Field edge side workstation based on artificial intelligence and use method thereof

Abstract

The invention relates to the technical field of edge computing, in particular to an artificial intelligence-based field edge side workstation and a use method thereof, and the workstation comprises a data acquisition module, a leaf state analysis module, an alarm module and an artificial intelligence prediction correction module. By integrating the functions of real-time data acquisition, intelligent analysis and prediction correction, the operation efficiency and safety of the wind power generation equipment in a complex environment are effectively improved, the starting wind speed is monitored in real time and dynamically adjusted, it is ensured that the equipment is started and operated under the optimal condition, and the service life of the equipment is prolonged. According to the method, the wind power generation efficiency is improved, power generation efficiency reduction and equipment damage caused by icing or wind speed fluctuation are avoided, in addition, the risk of equipment faults is remarkably reduced through an early warning and intelligent correction mechanism, the service life of the equipment is prolonged, the maintenance efficiency of the equipment is improved, and the problem that fault prediction is inaccurate due to single monitoring parameters of the wind power generation equipment is effectively solved.

Description

Technical Field

[0001] The present invention relates to the field of edge computing technology, and in particular to an on-site edge-side workstation based on artificial intelligence and a method of using the same. Background Art

[0002] As the world pays more attention to renewable energy, wind power generation is growing rapidly as a clean form of energy. However, wind power generation equipment faces various challenges in actual operation, especially in severe weather conditions, such as abnormal wind speed and icing, which may lead to reduced equipment efficiency or even failure. Therefore, it is particularly important to monitor and predict these potential problems in real time to ensure efficient operation of equipment and reduce maintenance costs.

[0003] The patent document with publication number CN117006000A discloses a separate wind turbine monitoring device. The device includes: a wind turbine operating component voiceprint data collection device, an edge computing device, a maintenance solution matching device and an alarm device, wherein the wind turbine component voiceprint data collection device is used to collect voiceprint data of wind turbine operating components, the edge computing device is used to control the wind turbine operating component voiceprint data collection device to collect voiceprint data of wind turbine operating components, and to identify abnormalities of wind turbine operating components according to the voiceprint data, the maintenance solution matching device is used to match maintenance solutions according to abnormal identification results, the alarm device is used to issue an alarm according to abnormal identification results, the wind turbine component voiceprint data collection device is connected to the edge computing device based on a transmission protocol, the edge computing device is connected to the maintenance solution matching device and the alarm device respectively, the wind turbine component voiceprint data collection device is installed at each operating component position of the wind turbine, and the edge computing device, the maintenance solution matching device and the alarm device are installed in a monitoring center.

[0004] It can be seen that the device mainly relies on the voiceprint data of the wind turbine operating components for anomaly identification, and the voiceprint data may be interfered by external noise, resulting in accuracy and reliability issues. This method has limited ability to detect physical conditions such as wind speed and icing, and may not be able to fully evaluate the operating status of the equipment. The maintenance plan matching device relies on the anomaly identification results to match the maintenance plan, but the voiceprint data may not be able to accurately distinguish different types of faults, which may lead to inaccurate maintenance plan selection. Summary of the invention

[0005] To this end, the present invention provides an artificial intelligence-based on-site edge-side workstation and a method of using the same, so as to overcome the problem of inaccurate fault prediction caused by the single monitoring parameter of wind power generation equipment in the prior art.

[0006] To achieve the above object, the present invention provides an artificial intelligence-based on-site edge workstation, comprising:

[0007] A data acquisition module, for monitoring the real-time wind speed, real-time ice area and real-time ice thickness at the blade surface of the wind power generation equipment during the process of rotating at a preset starting wind speed;

[0008] a blade state analysis module, connected to the data acquisition module, comprising a wind speed state determination unit and an icing state determination unit, wherein the wind speed state determination unit is used to determine an abnormal wind speed state according to the real-time wind speed and a preset standard wind speed, and to determine a short-term fluctuation mode or a long-term fluctuation mode of the abnormal wind speed state according to the abnormal wind speed state and a preset standard abnormal duration, and the icing state determination unit is used to determine a short-term fluctuation cause according to the short-term fluctuation mode, the real-time icing area, the preset standard icing area, and the preset standard ice surface change speed;

[0009] The icing state determination unit is also used to determine the cause of the long-term fluctuation according to the long-term fluctuation mode, the real-time icing area, the standard icing area, the real-time icing thickness and the preset standard thickness distribution;

[0010] An alarm module, connected to the blade state analysis module, for issuing a corresponding alarm according to the long-term fluctuation cause or the short-term fluctuation cause;

[0011] The artificial intelligence prediction and correction module is connected to the data acquisition module and the blade state analysis module respectively, and is used to predict the average predicted wind speed within a preset prediction time period based on all the real-time wind speeds within a preset correction time period, perform correction judgment based on the average predicted wind speed, the starting wind speed and a preset standard wind speed difference, form a correction judgment result, and increase the starting wind speed according to the correction judgment result.

