Yaw control method and system for wind turbine generator set

Abstract

The present invention relates to the field of wind power generation technology, and in particular to a yaw control method and system for a wind turbine generator set. The method comprises the following steps: performing wind data collection and historical data preprocessing on collected wind marker point data to generate standard historical wind data; performing seasonal wind data prediction on the standard historical wind data to generate predicted seasonal wind data; performing predicted seasonal wind data optimization on the predicted seasonal wind data to generate optimized predicted seasonal wind data; designing a seasonal yaw control method for the generator set based on the optimized predicted seasonal wind data to generate initial yaw control data; adjusting the yaw azimuth of the initial yaw control data based on real-time wind data to generate yaw control data; and optimizing the yaw direction of the yaw control data to generate optimized yaw control data. The present invention accurately controls the yaw of the wind turbine generator set, thereby reducing a large amount of wasted resources and time.

Description

Technical Field

[0001] The present invention relates to the technical field of wind power generation, and in particular to a yaw control method and system for a wind generator set. Background Art

[0002] Yaw control of wind turbines ensures that wind turbines always face the wind when wind direction changes. Through yaw control, wind turbines can maximize wind energy capture, improve power generation efficiency, and reduce unnecessary mechanical stress, thereby extending the service life of the equipment. However, traditional yaw control methods for wind turbines only control the surrounding wind direction. Each adjustment requires a lot of time and resources, and the adjusted yaw control of the wind turbine is not accurate enough. Summary of the Invention

[0003] Based on this, the present invention provides a yaw control method and system for a wind turbine generator set to solve at least one of the above technical problems.

[0004] To achieve the above object, a yaw control method for a wind turbine generator set includes the following steps:

[0005] Step S1: performing wind power point marking processing on the location of the wind turbine generator set to generate wind power mark point data; performing wind power data collection and historical data preprocessing on the wind power mark point data to generate standard historical wind power data;

[0006] Step S2: using an autoregressive neural network algorithm to perform seasonal wind data forecasting on standard historical wind data to generate forecasted seasonal wind data;

[0007] Step S3: performing historical change statistics of environmental impact parameters on the collected wind marker point data to generate historical environmental impact change parameters; performing optimized processing on the predicted seasonal wind data using the historical environmental impact change parameters to generate optimized predicted seasonal wind data; designing a seasonal yaw control method for the generator set based on the optimized predicted seasonal wind data to generate seasonal initial yaw control data;

[0008] Step S4: Using the Internet of Things device to collect wind data from the collected wind marker point data in real time to generate real-time wind data; adjusting the yaw azimuth of the initial yaw control data according to the real-time wind data to generate yaw control data for the wind direction generator set;

[0009] Step S5: Calculate the mutual influence parameters of the generator sets according to the yaw control data to generate the mutual influence parameters of the generator sets; and optimize and adjust the yaw direction of the yaw control data according to the mutual influence parameters of the generator sets to generate optimized yaw control data.

[0010] By collecting wind marker data, the present invention can obtain wind information for different time periods and locations. This is the basis for yaw control, as knowing the current wind direction and strength can help adjust the orientation of the generator set to maximize wind energy capture and power generation efficiency. By performing historical data preprocessing on the collected wind marker data, abnormal data and noise can be eliminated, resulting in more accurate and stable standard historical wind data. This can avoid interference from bad data in subsequent prediction and optimization processes, improving the reliability and accuracy of subsequent steps. An autoregressive neural network algorithm is used to predict standard historical wind data. This algorithm can learn the seasonal patterns and trends of historical wind data, thereby predicting seasonal wind conditions for a period of time in the future. Such predictions help prepare in advance, allowing the generator set to make yaw adjustments in advance to cope with future wind changes, thereby stabilizing power generation and reducing energy waste. By statistically analyzing the historical changes in environmental influencing parameters of the collected wind marker data, it is possible to understand the impact of environmental factors on wind power, including seasonal factors (such as temperature and air pressure) and other factors that may affect wind power. By statistically analyzing these historical change parameters, it is possible to better understand the patterns of environmental influence on wind power. Optimizing seasonal wind data based on historical environmental influence parameters allows for more accurate predictions of future seasonal wind conditions, making yaw control more precise and adaptable to environmental changes. Real-time wind data collection from wind markers using IoT devices provides real-time insights into current wind conditions. This provides real-time data support for yaw control, enabling turbines to make timely yaw adjustments based on the latest wind conditions, thereby improving wind energy capture. Initial yaw control data is adjusted based on real-time wind data to promptly respond to changing wind conditions, maintaining optimal wind energy capture and enhancing power generation efficiency. Interaction parameters between turbines are calculated using yaw control data. Interactions between turbines can cause changes in wind direction, so understanding these influencing parameters helps optimize yaw control strategies, reducing interference and improving overall wind farm efficiency. Further optimization of yaw control data based on interaction parameters can optimize wind turbine layout and yaw strategies, further improving overall power generation efficiency and reducing system losses. Therefore, the yaw control method of the wind turbine of the present invention performs preliminary control on the yaw through seasonal wind direction, and then adjusts it according to real-time wind data, saving a lot of time and resources for yaw control adjustment, and also considers the mutual wind influence between wind turbines to adjust the yaw control of the wind turbine, making the result more accurate.

[0011] Preferably, step S1 includes the following steps:

[0012] Step S11: performing wind power point marking processing on the location of the wind turbine generator set to generate wind power mark point data;

[0013] Step S12: using the Internet of Things device to collect wind data from the wind marker point data to generate initial wind data;

[0014] Step S13: performing data cleaning on the initial wind power data to generate cleaned wind power data;

[0015] Step S14: performing historical wind data statistics on the cleaning wind data to generate historical wind data;

[0016] Step S15: Standardize the historical wind data using the minimum-maximum standardization method to generate standard historical wind data.

[0017] By collecting wind marker data, the present invention can obtain wind information at different locations. This is crucial for yaw control of wind turbines, as only by understanding the wind direction and wind speed at the current location can corresponding yaw adjustments be made, ensuring that the turbine is always facing the wind direction, thereby maximizing wind energy capture. By collecting wind marker data, the location of the wind turbine can be accurately located, providing accurate location information for subsequent wind data collection and control. Using IoT devices for data collection can automate the data acquisition process, reduce manual intervention and errors, and thus improve the efficiency and accuracy of data collection. During the wind data collection process, there may be some outliers or noise data, such as sensor failures and environmental interference. Data cleaning can effectively remove these outliers, ensuring that the wind data used in subsequent steps is accurate and reliable. By performing historical data statistics on the cleaned wind data, wind conditions at different time periods can be obtained. This allows for analysis of wind seasonality and periodicity, helping to predict future wind trends and providing a basis for yaw control. The historical wind data are standardized using the minimum-maximum normalization method, scaling the data to a uniform range. The standardized data have zero mean and unit variance, which helps eliminate the dimensional differences between different wind data and makes the data easier to process and analyze.

[0018] Preferably, step S2 includes the following steps:

[0019] Step S21: dividing the standard historical wind data into seasonal wind data to generate seasonal wind data;

[0020] Step S22: using an autoregressive neural network algorithm to perform seasonal wind data prediction on the seasonal wind data to generate predicted seasonal wind data.

[0021] By seasonally segmenting standard historical wind data, the present invention can extract wind data for different seasons (e.g., spring, summer, autumn, and winter). This helps analyze and understand the seasonal patterns of wind power, namely, the changing trends and differences in wind power between seasons. An autoregressive neural network algorithm is used to predict seasonal wind data. This algorithm combines neural networks with time series prediction technology to capture the time dependence and seasonal patterns in wind data. Using this prediction model, wind power can be predicted for different future seasons, providing more accurate and reliable data support for yaw control.

[0022] Preferably, step S22 includes the following steps:

[0023] Step S221: using an autoregressive neural network algorithm to establish a mapping relationship for seasonal wind data prediction to generate an initial seasonal wind data prediction model;

[0024] Step S222: dividing the seasonal wind data into historical time series data to generate a seasonal wind training set and a seasonal wind test set;

[0025] Step S223: using the seasonal wind training set to perform model training on the initial seasonal wind data prediction model to generate a seasonal wind data prediction model;

[0026] Step S224: transmitting the seasonal wind test set to the seasonal wind data prediction model to perform seasonal wind data prediction to generate predicted seasonal wind data.

