Wind turbine yaw control method and device based on neural network

Abstract

The application provides a neural network-based yaw control method and device for a wind turbine, and relates to the technical field of wind turbine yaw control. The method comprises the following steps: obtaining state parameters of a target wind turbine; taking the state parameters of the target wind turbine as input, outputting a first predicted yaw angle of the wind turbine through a first yaw angle prediction model, and outputting a second predicted yaw angle of the wind turbine through a second yaw angle prediction model, wherein the first yaw angle prediction model and the second yaw angle prediction model are obtained by training different deep learning algorithms through historical state parameters and corresponding yaw angles of different wind turbines; determining a predicted yaw angle difference between the first predicted yaw angle and the second predicted yaw angle, and determining a target predicted yaw angle of the target wind turbine based on the first predicted yaw angle and the second predicted yaw angle according to the predicted yaw angle difference. The application improves the accuracy of yaw control for the wind turbine, and improves the efficiency and reliability of wind power generation.

Description

TECHNICAL FIELD

[0001] The present application relates to the yaw control technology field of wind turbines, in particular to a yaw control method for a wind turbine based on a neural network, a yaw control device for a wind turbine based on a neural network, a computer readable storage medium and a terminal device. BACKGROUND

[0002] At present, the yaw control of wind turbines usually adopts the following method: a wind direction sensor is installed on the wind turbine to monitor the changes in wind direction in real time, when the wind direction changes, the wind direction sensor transmits a signal to the control system, and the control system determines whether yaw operation is needed according to the included angle between the wind direction and the shaft line of the nacelle, if the included angle exceeds a certain threshold, the control system will start the yaw motor to rotate the nacelle until the included angle between the wind direction and the shaft line of the nacelle is within the allowable range. The existing system can obtain the basic operating parameters of the yaw control system, such as pressure, flow and temperature, etc., for basic state monitoring. However, due to the influence of sensor accuracy, response time, reliability, installation location, etc., the traditional yaw control method is easily affected by environmental temperature, humidity, air pressure, electromagnetic interference, lightning, vibration, surrounding buildings, trees and other obstacles, thereby reducing the measurement accuracy and response speed. At present, through deep learning and analysis of massive data, real-time monitoring and control of system state can also be achieved, for example, through learning of a large amount of historical data and real-time running data, the neural network can accurately perceive the wind direction change, the wind turbine running state and the working condition of the yaw control, so as to adjust the yaw angle in real time according to the external environment and the wind turbine running state, to adapt to different wind directions and wind conditions, and improve the power generation efficiency of the wind turbine. However, the existing yaw control based on deep learning usually adopts a single model, and the prediction accuracy and stability of the yaw angle are low. SUMMARY

[0003] The purpose of the embodiments of the present application is to provide a yaw control method for a wind turbine based on a neural network, a yaw control device for a wind turbine based on a neural network, a computer readable storage medium and a terminal device to solve the above problems.

[0004] In order to achieve the above purpose, the first aspect of the present application provides a yaw control method for a wind turbine based on a neural network, comprising:

[0005] obtaining state parameters of a target wind turbine, the state parameters at least including wind speed, wind direction, temperature, humidity, air pressure of the environment where the target wind turbine is located, and at least one of the rotational speed, torque and output power of the target wind turbine;

[0006] The state parameters of the target wind turbine are inputted, a first predicted yaw angle of the wind turbine is outputted through a first yaw angle prediction model, and the state parameters of the target wind turbine are inputted, a second predicted yaw angle of the wind turbine is outputted through a second yaw angle prediction model, wherein the first yaw angle prediction model and the second yaw angle prediction model are obtained by training different deep learning algorithms through historical state parameters and corresponding yaw angles of different wind turbines respectively;

[0007] A predicted yaw angle difference between the first predicted yaw angle and the second predicted yaw angle is determined, and a target predicted yaw angle of the target wind turbine is determined based on the first predicted yaw angle and the second predicted yaw angle according to the predicted yaw angle difference;

[0008] The target wind turbine is controlled to operate at the target predicted yaw angle.