[0012] Furthermore, the wind speed state determination unit is used to determine the abnormal wind speed state according to the real-time wind speed and the preset standard wind speed, including:

[0013] The wind speed state determination unit is used to determine that the situation where the real-time wind speed is greater than the standard wind speed is an abnormal wind speed state.

[0014] Further, the wind speed state determination unit is used to obtain the duration of the abnormal wind speed state, and determine that the abnormal wind speed state whose duration is less than or equal to the standard abnormal duration is a short-term fluctuation mode.

[0015] Further, the wind speed state determination unit is used to determine that the abnormal wind speed state whose duration is greater than the standard abnormal duration is a long-term fluctuation mode.

[0016] Furthermore, the icing state determination unit is used to calculate the change rate of the real-time icing area based on all real-time icing areas within the duration to form a real-time icing area change rate when the wind speed state determination unit determines that the abnormal wind speed state is in a short-term fluctuation mode and the real-time icing area is less than or equal to the standard icing area. The icing state determination unit determines that the short-term fluctuation reason why the real-time icing area change rate is greater than the standard ice surface change rate is a local cooling effect.

[0017] Furthermore, the icing state determination unit is used to determine that the abnormal wind speed state is a short-term fluctuation mode when the wind speed state determination unit determines that the real-time icing area is greater than the standard icing area, and the icing state determination unit determines that the cause of the short-term fluctuation is the influence of the extremely low temperature environment.

[0018] Furthermore, the icing state determination unit is used to calculate the standard deviation of all the real-time icing thicknesses within the duration when the wind speed state determination unit determines that the abnormal wind speed state is a long-term fluctuation mode, and the real-time icing area is larger than the standard icing area, to obtain the icing thickness distribution. When the icing thickness distribution is less than or equal to the standard thickness distribution, the icing state determination unit determines that the cause of the long-term fluctuation is the influence of the high humidity environment.

[0019] Furthermore, the icing state determination unit is used to determine that the cause of the long-term fluctuation is blade weathering and wear when the wind speed state determination unit determines that the abnormal wind speed state is a long-term fluctuation mode, and the real-time icing area is greater than the standard icing area, and the icing thickness distribution is greater than the standard thickness distribution.

[0020] Furthermore, a preset deep learning model is provided in the artificial intelligence prediction and correction module, and all the real-time wind speeds within the preset correction period are input into the deep learning model to predict all the predicted wind speeds within the prediction period, and the average value of all the predicted wind speeds within the prediction period is calculated to obtain the average predicted wind speed. The artificial intelligence prediction and correction module determines that the starting wind speed needs to be corrected when the average predicted wind speed is greater than the starting wind speed and the difference between the average predicted wind speed and the starting wind speed is greater than the standard wind speed difference. The artificial intelligence prediction and correction module increases the starting wind speed according to the difference between the average predicted wind speed and the starting wind speed, the standard wind speed difference and a preset correction coefficient.

[0021] A method for using an artificial intelligence-based on-site edge workstation, based on the artificial intelligence-based on-site edge workstation, comprising:

[0022] Monitoring the real-time wind speed, real-time ice area and real-time ice thickness at the blade surface of the wind power generation equipment during the process of rotating at a preset starting wind speed;

[0023] Determine an abnormal wind speed state according to the real-time wind speed and a preset standard wind speed, determine a short-term fluctuation mode or a long-term fluctuation mode of the abnormal wind speed state according to the abnormal wind speed state and a preset standard abnormal duration, and determine a short-term fluctuation cause according to the short-term fluctuation mode, the real-time icing area, the preset standard icing area, and the preset standard ice surface change speed;

[0024] Determining the cause of the long-term fluctuation according to the long-term fluctuation pattern, the real-time icing area, the preset standard icing area, the real-time icing thickness and the preset standard thickness distribution;

[0025] issuing corresponding alarms according to the long-term fluctuation reasons or the short-term fluctuation reasons;

[0026] Based on the average predicted wind speed within the preset prediction time period according to all the real-time wind speed predictions within the preset correction time period, a correction judgment is performed according to the average predicted wind speed, the starting wind speed and the preset standard wind speed difference to form a correction judgment result, and the starting wind speed is increased according to the correction judgment result.

[0027] Compared with the prior art, the beneficial effect of the present invention lies in that, by integrating real-time data collection, intelligent analysis and prediction and correction functions, the operating efficiency and safety of wind power generation equipment in complex environments are effectively improved, the starting wind speed is monitored in real time and dynamically adjusted, ensuring that the equipment is started and operated under optimal conditions, avoiding reduced power generation efficiency and equipment damage due to icing or wind speed fluctuations. In addition, the early alarm and intelligent correction mechanism significantly reduce the risk of equipment failure, improve the service life and maintenance efficiency of the equipment, and effectively solve the problem of inaccurate fault prediction due to the single monitoring parameters of wind power generation equipment.

