Control method, device, equipment, medium and program product of wind turbine generator system

Abstract

The application provides a kind of wind generating set control method, device, equipment, medium and program product. It relates to the technical field of generator set control. The method comprises: obtaining the current speed of the target wind generating set;When the current speed is in the first speed range, obtain the first wind speed data with a look-back step of a preset length from the current time;Input the first wind speed data into the pre-trained wind speed prediction model to obtain the first predicted wind speed data at the next time;According to the first predicted wind speed data, determine the wind speed change trend;According to the wind speed change trend, determine the power generation mode of the target wind generating set;Wherein, the power generation mode of the target wind generating set is double-fed mode or squirrel cage mode. Through the application, the reliability of wind generating set power generation and the stability of power grid are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of generator set control, and particularly relates to a control method, device, equipment, medium and program product of a wind turbine generator set. BACKGROUND

[0002] In recent years, with the increasing demand for clean energy, wind power as a renewable energy technology has been widely used. The doubly-fed wind turbine generator set has become an important part of the wind power field due to its simple structure and easy maintenance.

[0003] In related technologies, the doubly-fed wind turbine generator set usually adopts a double-mode power generation mode combining doubly-fed and squirrel cage, that is, the squirrel cage mode with higher power generation efficiency is used for power generation at low wind speed, and the doubly-fed mode with more stable output power is used for power generation at high wind speed.

[0004] However, the control strategy in the above scheme is realized based on the real-time speed of the fan as the basis for judgment. When the wind speed fluctuates back and forth, the doubly-fed wind turbine generator set will repeatedly switch between the two power generation modes, which seriously affects the reliability of power generation and the stability of the power grid. SUMMARY

[0005] The present application provides a control method, device, equipment, medium and program product of a wind turbine generator set to improve the reliability of wind turbine generator set power generation and the stability of the power grid.

[0006] In a first aspect, the present application provides a control method of a wind turbine generator set, comprising:

[0007] obtaining a current speed of a target wind turbine generator set;

[0008] when the current speed is within a first speed range, obtaining first wind speed data with a look-back step length of a preset length from the current time;

[0009] inputting the first wind speed data into a pre-trained wind speed prediction model to obtain first predicted wind speed data at the next time;

[0010] determining a wind speed change trend according to the first predicted wind speed data;

[0011] determining a power generation mode of the target wind turbine generator set according to the wind speed change trend; wherein the power generation mode of the target wind turbine generator set is a doubly-fed mode or a squirrel cage mode.

[0012] In one possible implementation, the training process of the wind speed prediction model comprises:

[0013] obtaining a training sample set, the training sample set comprising a plurality of training sample data, each of the training sample data comprising second wind speed data with a lookback step of a preset length from a first time point and third wind speed data at a second time point, the second time point being a next time point of the first time point;

[0014] for each of the training sample data, inputting the second wind speed data into the wind speed prediction model to obtain second predicted wind speed;

[0015] training the wind speed prediction model according to a preset loss function, the second predicted wind speed and the third wind speed data.

[0016] In a possible implementation, the training of the wind speed prediction model according to the preset loss function, the second predicted wind speed and the third wind speed data comprises:

[0017] calculating a loss value between the second predicted wind speed and the third wind speed data according to the preset loss function;

[0018] when the convergence condition is not met, adjusting parameters of the wind speed prediction model based on the loss value, and repeatedly performing the processing of inputting the second wind speed data into the wind speed prediction model to obtain the second predicted wind speed until the convergence condition is met;

[0019] when the convergence condition is met, obtaining the trained wind speed prediction model.

[0020] In a possible implementation, the obtaining of the training sample set comprises:

[0021] obtaining original wind speed data in a preset historical time period;

[0022] performing preprocessing on the original wind speed data to obtain historical wind speed time series data, wherein the preprocessing at least comprises missing value filling processing, data smoothing processing and normalization processing;

[0023] splitting the historical wind speed time series data into a plurality of training sample data to obtain the training sample set.

[0024] In a possible implementation, after obtaining the current rotating speed of the target wind turbine generator, the method further comprises:

[0025] when the current rotating speed is located in the second rotating speed range, determining that the power generation mode of the target wind turbine generator is the squirrel cage mode, wherein the maximum rotating speed of the second rotating speed range is less than the minimum rotating speed of the first rotating speed range;

[0026] when the current rotating speed is located in the third rotating speed range, determining that the power generation mode of the target wind turbine generator is the double-fed mode, wherein the minimum rotating speed of the third rotating speed range is greater than the maximum rotating speed of the first rotating speed range.

[0027] In a possible implementation, the power generation mode of the target wind turbine generator set is determined according to the wind speed change trend, including:

[0028] When the wind speed change trend is an upward trend, the power generation mode of the target wind turbine generator set is determined as a doubly-fed mode;

[0029] When the wind speed change trend is a downward trend, the power generation mode of the target wind turbine generator set is determined as a squirrel cage mode.