[0009] Optionally, the first yaw angle prediction model is obtained by training a first deep learning algorithm through historical state parameters and corresponding yaw angles of different wind turbines, and the training process of the first deep learning algorithm comprises:

[0010] At least one of wind speed, wind direction, temperature, humidity, air pressure of an environment where the wind turbine is located, and rotating speed, torque and output power of the wind turbine is determined as first historical state parameters;

[0011] First historical state parameters and corresponding actual yaw angles of different wind turbines at multiple sampling time points within a preset historical period are obtained;

[0012] The first historical state parameters at the multiple sampling time points within the preset historical period are inputted, corresponding predicted yaw angles are outputted through the first deep learning algorithm, errors between the predicted yaw angles outputted by the first deep learning algorithm and the corresponding actual yaw angles are calculated, algorithm parameters of the first deep learning algorithm are adjusted according to the comparison results, until the error between the predicted yaw angles outputted by the first deep learning algorithm and the actual yaw angles is lower than a preset error threshold or reaches a maximum iteration number, and the first yaw angle prediction model is obtained.

[0013] Optionally, the second yaw angle prediction model is obtained by training a second deep learning algorithm through historical state parameters and corresponding yaw angles of different wind turbines, and the training process of the second deep learning algorithm comprises:

[0014] At least one of wind speed, wind direction, temperature, humidity, air pressure of an environment where the wind turbine is located, and rotating speed, torque and output power of the wind turbine is determined as second historical state parameters, wherein at least one of the second historical state parameters is different from the first historical state parameters;

[0015] obtaining second historical state parameters of different wind turbines at multiple sampling time points in a preset historical period and corresponding actual yaw angles;

[0016] taking the second historical state parameters at the multiple sampling time points in the preset historical period as inputs, outputting corresponding predicted yaw angles through the second deep learning algorithm, calculating errors between the predicted yaw angles output by the second deep learning algorithm and the corresponding actual yaw angles, and adjusting algorithm parameters of the second deep learning algorithm according to the comparison results until the errors between the predicted yaw angles output by the second deep learning algorithm and the actual yaw angles are lower than a preset error threshold or a maximum iteration number is reached, to obtain the second yaw angle prediction model.

[0017] Optionally, at least one of wind speed, wind direction, temperature, humidity, air pressure of an environment where the wind turbine is located, and rotation speed, torque and output power of the wind turbine is determined as the first historical state parameter, comprising:

[0018] determining weight coefficients of the wind speed, the wind direction, the temperature, the humidity, the air pressure, the rotation speed, the torque and the output power, determining state parameters with weight coefficients belonging to a first weight coefficient interval as first state parameters, and determining state parameters with weight coefficients belonging to a second weight coefficient interval as second state parameters, wherein a minimum value of the first weight coefficient interval is not less than a maximum value of the second weight coefficient interval;

[0019] determining at least one first state parameter and at least one second state parameter as the first historical state parameter.

[0020] Optionally, at least one of wind speed, wind direction, temperature, humidity, air pressure of an environment where the wind turbine is located, and rotation speed, torque and output power of the wind turbine is determined as the second historical state parameter, comprising:

[0021] determining at least one first state parameter and at least one second state parameter as the second historical state parameter, and at least one first state parameter or second state parameter in the second historical state parameter is different from the first state parameter and the second state parameter in the first historical state parameter.

[0022] Optionally, according to the predicted yaw angle difference, a target predicted yaw angle of the target wind turbine is determined based on the first predicted yaw angle and the second predicted yaw angle, comprising:

[0023] if the predicted yaw angle difference is lower than a first difference value, taking the first predicted yaw angle or the second predicted yaw angle as the target predicted yaw angle of the target wind turbine;

[0024] If the predicted yaw angle difference is not less than a first difference, a first model weight of the first yaw angle prediction model and a second model weight of the second yaw angle prediction model are obtained, and the first predicted yaw angle and the second predicted yaw angle are weighted and summed based on the first model weight and the second model weight to obtain a target predicted yaw angle of the target wind turbine.

