Method and device for predicting output power of wind driven generator, storage medium and electronic device

Abstract

The invention discloses a prediction method and device for the output power of a wind driven generator, a storage medium and an electronic device, and relates to the field of wind power generation, and the prediction method for the output power of the wind driven generator comprises the steps: obtaining predicted meteorological data of a target time period after a current moment; dividing the target time period into a first time period and a second time period; correcting the predicted wind speed data corresponding to the first time period through a first wind speed correction model to obtain first corrected wind speed data; correcting the predicted wind speed data corresponding to the second time period through a second wind speed correction model to obtain second corrected wind speed data; predicting first power data corresponding to the wind driven generator of the target model in the first time period; predicting second power data corresponding to the wind driven generator of the target model in the second time period; and determining target power data according to the first power data and the second power data.

Description

Technical Field

[0001] The present application relates to the field of wind power generation, and in particular to a method and device for predicting the output power of a wind turbine generator, a storage medium, and an electronic device. Background Art

[0002] The uneven distribution of wind resources in time and space leads to the randomness and volatility of wind turbine output power. Large-scale wind power grid connection also poses great challenges to the safe and stable operation of the power grid. Timely and accurate prediction of wind turbine output power is a key technical means to deal with the volatility of wind turbine output power. Accurate prediction of wind turbine output power is conducive to power generation companies to formulate production plans and operation and maintenance arrangements, and at the same time helps to reduce the dispatching pressure of power grid dispatching agencies.

[0003] The existing technical routes for predicting the output power of wind turbines include physical methods, statistical methods and hybrid methods. The physical method is based on numerical weather forecast data, taking into account the terrain and surface characteristics of the wind farm, and calculating the wind speed prediction value at the hub height of the wind turbine based on fluid mechanics numerical simulation, and further obtaining the output power prediction value of the wind farm based on the power curve of the wind turbine. The statistical method mainly uses different statistical and artificial intelligence algorithms to establish a prediction model through learning and training of a large amount of historical data. This method can better capture complex nonlinear relationships and spatiotemporal changes, and realize wind power prediction more quickly. The hybrid method is a method that combines physical methods and statistical methods.

[0004] Traditional wind turbine output power prediction methods have been applied to a certain extent in wind power stations, but the power prediction effect of wind turbine cut-out is not good. Wind turbine cut-out refers to the process in which the wind turbine automatically stops working when the wind speed exceeds its designed upper limit of safe operation. This is part of the wind turbine protection system, which is designed to prevent excessive wind speed from causing damage to the wind turbine. When the wind speed reaches or exceeds the cut-out wind speed, the wind turbine control system will trigger a safety mechanism to disconnect the wind turbine from the grid and turn the blades to an angle that does not face the wind to reduce the impact of wind on the wind turbine. This state is called "cut-out". For areas with extremely rich wind resources, the average wind speed of the whole field during the strong wind season can reach 20 meters per second to 30 meters per second, while the cut-out wind speed of wind turbines is about 25 meters per second. If the wind speed fluctuates around the cut-out wind speed of wind turbines, a large number of wind turbines will frequently cut in and out of the wind turbine farm. When the wind turbine is not cut out, the wind speed has reached the wind speed required for rated power, so the wind turbine is in full or near full power state, and the output power is close to the installed capacity. When the wind turbine is cut out because the wind speed reaches the cut-out wind speed, the output power drops rapidly to 0. Small wind speed changes near the cut-out wind speed may cause the output power of the wind turbine to change from near full power to 0 in a short period of time, which poses a challenge to the accurate prediction of the output power of the wind turbine. Traditional wind turbine output power prediction methods usually do not take into account the power prediction problem near the cut-out wind speed. The physics-based prediction method will obtain the output power as the installed capacity value or 0 value based on whether the wind speed exceeds the cut-out wind speed. The statistical-based prediction method will predict the output power as an unspecified intermediate value near the cut-out wind speed based on the trained model, resulting in a low prediction accuracy.

[0005] Regarding the related art, the traditional method for predicting the output power of a wind turbine does not consider the cut-out condition of the wind turbine when predicting the output power of the wind turbine, resulting in a low accuracy rate in predicting the output power. Currently, no effective solution has been proposed.

[0006] Therefore, it is necessary to improve the related technology to overcome the above-mentioned defects in the related technology. Summary of the invention

[0007] The embodiments of the present application provide a method and device, a storage medium, and an electronic device for predicting the output power of a wind turbine, so as to at least solve the problem that the traditional method for predicting the output power of a wind turbine does not consider the cut-out situation of the wind turbine when predicting the output power of the wind turbine, resulting in low accuracy in predicting the output power.

[0008] According to one aspect of an embodiment of the present application, a method for predicting the output power of a wind turbine is provided, comprising: obtaining predicted meteorological data for a target time period after a current moment, wherein the predicted meteorological data for the target time period includes predicted wind speed data for the target time period; the predicted wind speed data for the target time period includes predicted wind speeds at multiple moments in the target time period; dividing the target time period into a first time period and a second time period based on the predicted meteorological data for the target time period, wherein there is at least one first moment in the first time period, and the probability that the actual wind speed at the first moment is greater than the cut-out wind speed of the wind turbine of the target model is greater than a preset threshold; correcting the predicted wind speed data corresponding to the first time period by using a first wind speed correction model corresponding to the first time period to obtain first corrected wind speed data; and correcting the predicted wind speed data corresponding to the second time period by using a second wind speed correction model corresponding to the second time period The first wind speed correction model and the second wind speed correction model are both correction models of the target model of wind turbine; based on the first corrected wind speed data, first power data corresponding to the target model of wind turbine in the first time period is predicted; and based on the second corrected wind speed data, second power data corresponding to the target model of wind turbine in the second time period is predicted, wherein the first power data includes the predicted output power of the target model of wind turbine at multiple moments in the first time period, and the second power data includes the predicted output power of the target model of wind turbine at multiple moments in the second time period; target power data is determined according to the first power data and the second power data, wherein the target power data includes the output power of the target model of wind turbine at multiple moments in the target time period.