[0030] In a second aspect, the present application provides a control device of a wind turbine generator set, including: an acquisition module configured to acquire a current rotating speed of a target wind turbine generator set; when the current rotating speed is within a first rotating speed range, acquire first wind speed data with a lookback step length of a preset length from a current time point;

[0031] a prediction module configured to input the first wind speed data into a wind speed prediction model pre-trained to obtain predicted wind speed data at a next time point;

[0032] a determination module configured to determine a wind speed change trend according to the predicted wind speed data, and determine a power generation mode of the target wind turbine generator set according to the wind speed change trend; wherein the power generation mode of the target wind turbine generator set is a doubly-fed mode or a squirrel cage mode.

[0033] In a possible implementation, the acquisition module is further configured to acquire a training sample set; the training sample set includes a plurality of training sample data, wherein each training sample data includes second wind speed data with a lookback step length of a preset length from a first time point and third wind speed data at a second time point; the second time point is a next time point of the first time point;

[0034] The prediction module is further configured to: for each training sample data, input the second wind speed data into the wind speed prediction model to obtain a second predicted wind speed;

[0035] The device further includes a training module configured to train the wind speed prediction model according to a preset loss function, the second predicted wind speed and the third wind speed data.

[0036] In a possible implementation, the training module is specifically configured to: calculate a loss value between the second predicted wind speed and the third wind speed data according to the preset loss function; when a convergence condition is not met, adjust parameters of the wind speed prediction model based on the loss value, and repeatedly execute the processing of inputting the second wind speed data into the wind speed prediction model to obtain the second predicted wind speed until the convergence condition is met; when the convergence condition is met, obtain the wind speed prediction model pre-trained.

[0037] In a possible implementation, the acquisition module is specifically configured to acquire original wind speed data in a preset historical time period; the device further comprises a preprocessing module configured to preprocess the original wind speed data to obtain historical wind speed time series data; wherein the preprocessing at least includes missing value filling processing, data smoothing processing and normalization processing; the device further comprises a pre-splitting module configured to split the historical wind speed time series data into a plurality of training sample data to obtain a training sample set.

[0038] In a possible implementation, the determination module is further configured to determine that the power generation mode of the target wind turbine generator is the squirrel cage mode when the current rotating speed is located in a second rotating speed range; wherein the maximum rotating speed of the second rotating speed range is less than the minimum rotating speed of the first rotating speed range; and determine that the power generation mode of the target wind turbine generator is the double-fed mode when the current rotating speed is located in a third rotating speed range; wherein the minimum rotating speed of the third rotating speed range is greater than the maximum rotating speed of the first rotating speed range.

[0039] In a possible implementation, when determining the power generation mode of the target wind turbine generator according to the wind speed change trend, the determination module is specifically configured to: determine that the power generation mode of the target wind turbine generator is the double-fed mode when the wind speed change trend is an upward trend; and determine that the power generation mode of the target wind turbine generator is the squirrel cage mode when the wind speed change trend is a downward trend.

[0040] In a third aspect, the present application provides an electronic device, comprising: a memory, a processor;

[0041] The memory stores computer execution instructions;

[0042] The processor executes the computer execution instructions stored in the memory, so that the processor executes the first aspect and / or various possible implementations of the first aspect as above.

[0043] In a fourth aspect, the present application provides a computer readable storage medium, the computer readable storage medium stores computer execution instructions, and the computer execution instructions are executed by the processor to implement the first aspect and / or various possible implementations of the first aspect as above.

[0044] In a fifth aspect, the present application provides a computer program product, comprising a computer program, which is executed by the processor to implement the first aspect and / or various possible implementations of the first aspect as above.

[0045] The wind turbine control method, device, equipment, medium and program product provided by the application, by dividing the speed of the wind turbine into three speed ranges, the first speed range is a speed range suitable for both the doubly-fed mode and the squirrel cage mode, the second speed range is a speed range suitable for low wind speed, and the squirrel cage mode is used for power generation, and the third speed range is a speed range suitable for high wind speed, and the doubly-fed mode is used for power generation. In actual application, the current speed of the target wind turbine is obtained, and it is judged whether the current speed is in the first speed range. If yes, the first wind speed data with a look-back step of a preset length from the current time is obtained, and the first wind speed data is input into a wind speed prediction model trained in advance to obtain first predicted wind speed data at the next time. At this time, the wind speed change trend can be determined in combination with the first predicted wind speed data. Then, the power generation mode of the target wind turbine at the future time can be determined on the basis of the current speed and the wind speed change trend, so that the target wind turbine can always run in the appropriate power generation mode under different wind speed conditions, the wind energy utilization rate is maximized, and the impact of frequent switching between the two modes on the power grid can be avoided, so as to achieve the purpose of improving the reliability of wind turbine power generation and the stability of the power grid. BRIEF DESCRIPTION OF DRAWINGS

[0046] The accompanying drawings, which are incorporated into and form a part of the specification, illustrate an embodiment consistent with the present application and, together with the description, serve to explain the principles of the application.