[0027] The present invention uses an autoregressive neural network algorithm to establish a mapping relationship for seasonal wind data prediction. The autoregressive neural network algorithm can capture the time dependency and seasonal patterns in wind data, enabling the prediction model to better predict wind changes in future seasons. The seasonal wind data is divided into a historical time series, and the data set is divided into a seasonal wind training set and a seasonal wind test set. The purpose of this is to provide independent data sets for model training and evaluation, avoid model overfitting and deviation in evaluation results, and help the model learn and predict time dependencies. By using the seasonal wind training set, the initial seasonal wind data prediction model is trained. During the model training process, the model learns the characteristics and patterns of seasonal wind data, optimizes model parameters, and enables the model to more accurately predict wind changes in future seasons. By using the seasonal wind test set to evaluate the prediction model, the prediction performance and accuracy of the model can be understood. The evaluation results can guide whether further model optimization is needed to ensure the reliability and accuracy of the model. In this way, the model can obtain predicted seasonal wind data. The predicted seasonal wind data can help predict wind conditions in future seasons and provide more accurate and reliable data support for yaw control.

[0028] Preferably, step S3 includes the following steps:

[0029] Step S31: Using the Internet of Things device to perform environmental impact parameter tagging on the collected wind power mark point data to generate environmental impact parameters;

[0030] Step S32: using a statistical analysis method to perform statistics on the change data of historical environmental impact parameters on the interference wind factor data to generate historical environmental impact change parameters;

[0031] Step S33: performing optimization calculation on the predicted seasonal wind data using the seasonal wind data optimization algorithm and the historical environmental impact change parameters to generate optimized predicted seasonal wind data;

[0032] Step S34: Designing a seasonal yaw control method for the generator set based on the optimized predicted seasonal wind data to generate seasonal initial yaw control data.

[0033] The present invention uses the Internet of Things device to mark the collected wind mark point data with environmental impact parameters, and can obtain parameter information related to the environment, which may have an impact on wind power, such as vegetation, humidity, mountain peaks and other problems. Through the marking process, the environmental impact parameters are associated with the wind data, providing a data basis for subsequent wind optimization. By performing statistical analysis on the interfering wind factor data, the degree of influence of different environmental factors on wind power and their historical change trends can be understood, which helps to understand the interference factors in the wind data, thereby more accurately predicting future wind power and optimizing the operation of the wind power generation system, taking into account the impact of different environmental factors on wind power, and improving the accuracy and stability of the prediction. The seasonal wind data optimization algorithm can take into account the impact of different environmental factors on wind power, and adjust the predicted data according to the changes in environmental factors, so that the prediction results are more accurate and reliable, which helps to improve the accuracy of the prediction and reduce the prediction error. Such an optimization process can better cope with wind power changes under different environmental conditions. The seasonal yaw control method of the generator set is designed based on the optimized predicted seasonal wind data, and the seasonal initial yaw control data is generated. This data will guide the yaw adjustment of the generator set, enabling the generator set to make corresponding initial control actions based on the predicted seasonal wind data, thereby improving the degree of wind energy capture. In the later stage, only fine-tuning is required for precise yaw control, saving a lot of time and resources.

[0034] Preferably, the seasonal wind data optimization algorithm in step S33 is as follows:

[0035]

[0036] Where P represents the optimized forecast seasonal wind data, N represents the coefficient of the forecast seasonal wind data, M represents the coefficient of the environmental impact parameter, and a j Expressed as wind disturbance data of the jth environmental impact parameter, b j It is expressed as the weight information of the wind interference data of the jth environmental impact parameter, p is expressed as the mean change rate of the wind interference data, v j It is expressed as the change rate of the jth environmental impact parameter, c is expressed as the seasonal change weight, k i It represents the wind data of the season where the i-th predicted seasonal wind data belongs, and τ represents the abnormal adjustment value of the optimized predicted seasonal wind data.

[0037] The present invention utilizes a seasonal wind data optimization algorithm, which fully considers the coefficient size N of the predicted seasonal wind data, the coefficient size M of the environmental impact parameter, the wind interference data a of the jth environmental impact parameter, and the wind speed of the wind speed data of the jth environmental impact parameter. j , the weight information b of the wind interference data of the jth environmental impact parameter j, the mean change rate p of wind interference data, the change rate v of the jth environmental impact parameter j , seasonal change weight c, wind data k of the season where the i-th predicted seasonal wind data is located i And the interaction between functions to form a functional relationship:

[0038] Right now. The size of the coefficient of the predicted seasonal wind data is expressed as the data volume of the predicted seasonal wind data. The larger the data volume, the more accurate the optimization effect. The size of the coefficient of the environmental impact parameter is expressed as the data volume of the environmental impact parameter. The larger the data volume, the more data basis is provided, and the better the optimization effect of the predicted seasonal wind data. The wind interference data of the jth environmental impact parameter represents the actual impact of the environmental impact parameter on the wind, and explains the degree of influence of the environmental impact parameter on the wind. The weight information of the wind interference data of the jth environmental impact parameter represents the weight importance of the environmental impact parameter on the wind, and explains the weight ratio of the environmental impact parameters. The mean change rate of the wind interference data represents the overall change trend of the wind interference data. The change rate of the jth environmental impact parameter represents the size of the change of the environmental impact parameter at different times, providing a basis for predicting subsequent change impact parameters, thereby optimizing the predicted seasonal wind data. The seasonal change weight represents the degree of influence of different seasons on the wind, and explains the degree of influence of seasonal factors on the wind. The wind data of the season in which the i-th predicted seasonal wind data is located represents the prediction result of the wind in the current season, and explains the wind strength in the current season. The seasonal variation weights in the optimization algorithm and the wind data for the season in which the i-th predicted seasonal wind data falls can effectively reflect the impact of seasonal variation on wind power. The characteristics of wind power variation vary from season to season, resulting in irregular changes in environmental influencing parameters. This algorithm can account for the impact of environmental variation, enabling more accurate predictions of wind power data for different seasons. The function relationship is adjusted and corrected using the anomaly adjustment value τ for the optimized predicted seasonal wind data, reducing the impact of errors caused by abnormal data or error terms. This allows for more accurate generation of the optimized predicted seasonal wind data P, improving the accuracy and reliability of the optimized calculation of the predicted seasonal wind data. Furthermore, the weight information and adjustment values ​​in this formula can be adjusted based on actual conditions and applied to different predicted seasonal wind data, enhancing the algorithm's flexibility and applicability.

[0039] Preferably, step S4 includes the following steps:

[0040] Step S41: using the Internet of Things device to collect wind mark point data in real time to generate real-time wind data;

[0041] Step S42: performing difference calculation between the real-time wind data and the optimized predicted seasonal wind data to generate difference wind data;

[0042] Step S43: adjusting the yaw azimuth of the initial yaw control data according to the differential wind data to generate yaw control data for the wind direction generator set.

[0043] The present invention utilizes Internet of Things devices to collect wind marker point data in real time and obtain real-time wind data of the current position. Real-time wind data is crucial for yaw control because the direction and intensity of the wind will change continuously. Timely data collection enables the generator set to make yaw adjustments in time to adapt to the changing wind conditions. The real-time wind data and the optimized predicted seasonal wind data are difference calculated to obtain differential wind data. The differential data reflects the deviation between the real-time wind and the predicted wind, helping the generator set to understand the difference between the current wind conditions and the prediction, and providing more accurate data support for yaw control. The initial yaw control data is adjusted for yaw azimuth according to the differential wind data to generate yaw control data for the wind direction generator set. Such yaw control data can enable the generator set to make yaw adjustments according to the real-time wind conditions to keep the generator set facing the direction of the wind, improve the degree of capturing wind energy, and achieve improved power generation efficiency and efficient use of energy.

[0044] Preferably, step S5 includes the following steps:

[0045] Step S51: performing yaw control on the wind turbine generator set according to the yaw control data, and collecting and processing wind power data of the wind turbine generator set using sensors to generate wind power data of the generator set;

[0046] Step S52: Calculating the generator group mutual influence parameters based on the generator group wind power data using the generator group wind power mutual influence calculation formula to generate the generator group mutual influence parameters;

[0047] Step S53: performing yaw direction optimization adjustment processing on the yaw control data according to the mutual influence parameters of the generator sets to generate optimized yaw control data.