[0028] Furthermore, by determining the abnormal wind speed status in real time, corresponding measures can be taken in a timely manner, such as adjusting equipment operating parameters or issuing alarms, to ensure that wind power generation equipment can effectively protect itself when the wind speed is abnormal, avoiding potential damage or efficiency loss.

[0029] Furthermore, by judging the duration of abnormal wind speed, it is possible to accurately distinguish short-term fluctuations from long-term abnormalities, and thus formulate corresponding response strategies. Timely identification of short-term fluctuation patterns helps to quickly adjust equipment operating parameters and reduce potential risks caused by short-term wind speed abnormalities.

[0030] Furthermore, by accurately identifying long-term fluctuation patterns, more stable and long-term countermeasures can be taken for persistent wind speed anomalies, thus preventing equipment from being damaged due to long-term wind speed instability.

[0031] Furthermore, by comparing the speed of ice surface change, the specific reasons for abnormal changes in ice area in a short period of time can be accurately identified, thereby effectively judging whether there is a local cooling effect, which helps to promptly discover and solve local icing problems of equipment and avoid equipment performance degradation or damage due to uneven cooling.

[0032] Furthermore, by comparing the real-time ice area, it is possible to accurately identify abnormal ice area caused by extremely low temperature environments, thereby helping to take prompt measures, such as adjusting operating conditions or performing environmental protection, to avoid equipment failure or performance degradation due to extreme ambient temperatures.

[0033] Furthermore, by evaluating the distribution of ice thickness, it is possible to identify whether the ice is uniform, which reflects the impact of ambient humidity on the ice formation process. A high humidity environment usually results in a more uniform ice layer with a smaller thickness distribution, while a low humidity environment may lead to uneven ice formation.

[0034] Furthermore, by comparing the distribution of ice thickness, it is possible to accurately identify icing anomalies caused by blade weathering and wear, helping to promptly detect and resolve blade surface damage problems.

[0035] Furthermore, by accurately predicting future wind speeds and adjusting the starting wind speed, it is possible to ensure that wind power generation equipment can adapt to environmental changes during actual operation and reduce the impact of wind speed fluctuations on equipment stability.

[0036] Furthermore, the abnormal causes of wind speed and icing conditions are determined through intelligent analysis. Short-term and long-term fluctuation patterns are identified through classification, and future wind speeds are predicted in combination with deep learning models, which enables effective adjustment of the preset start-up wind speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 Schematic diagram of an on-site edge workstation based on artificial intelligence in this embodiment;

[0038] Figure 2 A decision logic diagram for determining an abnormal wind speed state by the wind speed state determination unit of this embodiment;

[0039] Figure 3 A decision logic diagram for determining a long-term fluctuation mode and a short-term fluctuation mode by a wind speed state determination unit in this embodiment;

[0040] Figure 4 This is a decision logic diagram for the icing state determination unit of this embodiment to determine the cause of short-term fluctuation. DETAILED DESCRIPTION

[0041] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0042] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0043] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0044] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0045] See also Figure 1 As shown, it is a schematic diagram of an on-site edge workstation based on artificial intelligence in this embodiment;

[0046] This embodiment provides an artificial intelligence-based field edge workstation, which is installed near a wind power generation device to centrally process and analyze data collected from various components of the wind power generation device. The workstation includes:

[0047] A data acquisition module, for monitoring the real-time wind speed, real-time ice area and real-time ice thickness at the blade surface of the wind power generation equipment during the process of rotating at a preset starting wind speed;

[0048] The real-time wind speed is collected by a wind speed sensor, which is installed on or around the blades of the wind power generation equipment to monitor the real-time wind speed around the blades. These sensors usually use ultrasonic anemometers or thermal anemometers, which can accurately detect small changes in wind speed.

[0049] The real-time ice area can be collected by an optical sensor, which uses laser or infrared technology to monitor the ice area on the blade surface in real time. The optical sensor can detect changes in the reflectivity of the blade surface by reflecting or transmitting a light beam, thereby inferring the area of ​​the ice layer.

[0050] The real-time ice thickness is collected by a capacitive sensor, which measures the ice thickness by detecting changes in the capacitance of the blade surface. This sensor is very sensitive to changes in ice thickness and is suitable for fine monitoring.

[0051] a blade state analysis module, connected to the data acquisition module, comprising a wind speed state determination unit and an icing state determination unit, wherein the wind speed state determination unit is used to determine an abnormal wind speed state according to the real-time wind speed and a preset standard wind speed, and to determine a short-term fluctuation mode or a long-term fluctuation mode of the abnormal wind speed state according to the abnormal wind speed state and a preset standard abnormal duration, and the icing state determination unit is used to determine a short-term fluctuation cause according to the short-term fluctuation mode, the real-time icing area, the preset standard icing area, and the preset standard ice surface change speed;

[0052] The icing state determination unit is also used to determine the cause of the long-term fluctuation according to the long-term fluctuation mode, the real-time icing area, the standard icing area, the real-time icing thickness and the preset standard thickness distribution;

[0053] An alarm module, connected to the blade state analysis module, for issuing a corresponding alarm according to the long-term fluctuation cause or the short-term fluctuation cause;

[0054] The artificial intelligence prediction and correction module is connected to the data acquisition module and the blade state analysis module respectively, and is used to predict the average predicted wind speed within a preset prediction time period based on all the real-time wind speeds within a preset correction time period, perform correction judgment based on the average predicted wind speed, the starting wind speed and a preset standard wind speed difference, form a correction judgment result, and increase the starting wind speed according to the correction judgment result.