[0047] Figure 1 Flowchart of the wind turbine control method provided by the embodiment of the application Figure 1

[0048] Figure 2 Schematic diagram of the wind turbine control logic provided by the embodiment of the application

[0049] Figure 3 Schematic diagram of the training process of the wind speed prediction model in the wind turbine control method provided by the embodiment of the application

[0050] Figure 4 Structural schematic diagram of the wind speed prediction model of the embodiment of the application

[0051] Figure 5 Comparison diagram of the prediction effect of the wind speed prediction model of the embodiment of the application

[0052] Figure 6 Structural schematic diagram of the wind turbine control device provided by the embodiment of the application

[0053] Figure 7 Structural schematic diagram of the electronic device provided by the embodiment of the application ​

[0054] The specific embodiments of the application have been shown by way of example in the above figures, and will be described in more detail hereafter. These drawings and the written description are not intended to restrict the scope of the application in any way but are merely illustrative of specific embodiments of the application as presently contemplated. DETAILED DESCRIPTION

[0055] Exemplary embodiments will be described in detail with reference to the drawings, of which like references refer to like elements throughout. The following detailed description is not intended to restrict the scope of the application, as claimed, but rather the exemplary embodiments described herein are intended to explain the principles of the application as presently contemplated. It will thus be appreciated that those skilled in the art will be able to devise numerous alternative arrangements based on the description herein without departing from the scope of the application as set forth in the appended claims.

[0056] In recent years, with the increasing demand for clean energy worldwide, wind power generation as a clean and renewable energy technology has been widely used and rapidly developed worldwide. Doubly fed induction generator (DFIG) is widely used in medium and large wind power systems due to its relatively simple structure, low manufacturing cost, high operating efficiency and convenient maintenance.

[0057] In the traditional operation mode, doubly fed wind generators mainly use a dual-mode power generation operation strategy combining doubly fed mode and squirrel cage mode, i.e., in the low wind speed section, it is switched to squirrel cage mode operation, which can effectively improve the starting performance and power generation efficiency under low wind speed; while in the medium and high wind speed section, it is switched back to doubly fed mode, which uses its wide range of variable speed operation and flexible power regulation capability to ensure efficient and stable operation of the unit near the rated power, and to meet the demand of active / reactive power regulation of the power grid.

[0058] However, the dual-mode power generation mode generally uses a mode switching mechanism based on real-time speed threshold judgment. This control strategy, in the actual operating environment, will cause the control system to frequently issue mode switching instructions due to frequent fluctuations of wind speed around the mode switching threshold, causing the doubly fed wind generator set to repeatedly switch between doubly fed mode and squirrel cage mode.

[0059] This frequent mode switching not only reduces system reliability, but also affects power quality and threatens the stable operation of the power grid. Therefore, how to optimize the dual-mode switching logic on the basis of ensuring system safety and stability, reduce unnecessary frequent switching, and improve the adaptability and operating efficiency of the doubly fed wind generator set under complex wind conditions, has become a key problem in the current upgrading of wind power technology.

[0060] Based on this, the application provides a control method of a wind turbine generator set, in which the rotating speed is divided into three rotating speed ranges, wherein the first rotating speed range is a rotating speed range suitable for both the doubly-fed mode and the squirrel cage mode, the second rotating speed range is a rotating speed range suitable for low wind speed, and the squirrel cage mode is used for power generation, and the third rotating speed range is a rotating speed range suitable for high wind speed, and the doubly-fed mode is used for power generation. When the obtained current rotating speed is located in the first rotating speed range, the wind speed change trend at the next moment is predicted based on the wind speed data according to the pre-trained wind speed prediction model, so as to determine the power generation mode of the target wind turbine generator set at the future moment in combination with the predicted wind speed change trend. In this way, the target wind turbine generator set can always run in the appropriate power generation mode under different wind speed conditions, the wind energy utilization rate is maximized, meanwhile, the impact on the power grid caused by the frequent switching of the two modes can be avoided, so as to achieve the purpose of improving the reliability of the wind turbine generator set and the stability of the power grid.

[0061] The technical solutions of the application and how the technical solutions solve the above technical problems will be described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes can not be described again in some embodiments. The embodiments of the application will be described below with reference to the drawings.

[0062] Figure 1 Flowchart of the control method of the wind turbine generator set provided by the embodiments of the application Figure 1 As shown in Figure 1 , the method comprises:

[0063] Step 110, obtaining the current rotating speed of the target wind turbine generator set.

[0064] Step 120, when the current rotating speed is located in the first rotating speed range, obtaining the first wind speed data with a look-back step length of a preset length from the current moment.

[0065] Step 130, inputting the first wind speed data into the pre-trained wind speed prediction model to obtain the first predicted wind speed data at the next moment.

[0066] Step 140, determining the wind speed change trend according to the first predicted wind speed data.

[0067] Step 150, determining the power generation mode of the target wind turbine generator set according to the wind speed change trend; wherein the power generation mode of the target wind turbine generator set is the doubly-fed mode or the squirrel cage mode.

[0068] In actual application, the current rotating speed of the target wind turbine generator set is obtained in real time to determine the power generation mode of the target wind turbine generator set according to the current rotating speed.