[0048] The present invention performs yaw control on a wind turbine generator set according to yaw control data, ensuring that the generator set always faces the direction of the wind to maximize wind energy capture and improve power generation efficiency. Sensors are used to collect and process wind data of the wind turbine generator set to obtain wind information at the location of the generator set. These real-time collected wind data are very important for yaw control and subsequent calculation of mutual influence parameters. By utilizing the wind data of the generator set and applying the wind mutual influence calculation formula of the generator set, the mutual influence parameters of the generator set can be calculated, which reflect the degree of mutual influence of wind energy utilization between different generator sets, help understand the interaction between different generator sets, and provide a basis for subsequent yaw direction optimization. According to the mutual influence parameters of the generator sets, the yaw direction is optimized and adjusted for the yaw control data. By optimizing the yaw control data, multiple generator sets can work better together and avoid mutual interference, thereby improving the overall efficiency and stability of the entire wind power generation system. The optimized yaw control data will guide the operation of the generator set in a complex wind field, maximize the use of wind energy, and achieve efficient power generation.

[0049] Preferably, the calculation formula for the wind power interaction between the generator sets in step S52 is as follows:

[0050]

[0051] In the formula, K represents the mutual influence parameter of the generator set, n represents the coefficient size of the wind power data of the generator set, Expressed as The horizontal axis wind intensity of the wind data of each generator set is Expressed as The vertical axis wind intensity of the wind data of each generator set is Expressed as The wind power of each generator set’s wind data, Expressed as The wind power data of the wind turbines are separated by a distance, g is the rotor diameter of the wind turbine, k is the initial speed of the wind turbine, d is the air density, θ is the wind direction angle of the wind turbine, and δ is the abnormal adjustment value of the parameters that affect each other among the generators.

[0052] The present invention uses a generator set wind mutual influence calculation formula, which fully considers the coefficient size n of the generator set wind data, the first The horizontal axis wind intensity of the wind data of each generator set No. The vertical axis wind intensity of the wind data of each generator set No. The wind power of each generator set No. The distance between wind power data of each generator set The rotor diameter g of the wind turbine, the initial speed k of the wind turbine, the air density d, the wind direction angle θ of the wind turbine, and the interaction relationship between the functions form a functional relationship:

[0053] Right now, The coefficient of the wind power data of the generator set indicates the number of generator sets involved in the calculation. The more wind power data of the generator sets considered, the more comprehensive the calculation results will be. Expressed as The horizontal axis wind intensity of the wind data of the first generator set is The vertical axis wind intensity of the wind data of each generator set indicates the horizontal and vertical wind speeds of the generator set. This parameter is used to calculate the degree of mutual influence of wind forces. The wind power of the claw generator set wind data indicates the wind energy conversion efficiency of the generator set. The wind power plays an important role in the mutual influence of the generator set, and the wind strength is comprehensively considered. The distance between the wind power data of each generator set represents the spatial distance between the generator set and other generator sets, which is used to evaluate the positional relationship between the generator sets; the rotor diameter of the wind turbine set represents the rotor diameter of the generator set, which affects the ability to capture wind power; the initial speed of the wind turbine set represents the starting operating speed of the wind turbine set, which affects the response speed of the generator set; the air density affects the strength of the wind and the efficiency of wind energy conversion; the wind direction angle of the wind turbine set represents the direction facing the wind rotor, which is used to evaluate the orientation of the generator set. The wind power size and the wind power data of the claw generator set The distance between wind data from individual generators is a key factor in measuring the mutual influence between generators. Wind power reflects wind strength, while distance represents the spatial relationship between generators. By considering these two factors, the formula can more accurately assess the degree of mutual influence between different generators, facilitating the optimization of yaw control data. Assessing the degree of interaction between different generators provides a crucial basis for optimizing yaw control data and intelligently adjusting wind power systems. The functional relationship is adjusted using the abnormal adjustment value δ of the generator interaction parameter to reduce the impact of abnormal data or error terms, resulting in a more accurate generation of the generator interaction parameter K. This improves the accuracy and reliability of the calculation of the generator interaction parameter based on wind data. Furthermore, the weighting information and adjustment values ​​in this formula can be adjusted based on actual conditions and applied to wind data from different generators, enhancing the algorithm's flexibility and applicability.

[0054] In this specification, a yaw control system of a wind turbine generator set is provided, comprising:

[0055] The historical wind data collection module is used to collect wind power point markings at the location of the wind turbine generator set to generate collected wind power mark point data; collect wind power data and perform historical data preprocessing on the collected wind power mark point data to generate standard historical wind data;

[0056] Seasonal wind data prediction module, which uses autoregressive neural network algorithm to predict seasonal wind data based on standard historical wind data and generate predicted seasonal wind data;

[0057] The initial yaw control design module is used to perform historical changes in environmental impact parameters of collected wind marker point data to generate historical environmental impact change parameters; use historical environmental impact change parameters to optimize the predicted seasonal wind data to generate optimized predicted seasonal wind data; design the seasonal yaw control method of the generator set based on the optimized predicted seasonal wind data to generate seasonal initial yaw control data;

[0058] The yaw control adjustment module uses IoT devices to collect wind data from wind marker points in real time to generate real-time wind data; it adjusts the yaw direction of the initial yaw control data based on the real-time wind data to generate yaw control data for the wind direction generator set;

[0059] The yaw control optimization module is used to calculate the mutual influence parameters of the generator sets based on the yaw control data and generate the mutual influence parameters of the generator sets; and to optimize and adjust the yaw direction of the yaw control data based on the mutual influence parameters of the generator sets and generate optimized yaw control data.

[0060] The beneficial effect of the present application is that the present invention utilizes Internet of Things devices to collect and process wind data and environmental impact parameters in real time, obtains accurate and timely real-time wind data and environmental information, and provides a reliable data basis for the yaw control of wind turbines. By using an autoregressive neural network algorithm to predict historical wind data, and optimizing the prediction results based on historical environmental impact change parameters, accurate predictions of future seasonal wind power are achieved. Such optimized predictions contribute to the intelligent adjustment of generator sets and improve power generation efficiency. Based on the difference between real-time wind data and optimized predicted seasonal wind data, real-time yaw control of wind turbines is achieved. First, the yaw control data is preliminarily set using predicted seasonal wind direction data, and then adjusted based on real-time wind data, thereby reducing a large amount of lost resources and time. At the same time, the yaw direction is optimized using the parameters that influence each other among the generator sets, ensuring that the heading of the wind turbines can be more accurately controlled under the coordinated operation of multiple generator sets, thereby improving overall power generation efficiency and stability. Statistical analysis and calculation of the change rates of environmental influencing parameters and interfering wind factors enable the yaw control method to comprehensively consider the impact of different environmental factors, improve the ability of the power generation system to adapt to complex environments, maximize the use of renewable wind energy resources, reduce energy consumption, reduce environmental pollution, and achieve sustainable development of clean energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 A schematic flow chart of the steps of a yaw control method for a wind turbine generator set according to the present invention;

[0062] Figure 2 for Figure 1 Detailed implementation steps of step S3 in FIG.

[0063] Figure 3 for Figure 1 Detailed implementation steps of step S4 in FIG.

[0064] Figure 4 for Figure 1 Detailed implementation steps of step S5;

[0065] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0066] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the embodiments described are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of the present invention.

[0067] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0068] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0069] To achieve this, please refer to Figures 1 to 4 The present invention provides a yaw control method for a wind turbine generator set, comprising the following steps:

[0070] Step S1: performing wind power point marking processing on the location of the wind turbine generator set to generate wind power mark point data; performing wind power data collection and historical data preprocessing on the wind power mark point data to generate standard historical wind power data;

[0071] Step S2: using an autoregressive neural network algorithm to perform seasonal wind data forecasting on standard historical wind data to generate forecasted seasonal wind data;

[0072] Step S3: performing historical change statistics of environmental impact parameters on the collected wind marker point data to generate historical environmental impact change parameters; performing optimized processing on the predicted seasonal wind data using the historical environmental impact change parameters to generate optimized predicted seasonal wind data; designing a seasonal yaw control method for the generator set based on the optimized predicted seasonal wind data to generate seasonal initial yaw control data;

[0073] Step S4: Using the Internet of Things device to collect wind data from the collected wind marker point data in real time to generate real-time wind data; adjusting the yaw direction of the initial yaw control data according to the real-time wind data to generate yaw control data for the wind direction generator set;

[0074] Step S5: Calculate the mutual influence parameters of the generator sets according to the yaw control data to generate the mutual influence parameters of the generator sets; and optimize and adjust the yaw direction of the yaw control data according to the mutual influence parameters of the generator sets to generate optimized yaw control data.