[0009] In an exemplary embodiment, the target time period is divided into a first time period and a second time period based on the predicted meteorological data of the target time period, including: selecting M time periods from the target time period according to the length of the target time window, wherein M is an integer greater than or equal to 2; determining the specified probability of each time period being a reference time period based on the predicted meteorological data corresponding to each time period in the M time periods through a classification model, and determining the time period as the reference time period when the specified probability is greater than a preset probability, wherein there is at least one first moment in the reference time period; when there are P reference time periods in the M time periods, determining the union of the P reference time periods as the first time period, and determining the time period other than the first time period in the target time period as the second time period, wherein P is an integer greater than or equal to 1 and less than or equal to M.

[0010] In an exemplary embodiment, the predicted wind speed data corresponding to the first time period is corrected by a first wind speed correction model corresponding to the first time period to obtain first corrected wind speed data, including: determining the wind speed change trend corresponding to each reference time period in the P reference time periods by the classification model; correcting the predicted wind speed data corresponding to the pth reference time period by a wind speed correction model corresponding to the wind speed change trend corresponding to the pth reference time period in the P reference time periods to obtain the pth corrected wind speed data, wherein p is 1, ..., P-1, P, the first wind speed correction model includes wind speed correction models corresponding to different wind speed change trends, and the first corrected wind speed data includes the pth corrected wind speed data.

[0011] In an exemplary embodiment, predicting the first power data corresponding to the wind turbine of the target model in the first time period based on the first corrected wind speed data includes: determining the pth power data corresponding to the pth reference time period based on the pth corrected wind speed data and the predicted target meteorological element data corresponding to the pth reference time period through a power prediction model corresponding to the wind speed change trend corresponding to the pth reference time period, wherein the power prediction model is used to predict the output power of the wind turbine of the target model, and the power prediction models corresponding to different wind speed change trends are different, the predicted meteorological data of the target time period also includes the predicted target meteorological element data corresponding to the pth reference time period, and the pth power data includes the output power of the wind turbine of the target model at multiple moments in the pth reference time period; determining the first power data corresponding to the first time period according to P power data corresponding to the P reference time periods, wherein, in the case where there is a plurality of output powers at one moment in the P power data, the average of the multiple output powers is determined as the output power of the wind turbine of the target model at that moment.

[0012] In an exemplary embodiment, the method further includes: acquiring historical data of a historical time period, wherein the historical data includes: historical meteorological measured data, historical meteorological forecast data and historical output power of the target model of wind turbine generator at multiple moments in the historical time period, the historical meteorological measured data includes historical measured wind speed, and the historical meteorological forecast data includes historical predicted wind speed; in the case where it is determined according to the historical data that there is a target historical moment in the historical time period, adding a historical reference time period to the historical time period sequence, wherein the historical measured wind speed at the target historical moment is greater than the cut-out wind speed of the target model of wind turbine generator, and the historical reference time period is a time period of a specified time window length before and after the target historical moment in the historical time period, and the specified time window length is half of the target time window length; determining a first data set corresponding to the historical time period sequence based on the historical data, wherein the first data set The data set contains historical meteorological forecast data and historical output power at multiple moments in each historical reference time period in the historical time period sequence; the historical reference time periods in the historical time period sequence are classified by the classification model based on the first data set to obtain N clusters, wherein the wind speed change trend corresponding to the historical reference time period in each cluster of the N clusters is the same, the wind speed change trends corresponding to the N clusters are different, and N is an integer greater than or equal to 1; the general machine learning model is trained according to the historical forecast wind speed data and the historical measured wind speed data corresponding to the historical reference time period in each cluster to obtain a wind speed correction model corresponding to the wind speed change trend corresponding to each cluster, wherein the historical forecast wind speed data corresponding to the historical reference time period includes the historical forecast wind speeds at multiple moments in the historical reference time period, and the historical measured wind speed data corresponding to the historical reference time period includes the historical measured wind speeds at multiple moments in the historical reference time period.

[0013] In an exemplary embodiment, after obtaining the wind speed correction model corresponding to the wind speed change trend corresponding to each cluster, the method also includes: correcting the historical predicted wind speed data corresponding to the historical reference time period in each cluster by the wind speed correction model corresponding to the wind speed change trend corresponding to each cluster to obtain first historical corrected wind speed data; training a general machine learning model according to the first historical corrected wind speed data, the first historical predicted target meteorological element data and the historical output power data corresponding to the historical reference time period in each cluster to obtain a power prediction model corresponding to the wind speed change trend corresponding to each cluster, wherein the first historical predicted target meteorological element data includes historical predicted target meteorological element data at multiple moments in the historical reference time period, the historical meteorological forecast data includes the historical predicted target meteorological element data, and the historical output power data corresponding to the historical reference time period includes historical output powers at multiple moments in the historical reference time period.

[0014] In an exemplary embodiment, predicting the second power data corresponding to the wind turbine of the target model in the second time period based on the second corrected wind speed data includes: determining the second power data of the second time period based on the second corrected wind speed data and the predicted target meteorological element data corresponding to the second time period through a power prediction model of the wind speed level corresponding to the second time period; wherein the wind speed level corresponding to the second time period is different from the wind speed level corresponding to the first time period, the power prediction model is used to predict the output power of the wind turbine of the target model, different wind speed levels correspond to different power prediction models, and the predicted meteorological data of the target time period also includes the predicted target meteorological element data corresponding to the second time period.