[0069] In this embodiment, three rotating speed ranges are divided, as shown in the following table. Figure 2 The first rotating speed range is a rotating speed range suitable for both the doubly-fed mode and the squirrel cage mode, i.e. the rotating speed range from N1 to N2. Figure 2 In the rotating speed range from N1 to N2, the target wind turbine generator set can generate power in either the squirrel cage mode or the doubly-fed mode.

[0070] The second rotating speed range is a rotating speed range suitable for low wind speed, i.e. the rotating speed range from N3 to N1. Figure 2 As can be seen, the maximum rotating speed of the second rotating speed range is less than the minimum rotating speed of the first rotating speed range, and the squirrel cage mode is usually used in the second rotating speed range.

[0071] The third rotating speed range is a rotating speed range suitable for high wind speed, i.e. the rotating speed range greater than N2. Figure 2 As can be seen, the minimum rotating speed of the third rotating speed range is greater than the maximum rotating speed of the first rotating speed range, and the doubly-fed mode is used in the third rotating speed range.

[0072] After the current rotating speed is obtained, it is first determined in which rotating speed range the current rotating speed is located. If the current rotating speed is located in the second rotating speed range, it is determined that the current wind condition is low wind speed, and it is determined that the power generation mode to be used by the target wind turbine generator set is the squirrel cage mode. At this time, if it is determined according to the current power generation power that the current power generation mode of the target wind turbine generator set is the squirrel cage mode, the squirrel cage mode is continued to be maintained, and if it is determined according to the current power generation power that the current power generation mode of the target wind turbine generator set is the doubly-fed mode, the power generation mode of the target wind turbine generator set is directly switched from the doubly-fed mode to the squirrel cage mode.

[0073] If the current rotating speed is located in the third rotating speed range, it is determined that the current wind condition is high wind speed, and it is determined that the power generation mode to be used by the target wind turbine generator set is the doubly-fed mode. At this time, if it is determined according to the current power generation power that the current power generation mode of the target wind turbine generator set is the squirrel cage mode, the power generation mode of the target wind turbine generator set is switched from the squirrel cage mode to the doubly-fed mode, and if it is determined according to the current power generation power that the current power generation mode of the target wind turbine generator set is the doubly-fed mode, the doubly-fed mode is continued to be maintained.

[0074] If the current rotating speed is located in the first rotating speed range, it is determined that the target wind turbine generator set can generate power in either the squirrel cage mode or the doubly-fed mode at this time. In order to further determine which mode to generate power, the wind speed change trend at the next moment needs to be combined to determine.

[0075] Specifically, in order to realize the prediction of the wind speed change trend at the next moment, in the embodiment, a pre-trained wind speed prediction model is provided. When the wind speed prediction model is used to predict the wind speed change trend, first wind speed data with a preset length of lookback step from the current moment needs to be obtained, wherein the lookback step represents how many time steps of data are used to predict the future value.

[0076] In one example, the preset length of the lookback step is set to 10, that is, 10 historical time steps of wind speed data from the current moment are obtained as the first wind speed data, which is input into the wind speed prediction model to obtain the first predicted wind speed data at the next moment. It should be noted that the preset length of the lookback step can be determined according to actual use requirements, for example, it can be set to 5, 15, 20, etc., and the embodiment does not make specific limitation on the preset length of the lookback step.

[0077] Then, according to the first predicted wind speed data and the wind speed data at the current moment, the wind speed change trend can be determined. It can be understood that if the first predicted wind speed data is greater than the wind speed data at the current moment, it indicates that the wind speed presents an upward trend, if the first predicted wind speed data is less than the wind speed data at the current moment, it indicates that the wind speed presents a downward trend, and if the first predicted wind speed is similar to the wind speed data at the current moment, it indicates that the wind speed presents a stable trend.

[0078] In this way, the power generation mode of the target wind turbine generator can be determined according to the determined wind speed change trend. For example, when the wind speed change trend is an upward trend, the power generation mode of the target wind turbine generator is determined to be a double-fed mode. When the wind speed change trend is a downward trend, the power generation mode of the target wind turbine generator is determined to be a squirrel cage mode.

[0079] Specifically, when the wind speed change trend is an upward trend, if the current power generation mode of the target wind turbine generator is determined to be a squirrel cage mode according to the current power generation power, the power generation mode of the target wind turbine generator is switched from the squirrel cage mode to the double-fed mode; if the current power generation mode of the target wind turbine generator is determined to be a double-fed mode according to the current power generation power, the double-fed mode is continued to be maintained for power generation.

[0080] When the wind speed change trend is a downward trend, if the current power generation mode of the target wind turbine generator is determined to be a squirrel cage mode according to the current power generation power, the squirrel cage mode is continued to be maintained for power generation; if the current power generation mode of the target wind turbine generator is determined to be a double-fed mode according to the current power generation power, the power generation mode of the target wind turbine generator is switched from the double-fed mode to the squirrel cage mode for power generation.

[0081] It should be understood that when the wind speed change trend is a smooth trend, the power generation mode of the target wind turbine generator set does not need to be changed, that is, if the current power generation mode is the doubly-fed mode, the doubly-fed mode is continued to be used for power generation, and if the current power generation mode is the squirrel cage mode, the squirrel cage mode is continued to be used for power generation.