[0075] By collecting wind marker data, the present invention can obtain wind information for different time periods and locations. This is the basis for yaw control, as knowing the current wind direction and strength can help adjust the orientation of the generator set to maximize wind energy capture and power generation efficiency. By performing historical data preprocessing on the collected wind marker data, abnormal data and noise can be eliminated, resulting in more accurate and stable standard historical wind data. This can avoid interference from bad data in subsequent prediction and optimization processes, improving the reliability and accuracy of subsequent steps. An autoregressive neural network algorithm is used to predict standard historical wind data. This algorithm can learn the seasonal patterns and trends of historical wind data, thereby predicting seasonal wind conditions for a period of time in the future. Such predictions help prepare in advance, allowing the generator set to make yaw adjustments in advance to cope with future wind changes, thereby stabilizing power generation and reducing energy waste. By statistically analyzing the historical changes in environmental influencing parameters of the collected wind marker data, it is possible to understand the impact of environmental factors on wind power, including seasonal factors (such as temperature and air pressure) and other factors that may affect wind power. By statistically analyzing these historical change parameters, it is possible to better understand the patterns of environmental influence on wind power. Optimizing seasonal wind data based on historical environmental influence parameters allows for more accurate predictions of future seasonal wind conditions, making yaw control more precise and adaptable to environmental changes. Real-time wind data collection from wind markers using IoT devices provides real-time insights into current wind conditions. This provides real-time data support for yaw control, enabling turbines to make timely yaw adjustments based on the latest wind conditions, thereby improving wind energy capture. Initial yaw control data is adjusted based on real-time wind data to promptly respond to changing wind conditions, maintaining optimal wind energy capture and enhancing power generation efficiency. Interaction parameters between turbines are calculated using yaw control data. Interactions between turbines can cause changes in wind direction, so understanding these influencing parameters helps optimize yaw control strategies, reducing interference and improving overall wind farm efficiency. Further optimization of yaw control data based on interaction parameters can optimize wind turbine layout and yaw strategies, further improving overall power generation efficiency and reducing system losses. Therefore, the yaw control method of the wind turbine of the present invention performs preliminary control on the yaw through seasonal wind direction, and then adjusts it according to real-time wind data, saving a lot of time and resources for yaw control adjustment, and also considers the mutual wind influence between wind turbines to adjust the yaw control of the wind turbine, making the result more accurate.

[0076] In the embodiment of the present invention, reference Figure 1The above is a schematic flow chart of the steps of a yaw control method for a wind turbine generator set according to the present invention. In this embodiment, the yaw control method for a wind turbine generator set includes the following steps:

[0077] Step S1: performing wind power point marking processing on the location of the wind turbine generator set to generate wind power mark point data; performing wind power data collection and historical data preprocessing on the wind power mark point data to generate standard historical wind power data;

[0078] In this embodiment of the present invention, multiple wind measurement devices are installed at a wind farm located along a coastline, and the locations of the wind turbines are precisely marked. Each measurement device collects and records wind data over a period of time, including information such as wind speed, wind direction, and air density. This collected wind marker data is transmitted to a data center. After data cleaning, filtering, historical data collection, and normalization, standardized historical wind data is generated, covering wind conditions across multiple seasons and different environmental conditions.

[0079] Step S2: using an autoregressive neural network algorithm to perform seasonal wind data forecasting on standard historical wind data to generate forecasted seasonal wind data;

[0080] In an embodiment of the present invention, an autoregressive neural network algorithm is used to analyze and train standard historical wind data. The algorithm establishes a seasonal wind data prediction model based on the seasonal characteristics and periodic laws of the historical data. The model can predict the seasonal wind conditions for the same date and time period in the future based on the historical wind data of a specific date and time period. Through model prediction, predicted seasonal wind data for the next few days or weeks is generated.

[0081] Step S3: performing historical change statistics of environmental impact parameters on the collected wind marker point data to generate historical environmental impact change parameters; performing optimized processing on the predicted seasonal wind data using the historical environmental impact change parameters to generate optimized predicted seasonal wind data; designing a seasonal yaw control method for the generator set based on the optimized predicted seasonal wind data to generate seasonal initial yaw control data;

[0082] In an embodiment of the present invention, historical wind data and collected environmental impact parameters are used to perform statistical analysis on different environmental impact factors, including temperature, air pressure, altitude, etc. The seasonal wind data is optimized by combining the historical environmental impact change parameters with the predicted seasonal wind data to obtain optimized predicted seasonal wind data. The optimized predicted seasonal wind data is more accurately predicted seasonal wind data. The yaw control of the wind direction generator is adjusted according to the data to obtain a preliminary fixed yaw direction. When the position is subsequently precisely adjusted, since the wind speed and direction of the same season have a similar relationship, only a slight movement is required. Based on the optimized predicted seasonal wind data, a seasonal yaw control method for the generator set is designed. According to seasonal changes, environmental conditions and other factors, seasonal initial yaw control data is generated for use in wind direction adjustment.

[0083] Step S4: Using the Internet of Things device to collect wind data from the collected wind marker point data in real time to generate real-time wind data; adjusting the yaw direction of the initial yaw control data according to the real-time wind data to generate yaw control data for the wind direction generator set;

[0084] In this embodiment of the present invention, a wind measurement device transmits wind data in real time to a data center via an IoT device, generating real-time wind data. This real-time data is compared and calculated with initial yaw control data. Based on wind speed variations and seasonal yaw control methods, the yaw position of the generator set is adjusted to generate the latest yaw control data for the wind-direction generator set. This real-time adjustment enables the generator set to more accurately track wind direction, maximize wind energy capture, and improve power generation efficiency.

[0085] Step S5: Calculate the mutual influence parameters of the generator sets according to the yaw control data to generate the mutual influence parameters of the generator sets; and optimize and adjust the yaw direction of the yaw control data according to the mutual influence parameters of the generator sets to generate optimized yaw control data.

[0086] In an embodiment of the present invention, mutual influence parameters are calculated based on the yaw control data between wind turbine generator sets, taking into account factors such as the spatial relationship between the wind turbine generator sets and the wind direction, so as to evaluate the degree of interaction between the generator sets. Based on the calculated mutual influence parameters of the generator sets, the yaw control data is optimized and adjusted. With the optimized yaw control data, the wind turbine generator sets can operate more intelligently in a coordinated manner, avoid mutual interference, and improve the overall power generation efficiency.

[0087] Preferably, step S1 includes the following steps:

[0088] Step S11: performing wind power point marking processing on the location of the wind turbine generator set to generate wind power mark point data;

[0089] Step S12: using the Internet of Things device to collect wind data from the wind marker point data to generate initial wind data;

[0090] Step S13: performing data cleaning on the initial wind power data to generate cleaned wind power data;

[0091] Step S14: performing historical wind data statistics on the cleaning wind data to generate historical wind data;

[0092] Step S15: Standardize the historical wind data using the minimum-maximum standardization method to generate standard historical wind data.

[0093] By collecting wind marker data, the present invention can obtain wind information at different locations. This is crucial for yaw control of wind turbines, as only by understanding the wind direction and wind speed at the current location can corresponding yaw adjustments be made, ensuring that the turbine is always facing the wind direction, thereby maximizing wind energy capture. By collecting wind marker data, the location of the wind turbine can be accurately located, providing accurate location information for subsequent wind data collection and control. Using IoT devices for data collection can automate the data acquisition process, reduce manual intervention and errors, and thus improve the efficiency and accuracy of data collection. During the wind data collection process, there may be some outliers or noise data, such as sensor failures and environmental interference. Data cleaning can effectively remove these outliers, ensuring that the wind data used in subsequent steps is accurate and reliable. By performing historical data statistics on the cleaned wind data, wind conditions at different time periods can be obtained. This allows for analysis of wind seasonality and periodicity, helping to predict future wind trends and providing a basis for yaw control. The historical wind data are standardized using the minimum-maximum normalization method, scaling the data to a uniform range. The standardized data have zero mean and unit variance, which helps eliminate the dimensional differences between different wind data and makes the data easier to process and analyze.