[0015] In an exemplary embodiment, the method further includes: acquiring historical data of a historical time period, wherein the historical data includes: historical meteorological measured data, historical meteorological forecast data and historical output power of the target model of wind turbine generator at multiple moments in the historical time period, the historical meteorological measured data includes historical measured wind speed, and the historical meteorological forecast data includes historical predicted wind speed; in the case where it is determined according to the historical data that there is a target historical moment in the historical time period, determining a historical reference time period according to the target historical moment, wherein the historical measured wind speed at the target historical moment is greater than the cut-out wind speed of the target model of wind turbine generator, and the historical reference time period is a time period of a specified time window length before and after the target historical moment in the historical time period; determining a time period in the historical time period other than the historical reference time period to obtain a historical specified time period; training a general machine learning model according to the historical predicted wind speed data and the historical measured wind speed data corresponding to the historical specified time period to obtain the second wind speed correction model, wherein the historical predicted wind speed data corresponding to the historical specified time period includes the historical predicted wind speeds at multiple moments in the historical specified time period, and the historical measured wind speed data corresponding to the historical specified time period includes the historical measured wind speeds at multiple moments in the historical specified time period.

[0016] In an exemplary embodiment, after obtaining the second wind speed correction model, the method further includes: correcting the historical predicted wind speed data corresponding to the historical specified time period through the second wind speed correction model to obtain second historical corrected wind speed data; training a general machine learning model according to the second historical corrected wind speed data, second historical predicted target meteorological element data and historical output power data corresponding to the historical specified time period to obtain a power prediction model corresponding to the wind speed level corresponding to the second time period, wherein the second historical predicted target meteorological element data includes historical predicted target meteorological element data at multiple moments in the historical specified time period, the historical meteorological forecast data includes the historical predicted target meteorological element data, and the historical output power data corresponding to the historical specified time period includes the historical output power at multiple moments in the historical specified time period.

[0017] According to another aspect of an embodiment of the present application, a device for predicting the output power of a wind turbine is also provided, including: an acquisition module, used to acquire predicted meteorological data for a target time period after a current moment, wherein the predicted meteorological data for the target time period includes predicted wind speed data for the target time period; the predicted wind speed data for the target time period includes predicted wind speeds at multiple moments in the target time period; a division module, used to divide the target time period into a first time period and a second time period based on the predicted meteorological data for the target time period, wherein there is at least one first moment in the first time period, and the probability that the actual wind speed at the first moment is greater than the cut-out wind speed of the wind turbine of the target model is greater than a preset threshold; a correction module, used to correct the predicted wind speed data corresponding to the first time period by using a first wind speed correction model corresponding to the first time period to obtain first corrected wind speed data; and correct the predicted wind speed data for the second time period by using a second wind speed correction model corresponding to the second time period The predicted wind speed data is corrected to obtain second corrected wind speed data, wherein the first wind speed correction model and the second wind speed correction model are both correction models of the wind turbine of the target model; a prediction module, used to predict the first power data corresponding to the wind turbine of the target model in the first time period based on the first corrected wind speed data; and predict the second power data corresponding to the wind turbine of the target model in the second time period based on the second corrected wind speed data, wherein the first power data includes the predicted output power of the wind turbine of the target model at multiple moments in the first time period, and the second power data includes the predicted output power of the wind turbine of the target model at multiple moments in the second time period; a determination module, used to determine target power data according to the first power data and the second power data, wherein the target power data includes the output power of the wind turbine of the target model at multiple moments in the target time period.

[0018] According to another aspect of the embodiments of the present application, a computer-readable storage medium is provided, the computer-readable storage medium comprising a stored program, wherein the program is configured to execute the above-mentioned method for predicting the output power of the wind turbine when running.

[0019] According to another aspect of the embodiments of the present application, there is also provided an electronic device, including a memory and a processor, wherein the memory stores a computer program, wherein the processor is configured to execute the above-mentioned method for predicting the output power of the wind turbine through the computer program.

[0020] According to another aspect of the embodiments of the present application, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the method for predicting the output power of the wind turbine is implemented.

[0021] In the present application, the predicted meteorological data of the target time period is first obtained, and then the target time period is divided into a first time period and a second time period based on the predicted meteorological data, wherein there is a moment in the first time period when the predicted wind speed exceeds the wind turbine cut-out wind speed, and then the predicted wind speed data is corrected by applying the first wind speed correction model and the second wind speed correction model respectively for different time periods, and the output power of the wind turbine in different time periods is predicted based on the corrected wind speed data, and finally, the target power data of the wind turbine is determined according to the predicted first power data and the second power data. Since the situation that the wind speed exceeds the wind turbine cut-out wind speed is taken into account in the process of predicting the output power of the wind turbine, that is, the cut-out condition of the wind turbine is effectively identified and processed, thereby improving the accuracy of predicting the output power of the wind turbine, and solving the problem that the traditional method for predicting the output power of the wind turbine does not take the cut-out condition of the wind turbine into account when predicting the output power of the wind turbine, resulting in a low accuracy of predicting the output power. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0023] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0024] Figure 1 It is a hardware structure block diagram of a mobile terminal of a method for predicting output power of a wind turbine generator according to an embodiment of the present application;

[0025] Figure 2 is a flow chart of a method for predicting output power of a wind turbine according to an embodiment of the present application;

[0026] Figure 3 is an optional flow chart for predicting the output power of a wind turbine generator according to an embodiment of the present application;