[0025] Optionally, after determining the target predicted yaw angle of the target wind turbine, the method further comprises:

[0026] In response to the weight adjustment instruction, a plurality of historical state parameters and corresponding actual yaw angles of the target wind turbine in a preset time period before a current time are randomly obtained;

[0027] The plurality of obtained historical state parameters are input into the first yaw angle prediction model to output a third predicted yaw angle, and the plurality of obtained historical state parameters are input into the second yaw angle prediction model to output a fourth predicted yaw angle;

[0028] A first accuracy rate of the first yaw angle prediction model is determined according to the plurality of obtained third predicted yaw angles and corresponding actual yaw angles, and a second accuracy rate of the second yaw angle prediction model is determined according to the plurality of obtained fourth predicted yaw angles and corresponding actual yaw angles;

[0029] If the first accuracy rate is greater than the second accuracy rate, the first model weight is increased by a set step, and the second model weight is decreased by the set step;

[0030] If the first accuracy rate is less than the second accuracy rate, the first model weight is decreased by the set step, and the second model weight is increased by the set step.

[0031] In a second aspect, the application provides a wind turbine yaw control device based on a neural network, comprising:

[0032] A data acquisition module configured to obtain state parameters of a target wind turbine, the state parameters at least including wind speed, wind direction, temperature, humidity, air pressure of an environment in which the target wind turbine is located, and at least one of rotational speed, torque and output power of the target wind turbine;

[0033] The initial yaw angle prediction model is configured to output, by a first yaw angle prediction model, a first predicted yaw angle of the target wind turbine as input of state parameters of the target wind turbine, and output, by a second yaw angle prediction model, a second predicted yaw angle of the target wind turbine as input of state parameters of the target wind turbine, wherein the first yaw angle prediction model and the second yaw angle prediction model are obtained by training different deep learning algorithms through historical state parameters and corresponding yaw angles of different wind turbines respectively;

[0034] The target yaw angle prediction model is configured to determine a predicted yaw angle difference between the first predicted yaw angle and the second predicted yaw angle, and determine a target predicted yaw angle of the target wind turbine based on the first predicted yaw angle and the second predicted yaw angle according to the predicted yaw angle difference.

[0035] The control module is configured to control the target wind turbine to operate at the target predicted yaw angle.

[0036] In a third aspect, the present application provides a computer readable storage medium storing a computer program which, when executed by a processor, causes the processor to perform the neural network-based yaw control method of a wind turbine as described above.

[0037] In a fourth aspect, the present application provides a terminal device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the neural network-based yaw control method of a wind turbine when executing the computer program.

[0038] The embodiments provided by the present application have the following beneficial effects:

[0039] The present application effectively improves the accuracy of yaw control of the wind turbine, thereby improving the efficiency and reliability of wind power generation.

[0040] Other features and advantages of the embodiments or the implementation of the present application will be described in detail in the following specific implementation part. DETAILED DESCRIPTION

[0041] The accompanying drawings are included to provide a further understanding of the embodiments of the present application, and constitute a part of the specification, and are used together with the following specific implementation to explain the embodiments of the present application, but do not constitute a limitation on the embodiments of the present application. In the drawings:

[0042] Figure 1 The method flowchart of the neural network-based yaw control method of a wind turbine of the embodiments of the present application is schematically shown;

[0043] Figure 2A system structure schematic diagram of a wind turbine yaw control system according to an embodiment of the present application is shown schematically.

[0044] Figure 3 A schematic block diagram of a neural network-based wind turbine yaw control device according to an embodiment of the present application is shown schematically.

[0045] Figure 4 A terminal device structure schematic diagram according to an embodiment of the present application is shown schematically.

[0046] Reference Signs List

[0047] 10 - terminal device, 100 - processor, 101 - memory, 102 - computer program. DETAILED DESCRIPTION

[0048] The specific embodiments of the present application are described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely intended to illustrate and explain the present application, and are not intended to limit the present application.

[0049] The wind turbine blade airfoil has a direct impact on the operation of the wind turbine. At present, when designing the airfoil of the wind turbine blade, only the influence of icing on the mass of the blade is considered when simulating the load under the icing condition of the blade, and the influence of icing on the aerodynamic performance of the blade is not considered. Therefore, when the designed wind turbine blade operates under the icing condition, it may affect the operation of the wind turbine and exist a safety hazard.