[0094] In an embodiment of the present invention, multiple wind measurement devices are installed in a wind farm. These devices are located around the wind turbine generators and are precisely marked. Each measurement device collects data such as wind speed and direction in real time and records the location information of the data collection using GPS. This data forms collected wind marker point data, which contains information such as wind speed, wind direction, and a timestamp for each location. The collected wind marker point data is transmitted to a data center via an IoT device. After receiving this data, the data center processes and merges the wind data for each location to obtain initial wind data, which contains information such as wind speed and wind direction for analyzing and predicting wind conditions. The initial wind data is then cleaned, including operations such as removing outliers, filling missing values, and filtering out noise. The cleaned wind data becomes more accurate and reliable, free from interference or erroneous data. The cleaned wind data is used to perform statistical analysis of wind conditions over historical time periods. By calculating data such as the average wind speed, maximum wind speed, and minimum wind speed for each time period, historical wind data is generated. This data helps understand the seasonal changes and cyclical patterns of wind conditions. The historical wind data is processed using the minimum-maximum normalization method, converting data such as wind speed and direction into standardized values ​​between 0 and 1. This is done to eliminate the impact of different data scales so that historical wind data can be compared and analyzed uniformly. The standardized data is conducive to establishing prediction models and optimization algorithms, achieving more accurate seasonal wind data prediction and generator set yaw control.

[0095] Preferably, step S2 includes the following steps:

[0096] Step S21: dividing the standard historical wind data into seasonal wind data to generate seasonal wind data;

[0097] Step S22: using an autoregressive neural network algorithm to perform seasonal wind data prediction on the seasonal wind data to generate predicted seasonal wind data.

[0098] By seasonally segmenting standard historical wind data, the present invention can extract wind data for different seasons (e.g., spring, summer, autumn, and winter). This helps analyze and understand the seasonal patterns of wind power, namely, the changing trends and differences in wind power between seasons. An autoregressive neural network algorithm is used to predict seasonal wind data. This algorithm combines neural networks with time series prediction technology to capture the time dependence and seasonal patterns in wind data. Using this prediction model, wind power can be predicted for different future seasons, providing more accurate and reliable data support for yaw control.

[0099] In an embodiment of the present invention, the standard historical wind data is divided according to seasonality, for example, according to the four seasons of spring, summer, autumn and winter, or according to months. For each season or month, the wind data within the corresponding time period is extracted to obtain a seasonal wind data subset. An autoregressive neural network (AR-NN) algorithm is used to train and predict each seasonal wind data subset. AR-NN is a prediction algorithm that combines an autoregressive model and a neural network, and can effectively capture the seasonal and cyclical changes in time series data. In the training phase, the seasonal wind data subset is used as the input sequence, and the wind conditions for a period of time in the season are used as the target sequence. The neural network model is used for training, and the model parameters are optimized so that it can better fit the seasonal changes in the historical wind data. In the prediction phase, the trained AR-NN model is applied to the new seasonal wind data subset, and the wind data of the current time period is input to obtain the predicted seasonal wind data for a period of time in the future. Through the prediction of the AR-NN algorithm, the predicted seasonal wind data for each season or month can be obtained.

[0100] Preferably, step S22 includes the following steps:

[0101] Step S221: using an autoregressive neural network algorithm to establish a mapping relationship for seasonal wind data prediction to generate an initial seasonal wind data prediction model;

[0102] Step S222: dividing the seasonal wind data into historical time series data to generate a seasonal wind training set and a seasonal wind test set;

[0103] Step S223: using the seasonal wind training set to perform model training on the initial seasonal wind data prediction model to generate a seasonal wind data prediction model;

[0104] Step S224: transmitting the seasonal wind test set to the seasonal wind data prediction model to perform seasonal wind data prediction to generate predicted seasonal wind data.

[0105] The present invention uses an autoregressive neural network algorithm to establish a mapping relationship for seasonal wind data prediction. The autoregressive neural network algorithm can capture the time dependency and seasonal patterns in wind data, enabling the prediction model to better predict wind changes in future seasons. The seasonal wind data is divided into a historical time series, and the data set is divided into a seasonal wind training set and a seasonal wind test set. The purpose of this is to provide independent data sets for model training and evaluation, avoid model overfitting and deviation in evaluation results, and help the model learn and predict time dependencies. By using the seasonal wind training set, the initial seasonal wind data prediction model is trained. During the model training process, the model learns the characteristics and patterns of seasonal wind data, optimizes model parameters, and enables the model to more accurately predict wind changes in future seasons. By using the seasonal wind test set to evaluate the prediction model, the prediction performance and accuracy of the model can be understood. The evaluation results can guide whether further model optimization is needed to ensure the reliability and accuracy of the model. In this way, the model can obtain predicted seasonal wind data. The predicted seasonal wind data can help predict wind conditions in future seasons and provide more accurate and reliable data support for yaw control.

[0106] In an embodiment of the present invention, historical standard wind data is used to establish a mapping relationship for seasonal wind data prediction. This involves selecting an appropriate autoregressive neural network model structure and parameter settings, as well as determining input features and output targets. Wind data from several past time steps is used as the input sequence, and wind data for a period of time in the future is used as the output target. Seasonal wind data is partitioned and processed according to the historical time series, into a seasonal wind training set and a seasonal wind test set. The training set is used for model training and parameter optimization, while the test set is used to evaluate the model's prediction performance. The seasonal wind training set is used to train an initial seasonal wind data prediction model. During the training process, the model continuously adjusts its weights and biases through multiple iterations to minimize the error between the predicted and actual values. The seasonal wind test set is then transferred to the trained seasonal wind data prediction model for seasonal wind data prediction. Based on the input test set data, the model outputs predicted seasonal wind data for a period of time in the future. This predicted data is used in subsequent optimization processes to assist in yaw control decisions for wind turbines.

[0107] Preferably, step S3 includes the following steps:

[0108] Step S31: Using the Internet of Things device to perform environmental impact parameter tagging on the collected wind power mark point data to generate environmental impact parameters;

[0109] Step S32: using a statistical analysis method to perform statistics on the change data of historical environmental impact parameters on the interference wind factor data to generate historical environmental impact change parameters;

[0110] Step S33: performing optimization calculation on the predicted seasonal wind data using the seasonal wind data optimization algorithm and the historical environmental impact change parameters to generate optimized predicted seasonal wind data;

[0111] Step S34: Designing a seasonal yaw control method for the generator set based on the optimized predicted seasonal wind data to generate seasonal initial yaw control data.

[0112] The present invention uses the Internet of Things device to mark the collected wind mark point data with environmental impact parameters, and can obtain parameter information related to the environment, which may have an impact on wind power, such as vegetation, humidity, mountain peaks and other problems. Through the marking process, the environmental impact parameters are associated with the wind data, providing a data basis for subsequent wind optimization. By performing statistical analysis on the interfering wind factor data, the degree of influence of different environmental factors on wind power and their historical change trends can be understood, which helps to understand the interference factors in the wind data, thereby more accurately predicting future wind power and optimizing the operation of the wind power generation system, taking into account the impact of different environmental factors on wind power, and improving the accuracy and stability of the prediction. The seasonal wind data optimization algorithm can take into account the impact of different environmental factors on wind power, and adjust the predicted data according to the changes in environmental factors, so that the prediction results are more accurate and reliable, which helps to improve the accuracy of the prediction and reduce the prediction error. Such an optimization process can better cope with wind power changes under different environmental conditions. The seasonal yaw control method of the generator set is designed based on the optimized predicted seasonal wind data, and the seasonal initial yaw control data is generated. This data will guide the yaw adjustment of the generator set, enabling the generator set to make corresponding initial control actions based on the predicted seasonal wind data, thereby improving the degree of wind energy capture. In the later stage, only fine-tuning is required for precise yaw control, saving a lot of time and resources.