[0082] The control method of the wind turbine generator set provided by the embodiment of the present application divides the rotating speed into three rotating speed ranges, wherein the first rotating speed range is a rotating speed range suitable for both the doubly-fed mode and the squirrel cage mode, the second rotating speed range is a rotating speed range suitable for low wind speed and the squirrel cage mode is used for power generation, and the third rotating speed range is a rotating speed range suitable for high wind speed and the doubly-fed mode is used for power generation. In actual application, the current rotating speed of the target wind turbine generator set is obtained, and it is judged whether the current rotating speed is located in the first rotating speed range. If yes, the first wind speed data with a look-back step of a preset length from the current time is obtained, and the first wind speed data is input into the wind speed prediction model pre-trained to obtain the first predicted wind speed data of the next time. At this time, the wind speed change trend can be determined on the basis of the current wind speed data combined with the first predicted wind speed data. Then, the power generation mode of the target wind turbine generator set at the future time can be determined according to the wind speed change trend, so that the target wind turbine generator set can always run in the appropriate power generation mode under different wind speed conditions, the wind energy utilization rate is maximized, and at the same time, the impact on the power grid caused by frequent switching between the two modes can be avoided, so as to achieve the purpose of improving the reliability of the wind turbine generator set and the stability of the power grid.

[0083] The control method of the wind turbine generator set provided by the embodiment of the present application divides the rotating speed into three rotating speed ranges, wherein the first rotating speed range is a rotating speed range suitable for both the doubly-fed mode and the squirrel cage mode, the second rotating speed range is a rotating speed range suitable for low wind speed and the squirrel cage mode is used for power generation, and the third rotating speed range is a rotating speed range suitable for high wind speed and the doubly-fed mode is used for power generation. In actual application, the current rotating speed of the target wind turbine generator set is obtained, and it is judged whether the current rotating speed is located in the first rotating speed range. If yes, the first wind speed data with a look-back step of a preset length from the current time is obtained, and the first wind speed data is input into the wind speed prediction model pre-trained to obtain the first predicted wind speed data of the next time. At this time, the wind speed change trend can be determined on the basis of the current wind speed data combined with the first predicted wind speed data. Then, the power generation mode of the target wind turbine generator set at the future time can be determined according to the wind speed change trend, so that the target wind turbine generator set can always run in the appropriate power generation mode under different wind speed conditions, the wind energy utilization rate is maximized, and at the same time, the impact on the power grid caused by frequent switching between the two modes can be avoided, so as to achieve the purpose of improving the reliability of the wind turbine generator set and the stability of the power grid. Figure 3 and Figure 4 The training process of the wind speed prediction model provided by the embodiment of the present application is described in detail.

[0084] Figure 3 The training process of the wind speed prediction model provided by the embodiment of the present application is described in detail. Figure 3 The training process of the wind speed prediction model provided by the embodiment of the present application is described in detail.

[0085] Step 310, obtaining a training sample set; the training sample set includes a plurality of training sample data, wherein each training sample data includes second wind speed data with a look-back step of a preset length from a first time and third wind speed data of a second time; the second time is the next time of the first time.

[0086] Step 320, for each training sample data, inputting the second wind speed data into the wind speed prediction model to obtain a second predicted wind speed.

[0087] Step 330, training the wind speed prediction model according to a preset loss function, the second predicted wind speed and the third wind speed data.

[0088] In acquiring the training sample set, the original wind speed data in a preset historical time period can be acquired first. The historical time period can be selected according to actual needs, for example, the original wind speed data recorded every 5 minutes in the past two months can be collected. Then, the original wind speed data is preprocessed to obtain historical wind speed time series data; wherein the preprocessing at least includes missing value filling processing, data smoothing processing and normalization processing.

[0089] The missing value filling processing can be forward filling and / or average value filling. If the missing values are few, the rows or columns containing the missing values can be directly deleted.

[0090] It should be noted that in the original wind speed data, noise is inevitable, which can mask the true trend and pattern of the original wind speed data, resulting in inaccurate analysis results. The original wind speed data contains long-term trends, seasonal patterns or periodic changes, but these features can be masked by short-term fluctuations. If the original wind speed data is directly analyzed, these important trends and patterns can be ignored. Data smoothing processing can remove short-term fluctuations, making long-term trends and patterns more obvious, which is convenient for further analysis. Therefore, data smoothing processing is needed to reduce the influence of noise and make the data clearer and more stable.

[0091] In this embodiment, wavelet transform algorithm is used to smooth the original score data. The mathematical principle of wavelet transform is to extract the characteristics of the signal at different scales by convolving the signal with a series of scaled and translated wavelet functions. The specific expression formula is as follows:

[0092]

[0093] Wherein: is the signal to be analyzed, is the wavelet mother function, a is the scale parameter, which controls the stretching of the wavelet, b is the translation parameter, which controls the movement of the wavelet on the time axis, is the conjugate of the wavelet mother function, t is the time. In this way, the original wind speed data can be decomposed into components of different frequencies by wavelet transform, and then the high-frequency noise can be removed, and the long-term trend and seasonal pattern can be retained.