[0027] Figure 4 It is a structural block diagram of a device for predicting the output power of a wind turbine according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0029] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0030] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 1 is a hardware structure block diagram of a mobile terminal for predicting the output power of a wind turbine generator according to an embodiment of the present application. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor (MP) or a programmable logic device (FPGA)) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0031] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the method for predicting the output power of the wind turbine in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, to implement the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the mobile terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0032] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0033] In order to solve the above problem, this embodiment provides a method for predicting the output power of a wind turbine, including but not limited to being applied to the above mobile terminal. Figure 2 is a flow chart of a method for predicting the output power of a wind turbine according to an embodiment of the present application, such as Figure 2 As shown, the process includes the following steps S202-S210:

[0034] Step S202: Acquire forecasted meteorological data for a target time period after the current moment, wherein the forecasted meteorological data for the target time period includes forecasted wind speed data for the target time period; the forecasted wind speed data for the target time period includes forecasted wind speeds at multiple moments in the target time period;

[0035] Optionally, the predicted meteorological data includes: predicted wind speed data of the target wind power station within the target time period, predicted wind direction data of the target wind power station within the target time period, predicted temperature data of the environment of the target wind power station within the target time period, predicted humidity data of the environment of the target wind power station within the target time period, and predicted atmospheric pressure data of the environment of the target wind power station within the target time period.

[0036] It should be noted that each type of forecasted meteorological data includes forecasted meteorological values ​​at multiple moments. For example, within the next 24 hours (target time period), starting from the moment when the output power of the wind turbine is predicted, 24*60 moments are determined with 1 minute as the division standard, and the forecasted wind speed at these 24*60 moments is obtained.

[0037] Step S204: dividing the target time period into a first time period and a second time period based on the forecast meteorological data of the target time period, wherein there is at least one first moment in the first time period, and the probability that the actual wind speed at the first moment is greater than the cut-out wind speed of the wind turbine of the target model is greater than a preset threshold;

[0038] That is, the first time period is a high wind speed time period when the wind turbine is expected to be switched off due to excessively high wind speed, and the second time period is the rest of the time except the first time period, that is, a normal wind speed time period when the wind speed is predicted not to cause the wind turbine to be switched off.

[0039] It should be noted that the first time period plus the second time period is the target time period, wherein the first time period and the second time period do not overlap.

[0040] Step S206: Correcting the predicted wind speed data corresponding to the first time period by using the first wind speed correction model corresponding to the first time period to obtain first corrected wind speed data; and correcting the predicted wind speed data corresponding to the second time period by using the second wind speed correction model corresponding to the second time period to obtain second corrected wind speed data, wherein the first wind speed correction model and the second wind speed correction model are both correction models of the wind turbine of the target model;

[0041] It should be noted that the first wind speed correction model is used for the first time period. The first wind speed correction model is a correction model for the difference between the predicted wind speed and the actual wind speed in a high wind speed environment. The purpose is to reduce the deviation of the predicted wind speed near the wind turbine cut-out wind speed and improve the prediction accuracy. The second wind speed correction model is used for the second time period. The second wind speed correction model is used to correct the predicted wind speed within the normal wind speed range to more accurately reflect the operating status of the wind turbine under non-cut-out conditions.

[0042] Step S208: predicting first power data corresponding to the wind turbine of the target model in the first time period based on the first corrected wind speed data; and predicting second power data corresponding to the wind turbine of the target model in the second time period based on the second corrected wind speed data, wherein the first power data includes the predicted output power of the wind turbine of the target model at multiple moments in the first time period, and the second power data includes the predicted output power of the wind turbine of the target model at multiple moments in the second time period;

[0043] Optionally, the cut-out wind speeds of wind turbines of the same model are the same, and the cut-out wind speeds of wind turbines of different models are different. The wind turbine of the target model includes: one or more wind turbines of the target model.

[0044] It should be noted that the power prediction model can be used to predict the output power of the wind turbine. Optionally, the power prediction model based on the first corrected wind speed data predicting the first power data corresponding to the target model of wind turbine in the first time period is different from the power prediction model based on the second corrected wind speed data predicting the second power data corresponding to the target model of wind turbine in the second time period.

[0045] Step S210: determining target power data according to the first power data and the second power data, wherein the target power data includes the output power of the wind turbine of the target model at multiple moments in the target time period.

[0046] Optionally, the first power data and the second power data of the target model of wind turbine are spliced ​​in chronological order to obtain the output power of the target model of wind turbine that is continuously predicted within the target time period, that is, the target power data.

[0047] Optionally, the target power data of the entire target wind farm may be obtained by adding the target power data of wind turbines of all models in the target wind farm, wherein the wind farm may have multiple target models.

[0048] In the above steps, the predicted meteorological data of the target time period is first obtained, and then the target time period is divided into a first time period and a second time period based on the predicted meteorological data, wherein there is a moment in the first time period when the predicted wind speed exceeds the wind turbine cut-out wind speed, and then the predicted wind speed data is corrected by applying the first wind speed correction model and the second wind speed correction model respectively for different time periods, and the output power of the wind turbine in different time periods is predicted based on the corrected wind speed data, and finally, the target power data of the wind turbine is determined according to the predicted first power data and the second power data. Since the situation that the wind speed exceeds the wind turbine cut-out wind speed is taken into account in the process of predicting the output power of the wind turbine, that is, the cut-out condition of the wind turbine is effectively identified and processed, thereby improving the accuracy of predicting the output power of the wind turbine, and solving the problem that the traditional method for predicting the output power of the wind turbine does not take the cut-out condition of the wind turbine into account when predicting the output power of the wind turbine, resulting in a low accuracy rate in predicting the output power.