[0113] As an example of the present invention, refer to Figure 2 As shown, Figure 1 Detailed implementation steps of step S3 are shown in the flowchart. In this example, step S3 includes:

[0114] Step S31: Using the Internet of Things device to perform environmental impact parameter tagging on the collected wind power mark point data to generate environmental impact parameters;

[0115] In an embodiment of the present invention, a plurality of environmental impact parameter sensors are installed in the wind farm. These sensors can sense the influencing factors in the environment, such as temperature, humidity, altitude, etc. These environmental impact parameters will be collected in real time near the wind measurement point and marked corresponding to the wind data. The environmental impact parameters for each collected wind marking point are obtained for subsequent analysis and optimization prediction.

[0116] Step S32: using a statistical analysis method to perform statistics on the change data of historical environmental impact parameters on the interference wind factor data to generate historical environmental impact change parameters;

[0117] In an embodiment of the present invention, historical environmental impact parameter data and wind data are used, and statistical analysis methods such as regression analysis and time series analysis are adopted to study the relationship between environmental impact parameters and wind power, and obtain the degree of influence of interfering wind factors on wind power. By statistically analyzing the change data of historical environmental impact parameters, the change trend and periodic law of the environmental impact parameters can be obtained.

[0118] Step S33: performing optimization calculation on the predicted seasonal wind data using the seasonal wind data optimization algorithm and the historical environmental impact change parameters to generate optimized predicted seasonal wind data;

[0119] In an embodiment of the present invention, the seasonal wind data optimization algorithm and the historical environmental impact change parameters are combined to optimize the calculation of the predicted seasonal wind data. The algorithm takes into account the impact of the environmental impact parameters on wind power and the change trend of the historical environmental impact parameters, and then adjusts the predicted seasonal wind data. For example, through weighted averaging or compensation calculation, the seasonal wind data is optimized so that the predicted data is more in line with the actual situation.

[0120] Step S34: Designing a seasonal yaw control method for the generator set based on the optimized predicted seasonal wind data to generate seasonal initial yaw control data.

[0121] In this embodiment of the present invention, a seasonal yaw control method is designed based on optimized and predicted seasonal wind data and the characteristics of the wind turbine. For example, when the predicted seasonal wind data indicates strong winds in a certain season, the yaw direction of the wind turbine can be adjusted to better capture wind energy and improve power generation efficiency. Once the seasonal initial yaw control data is generated, it serves as the basis for the control system, which is further optimized and adjusted to ensure optimal yaw control of the wind turbine in different seasons.

[0122] Preferably, the seasonal wind data optimization algorithm in step S33 is as follows:

[0123]

[0124] Where P represents the optimized forecast seasonal wind data, N represents the coefficient of the forecast seasonal wind data, M represents the coefficient of the environmental impact parameter, and a j Expressed as wind disturbance data of the jth environmental impact parameter, b j It is expressed as the weight information of the wind interference data of the jth environmental impact parameter, p is expressed as the mean change rate of the wind interference data, v j It is expressed as the change rate of the jth environmental impact parameter, c is expressed as the seasonal change weight, k i It represents the wind data of the season where the i-th predicted seasonal wind data belongs, and τ represents the abnormal adjustment value of the optimized predicted seasonal wind data.

[0125] The present invention utilizes a seasonal wind data optimization algorithm, which fully considers the coefficient size N of the predicted seasonal wind data, the coefficient size M of the environmental impact parameter, the wind interference data a of the jth environmental impact parameter, and the wind speed of the wind speed data of the jth environmental impact parameter. j , the weight information b of the wind interference data of the jth environmental impact parameter j , the mean change rate p of wind interference data, the change rate v of the jth environmental impact parameter j , seasonal change weight c, wind data k of the season where the i-th predicted seasonal wind data is located i And the interaction between functions to form a functional relationship:

[0126] Right now, The size of the coefficient of the predicted seasonal wind data is expressed as the data volume of the predicted seasonal wind data. The larger the data volume, the more accurate the optimization effect. The size of the coefficient of the environmental impact parameter is expressed as the data volume of the environmental impact parameter. The larger the data volume, the more data basis is provided, and the better the optimization effect of the predicted seasonal wind data. The wind interference data of the jth environmental impact parameter represents the actual impact of the environmental impact parameter on the wind, and explains the degree of influence of the environmental impact parameter on the wind. The weight information of the wind interference data of the jth environmental impact parameter represents the weight importance of the environmental impact parameter on the wind, and explains the weight ratio of the environmental impact parameters. The mean change rate of the wind interference data represents the overall change trend of the wind interference data. The change rate of the jth environmental impact parameter represents the size of the change of the environmental impact parameter at different times, providing a basis for predicting subsequent change impact parameters, thereby optimizing the predicted seasonal wind data. The seasonal change weight represents the degree of influence of different seasons on the wind, and explains the degree of influence of seasonal factors on the wind. The wind data of the season in which the i-th predicted seasonal wind data is located represents the prediction result of the wind in the current season, and explains the wind strength in the current season. The seasonal variation weights in the optimization algorithm and the wind data for the season in which the i-th predicted seasonal wind data falls can effectively reflect the impact of seasonal variation on wind power. The characteristics of wind power variation vary from season to season, resulting in irregular changes in environmental influencing parameters. This algorithm can account for the impact of environmental variation, enabling more accurate predictions of wind power data for different seasons. The function relationship is adjusted and corrected using the anomaly adjustment value τ for the optimized predicted seasonal wind data, reducing the impact of errors caused by abnormal data or error terms. This allows for more accurate generation of the optimized predicted seasonal wind data P, improving the accuracy and reliability of the optimized calculation of the predicted seasonal wind data. Furthermore, the weight information and adjustment values ​​in this formula can be adjusted based on actual conditions and applied to different predicted seasonal wind data, enhancing the algorithm's flexibility and applicability.

[0127] Preferably, step S4 includes the following steps:

[0128] Step S41: using the Internet of Things device to collect wind mark point data in real time to generate real-time wind data;

[0129] Step S42: performing difference calculation between the real-time wind data and the optimized predicted seasonal wind data to generate difference wind data;

[0130] Step S43: adjusting the yaw azimuth of the initial yaw control data according to the differential wind data to generate yaw control data for the wind direction generator set.

[0131] The present invention utilizes Internet of Things devices to collect wind marker point data in real time and obtain real-time wind data of the current position. Real-time wind data is crucial for yaw control because the direction and intensity of the wind will change continuously. Timely data collection enables the generator set to make yaw adjustments in time to adapt to the changing wind conditions. The real-time wind data and the optimized predicted seasonal wind data are difference calculated to obtain differential wind data. The differential data reflects the deviation between the real-time wind and the predicted wind, helping the generator set to understand the difference between the current wind conditions and the prediction, and providing more accurate data support for yaw control. The initial yaw control data is adjusted for yaw azimuth according to the differential wind data to generate yaw control data for the wind direction generator set. Such yaw control data can enable the generator set to make yaw adjustments according to the real-time wind conditions to keep the generator set facing the direction of the wind, improve the degree of capturing wind energy, and achieve improved power generation efficiency and efficient use of energy.

[0132] As an example of the present invention, refer to Figure 3 As shown, Figure 1 Detailed implementation steps of step S4 are shown in the flowchart. In this example, step S4 includes:

[0133] Step S41: using the Internet of Things device to collect wind mark point data in real time to generate real-time wind data;

[0134] In an embodiment of the present invention, Internet of Things devices are installed in the wind farm. These devices can collect wind marker point data in real time, including information such as wind speed and wind direction at the current moment. When the wind turbine is running, the Internet of Things devices will continuously collect wind data and transmit it to the data center, obtaining real-time wind data for comparison and analysis with previously predicted seasonal wind data.

[0135] Step S42: performing difference calculation between the real-time wind data and the optimized predicted seasonal wind data to generate difference wind data;

[0136] In an embodiment of the present invention, the real-time wind data is compared with the seasonal wind data that was previously optimized and predicted, and the difference between them is calculated. The difference wind data reflects the deviation between the prediction and the actual, that is, the difference between the predicted data and the actual measured data. The difference may be positive or negative, indicating that the difference between the predicted wind and the actual wind in wind direction is too high or too low.

[0137] Step S43: adjusting the yaw azimuth of the initial yaw control data according to the differential wind data to generate yaw control data for the wind direction generator set.