[0055] The artificial intelligence-based field edge workstation provided in this embodiment is mainly used in wind power generation scenarios in cold areas. Especially in winter or high altitude areas, the blades of wind power generation equipment are easily affected by low temperatures and freezing conditions, resulting in icing, which affects the power generation efficiency and equipment safety. Through this workstation, the wind speed, ice area and ice thickness on the blade surface can be monitored in real time, and these data can be analyzed and processed to ensure the stable operation of wind power generation equipment in complex environments.

[0056] The preset starting wind speed refers to the minimum wind speed at which the wind power generation equipment starts to work, which is usually set at 3-5m / s. In this embodiment, it is set to 4m / s to ensure that the equipment can start stably and avoid instability at low wind speeds.

[0057] The preset standard wind speed is the ideal wind speed range for normal operation of the equipment, which is usually 6-12m / s. In this embodiment, it is set to 8-10m / s, which can ensure that the equipment operates within the optimal performance range and improve power generation efficiency and safety.

[0058] The abnormal wind speed state refers to the state where the wind speed exceeds the normal working range, usually lower than the lower limit of the standard wind speed or higher than the upper limit. In this embodiment, it is set to be lower than 6m / s or higher than 12m / s to detect abnormalities in time and prevent equipment damage.

[0059] The standard abnormal duration refers to the duration of the abnormal wind speed state, which is usually set to 10-30 minutes. In this embodiment, it is set to 20 minutes, which helps to distinguish short-term fluctuations from long-term abnormalities and optimize response measures.

[0060] The short-term fluctuation mode refers to the fluctuation of wind speed in a short period of time, usually a change within 1-5 minutes. In this embodiment, it is set to fluctuate within 3 minutes to capture short-term wind speed changes and adjust the equipment status in time.

[0061] The long-term fluctuation mode refers to the change of wind speed over a long period of time, and usually takes into account the long-term trend of wind speed change. This embodiment provides long-term stability analysis based on the preset long-term wind speed fluctuation analysis.

[0062] The preset standard icing area refers to the normal range of icing on the blades, which is usually set to the standard value of the blade surface. This embodiment sets a reasonable icing area threshold based on the equipment working conditions and environment. It is usually set to 10-30 square meters, and the specific value depends on the equipment and environmental conditions. This embodiment is set to 20 square meters, which can cover the icing area under most normal operating conditions and effectively evaluate the impact of icing on the equipment.

[0063] The preset standard ice surface change speed refers to the normal speed of ice layer change, which is usually set to a change speed within a certain range. This embodiment sets a suitable change speed based on the actual environment to judge the icing situation. It is usually set to 0.5-2 mm / hour, and this embodiment is set to 1 mm / hour. This range can reflect the normal icing rate and is convenient for comparison with the actual icing rate to judge whether it is abnormal.

[0064] The preset standard thickness distribution refers to the uniformity of ice thickness on the blade, which is usually set to a specific distribution range. This embodiment is set according to equipment requirements and environment to accurately evaluate the impact of ice. It is usually set to 90%-95% of the ice thickness uniformity. This embodiment is set to 92% to ensure that the uniformity of ice thickness can be accurately evaluated and abnormal conditions of uneven distribution can be found.

[0065] The preset correction time refers to the time range for wind speed correction, which is usually the wind speed data collection period within a certain period of time. This embodiment is set to a reasonable correction period to ensure the accuracy of the correction result. It is usually 30-60 minutes, and this embodiment is set to 45 minutes. This time can cover the fluctuation range of wind speed and make effective corrections to ensure the accuracy of the correction result.

[0066] The preset prediction duration refers to the time range used to predict future wind speeds, which is usually set to the time period for data prediction. This embodiment is set to the effective prediction period to provide a timely basis for wind speed adjustment. The preset standard wind speed difference refers to the threshold used to determine wind speed corrections, which is usually set to a certain range of wind speed differences. This embodiment is set to ensure the rationality and accuracy of wind speed adjustments. It is generally set to 1-2 hours, and this embodiment is set to 1.5 hours, so as to provide timely wind speed forecasts and support the implementation of wind speed corrections and adjustment measures.

[0067] The preset standard wind speed difference refers to the threshold used to determine whether the wind speed needs to be adjusted during the wind speed correction process. Due to the different sensitivities of different wind power generation equipment to wind speed changes, the setting of the wind speed difference should take into account the design and operating characteristics of the equipment. It is usually set to 2-5m / s, and is set to 3m / s in this embodiment to ensure the rationality and accuracy of the wind speed correction and prevent excessive impact of wind speed fluctuations on equipment performance.