[0049] In an exemplary embodiment, dividing the target time period into a first time period and a second time period based on the forecasted meteorological data of the target time period can be achieved by the following steps S11-S13:

[0050] Step S11: selecting M time periods from the target time period according to the target time window length, where M is an integer greater than or equal to 2;

[0051] Optionally, if the target time window length is 4 hours, the time length of each of the M time periods is 4 hours.

[0052] Step S12: determining the specified probability of each time period being a reference time period based on the forecast meteorological data corresponding to each time period in the M time periods through a classification model, and determining the time period as the reference time period when the specified probability is greater than a preset probability, wherein at least one first moment exists in the reference time period;

[0053] Optionally, taking the preset probability of 80 percent as an example, the classification model is used to specify the probability of each time period based on the predicted meteorological data of each of the M time periods, and when the specified probability is greater than the preset probability, the time period is determined as the reference time period.

[0054] Step S13: When there are P reference time periods in the M time periods, the union of the P reference time periods is determined as the first time period, and the time period in the target time period except the first time period is determined as the second time period, where P is an integer greater than or equal to 1 and less than or equal to M.

[0055] Optionally, taking the target time window length as 4 hours and the target time period as the next 24 hours as an example, among the M time periods selected, 2 time periods are reference time periods (for example, the next 1-5 hours and 4-8 hours), then these two reference time periods are unioned to determine as the first time period (the next 1-8 hours), and the M-2 time periods other than the two reference time periods among the selected M time periods are unioned to determine as the second time period (the next 0-1 hour, 9-24 hours).

[0056] In an exemplary embodiment, the predicted wind speed data corresponding to the first time period is corrected by the first wind speed correction model corresponding to the first time period to obtain the first corrected wind speed data, which can be achieved by the following steps S21-S22:

[0057] Step S21: determining the wind speed change trend corresponding to each reference time period in the P reference time periods by using the classification model;

[0058] Optionally, the classification model is a pre-trained model, and the wind speed change trend includes: a wind speed change trend from low to high, a wind speed change trend from high to low, etc.

[0059] It should be noted that the first time period is a time period of a first wind speed level, the second time period is a time period of a second constant wind speed, and the first wind speed level (eg, a high wind speed level) is higher than the second wind speed level (eg, a normal wind speed level).

[0060] It should be noted that, when the first time period is a time period of the first wind speed level, the first time period can be divided into P reference time periods, and each reference time period corresponds to a wind speed change trend.

[0061] Step S22: Correct the predicted wind speed data corresponding to the p-th reference time period through the wind speed correction model corresponding to the wind speed change trend corresponding to the p-th reference time period among the P reference time periods to obtain the p-th corrected wind speed data, wherein p is 1, ..., P-1, P, the first wind speed correction model includes wind speed correction models corresponding to different wind speed change trends, and the first corrected wind speed data includes the p-th corrected wind speed data.

[0062] It should be noted that different wind speed change trends correspond to different wind speed correction models, and then different wind speed correction models are used for reference time periods with different wind speed change trends to correct the predicted wind speed data of the reference time period.

[0063] Optionally, the wind speed change trend corresponding to the first reference time period is a wind speed change trend from low to high, and the wind speed correction model corresponding to the first reference time period is the wind speed correction model corresponding to the wind speed change trend from low to high.

[0064] Optionally, the predicted wind speed data for P reference time periods are input into the corresponding wind speed correction model, and the wind speed correction model corrects the predicted wind speed data according to the deviation pattern between the historically learned wind speed and the actual wind speed to obtain P corrected wind speed data. By using the wind speed correction model corresponding to the wind speed change trend, the predicted wind speed data can be corrected more accurately.

[0065] It should be noted that the above steps S21-S22 use the classification model to identify the wind speed change trends corresponding to different reference time periods, and then use different wind speed correction models corresponding to different wind speed change trends to correct the predicted wind speed data of these reference time periods, so as to improve the prediction accuracy of the output power of the wind turbine generator under high wind speed conditions, especially when it is greater than the wind turbine cut-out wind speed. This process is of great significance for the refined management and scheduling of wind farms, and helps to avoid power fluctuations and scheduling problems caused by inaccurate predictions.

[0066] In an exemplary embodiment, predicting the first power data corresponding to the wind turbine of the target model in the first time period based on the first corrected wind speed data can be achieved by the following steps S31-S32:

[0067] Step S31: determining the p-th power data corresponding to the p-th reference time period based on the p-th corrected wind speed data and the predicted target meteorological element data corresponding to the p-th reference time period through a power prediction model corresponding to the wind speed change trend corresponding to the p-th reference time period, wherein the power prediction model is used to predict the output power of the wind turbine of the target model, different wind speed change trends correspond to different power prediction models, the predicted meteorological data of the target time period also includes the predicted target meteorological element data corresponding to the p-th reference time period, and the p-th power data includes the output power of the wind turbine of the target model at multiple moments in the p-th reference time period;

[0068] Optionally, the predicted target meteorological element data corresponding to the pth reference time period includes: target meteorological element data at multiple moments in the pth reference time period. The target meteorological element data includes: wind direction, temperature, humidity, etc.

[0069] Optionally, the power prediction model is used to predict the total output power of one or more wind turbines of the target model.

[0070] Optionally, if the wind speed change trend corresponding to the first reference time period is a wind speed change trend from low to high, then the power prediction model corresponding to the first reference time period is the power prediction model corresponding to the wind speed change trend from low to high, and the first corrected wind speed data and the predicted target meteorological element data corresponding to the first reference time period are input into the power prediction model corresponding to the wind speed change trend from low to high to obtain the first power data corresponding to the first reference time period, wherein the first power data includes the output power at multiple moments in the first time period.