[0094] After smoothing, the original wind speed data is normalized to obtain the historical wind speed time series data, so that the original wind speed data can meet the requirements of model input. Then, the time prediction problem is converted into a supervised learning problem, that is, the wind speed data at a preset length of time is used to obtain the predicted wind speed at the next time.

[0095] Specifically, the historical wind speed time series data is split into multiple training sample data to obtain a training sample set. It can be understood that when the model is trained, the training sample data is usually divided into a training sample set and a test set according to a preset proportion, wherein the training sample set is used for training the model, and the test set is used for testing whether the prediction accuracy of the trained model meets the requirements. In this embodiment, 80% of the multiple training sample data is divided into data in the training sample set, and 20% is divided into data in the test set.

[0096] Then, a wind speed prediction model is constructed. In this embodiment, a long short-term memory network (LSTM) is used as the basic model of the wind speed prediction model, and relevant parameters are set. For example, the number of LSTM hidden units is set to 64, the number of LSTM layers is set to 2, and the rectified linear unit (ReLU) is selected as the activation function.

[0097] Generally, the LSTM model introduces a gating mechanism to solve the gradient vanishing and gradient explosion problems of RNN when processing long sequences. As shown in Figure 4 , the basic unit of LSTM includes three gates (input gate , forget gate , output gate ) and a cell state . Specifically:

[0098] Forget gate :

[0099] The forget gate is used to determine which information in the input second wind speed data is forgotten from the cell state. Among them, and are weight parameters, is the hidden state at time step t-1, is the bias term.

[0100] Input gate :

[0101] The input gate determines which new information in the input second wind speed data is added to the cell state. Among them, represents the Sigmoid activation function, and the output value interval is [0, 1], and are weight parameters, is the bias term.

[0102] Cell state : ; wherein, and are weight parameters, is a bias term.

[0103] Output Gate decides which information is output from the cell state, where, ; where, denotes the candidate state.

[0104] The hidden state is the output of the LSTM unit, which is used as the input for the next time step. Where, ; where, denotes the hidden state, and are weight parameters, is a bias term.

[0105] The process begins with the forget gate, which uses a sigmoid activation function to filter information, converting data to a value between 0 and 1 (i.e., normalization). A value close to 0 indicates that the information is discarded, while a value close to 1 means that the information is retained, which is called the gating mechanism. Then, the role of the input gate is to refresh the current memory cell c t . To do this, it processes the hidden state h t-1 of the previous time step and the input x t of the current time step through a sigmoid activation function. At the same time, x t and h t-1 generate a candidate memory state through a tanh activation function. Combining the operations of the forget gate and the input gate, the memory cell c t can be updated. The update of the memory cell c t depends on the memory c t-1 of the previous time step and the candidate memory state of the current time step.

[0106] The data used in this embodiment is continuous data with a time interval of 5 minutes within 2 months, which is continuously distributed in the time dimension, meeting the input requirements of the LSTM neural model. After setting the initial wind speed prediction model, the training sample data in the training sample set is used to train the wind speed prediction model. Specifically, in the training process, the second wind speed data is input into the wind speed prediction model to obtain the second predicted wind speed, and then the loss value between the second predicted wind speed and the third wind speed data is calculated according to the preset loss function. In one example, the preset loss function is mean square error, and Adam is used as the optimizer, the learning rate of the wind speed prediction model is set to 0.001, and several rounds of training are performed.

[0107] After the loss value is calculated, it is determined whether a preset convergence condition is met. When the convergence condition is not met, the parameters of the wind speed prediction model are adjusted based on the loss value, and the process of inputting the second wind speed data into the wind speed prediction model to obtain the second predicted wind speed is repeatedly performed until the convergence condition is met; wherein, in the training loop process, the training loop process of forward propagation, loss calculation, backward propagation and optimization is performed. When the convergence condition is met, the trained wind speed prediction model is obtained. It can be understood that the preset convergence condition can be that the loss value meets a preset threshold, or the number of training times reaches a preset number.

[0108] After obtaining the trained wind speed prediction model, the prediction accuracy of the wind speed prediction model is evaluated using the training sample data in the test set to determine whether the prediction accuracy of the wind speed prediction model meets the requirements. As shown in Figure 5 The wind speed prediction model trained in the embodiment has good generalization ability and can meet the prediction accuracy requirements.

[0109] According to the scheme of the embodiment, the wind speed prediction model capable of accurately predicting the wind speed data at the next moment can be obtained, so that the wind turbine generator set can dynamically adjust the power generation mode according to the predicted wind speed trend, maximize the wind energy utilization efficiency, reduce the influence on the stability of the power grid, and ensure the safe operation of the wind turbine generator set.

[0110] Figure 6 The structure schematic diagram of the control device of the wind turbine generator set provided by the embodiment of the application is shown in Figure 6 As shown in the figure, the control device 60 of the wind turbine generator set provided by the embodiment can include an acquisition module 61, a prediction module 62 and a determination module 63.