[0068] First, the data acquisition module monitors the wind speed, ice area and ice thickness on the surface of the wind turbine blades in real time; then, the blade status analysis module uses the wind speed status determination unit to identify abnormal wind speed status based on the real-time data collected, and determines the wind speed fluctuation mode (short-term or long-term) according to the duration of the abnormality. Next, the ice state determination unit analyzes the cause of ice according to different fluctuation modes and changes in the ice surface; at the same time, the alarm module will issue corresponding alarms based on the analysis results. Finally, the artificial intelligence prediction and correction module predicts future wind speed changes based on historical wind speed data, and dynamically adjusts the startup wind speed by correcting the judgment results to optimize the equipment operation status.

[0069] By integrating real-time data collection, intelligent analysis and prediction and correction functions, the operating efficiency and safety of wind power generation equipment in complex environments are effectively improved. Real-time monitoring and dynamic adjustment of the starting wind speed ensure that the equipment starts and operates under optimal conditions, avoiding reduced power generation efficiency and equipment damage caused by icing or wind speed fluctuations. In addition, early alarms and intelligent correction mechanisms significantly reduce the risk of equipment failure, improve the service life and maintenance efficiency of the equipment, and effectively solve the problem of inaccurate fault prediction due to the single monitoring parameters of wind power generation equipment.

[0070] Please continue reading Figure 2 As shown, it is a decision logic diagram of the wind speed state determination unit of this embodiment for determining an abnormal wind speed state;

[0071] Specifically, the wind speed state determination unit is used to determine the abnormal wind speed state according to the real-time wind speed and the preset standard wind speed, including:

[0072] The wind speed state determination unit is used to determine that the situation where the real-time wind speed is greater than the standard wind speed is an abnormal wind speed state.

[0073] The wind speed status determination unit compares the real-time monitored wind speed with the preset standard wind speed. If the real-time wind speed exceeds the standard wind speed, it is determined to be an abnormal wind speed state. This process ensures that the system can immediately identify wind speed anomalies and prevent the equipment from operating under unsuitable wind speed conditions.

[0074] By determining abnormal wind speed conditions in real time, appropriate measures can be taken in a timely manner, such as adjusting equipment operating parameters or issuing alarms, to ensure that wind power generation equipment can effectively protect itself when wind speeds are abnormal, avoiding potential damage or efficiency loss.

[0075] Please continue reading Figure 3 As shown, it is a decision logic diagram of the wind speed state determination unit in this embodiment for determining the long-term fluctuation mode and the short-term fluctuation mode;

[0076] Specifically, the wind speed state determination unit is used to obtain the duration of the abnormal wind speed state, and determine that the abnormal wind speed state whose duration is less than or equal to the standard abnormal duration is a short-term fluctuation mode.

[0077] The wind speed state determination unit monitors the duration of the abnormal wind speed state and compares it with the preset standard abnormal duration. If the duration is less than or equal to the standard abnormal duration, it is determined to be a short-term fluctuation mode. This detection ensures the timeliness of the wind speed anomaly analysis and helps to timely identify the wind speed changes in the short term.

[0078] By determining the duration of abnormal wind speed, it is possible to accurately distinguish between short-term fluctuations and long-term abnormalities, and thus formulate corresponding response strategies. Timely identification of short-term fluctuation patterns helps to quickly adjust equipment operating parameters and reduce potential risks caused by short-term wind speed abnormalities.

[0079] Specifically, the wind speed state determination unit is used to determine that the abnormal wind speed state whose duration is greater than the standard abnormal duration is a long-term fluctuation mode.

[0080] The wind speed state determination unit monitors the duration of the abnormal wind speed state and compares the duration with the preset standard abnormal duration. If the duration is greater than the standard abnormal duration, it is determined to be a long-term fluctuation mode. This determination method is used to identify long-term wind speed anomalies for in-depth analysis and adjustment.

[0081] By accurately identifying long-term fluctuation patterns, more stable and long-term responses can be taken to persistent wind speed anomalies, thus preventing equipment from being damaged by long-term wind speed instability.

[0082] Please continue reading Figure 4 As shown, it is a decision logic diagram of the icing state determination unit of this embodiment for determining the cause of short-term fluctuation;

[0083] Specifically, the icing state determination unit is used to calculate the change rate of the real-time icing area based on all real-time icing areas within the duration to form the real-time icing area change rate when the wind speed state determination unit determines that the abnormal wind speed state is in a short-term fluctuation mode and the real-time icing area is less than or equal to the standard icing area. The icing state determination unit determines that the short-term fluctuation reason why the real-time icing area change rate is greater than the standard ice surface change rate is a local cooling effect.