[0071] Step S32: Determine the first power data corresponding to the first time period based on the P power data corresponding to the P reference time periods, wherein, when there is a moment with multiple output powers in the P power data, the average of the multiple output powers is determined as the output power of the wind turbine of the target model at that moment.

[0072] Optionally, P power data corresponding to P reference time periods are integrated in chronological order (during the integration process, if there are multiple output powers at one moment, the average of the multiple output powers is determined as the output power of the wind turbine of the target model at that moment), thereby determining the first power data corresponding to the first time period.

[0073] In an exemplary embodiment, the method further comprises the following steps S41-S45:

[0074] Step S41: Acquire historical data of a historical time period, wherein the historical data includes: historical meteorological measured data, historical meteorological forecast data and historical output power of the wind turbine of the target model at multiple moments in the historical time period, the historical meteorological measured data includes historical measured wind speed, and the historical meteorological forecast data includes historical forecast wind speed;

[0075] It should be noted that historical data refers to data over a historical period of time, such as data within the past 24 hours.

[0076] Optionally, the historical meteorological measured data includes: historical measured wind speed, historical measured wind direction, historical measured temperature and historical measured humidity, etc.

[0077] Optionally, the historical meteorological forecast data includes: historical forecast wind speed, historical forecast wind direction, historical forecast temperature and historical forecast humidity, etc.

[0078] Step S42: when it is determined according to the historical data that there is a target historical moment in the historical time period, a historical reference time period is added to the historical time period sequence, wherein the historical measured wind speed at the target historical moment is greater than the cut-out wind speed of the wind turbine of the target model, and the historical reference time period is a time period of a specified time window length before and after the target historical moment in the historical time period, and the specified time window length is half of the target time window length;

[0079] Optionally, taking the target time window length of 4 hours as an example, the specified time window length is 2 hours. When there is a moment (target historical moment) in a historical time period when the historical measured wind speed is greater than the cut-out wind speed of the target model of wind turbine, the time period of 2 hours before and 2 hours after the moment when the historical measured wind speed is greater than the cut-out wind speed of the target model of wind turbine in the historical time period is confirmed as the historical reference time period, and the determined historical reference time period is added to the historical time period sequence.

[0080] Step S43: determining a first data set corresponding to the historical time period sequence based on the historical data, wherein the first data set includes historical meteorological forecast data and historical output power at multiple moments in each historical reference time period in the historical time period sequence;

[0081] It should be noted that the historical meteorological forecast data at each moment in each reference time period in the historical time period sequence corresponds one-to-one to the historical output power at each moment.

[0082] Step S44: classifying the historical reference time periods in the historical time period sequence based on the first data set by using the classification model to obtain N clusters, wherein the wind speed change trend corresponding to the historical reference time period in each cluster of the N clusters is the same, and the wind speed change trends corresponding to the N clusters are different from each other, and N is an integer greater than or equal to 1;

[0083] Optionally, a clustering algorithm is used through a classification model to classify historical reference time periods in the historical time period sequence based on the first data set to obtain N clusters, wherein the clustering algorithm includes but is not limited to: K-means algorithm, density-based clustering algorithm, Gaussian mixture model, etc.; clustering features include but are not limited to predicted wind speed and predicted wind direction.

[0084] It should be noted that the K-means algorithm is an iterative clustering method that attempts to divide a data set into K different clusters or groups, where K is the number of predefined clusters. The algorithm assigns points to different clusters by calculating the distance from the data point to the cluster center, and then recalculates the center of each cluster until the cluster center no longer changes or the maximum number of iterations is reached. The goal of the K-means algorithm is to minimize the sum of the squared distances from each data point to the center of its cluster, that is, the compactness of the data points within the cluster.

[0085] Density-Based Spatial Clustering ofApplications with Noise (DBSCAN) is a density-based clustering algorithm that does not require a preset number of clusters, but determines the formation of clusters based on the density of data points. DBSCAN identifies clusters by defining two key parameters: the neighborhood radius and the minimum number of points in the neighborhood. If a point has at least the minimum number of points within its neighborhood radius, then the point is considered a core point. If a point does not have enough points in its neighborhood, but the point is within the neighborhood of another core point, the point is considered a boundary point. Points other than core points and boundary points are considered noise points. DBSCAN can discover clusters of arbitrary shapes from complex data sets and can identify and remove noise points.

[0086] Gaussian Mixture Model (GMM) is a statistical model used to describe a mixture of data points from multiple normal distributions. The algorithm model assumes that each data point in the data set can belong to a different Gaussian distribution, and each Gaussian distribution represents a potential cluster. GMM uses the expectation maximization algorithm to iteratively estimate the parameters (mean, variance, and mixing weight) of each cluster until the parameters converge. GMM can handle non-spherical clusters and clusters of different sizes and densities, and can estimate the uncertainty in the data set.

[0087] Step S45: Train a general machine learning model according to the historical predicted wind speed data and the historical measured wind speed data corresponding to the historical reference time period in each cluster to obtain a wind speed correction model corresponding to the wind speed change trend corresponding to each cluster, wherein the historical predicted wind speed data corresponding to the historical reference time period includes the historical predicted wind speeds at multiple moments in the historical reference time period, and the historical measured wind speed data corresponding to the historical reference time period includes the historical measured wind speeds at multiple moments in the historical reference time period.

[0088] It should be noted that different wind speed change trends correspond to different wind speed correction models.