[0111] Specifically, the acquisition module 61 is configured to acquire the current speed of the target wind turbine generator set; when the current speed is within a first speed range, the first wind speed data with a look-back step of a preset length from the current moment is acquired.

[0112] The prediction module 62 is configured to input the first wind speed data into a pre-trained wind speed prediction model to obtain predicted wind speed data at the next moment.

[0113] The determination module 63 is configured to determine a wind speed change trend according to the predicted wind speed data, and determine a power generation mode of the target wind turbine generator set according to the wind speed change trend; wherein, the power generation mode of the target wind turbine generator set is a double-fed mode or a squirrel cage mode.

[0114] In a possible implementation, the acquisition module 61 is further configured to acquire a training sample set, and the training sample set includes a plurality of training sample data, each of which includes second wind speed data with a preset length of lookback step from a first time and third wind speed data at a second time, and the second time is a next time of the first time. The prediction module 62 is further configured to, for each training sample data, input the second wind speed data into the wind speed prediction model to obtain second predicted wind speed. The apparatus further includes a training module configured to train the wind speed prediction model according to a preset loss function, the second predicted wind speed, and the third wind speed data.

[0115] In a possible implementation, the training module is specifically configured to: calculate a loss value between the second predicted wind speed and the third wind speed data according to a preset loss function; when a convergence condition is not met, adjust parameters of the wind speed prediction model based on the loss value, and repeatedly perform the processing of inputting the second wind speed data into the wind speed prediction model to obtain the second predicted wind speed until the convergence condition is met; and when the convergence condition is met, obtain the trained wind speed prediction model.

[0116] In a possible implementation, the acquisition module 61 is specifically configured to acquire original wind speed data in a preset historical time period, and the apparatus further includes a preprocessing module configured to preprocess the original wind speed data to obtain historical wind speed time series data, and the preprocessing at least includes missing value filling processing, data smoothing processing, and normalization processing. The apparatus further includes a pre-splitting module configured to split the historical wind speed time series data into a plurality of training sample data to obtain the training sample set.

[0117] In a possible implementation, the determination module 63 is further configured to determine that the power generation mode of the target wind turbine generator is the squirrel cage mode when the current rotating speed is located in a second rotating speed range, and the maximum rotating speed of the second rotating speed range is less than the minimum rotating speed of the first rotating speed range; and determine that the power generation mode of the target wind turbine generator is the doubly-fed mode when the current rotating speed is located in a third rotating speed range, and the minimum rotating speed of the third rotating speed range is greater than the maximum rotating speed of the first rotating speed range.

[0118] In a possible implementation, when determining the power generation mode of the target wind turbine generator according to the wind speed change trend, the determination module 63 is specifically configured to: determine that the power generation mode of the target wind turbine generator is the doubly-fed mode when the wind speed change trend is an upward trend; and determine that the power generation mode of the target wind turbine generator is the squirrel cage mode when the wind speed change trend is a downward trend.

[0119] The control apparatus of the wind turbine generator provided in this embodiment can be used to execute the method provided in the method embodiments, and has similar implementation principles and technical effects, which will not be described here in detail.

[0120] Figure 7A structural schematic diagram of an electronic device is provided in the embodiments of the present application. As shown in Figure 7 The electronic device 70 provided in the embodiments includes at least one processor 701 and a memory 702. Optionally, the electronic device 70 further includes a communication component 703. The processor 701, the memory 702 and the communication component 703 are connected through a bus 704.

[0121] In the implementation process, the at least one processor 701 executes the computer execution instructions stored in the memory 702, so that the at least one processor 701 executes the above-mentioned method.

[0122] The specific implementation process of the processor 701 can refer to the method embodiments described above, which have similar implementation principles and technical effects, and will not be described here again in the embodiments.

[0123] In the above embodiments, it should be understood that the processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the application can be directly embodied as execution completed by a hardware processor, or executed by a combination of hardware and software modules in the processor.

[0124] The memory can include a random access memory (RAM), and can also include a non-volatile memory (NVM), for example, at least one disk memory.

[0125] The bus can be an industry standard architecture (ISA) bus, a peripheral component (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the convenience of representation, the bus in the drawings of the embodiments of the present application is not limited to only one bus or one type of bus.

[0126] The embodiments of the present application further provide a computer program product, including a computer program, which is executed by a processor to implement the above-mentioned method.

[0127] The embodiment of the present application further provides a computer readable storage medium, and the computer readable storage medium stores computer execution instructions. When the computer execution instructions are executed, any method described above is implemented.

[0128] The readable storage medium described above can be implemented by any type of volatile or nonvolatile storage device or a combination thereof, such as static random-access memory (SRAM), electrically-erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The readable storage medium can be any available medium that can be accessed by a general or special purpose computer.

[0129] An exemplary readable storage medium is coupled to the processor, so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be an integral part of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (ASIC). Of course, the processor and the readable storage medium can also exist as discrete components in the device.

[0130] The division of units is only a logical function division, and in actual implementation, there can be another division mode. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interfaces, devices or units, and can be electrical, mechanical or other forms.

[0131] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, can be located in one place, or can be distributed on multiple network units. According to actual needs, part or all of the units can be selected to achieve the purpose of the embodiment.