[0050] To solve the above problems, as shown in Figure 1 The first aspect of the present application provides a neural network-based wind turbine yaw control method, comprising: obtaining state parameters of a target wind turbine, the state parameters at least including wind speed, wind direction, temperature, humidity, air pressure of an environment where the target wind turbine is located, and at least one of rotational speed, torque and output power of the target wind turbine; inputting the state parameters of the target wind turbine to output a first predicted yaw angle of the wind turbine through a first yaw angle prediction model, and inputting the state parameters of the target wind turbine to output a second predicted yaw angle of the wind turbine through a second yaw angle prediction model, wherein the first yaw angle prediction model and the second yaw angle prediction model are obtained by training different deep learning algorithms through historical state parameters and corresponding yaw angles of different wind turbines; determining a predicted yaw angle difference between the first predicted yaw angle and the second predicted yaw angle, determining a target predicted yaw angle of the target wind turbine based on the first predicted yaw angle and the second predicted yaw angle according to the predicted yaw angle difference; and controlling the target wind turbine to operate at the target predicted yaw angle.

[0051] Therefore, the application effectively improves the accuracy of yaw control of the wind turbine, thereby improving the efficiency and reliability of wind power generation.

[0052] It can be understood that the yaw system is an important part of the wind turbine, and when the wind speed vector changes to a certain extent, the yaw system drives the wind turbine to wind. However, the wind is uncertain, and the wind direction also fluctuates frequently. Therefore, whether the yaw system can wind becomes a key factor to improve power generation and reduce the load of the unit. The yaw angle prediction model is constructed by the deep learning algorithm, which can predict the optimal yaw angle of the wind turbine according to the real-time state parameters of the wind turbine, and control the wind turbine to operate at the optimal yaw angle. As shown in Figure 2 The method of the application is implemented based on the control system constructed by the application. The system includes a server connected to the PLC of the entire wind farm through a network, realizing real-time acquisition of main parameters of the yaw control system of the entire wind farm unit. The server is used for processing, storing and analyzing the data acquired from the PLC of the entire wind farm. The system of the application can realize fast storage of a large amount of data, real-time analysis of the data through an efficient algorithm, and provides support for training and model optimization of the neural network algorithm. In addition, the scheme stores the results of real-time analysis in the cloud, realizing remote access and management of the data.

[0053] In the application, the first yaw angle prediction model is obtained by training the first deep learning algorithm based on the historical state parameters and the corresponding yaw angle of different wind turbines. The training process of the first deep learning algorithm includes:

[0054] At least one of the wind speed, wind direction, temperature, humidity, air pressure of the environment where the wind turbine is located, and the rotational speed, torque and output power of the wind turbine is determined as the first historical state parameter, for example, the wind speed, wind direction, rotational speed and output power are used as the input parameters of the deep learning algorithm; the first historical state parameters and the corresponding actual yaw angles of different wind turbines at multiple sampling time points within a preset historical period are obtained, for example, the historical state parameters of different wind turbines in the past six months and the actual yaw angles corresponding to each parameter are obtained as the training samples of the deep learning algorithm, so as to train the deep learning algorithm. It can be understood that the actual yaw angle corresponding to the state parameter is the yaw angle of the yaw system accurately wind under the state parameter;

[0055] The first historical state parameter of a plurality of sampling time points in a preset historical period is input, and the corresponding predicted yaw angle is output by the first deep learning algorithm. The error between the predicted yaw angle output by the first deep learning algorithm and the corresponding actual yaw angle is calculated, and the algorithm parameters of the first deep learning algorithm are adjusted according to the comparison result until the error between the predicted yaw angle output by the first deep learning algorithm and the actual yaw angle is lower than a preset error threshold or reaches a maximum iteration number, and the first yaw angle prediction model is obtained. The first deep learning algorithm includes but is not limited to BP neural network, CNN neural network, long-short time neural network, etc.