[0089] In an exemplary embodiment, after obtaining the wind speed correction model corresponding to the wind speed change trend corresponding to each cluster, the method further includes the following steps S51-S52:

[0090] Step S51: correcting the historical forecast wind speed data corresponding to the historical reference time period in each cluster by using the wind speed correction model corresponding to the wind speed change trend corresponding to each cluster to obtain first historical corrected wind speed data;

[0091] It should be noted that the first historical correction data more accurately reflects the actual wind speed condition under a specific wind speed change trend (such as a wind speed change trend from low to high).

[0092] Step S52: Train a general machine learning model according to the first historical corrected wind speed data, the first historical prediction target meteorological element data and the historical output power data corresponding to the historical reference time period in each cluster to obtain a power prediction model corresponding to the wind speed change trend corresponding to each cluster, wherein the first historical prediction target meteorological element data includes the historical prediction target meteorological element data at multiple moments in the historical reference time period, the historical meteorological forecast data includes the historical prediction target meteorological element data, and the historical output power data corresponding to the historical reference time period includes the historical output power at multiple moments in the historical reference time period.

[0093] Optionally, the historical forecast target meteorological element data includes: historical forecast target wind speed, historical forecast target wind direction, historical forecast target temperature and historical forecast target humidity, etc.

[0094] It should be noted that the historical meteorological forecast data at a moment includes the historical forecast target meteorological element data at a moment.

[0095] It should be noted that, through the above steps, a power prediction model corresponding to the wind speed change trend corresponding to each cluster can be obtained, so that the output power of the wind turbine generator under different wind speed change trends can be predicted more accurately.

[0096] It should be noted that the wind speed correction model and the power prediction model in the embodiments of the present application are both obtained by supervised learning training.

[0097] In an exemplary embodiment, predicting the second power data corresponding to the wind turbine of the target model in the second time period based on the second corrected wind speed data can be achieved through the following steps: determining the second power data of the second time period based on the second corrected wind speed data and the predicted target meteorological element data corresponding to the second time period through a power prediction model of the wind speed level corresponding to the second time period; wherein the wind speed level corresponding to the second time period is different from the wind speed level corresponding to the first time period, the power prediction model is used to predict the output power of the wind turbine of the target model, different wind speed levels correspond to different power prediction models, and the predicted meteorological data of the target time period also includes the predicted target meteorological element data corresponding to the second time period.

[0098] Optionally, the predicted target meteorological elements include: predicted target wind direction, predicted target temperature, predicted target humidity, etc.

[0099] Optionally, the wind speed level corresponding to the second time period is a normal wind speed level (different from the wind speed level in the first time period, with a wind speed between 0 meters per second and 5 meters per second), then the power prediction model corresponding to the second time period is the power prediction model corresponding to the normal wind speed level, and the second corrected wind speed data and the predicted target meteorological element data corresponding to the second time period are input into the power prediction model corresponding to the normal wind speed level to obtain the second power data corresponding to the target model of wind turbine in the second time period.

[0100] In an exemplary embodiment, the method further includes the following steps S61-S64:

[0101] Step S61: Acquire historical data of a historical time period, wherein the historical data includes: historical meteorological measured data, historical meteorological forecast data and historical output power of the wind turbine of the target model at multiple moments in the historical time period, the historical meteorological measured data includes historical measured wind speed, and the historical meteorological forecast data includes historical forecast wind speed;

[0102] Optionally, the historical meteorological measured data includes: historical measured wind speed, historical measured wind direction, historical measured temperature and historical measured humidity, etc.

[0103] Optionally, the historical meteorological forecast data includes: historical forecast wind speed, historical forecast wind direction, historical forecast temperature and historical forecast humidity, etc.

[0104] Step S62: when it is determined according to the historical data that there is a target historical moment in the historical time period, a historical reference time period is determined according to the target historical moment, wherein the historical measured wind speed at the target historical moment is greater than the cut-out wind speed of the wind turbine of the target model, and the historical reference time period is a time period of the historical time period with a specified time window length before and after the target historical moment;

[0105] Optionally, taking the indication time window as 2 hours as an example, when there is a moment (target historical moment) in a historical time period when the historical measured wind speed is greater than the cut-out wind speed of the target model of wind turbine, the time period of 2 hours before and 2 hours after the moment when the historical measured wind speed is greater than the cut-out wind speed of the target model of wind turbine in the historical time period is confirmed as the historical reference time period.

[0106] Step S63: determining a time period in the historical time period except the historical reference time period to obtain a historical designated time period;

[0107] Optionally, assuming that a historical time period is from 9:00 to 17:00 in the past day, and the historical reference time period in the historical time period is from 9:00 to 13:00 in the past day, then the historical designated time period is from 13:00 to 17:00 in the past day.

[0108] Step S64: Train the general machine learning model according to the historical predicted wind speed data and the historical measured wind speed data corresponding to the historical specified time period to obtain the second wind speed correction model, wherein the historical predicted wind speed data corresponding to the historical specified time period includes the historical predicted wind speeds at multiple moments in the historical specified time period, and the historical measured wind speed data corresponding to the historical specified time period includes the historical measured wind speeds at multiple moments in the historical specified time period.

[0109] Optionally, a general machine learning model is trained using a machine learning algorithm (such as a multi-layer perceptron, a support vector machine, a k-nearest neighbor algorithm, etc.) based on historical predicted wind speed data and historical measured wind speed data corresponding to a specified historical time period to obtain a second wind speed correction model.