[0084] In the short-term fluctuation mode, if the real-time icing area is less than or equal to the standard icing area, the icing state determination unit will calculate the change rate of the real-time icing area. By comparing this change rate with the standard ice surface change rate, it is determined whether there is a local cooling effect. If the real-time icing area change rate is greater than the standard ice surface change rate, it is determined to be a local cooling effect.

[0085] By comparing the speed of ice surface change, the specific reasons for abnormal changes in ice area in a short period of time can be accurately identified, thereby effectively judging whether there is a local cooling effect, which helps to promptly discover and solve local icing problems of equipment and avoid equipment performance degradation or damage due to uneven cooling.

[0086] Specifically, the icing state determination unit is used to determine that the abnormal wind speed state is a short-term fluctuation mode when the wind speed state determination unit determines that the real-time icing area is larger than the standard icing area, and the icing state determination unit determines that the cause of the short-term fluctuation is the influence of the extremely low temperature environment.

[0087] In the short-term fluctuation mode, if the real-time icing area is larger than the standard icing area, the icing state determination unit will determine the cause of the short-term fluctuation. If the real-time icing area exceeds the standard value, it means that the icing speed is fast and cannot be explained by the local cooling effect. At this time, the icing state determination unit will determine that it is affected by the extremely low temperature environment.

[0088] By comparing the real-time ice area, it is possible to accurately identify abnormal ice area caused by extremely low temperature environment, thus helping to take prompt measures, such as adjusting operating conditions or performing environmental protection, to avoid equipment failure or performance degradation caused by extreme ambient temperature.

[0089] Specifically, the icing state determination unit is used to calculate the standard deviation of all the real-time icing thicknesses within the duration when the wind speed state determination unit determines that the abnormal wind speed state is a long-term fluctuation mode, and the real-time icing area is larger than the standard icing area, to obtain the icing thickness distribution. When the icing thickness distribution is less than or equal to the standard thickness distribution, the icing state determination unit determines that the cause of the long-term fluctuation is the influence of the high humidity environment.

[0090] In the long-term fluctuation mode, if the real-time icing area is greater than the standard icing area, the icing state determination unit will calculate the standard deviation of all real-time icing thicknesses within the duration to obtain the icing thickness distribution. If the calculated icing thickness distribution is less than or equal to the standard thickness distribution, the icing state determination unit will determine that the cause of the long-term fluctuation is the influence of the high humidity environment.

[0091] By evaluating the distribution of ice thickness, it is possible to identify whether the ice is uniform, which reflects the impact of ambient humidity on the ice formation process. A high humidity environment usually results in a more uniform ice layer with a smaller thickness distribution, while a low humidity environment may lead to uneven ice formation.

[0092] Specifically, the icing state determination unit is used to determine that the cause of the long-term fluctuation is blade weathering and wear when the wind speed state determination unit determines that the abnormal wind speed state is a long-term fluctuation mode, and the real-time icing area is greater than the standard icing area, and the icing thickness distribution is greater than the standard thickness distribution.

[0093] When the wind speed state determination unit identifies a long-term fluctuation pattern, the icing state determination unit further analyzes the real-time icing area and icing thickness distribution. If the real-time icing area exceeds the standard range and the icing thickness distribution also exceeds the standard, the system will determine that the cause of the icing phenomenon is the weathering wear of the blades. This is because weathering wear causes uneven distribution of the ice layer, which affects the overall thickness and distribution of the ice layer.

[0094] By comparing the ice thickness distribution, we can accurately identify icing anomalies caused by blade weathering and wear, helping to promptly detect and resolve blade surface damage problems.

[0095] Specifically, a preset deep learning model is provided in the artificial intelligence prediction and correction module, and all the real-time wind speeds within the preset correction time are input into the deep learning model to predict all the predicted wind speeds within the prediction time, and the average value of all the predicted wind speeds within the prediction time is calculated to obtain the average predicted wind speed. When the average predicted wind speed is greater than the starting wind speed and the difference between the average predicted wind speed and the starting wind speed is greater than the standard wind speed difference, the artificial intelligence prediction and correction module determines that the starting wind speed needs to be corrected. The artificial intelligence prediction and correction module increases the starting wind speed according to the difference between the average predicted wind speed and the starting wind speed, the standard wind speed difference and a preset correction coefficient, wherein Vn=V+k×ΔV, Vn is the increased starting wind speed, V is the starting wind speed, k is the correction coefficient, and ΔV is the difference between the average predicted wind speed and the starting wind speed.

[0096] The preset correction coefficient is a parameter used to adjust the starting wind speed, and its value depends on the actual needs of the system and the operating environment. Usually, the correction coefficient is set to a value between 0 and 1, and the specific value depends on the characteristics of the wind power generation system and the prediction accuracy. The common setting range is 0.1 to 0.5. Setting it too low may result in a slow response, and setting it too high may result in over-adjustment. In this embodiment, the correction coefficient is set to 0.3. This value is a balanced choice that can provide sufficient adjustment range when correcting the wind speed while avoiding the risks of over-correction.