[0056] The second yaw angle prediction model is obtained by training the second deep learning algorithm with the historical state parameters and the corresponding yaw angles of different wind turbines. The training process of the second deep learning algorithm includes:

[0057] At least one of the wind speed, wind direction, temperature, humidity, air pressure of the environment where the wind turbine is located, and the rotating speed, torque and output power of the wind turbine is determined as the second historical state parameter, wherein the second historical state parameter is different from at least one of the first historical state parameter. For example, the first historical state parameter is wind speed, wind direction, rotating speed and output power, and the second historical state parameter can be wind speed, wind direction, temperature, rotating speed, torque and output power, that is, the second historical state parameter is different from at least one of the first historical state parameter.

[0058] The second historical state parameter and the corresponding actual yaw angle of a plurality of sampling time points in a preset historical period of different wind turbines are obtained. The second historical state parameter of a plurality of sampling time points in a preset historical period is input, and the corresponding predicted yaw angle is output by the second deep learning algorithm. The error between the predicted yaw angle output by the second deep learning algorithm and the corresponding actual yaw angle is calculated, and the algorithm parameters of the second deep learning algorithm are adjusted according to the comparison result until the error between the predicted yaw angle output by the second deep learning algorithm and the actual yaw angle is lower than a preset error threshold or reaches a maximum iteration number, and the second yaw angle prediction model is obtained. The second deep learning algorithm includes but is not limited to BP neural network, CNN neural network, long-short time neural network, etc., and the second deep learning algorithm is different from the first deep learning algorithm. For example, the first deep learning algorithm is BP neural network, and the second deep learning algorithm can be long-short time neural network.

[0059] In the present application, the at least one of the wind speed, the wind direction, the temperature, the humidity, the air pressure of the environment where the wind turbine generator is located, and the rotating speed, the torque and the output power of the wind turbine generator is determined as the first historical state parameter, comprising: determining the weight coefficients of the wind speed, the wind direction, the temperature, the humidity, the air pressure, the rotating speed, the torque and the output power, determining the state parameters whose weight coefficients belong to the first weight coefficient interval as the first state parameters, determining the state parameters whose weight coefficients belong to the second weight coefficient interval as the second state parameters, the minimum value of the first weight coefficient interval is not less than the maximum value of the second weight coefficient interval; determining the at least one first state parameter and the at least one second state parameter as the first historical state parameter. For example, the corresponding weight coefficients of each parameter can be configured according to the importance of the above-mentioned parameters to the yaw angle, and each parameter can be divided into multiple groups according to the size of the weight coefficients, for example, the weight coefficients of the wind speed and the wind direction are the largest, so the wind speed and the wind direction are divided into 1 group, the weight coefficients of the rotating speed, the torque and the output power are less than the wind speed and the wind direction, so the rotating speed, the torque and the output power are divided into 1 group, the weight coefficients of the temperature, the humidity and the air pressure are the smallest, so the temperature, the humidity and the air pressure are divided into 1 group; or, the wind speed, the wind direction, the rotating speed and the output power of the wind turbine generator are divided into 1 group, and the temperature, the humidity, the air pressure and the torque are divided into 1 group, which is not limited here. When determining the first historical state parameter, at least one parameter in each group is determined as the first historical state parameter, for example, in the first group, the wind speed and the wind direction have the greatest influence on the yaw system, so the wind speed and the wind direction are determined as the first historical state parameter, in the second group, the rotating speed and the output power are determined as the first historical state parameter, in the third group, the air pressure is determined as the first historical state parameter, and finally the wind speed, the wind direction, the rotating speed, the output power and the air pressure are determined as the first historical state parameter.

[0060] Similarly, the at least one of the wind speed, the wind direction, the temperature, the humidity, the air pressure of the environment where the wind turbine generator is located, and the rotating speed, the torque and the output power of the wind turbine generator is determined as the second historical state parameter, comprising: determining the at least one first state parameter and the at least one second state parameter as the second historical state parameter, and the at least one first state parameter or the second state parameter in the second historical state parameter is different from the first state parameter and the second state parameter in the first historical state parameter. For example, the second historical state parameter can include the wind speed, the wind direction, the rotating speed, the torque, the output power, the temperature. In this way, different deep learning algorithms use different parameters as training data, which can effectively improve the prediction accuracy of the final target yaw angle.