[0132] In addition, each function unit in each embodiment of the present application 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.

[0133] If the functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the part of the present application that essentially contributes to the prior art or the part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the embodiments of the method of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various program code storage media.

[0134] Those skilled in the art can understand that all or part of the steps of the above-mentioned method embodiments can be completed by program instruction related hardware. The aforementioned program can be stored in a computer readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk, and various program code storage media.

[0135] Finally, it should be noted that: those skilled in the art will easily think of other embodiments of the present application after considering the specification and practicing the application disclosed herein. The present application is intended to cover any variations, uses or adaptations of the present application which follow the general principles of the present application and include known or customary technical means in the art which are not disclosed in the present application, and is not limited to the precise structure described above and shown in the drawings, and various modifications and changes can be made without departing from the scope thereof. The scope of the present application is only limited by the appended claims.

Claims

1. A control method of a wind turbine generator system, characterized by, The method comprises: obtaining a current rotating speed of a target wind turbine generator set; when the current rotating speed is within a first rotating speed range, obtaining first wind speed data with a lookback step length of a preset length from a current time point; inputting the first wind speed data into a pre-trained wind speed prediction model to obtain first predicted wind speed data of a next time point; determining a wind speed change trend according to the first predicted wind speed data; when the wind speed change trend is an upward trend, determining that a power generation mode of the target wind turbine generator set is a doubly-fed mode; and when the wind speed change trend is a downward trend, determining that the power generation mode of the target wind turbine generator set is a squirrel cage mode; a training process of the wind speed prediction model comprises: obtaining a training sample set; the training sample set comprises a plurality of training sample data, wherein each training sample data comprises second wind speed data with a lookback step length of the preset length from a first time point and third wind speed data of a second time point; the second time point is a next time point of the first time point; for each training sample data, inputting the second wind speed data into the wind speed prediction model to obtain second predicted wind speed; calculating a loss value between the second predicted wind speed and the third wind speed data according to a preset loss function; when a convergence condition is not met, adjusting parameters of the wind speed prediction model based on the loss value, and repeatedly performing the processing of inputting the second wind speed data into the wind speed prediction model to obtain the second predicted wind speed until the convergence condition is met; when the convergence condition is met, obtaining the trained wind speed prediction model.

2. The method of claim 1, wherein, The obtaining of the training sample set comprises: obtaining original wind speed data in a preset historical time period; preprocessing the original wind speed data to obtain historical wind speed time series data; wherein the preprocessing at least comprises missing value filling processing, data smoothing processing and normalization processing; splitting the historical wind speed time series data into a plurality of training sample data to obtain the training sample set.

3. The method of claim 1, wherein, After the obtaining of the current rotating speed of the target wind turbine generator set, the method further comprises: when the current rotating speed is within a second rotating speed range, determining that the power generation mode of the target wind turbine generator set is the squirrel cage mode; wherein a maximum rotating speed of the second rotating speed range is less than a minimum rotating speed of the first rotating speed range; when the current rotating speed is within a third rotating speed range, determining that the power generation mode of the target wind turbine generator set is the doubly-fed mode; wherein a minimum rotating speed of the third rotating speed range is greater than a maximum rotating speed of the first rotating speed range.

4. A control device for a wind turbine generator system, characterized by comprising: comprise: an obtaining module, configured to obtain a current rotating speed of a target wind turbine generator set; when the current rotating speed is within a first rotating speed range, obtaining first wind speed data with a lookback step length of a preset length from a current time point; a prediction module, configured to input the first wind speed data into a pre-trained wind speed prediction model to obtain predicted wind speed data of a next time point; a determination module, configured to determine a wind speed change trend according to the predicted wind speed data; When the wind speed change trend is an upward trend, it is determined that the power generation mode of the target wind turbine generator unit is a doubly-fed mode; and when the wind speed change trend is a downward trend, it is determined that the power generation mode of the target wind turbine generator unit is a squirrel cage mode. The acquisition module is further configured to acquire a training sample set, and the training sample set includes a plurality of training sample data, wherein each training sample data includes second wind speed data with a lookback step of the preset length from a first time point and third wind speed data at a second time point, and the second time point is a next time point of the first time point. The prediction module is further configured to input the second wind speed data into the wind speed prediction model to obtain second predicted wind speed for each training sample data. The device further includes a training module configured to calculate a loss value between the second predicted wind speed and the third wind speed data according to a preset loss function, adjust parameters of the wind speed prediction model based on the loss value when a convergence condition is not met, and repeatedly perform the processing of inputting the second wind speed data into the wind speed prediction model to obtain second predicted wind speed until the convergence condition is met, and obtain the trained wind speed prediction model when the convergence condition is met.

5. An electronic device, comprising: including: a memory, a processor; the memory stores computer execution instructions; the processor executes the computer execution instructions stored in the memory, so that the processor executes the method of any one of claims 1-3.

6. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer execution instructions, and the computer execution instructions are executed to implement the method of any one of claims 1-3.

7. A computer program product, characterised in that, including a computer program, the computer program is executed to implement the method of any one of claims 1-3.

Images