[0138] In this embodiment of the present invention, the initial yaw control data is adjusted based on the differential wind data to optimize the yaw direction of the wind turbine. If the differential wind data indicates that the predicted value is too high, indicating that the predicted wind speed is stronger than the actual situation, the yaw direction can be slightly adjusted toward the direction of stronger wind speed to better utilize wind energy. If the differential wind data indicates that the predicted value is too low, the yaw direction can be slightly adjusted away from the direction of stronger wind speed to avoid the risks associated with excessive wind speed.

[0139] Preferably, step S5 includes the following steps:

[0140] Step S51: performing yaw control on the wind turbine generator set according to the yaw control data, and collecting and processing wind power data of the wind turbine generator set using sensors to generate wind power data of the generator set;

[0141] Step S52: Calculating the generator group mutual influence parameters based on the generator group wind power data using the generator group wind power mutual influence calculation formula to generate the generator group mutual influence parameters;

[0142] Step S53: performing yaw direction optimization adjustment processing on the yaw control data according to the mutual influence parameters of the generator sets to generate optimized yaw control data.

[0143] The present invention performs yaw control on a wind turbine generator set according to yaw control data, ensuring that the generator set always faces the direction of the wind to maximize wind energy capture and improve power generation efficiency. Sensors are used to collect and process wind data of the wind turbine generator set to obtain wind information at the location of the generator set. These real-time collected wind data are very important for yaw control and subsequent calculation of mutual influence parameters. By utilizing the wind data of the generator set and applying the wind mutual influence calculation formula of the generator set, the mutual influence parameters of the generator set can be calculated, which reflect the degree of mutual influence of wind energy utilization between different generator sets, help understand the interaction between different generator sets, and provide a basis for subsequent yaw direction optimization. According to the mutual influence parameters of the generator sets, the yaw direction is optimized and adjusted for the yaw control data. By optimizing the yaw control data, multiple generator sets can work better together and avoid mutual interference, thereby improving the overall efficiency and stability of the entire wind power generation system. The optimized yaw control data will guide the operation of the generator set in a complex wind field, maximize the use of wind energy, and achieve efficient power generation.

[0144] As an example of the present invention, refer to Figure 4 As shown, Figure 1 Detailed implementation steps of step S5 are shown in the flowchart. In this example, step S5 includes:

[0145] Step S51: performing yaw control on the wind turbine generator set according to the yaw control data, and collecting and processing wind power data of the wind turbine generator set using sensors to generate wind power data of the generator set;

[0146] In an embodiment of the present invention, the yaw control of the wind turbine is performed based on the yaw control data obtained by the previous optimization, and the direction of the turbine is adjusted so that it faces the wind direction, thereby capturing wind energy to the greatest extent. During the yaw control process, the wind data of the wind turbine is collected and monitored in real time using the equipped sensors. These sensors can measure parameters such as wind speed, wind direction, and wind power, and transmit the collected data to the control system. The collected wind data can be used to understand the wind conditions in real time, providing an important reference basis for yaw control.

[0147] Step S52: Calculating the generator group mutual influence parameters based on the generator group wind power data using the generator group wind power mutual influence calculation formula to generate the generator group mutual influence parameters;

[0148] In this embodiment of the present invention, the degree of mutual influence between generator sets is calculated based on wind data from the generator sets and a formula for calculating wind interaction. This formula takes into account factors such as the distance between generator sets, rotor diameter, and wind speed to produce a generator set mutual influence parameter, which reflects the degree of mutual interference between the generator sets. The calculated mutual influence parameter provides key information for the next step of yaw control optimization.

[0149] Step S53: performing yaw direction optimization adjustment processing on the yaw control data according to the mutual influence parameters of the generator sets to generate optimized yaw control data.

[0150] In an embodiment of the present invention, the yaw control data is optimized and adjusted according to the mutual influence parameters of the generator sets. By analyzing the degree of mutual influence between the generator sets, mutual interference between multiple wind turbines in adjacent positions can be avoided, thereby improving the overall power generation efficiency. The optimized yaw control data will more accurately guide the yaw control of the wind turbine set, enabling it to better adapt to the complex and changeable wind field environment.

[0151] Preferably, the calculation formula for the wind power interaction between the generator sets in step S52 is as follows:

[0152]

[0153] In the formula, K represents the mutual influence parameter of the generator set, n represents the coefficient size of the wind power data of the generator set, Expressed as The horizontal axis wind intensity of the claw generator wind data, Expressed as The vertical axis wind intensity of the wind data of each generator set is Expressed as The wind power of each generator set’s wind data, Expressed as The wind power data of the wind turbines are separated by a distance, g is the rotor diameter of the wind turbine, k is the initial speed of the wind turbine, d is the air density, θ is the wind direction angle of the wind turbine, and δ is the abnormal adjustment value of the parameters that affect each other among the generators.

[0154] The present invention uses a generator set wind mutual influence calculation formula, which fully considers the coefficient size n of the generator set wind data, the first The horizontal axis wind intensity of the wind data of each generator set No. The vertical axis wind intensity of the wind data of each generator set No. The wind power of each generator set No. The distance between wind power data of each generator set The rotor diameter g of the wind turbine, the initial speed k of the wind turbine, the air density d, the wind direction angle θ of the wind turbine, and the interaction relationship between the functions form a functional relationship:

[0155] Right now. The coefficient of the wind power data of the generator set indicates the number of generator sets involved in the calculation. The more wind power data of the generator sets considered, the more comprehensive the calculation results will be. Expressed as The horizontal axis wind intensity of the wind data of the first generator set is The vertical axis wind intensity of the wind data of each generator set indicates the horizontal and vertical wind speeds of the generator set. This parameter is used to calculate the degree of mutual influence of wind forces. The wind power of each generator set indicates the wind energy conversion efficiency of the generator set. The wind power plays an important role in the mutual influence of the generator sets, and the wind strength is comprehensively considered. The distance between the wind power data of each generator set represents the spatial distance between the generator set and other generator sets, which is used to evaluate the positional relationship between the generator sets; the rotor diameter of the wind turbine set represents the rotor diameter of the generator set, which affects the ability to capture wind power; the initial speed of the wind turbine set represents the starting operating speed of the wind turbine set, which affects the response speed of the generator set; the air density affects the strength of the wind and the efficiency of wind energy conversion; the wind direction angle of the wind turbine set represents the direction facing the wind rotor, which is used to evaluate the orientation of the generator set. The wind power size and the wind power data of the claw generator set The distance between wind data from individual generators is a key factor in measuring the mutual influence between generators. Wind power reflects wind strength, while distance represents the spatial relationship between generators. By considering these two factors, the formula can more accurately assess the degree of mutual influence between different generators, facilitating the optimization of yaw control data. Assessing the degree of interaction between different generators provides a crucial basis for optimizing yaw control data and intelligently adjusting wind power systems. The functional relationship is adjusted using the abnormal adjustment value δ of the generator interaction parameter to reduce the impact of abnormal data or error terms, resulting in a more accurate generation of the generator interaction parameter K. This improves the accuracy and reliability of the calculation of the generator interaction parameter based on wind data. Furthermore, the weighting information and adjustment values ​​in this formula can be adjusted based on actual conditions and applied to wind data from different generators, enhancing the algorithm's flexibility and applicability.

[0156] In this specification, a yaw control system of a wind turbine generator set is provided, comprising:

[0157] The historical wind data collection module is used to collect wind power point markings at the location of the wind turbine generator set to generate collected wind power mark point data; collect wind power data and perform historical data preprocessing on the collected wind power mark point data to generate standard historical wind data;

[0158] Seasonal wind data prediction module, which uses autoregressive neural network algorithm to predict seasonal wind data based on standard historical wind data and generate predicted seasonal wind data;

[0159] The initial yaw control design module is used to perform historical changes in environmental impact parameters of collected wind marker point data to generate historical environmental impact change parameters; use historical environmental impact change parameters to optimize the predicted seasonal wind data to generate optimized predicted seasonal wind data; design the seasonal yaw control method of the generator set based on the optimized predicted seasonal wind data to generate seasonal initial yaw control data;

[0160] The yaw control adjustment module uses IoT devices to collect wind data from wind marker points in real time to generate real-time wind data; it adjusts the yaw direction of the initial yaw control data based on the real-time wind data to generate yaw control data for the wind direction generator set;

[0161] The yaw control optimization module is used to calculate the mutual influence parameters of the generator sets based on the yaw control data and generate the mutual influence parameters of the generator sets; and to optimize and adjust the yaw direction of the yaw control data based on the mutual influence parameters of the generator sets and generate optimized yaw control data.