[0097] In the artificial intelligence prediction and correction module, a long short-term memory network (LSTM) is used as a deep learning model for wind speed prediction. The model predicts the wind speed for a period of time in the future by analyzing the real-time wind speed data within the preset correction time, and calculates the average value of these predicted wind speeds. If the average predicted wind speed is higher than the starting wind speed, and the difference between the two is greater than the standard wind speed difference, it is determined that the starting wind speed needs to be corrected. In this way, the LSTM model can capture the long-term trend and time dependence of wind speed data, provide accurate prediction results, thereby optimizing wind speed correction and improving the operating efficiency and stability of wind power generation equipment.

[0098] The AI ​​prediction and correction module uses a deep learning model to process real-time wind speed data within a preset correction period, predict wind speeds for a period of time in the future, and calculate the average of these predicted wind speeds. When this average predicted wind speed exceeds the preset start-up wind speed, and the difference between it and the start-up wind speed exceeds the standard wind speed difference, the module determines that the start-up wind speed needs to be corrected. Then, based on the difference between the predicted wind speed and the start-up wind speed, the standard wind speed difference, and the correction factor, the module adjusts and increases the start-up wind speed.

[0099] By accurately predicting future wind speeds and adjusting the starting wind speed, it is possible to ensure that wind power generation equipment can adapt to environmental changes during actual operation and reduce the impact of wind speed fluctuations on equipment stability.

[0100] A method for using an artificial intelligence-based on-site edge workstation, based on the artificial intelligence-based on-site edge workstation, comprising:

[0101] Monitoring the real-time wind speed, real-time ice area and real-time ice thickness at the blade surface of the wind power generation equipment during the process of rotating at a preset starting wind speed;

[0102] Determine an abnormal wind speed state according to the real-time wind speed and a preset standard wind speed, determine a short-term fluctuation mode or a long-term fluctuation mode of the abnormal wind speed state according to the abnormal wind speed state and a preset standard abnormal duration, and determine a short-term fluctuation cause according to the short-term fluctuation mode, the real-time icing area, the preset standard icing area, and the preset standard ice surface change speed;

[0103] Determining the cause of the long-term fluctuation according to the long-term fluctuation pattern, the real-time icing area, the preset standard icing area, the real-time icing thickness and the preset standard thickness distribution;

[0104] issuing corresponding alarms according to the long-term fluctuation reasons or the short-term fluctuation reasons;

[0105] Based on the average predicted wind speed within the preset prediction time period according to all the real-time wind speed predictions within the preset correction time period, a correction judgment is performed according to the average predicted wind speed, the starting wind speed and the preset standard wind speed difference to form a correction judgment result, and the starting wind speed is increased according to the correction judgment result.

[0106] When using the AI-based field edge workstation, first, the real-time wind speed, ice area and ice thickness on the surface of the wind turbine blades are monitored through the data acquisition module. Then, based on the comparison between the real-time wind speed and the preset standard wind speed, it is determined whether there is an abnormal wind speed state, and whether it is a short-term or long-term fluctuation mode based on the duration of the abnormal wind speed state. Next, the icing state in the short-term fluctuation mode is analyzed. If the icing area and the speed of change are abnormal, it is identified as a local cooling effect; if the icing area in the long-term fluctuation mode is large and the thickness distribution exceeds the standard, it is identified as blade weathering wear. Finally, the future wind speed is predicted through the deep learning model. If the predicted wind speed exceeds the starting wind speed and the difference exceeds the standard wind speed difference, the starting wind speed is adjusted.

[0107] Intelligent analysis is used to determine the causes of abnormal wind speed and icing conditions. Classification is used to identify short-term and long-term fluctuation patterns, and deep learning models are used to predict future wind speeds, enabling effective adjustment of the preset start-up wind speed.

[0108] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0109] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An artificial intelligence-based on-site edge workstation, characterized in that: include: A data acquisition module, for monitoring the real-time wind speed, real-time ice area and real-time ice thickness at the blade surface of the wind power generation equipment during the process of rotating at a preset starting wind speed; a blade state analysis module, connected to the data acquisition module, comprising a wind speed state determination unit and an icing state determination unit, wherein the wind speed state determination unit is used to determine an abnormal wind speed state according to the real-time wind speed and a preset standard wind speed, and to determine a short-term fluctuation mode or a long-term fluctuation mode of the abnormal wind speed state according to the abnormal wind speed state and a preset standard abnormal duration, and the icing state determination unit is used to determine a short-term fluctuation cause according to the short-term fluctuation mode, the real-time icing area, the preset standard icing area, and the preset standard ice surface change speed; The icing state determination unit is also used to determine the cause of the long-term fluctuation according to the long-term fluctuation mode, the real-time icing area, the standard icing area, the real-time icing thickness and the preset standard thickness distribution; An alarm module, connected to the blade state analysis module, for issuing a corresponding alarm according to the long-term fluctuation cause or the short-term fluctuation cause; The artificial intelligence prediction and correction module is connected to the data acquisition module and the blade state analysis module respectively, and is used to predict the average predicted wind speed within a preset prediction time period based on all the real-time wind speeds within a preset correction time period, perform correction judgment based on the average predicted wind speed, the starting wind speed and a preset standard wind speed difference, form a correction judgment result, and increase the starting wind speed according to the correction judgment result.

