuity.

3. The wind farm monitoring system integrating historical data and AI technology according to claim 1 is characterized in that: The historical data integration module is based on the constructed historical data warehouse to centrally store and manage the massive data accumulated by the wind farm in the past. The historical data includes design parameters and equipment selection data from the initial construction of the wind farm, as well as operation records, maintenance logs, fault reports over the years, as well as detailed meteorological data and power generation data in different seasons and weather conditions.

4. The wind farm monitoring system integrating historical data and AI technology according to claim 1 is characterized in that: The preprocessing module removes noise, fills in gaps, processes outliers and performs standardization conversion on historical data based on data cleaning and preprocessing technology. It identifies and corrects abnormal data points caused by sensor failure and severe weather conditions through data mining algorithms, and uses time series analysis methods to reasonably interpolate and fill in missing data, so as to make historical data have higher quality and availability.

5. The wind farm monitoring system integrating historical data and AI technology according to claim 1 is characterized in that: The AI ​​application module includes a fault simulation and diagnosis enhancement module based on a generative adversarial network, an intelligent wind turbine group collaborative control module driven by reinforcement learning, and a cross-wind farm performance optimization module based on transfer learning. The fault simulation and diagnosis enhancement module based on a generative adversarial network is based on GAN technology. By learning from historical fault data and normal operating data, it generates highly realistic simulated fault data, including various types of wind turbine faults and equipment operating status data under complex meteorological conditions. This data is used to expand the training data set of the fault diagnosis model and enhance the model's ability to identify rare faults and complex working conditions; a fault diagnosis model based on a combination of a convolutional neural network and a long short-term memory network is constructed. The convolutional neural network is used to extract features from the image data of the wind turbine equipment and automatically identify potential physical damage and structural defects; the long short-term memory network models the operating parameter data of the time series to capture the dynamic change trend of the equipment operating status, thereby achieving early warning and accurate diagnosis of faults.

6. A wind farm monitoring system integrating historical data and AI technology according to claim 5, characterized in that: The reinforcement learning-driven intelligent wind turbine group collaborative control module is used to regard the wind turbine group of the entire wind farm as a multi-agent system. Each wind turbine is trained as an independent agent through a reinforcement learning algorithm to autonomously decide on the optimal operating parameters based on local wind speed and wind direction information and the operating status of adjacent wind turbines, including but not limited to pitch angle adjustment and speed control, so as to maximize the overall power generation efficiency and load balance of the wind turbine group. A distributed collaborative reinforcement learning framework is introduced to enable information exchange and collaboration between wind turbines to avoid wake interference between each other, optimize the wind flow field distribution in the wind farm, and improve wind energy utilization.

7. The wind farm monitoring system integrating historical data and AI technology according to claim 5 is characterized in that: The cross-wind farm performance optimization module based on transfer learning is used to target the similarities and differences between different wind farms. By using transfer learning technology, the performance optimization model trained in a mature wind farm can be quickly migrated to a newly constructed wind farm with similar operating conditions, so as to reduce the time and amount of data required for model training of new wind farms. At the same time, a meta-model for wind farm performance evaluation is established. By comprehensively analyzing the historical data of multiple wind farms, common performance influencing factors and key indicators are extracted, including the influence of topography on wind speed distribution and the adaptability of different wind turbine models in different climate zones. Based on meta-knowledge, decision support is provided for the planning, construction and operation of new wind farms.

8. The wind farm monitoring system integrating historical data and AI technology according to claim 1 is characterized in that: The visualization module constructs an immersive 3D visualization monitoring interface for the wind farm based on virtual reality and augmented reality technologies. Operation and maintenance personnel can view the real-time operating status of the wind farm by wearing VR helmets and using AR smart glasses, including but not limited to the rotation of wind turbines, changes in meteorological conditions, and key parameter information of equipment.

9. The wind farm monitoring system integrating historical data and AI technology according to claim 1, characterized in that: The intelligent decision-making module develops an intelligent decision-making assistance platform based on the results of the AAI application module. It is used to generate multiple feasible decision-making plans based on different decision-making scenarios, including wind turbine maintenance plan formulation, power generation scheduling optimization, and equipment upgrade and transformation decisions, using machine learning algorithms, and quantitatively evaluates and compares the potential risks, economic benefits, and implementation difficulties of each plan.

Images