[0061] In the present application, the target yaw angle of the target wind turbine is determined based on the first predicted yaw angle and the second predicted yaw angle according to the predicted yaw angle difference. If the predicted yaw angle difference is lower than the first difference, the first predicted yaw angle or the second predicted yaw angle is taken as the target predicted yaw angle of the target wind turbine. If the predicted yaw angle difference is not lower than the first difference, the first model weight of the first yaw angle prediction model and the second model weight of the second yaw angle prediction model are obtained, and the first predicted yaw angle and the second predicted yaw angle are weighted and summed based on the first model weight and the second model weight to obtain the target predicted yaw angle of the target wind turbine. For example, different weights are configured for different prediction models in advance, wherein the model weight of the prediction model can be configured according to the prediction accuracy of the yaw angle by different deep learning algorithms, and the model with higher accuracy has higher model weight. If the first predicted yaw angle and the second predicted yaw angle are small, any predicted yaw angle can be taken as the target yaw angle; if the difference between the two is large, the output results are weighted and summed to eliminate errors.

[0062] In the present application, after determining the target predicted yaw angle of the target wind turbine, the method further comprises:

[0063] In response to the weight adjustment instruction, a plurality of historical state parameters and corresponding actual yaw angles of the target wind turbine within a preset time period before the current time are randomly obtained. In the present application, the model weight of each prediction model is dynamically adjusted during the operation of the wind turbine, wherein the weight adjustment instruction can be a timing instruction, for example, the weight adjustment is triggered once a week. After activating the weight adjustment, the historical state parameters and corresponding actual yaw angles of the target wind turbine correctly pointing to the wind in the recent period are randomly obtained, and the obtained data is taken as test data. It can be understood that the historical state parameters and corresponding actual yaw angles of the target wind turbine correctly pointing to the wind can be determined according to the wind direction collected by the wind vane sensor at the same sampling time, for example, at a certain historical time, the actual yaw angle corresponding to the state parameter group 1 is X1°, and the optimal yaw angle determined according to the wind direction collected by the wind vane sensor at this time is X2°. If the difference between X1 and X2 is lower than the threshold, it is considered that the yaw system correctly points to the wind at this time.

[0064] The third predicted yaw angle is output by the first yaw angle prediction model with the obtained plurality of historical state parameters as input, and the fourth predicted yaw angle is output by the second yaw angle prediction model with the obtained plurality of historical state parameters as input. The obtained historical state parameters are input into the two models to obtain the corresponding prediction results.

[0065] The first accuracy of the first yaw angle prediction model is determined according to the obtained plurality of third predicted yaw angles and corresponding actual yaw angles, and the second accuracy of the second yaw angle prediction model is determined according to the obtained plurality of fourth predicted yaw angles and corresponding actual yaw angles; and the accuracy of each prediction model is determined according to the actual yaw angle corresponding to each historical state parameter in the test data and the prediction result;

[0066] If the first accuracy is greater than the second accuracy, the first model weight is increased by a set step, and the second model weight is decreased by a set step; if the first accuracy is less than the second accuracy, the first model weight is decreased by a set step, and the second model weight is increased by a set step, that is, according to the accuracy, the model weight of the model with lower accuracy is reduced by a set step, and the model weight of the model with higher accuracy is increased by a set step, in this way, the different prediction models are tested according to the actual data in the recent period of time, and the model weight of each model is dynamically adjusted, which can further improve the prediction accuracy of the yaw angle.