2. The artificial intelligence-based on-site edge workstation according to claim 1, characterized in that: The wind speed state determination unit is used to determine the abnormal wind speed state according to the real-time wind speed and the preset standard wind speed, including: The wind speed state determination unit is used to determine that the situation where the real-time wind speed is greater than the standard wind speed is an abnormal wind speed state.

3. The artificial intelligence-based on-site edge workstation according to claim 2, characterized in that: The wind speed state determination unit is used to obtain the duration of the abnormal wind speed state, and determine that the abnormal wind speed state whose duration is less than or equal to the standard abnormal duration is a short-term fluctuation mode.

4. The artificial intelligence-based on-site edge workstation according to claim 3, characterized in that: The wind speed state determination unit is used to determine that the abnormal wind speed state in which the duration is greater than the standard abnormal duration is a long-term fluctuation mode.

5. The artificial intelligence-based on-site edge workstation according to claim 4, characterized in that: The icing state determination unit is used to calculate the change speed of the real-time icing area based on all real-time icing areas within the duration to form the real-time icing area change speed when the wind speed state determination unit determines that the abnormal wind speed state is in a short-term fluctuation mode and the real-time icing area is less than or equal to the standard icing area. The icing state determination unit determines that the short-term fluctuation reason why the real-time icing area change speed is greater than the standard ice surface change speed is a local cooling effect.

6. The artificial intelligence-based on-site edge workstation according to claim 5, characterized in that: The icing state determination unit is used to determine that the abnormal wind speed state is a short-term fluctuation mode when the wind speed state determination unit determines that the real-time icing area is greater than the standard icing area, and the icing state determination unit determines that the cause of the short-term fluctuation is the influence of the extremely low temperature environment.

7. The artificial intelligence-based on-site edge workstation according to claim 6, characterized in that: The icing state determination unit is used to calculate the standard deviation of all the real-time icing thicknesses within the duration when the wind speed state determination unit determines that the abnormal wind speed state is a long-term fluctuation mode, and when the real-time icing area is larger than the standard icing area, to obtain the icing thickness distribution. When the icing thickness distribution is less than or equal to the standard thickness distribution, the icing state determination unit determines that the cause of the long-term fluctuation is the influence of the high humidity environment.

8. The artificial intelligence-based on-site edge workstation according to claim 7, characterized in that: The icing state determination unit is used to determine that the cause of the long-term fluctuation is blade weathering and wear when the wind speed state determination unit determines that the abnormal wind speed state is a long-term fluctuation mode, and the real-time icing area is greater than the standard icing area, and the icing thickness distribution is greater than the standard thickness distribution.

9. The artificial intelligence-based on-site edge workstation according to claim 8, characterized in that: A preset deep learning model is provided in the artificial intelligence prediction and correction module. All the real-time wind speeds within the preset correction period are input into the deep learning model to predict all the predicted wind speeds within the prediction period, and the average value of all the predicted wind speeds within the prediction period is calculated to obtain the average predicted wind speed. The artificial intelligence prediction and correction module determines that the startup wind speed needs to be corrected when the average predicted wind speed is greater than the startup wind speed and the difference between the average predicted wind speed and the startup wind speed is greater than the standard wind speed difference. The artificial intelligence prediction and correction module increases the startup wind speed according to the difference between the average predicted wind speed and the startup wind speed, the standard wind speed difference and a preset correction coefficient.

10. A method for using an artificial intelligence-based on-site edge workstation, based on the artificial intelligence-based on-site edge workstation according to claims 1-9, characterized in that: include: Monitoring the real-time wind speed, real-time ice area and real-time ice thickness at the blade surface of the wind power generation equipment during the process of rotating at a preset starting wind speed; Determine an abnormal wind speed state according to the real-time wind speed and a preset standard wind speed, determine a short-term fluctuation mode or a long-term fluctuation mode of the abnormal wind speed state according to the abnormal wind speed state and a preset standard abnormal duration, and determine a short-term fluctuation cause according to the short-term fluctuation mode, the real-time icing area, the preset standard icing area, and the preset standard ice surface change speed; Determining the cause of the long-term fluctuation according to the long-term fluctuation pattern, the real-time icing area, the preset standard icing area, the real-time icing thickness and the preset standard thickness distribution; issuing corresponding alarms according to the long-term fluctuation reasons or the short-term fluctuation reasons; Based on the average predicted wind speed within the preset prediction time period according to all the real-time wind speed predictions within the preset correction time period, a correction judgment is performed according to the average predicted wind speed, the starting wind speed and the preset standard wind speed difference to form a correction judgment result, and the starting wind speed is increased according to the correction judgment result.

Images