[0162] The beneficial effect of the present application is that the present invention utilizes Internet of Things devices to collect and process wind data and environmental impact parameters in real time, obtains accurate and timely real-time wind data and environmental information, and provides a reliable data basis for the yaw control of wind turbines. By using an autoregressive neural network algorithm to predict historical wind data, and optimizing the prediction results based on historical environmental impact change parameters, accurate predictions of future seasonal wind power are achieved. Such optimized predictions contribute to the intelligent adjustment of generator sets and improve power generation efficiency. Based on the difference between real-time wind data and optimized predicted seasonal wind data, real-time yaw control of wind turbines is achieved. First, the yaw control data is preliminarily set using predicted seasonal wind direction data, and then adjusted based on real-time wind data, thereby reducing a large amount of lost resources and time. At the same time, the yaw direction is optimized using the parameters that influence each other among the generator sets, ensuring that the heading of the wind turbines can be more accurately controlled under the coordinated operation of multiple generator sets, thereby improving overall power generation efficiency and stability. Statistical analysis and calculation of the change rates of environmental influencing parameters and interfering wind factors enable the yaw control method to comprehensively consider the impact of different environmental factors, improve the ability of the power generation system to adapt to complex environments, maximize the use of renewable wind energy resources, reduce energy consumption, reduce environmental pollution, and achieve sustainable development of clean energy.

[0163] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced therein.

[0164] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A yaw control method for a wind turbine generator set, characterized in that: The following steps are involved: Step S1: performing wind power point marking processing on the location of the wind turbine generator set to generate wind power mark point data; Perform wind data collection and historical data preprocessing on the collected wind mark point data to generate standard historical wind data; Step S2: using an autoregressive neural network algorithm to perform seasonal wind data forecasting on standard historical wind data to generate forecasted seasonal wind data; Step S3: performing historical change statistics of environmental impact parameters on the collected wind marker point data to generate historical environmental impact change parameters; performing seasonal wind data optimization processing on the predicted seasonal wind data using the historical environmental impact change parameters to generate optimized predicted seasonal wind data; The seasonal yaw control mode of the generator set is designed based on the optimized predicted seasonal wind data to generate seasonal initial yaw control data. Step S3 includes the following steps: Step S31: Using the Internet of Things device to perform environmental impact parameter tagging on the collected wind power mark point data to generate environmental impact parameters; Step S32: using a statistical analysis method to perform statistics on the change data of historical environmental impact parameters on the interference wind factor data to generate historical environmental impact change parameters; Step S33: Optimize the seasonal wind data using the seasonal wind data optimization algorithm and the historical environmental impact change parameters to generate optimized seasonal wind data. The seasonal wind data optimization algorithm in step S33 is as follows: ； Where, Represents the optimization of forecasting seasonal wind data, It is expressed as the coefficient size of the predicted seasonal wind data, Expressed as the coefficient size of the environmental impact parameter, Expressed as Wind disturbance data of environmental impact parameters, Expressed as The weight information of wind disturbance data of environmental impact parameters, Expressed as the mean change rate of wind disturbance data, Expressed as The rate of change of environmental impact parameters, Expressed as seasonal variation weight, Expressed as The wind data for the season in which the seasonal wind data is predicted, represents the anomaly adjustment value for optimizing the forecast of seasonal wind data; Step S34: Designing a seasonal yaw control method for the generator set based on the optimized predicted seasonal wind data to generate seasonal initial yaw control data; Step S4: Using the Internet of Things device to collect wind data from the collected wind marker point data in real time to generate real-time wind data; adjusting the yaw direction of the initial yaw control data according to the real-time wind data to generate yaw control data for the wind direction generator set; Step S5: Calculating the mutual influence parameters of the generator sets according to the yaw control data to generate the mutual influence parameters of the generator sets; performing yaw direction optimization adjustment processing on the yaw control data according to the mutual influence parameters of the generator sets to generate optimized yaw control data. Step S5 includes the following steps: Step S51: performing yaw control on the wind turbine generator set according to the yaw control data, and collecting and processing wind power data of the wind turbine generator set using sensors to generate wind power data of the generator set; Step S52: Calculate the generator group mutual influence parameters based on the generator group wind data using the generator group wind force mutual influence calculation formula to generate the generator group mutual influence parameters. The generator group wind force mutual influence calculation formula in step S52 is as follows: ； Where, Expressed as the mutual influence parameter of the generator sets, It is expressed as the coefficient of wind power data of the generator set. Expressed as The horizontal axis wind intensity of the wind data of each generator set is Expressed as The vertical axis wind intensity of the wind data of each generator set is Expressed as The wind power of each generator set’s wind data, Expressed as The distance between the wind power data of each generator set, It is expressed as the rotor diameter of the wind turbine generator set, Expressed as the initial speed of the wind turbine, Expressed as air density, It is represented by the wind direction angle of the wind turbine. It is expressed as abnormal adjustment value of the mutual influence parameters of the generator sets; Step S53: performing yaw direction optimization adjustment processing on the yaw control data according to the mutual influence parameters of the generator sets to generate optimized yaw control data.

2. The yaw control method of a wind turbine generator set according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: performing wind power point marking processing on the location of the wind turbine generator set to generate wind power mark point data; Step S12: using the Internet of Things device to collect wind data from the wind marker point data to generate initial wind data; Step S13: performing data cleaning on the initial wind power data to generate cleaned wind power data; Step S14: performing historical wind data statistics on the cleaning wind data to generate historical wind data; Step S15: Standardize the historical wind data using the minimum-maximum standardization method to generate standard historical wind data.

3. The yaw control method of a wind turbine generator set according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: dividing the standard historical wind data into seasonal wind data to generate seasonal wind data; Step S22: using an autoregressive neural network algorithm to perform seasonal wind data prediction on the seasonal wind data to generate predicted seasonal wind data.

4. The yaw control method of a wind turbine generator set according to claim 3, characterized in that: Step S22 includes the following steps: Step S221: using an autoregressive neural network algorithm to establish a mapping relationship for seasonal wind data prediction to generate an initial seasonal wind data prediction model; Step S222: dividing the seasonal wind data into historical time series data to generate a seasonal wind training set and a seasonal wind test set; Step S223: using the seasonal wind training set to perform model training on the initial seasonal wind data prediction model to generate a seasonal wind data prediction model; Step S224: transmitting the seasonal wind test set to the seasonal wind data prediction model to perform seasonal wind data prediction to generate predicted seasonal wind data.

5. The yaw control method of a wind turbine generator set according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: using the Internet of Things device to collect wind mark point data in real time to generate real-time wind data; Step S42: performing difference calculation between the real-time wind data and the optimized predicted seasonal wind data to generate difference wind data; Step S43: adjusting the yaw azimuth of the initial yaw control data according to the differential wind data to generate yaw control data for the wind direction generator set.

6. A yaw control system for a wind turbine generator set, characterized in that: For executing the yaw control method of a wind turbine generator set according to claim 1, the yaw control system of the wind turbine generator set comprises: The historical wind data collection module is used to collect wind power point markings at the location of the wind turbine generator set to generate collected wind power mark point data; collect wind power data and perform historical data preprocessing on the collected wind power mark point data to generate standard historical wind data; Seasonal wind data prediction module, which uses autoregressive neural network algorithm to predict seasonal wind data based on standard historical wind data and generate predicted seasonal wind data; The initial yaw control design module is used to perform historical change statistics on environmental impact parameters of collected wind marker point data to generate historical environmental impact change parameters; use historical environmental impact change parameters to optimize the predicted seasonal wind data to generate optimized predicted seasonal wind data; design the seasonal yaw control method of the generator set based on the optimized predicted seasonal wind data to generate seasonal initial yaw control data; The yaw control adjustment module uses IoT devices to collect wind data from wind marker points in real time to generate real-time wind data; it adjusts the yaw direction of the initial yaw control data based on the real-time wind data to generate yaw control data for the wind direction generator set; The yaw control optimization module is used to calculate the mutual influence parameters of the generator sets based on the yaw control data and generate the mutual influence parameters of the generator sets; and to optimize and adjust the yaw direction of the yaw control data based on the mutual influence parameters of the generator sets and generate optimized yaw control data.

Images