[0067] As shown in Figure 3 the second aspect of the present application, a wind turbine yaw control device based on a neural network is provided, comprising:

[0068] The data acquisition module is configured to obtain state parameters of a target wind turbine, the state parameters at least including wind speed, wind direction, temperature, humidity, air pressure of an environment in which the target wind turbine is located, and at least one of rotational speed, torque and output power of the target wind turbine;

[0069] The initial yaw angle prediction model is configured to input the state parameters of the target wind turbine, output a first predicted yaw angle of the wind turbine through the first yaw angle prediction model, and input the state parameters of the target wind turbine, output a second predicted yaw angle of the wind turbine through the second yaw angle prediction model, wherein the first yaw angle prediction model and the second yaw angle prediction model are obtained by training different deep learning algorithms through historical state parameters and corresponding yaw angles of different wind turbines;

[0070] The target yaw angle prediction model is configured to determine a predicted yaw angle difference between the first predicted yaw angle and the second predicted yaw angle, and determine a target predicted yaw angle of the target wind turbine based on the first predicted yaw angle and the second predicted yaw angle according to the predicted yaw angle difference;

[0071] The control module is configured to control the target wind turbine to operate at the target predicted yaw angle.

[0072] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of each functional unit and module is exemplified, and in actual application, the above-mentioned functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the above-described functions. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be realized in the form of hardware or software. In addition, the specific name of each functional unit and module is only for easy distinction, and does not limit the protection scope of the present application. The specific working process of the unit and module in the above system can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0073] In a third aspect, the present application provides a computer readable storage medium storing a computer program which, when executed by a processor, causes the processor to perform the neural network-based yaw control method of a wind turbine generator set as described above.

[0074] In a fourth aspect, the present application provides a terminal device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the neural network-based yaw control method of a wind turbine generator set when executing the computer program.

[0075] As shown in Figure 4 is a schematic diagram of a terminal device provided by an embodiment of the present application. As shown in Figure 4 The terminal device 10 of this embodiment comprises a processor 100, a memory 101, and a computer program 102 stored in the memory 101 and executable on the processor 100. The processor 100 implements the steps in the method embodiments described above when executing the computer program 102. Alternatively, the processor 100 implements the functions of each module / unit in the device embodiments described above when executing the computer program 102.

[0076] For example, the computer program 102 can be divided into one or more modules / units, which are stored in the memory 101 and executed by the processor 100 to complete the present application. One or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program 102 in the terminal device 10.

[0077] The terminal device 10 can be a desktop computer, a notebook computer, a palm computer, a cloud server, and other computing devices. The terminal device 10 can include, but is not limited to, the processor 100 and the memory 101. Those skilled in the art can understand that the terminal device 10 can further include other components necessary for the terminal device 10 to perform the functions described above.Figure 4 The terminal device 10 is merely an example and does not constitute a limitation on the terminal device 10, and can include more or fewer components than shown, or combine some components, or different components, for example, the terminal device can also include an input / output device, a network access device, a bus, etc.

[0078] The processor 100 can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0079] The memory 101 can be an internal storage unit of the terminal device 10, for example, a hard disk or a memory of the terminal device 10. The memory 101 can also be an external storage device of the terminal device 10, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the memory 101 can include both the internal storage unit and the external storage device of the terminal device 10. The memory 101 is used to store computer programs and other programs and data required by the terminal device 10. The memory 101 can also be used to temporarily store data that has been output or will be output.

[0080] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) containing computer-usable program code.

[0081] It should also be noted that the terms "comprising", "comprises" or other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0082] The above embodiments are only used to illustrate the present application, but not to limit it. Instead of the above, various modifications and changes can be made to the application by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the application shall fall into the scope of the claims of the application.

Claims

1. A neural network-based yaw control method for a wind turbine generator system, characterized by, The method comprises: obtaining state parameters of a target wind turbine; outputting a first predicted yaw angle of the wind turbine through a first yaw angle prediction model taking the state parameters of the target wind turbine as input, and outputting a second predicted yaw angle of the wind turbine through a second yaw angle prediction model taking the state parameters of the target wind turbine as input, wherein the first yaw angle prediction model and the second yaw angle prediction model are obtained by training different deep learning algorithms through historical state parameters and corresponding yaw angles of different wind turbines; determining a predicted yaw angle difference between the first predicted yaw angle and the second predicted yaw angle, and determining a target predicted yaw angle of the target wind turbine based on the first predicted yaw angle and the second predicted yaw angle according to the predicted yaw angle difference; controlling the target wind turbine to operate at the target predicted yaw angle.

2. The neural network-based wind turbine yaw control method according to claim 1, characterized in that, The first yaw angle prediction model is obtained by training a first deep learning algorithm through historical state parameters and corresponding yaw angles of different wind turbines, and the training process of the first deep learning algorithm comprises: determining at least one of wind speed, wind direction, temperature, humidity, air pressure of an environment where the wind turbine is located, and rotational speed, torque and output power of the wind turbine as first historical state parameters; obtaining first historical state parameters and corresponding actual yaw angles of different wind turbines at multiple sampling time points within a preset historical period; taking the first historical state parameters at the multiple sampling time points within the preset historical period as input, outputting corresponding predicted yaw angles through the first deep learning algorithm, calculating errors between the predicted yaw angles output by the first deep learning algorithm and the corresponding actual yaw angles, and adjusting algorithm parameters of the first deep learning algorithm according to the comparison results until the errors between the predicted yaw angles output by the first deep learning algorithm and the actual yaw angles are lower than a preset error threshold or a maximum iteration number is reached, thereby obtaining the first yaw angle prediction model.

3. The neural network-based wind turbine yaw control method according to claim 2, characterized in that, The second yaw angle prediction model is obtained by training a second deep learning algorithm through historical state parameters and corresponding yaw angles of different wind turbines, and the training process of the second deep learning algorithm comprises: determining at least one of wind speed, wind direction, temperature, humidity, air pressure of an environment where the wind turbine is located, and rotational speed, torque and output power of the wind turbine as second historical state parameters, wherein at least one of the second historical state parameters is different from the first historical state parameters; obtaining second historical state parameters and corresponding actual yaw angles of different wind turbines at multiple sampling time points within a preset historical period; taking the second historical state parameters at the multiple sampling time points within the preset historical period as input, outputting corresponding predicted yaw angles through the second deep learning algorithm, calculating errors between the predicted yaw angles output by the second deep learning algorithm and the corresponding actual yaw angles, and adjusting algorithm parameters of the second deep learning algorithm according to the comparison results until the errors between the predicted yaw angles output by the second deep learning algorithm and the actual yaw angles are lower than a preset error threshold or a maximum iteration number is reached, thereby obtaining the second yaw angle prediction model.

4. The neural network-based wind turbine yaw control method according to claim 3, characterized in that, The method comprises the following steps: The method comprises the following steps: The method comprises the following steps:

5. The neural network-based wind turbine group yaw control method according to claim 4, characterized in that, The method comprises the following steps: The method comprises the following steps:

6. The wind turbine generator system yaw control method of neural network according to claim 1, characterized in that, The method comprises the following steps: The method comprises the following steps: The method comprises the following steps:

7. The neural network-based wind turbine yaw control method according to claim 6, characterized in that, The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps:

8. A neural network-based yaw control device for a wind turbine generator system, characterized by, The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the a data collection module configured to obtain state parameters of a target wind turbine, the state parameters comprising at least one of wind speed, wind direction, temperature, humidity, air pressure of an environment where the target wind turbine is located, and at least one of rotation speed, torque and output power of the target wind turbine; an initial yaw angle prediction model configured to output a first predicted yaw angle of the wind turbine via a first yaw angle prediction model with the state parameters of the target wind turbine as input, and output a second predicted yaw angle of the wind turbine via a second yaw angle prediction model with the state parameters of the target wind turbine as input, wherein the first yaw angle prediction model and the second yaw angle prediction model are obtained by training different deep learning algorithms with historical state parameters and corresponding yaw angles of different wind turbines respectively; a target yaw angle prediction model configured to determine a predicted yaw angle difference between the first predicted yaw angle and the second predicted yaw angle, and determine a target predicted yaw angle of the target wind turbine based on the first predicted yaw angle and the second predicted yaw angle according to the predicted yaw angle difference; a control module configured to control the target wind turbine to operate at the target predicted yaw angle.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program which, when executed by the processor, causes the processor to perform the neural network-based wind turbine yaw control method according to any one of claims 1-7.

10. A terminal device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor implements the neural network-based wind turbine yaw control method according to any one of claims 1-7 when executing the computer program.

Images