[0110] In an exemplary embodiment, after obtaining the second wind speed correction model, the method further includes the following steps S71-S72:

[0111] Step S71: correcting the historical predicted wind speed data corresponding to the historical specified time period by using the second wind speed correction model to obtain second historical corrected wind speed data;

[0112] Step S72: Train the general machine learning model according to the second historical corrected wind speed data, the second historical predicted target meteorological element data and the historical output power data corresponding to the historical specified time period to obtain a power prediction model corresponding to the wind speed level corresponding to the second time period, wherein the second historical predicted target meteorological element data includes the historical predicted target meteorological element data at multiple moments in the historical specified time period, the historical meteorological forecast data includes the historical predicted target meteorological element data, and the historical output power data corresponding to the historical specified time period includes the historical output power at multiple moments in the historical specified time period.

[0113] Optionally, a general machine learning model is trained using a machine learning algorithm (such as a multi-layer perceptron, a support vector machine, a k-nearest neighbor algorithm, etc.) based on the second historical corrected wind speed data, the second historical predicted target meteorological element data, and the historical output power data corresponding to the historical specified time period to obtain a power prediction model corresponding to the wind speed level corresponding to the second time period.

[0114] It should be noted that by inputting the second historical corrected wind speed data, the second historical predicted target meteorological element data and the historical output power data corresponding to the historical specified time period into the general machine learning model for training, the general machine learning model can learn the relationship between the corrected wind speed and other meteorological elements and the actual output power of the wind turbine under normal wind speed conditions, so that the power prediction model corresponding to the wind speed level corresponding to the second time period can more accurately predict the output power of the wind turbine under normal wind speed conditions.

[0115] Obviously, the above-described embodiments are only embodiments of a part of the present invention, rather than all embodiments. In order to better understand the above method, the above process is described below in conjunction with embodiments, but it is not intended to limit the technical solutions of the embodiments of the present invention, specifically:

[0116] When the wind speed of a wind farm reaches the cut-out wind speed, it is considered extreme wind weather, which is only likely to occur in a few areas with extremely rich wind resources across the country. Most weather forecasting methods use a unified physical model to carry out forecasts for a larger area, and cannot predict extremely high wind speeds of 20 meters per second to 30 meters per second. Therefore, when dividing the data, the actual wind speed of the wind farm is greater than the cut-out wind speed as the basis for division. Figure 3 An optional process for predicting the output power of a wind turbine generator is illustrated (S301-S308):

[0117] S301: Acquire historical data and align them;

[0118] S302: Divide the data set according to different models of wind turbines;

[0119] S303: Identify and separate high wind speed periods and normal wind speed periods;

[0120] S304: Clustering high wind speed periods;

[0121] S305: Classify and establish a correction model for predicting wind speed;

[0122] S306: Classify and establish a short-term power prediction statistical model;

[0123] S307: Correcting the predicted wind speed using the correction model, and obtaining the predicted output power of various types of wind turbines according to the corrected predicted wind speed;

[0124] S308: Obtain the predicted output power of the entire wind turbine farm.

[0125] The specific steps are as follows:

[0126] 1. Data collection and division;

[0127] Step 1: Obtain the wind turbine model information of the wind farm. The key is to obtain the cut-out wind speed of different models. Assume that there are n types of wind turbines in the wind farm, and the cut-out wind speeds are, v i , cut out, i = 1, 2, ..., n;

[0128] Step 2: Obtain historical data of the wind farm, including weather forecast data, weather measured data, active power data, etc., and align all data in terms of time scale and resolution ( Figure 3 Step S301 in the above example, where the weather forecast data and the weather measured data include wind speed, wind direction, temperature, humidity, pressure and other elements. The actual wind speed data is divided into n sets according to n types of aircraft ( Figure 3 Step S302), respectively a i , each set includes one type of machine, corresponding to the same cut-out wind speed. i The actual wind speed in is the average value of the wind speed measured by the cabin wind measuring device of this aircraft model;

[0129] Step 3: For each actual wind speed data set a i , filter out data points where the actual wind speed is greater than the cut-out wind speed. In order to fully reflect the change in wind speed and its corresponding relationship with power, for each filtered actual wind speed data point, combine the point with the data points of a period of time before and after its corresponding moment to form a time series of actual wind speed data. The characteristic of this sequence is that there is at least one point in the sequence where the wind speed is greater than the cut-out wind speed of the corresponding wind turbine model. The length of time that constitutes this sequence can be determined according to the frequency and amplitude of the wind speed change in the wind farm, and can generally be taken as 4 hours. Historical data set a iEach actual wind speed data point selected in the above formula can obtain a corresponding time series of actual wind speed. These time series together constitute the high wind speed data set b. i For the historical data set a i There is no high wind speed data set b i The data points appearing in are classified into the normal wind speed data set c i ( Figure 3 Step S303);

[0130] Step 4: For set b i For each actual wind speed data sequence in the data set, the corresponding data at the corresponding time is taken out from the meteorological forecast data and active power data to form the high wind speed period data set B i The active power data is the sum of the active power of each unit of this model;

[0131] Step 5: For set c i For each actual wind speed data point in the data, the corresponding data at the corresponding time is taken out from the meteorological forecast data and active power data to form the normal wind speed period data set C i The active power data is the sum of the active powers of all units of this model.

[0132] Through the above steps, the data division of high wind speed period and normal wind speed period is completed. The data processing and power prediction will be carried out respectively below.

[0133] 2. Clustering and correction of data;

[0134] The wind speed changes quickly and fluctuates violently during high wind speed periods. The wind speed change patterns in different time periods are different. The high wind speed period data set needs to be clustered and processed separately. The wind speed pattern of the normal wind speed period data set changes less, so the data is processed as a whole. At the same time, there is a prediction deviation in the predicted wind speed data, which needs to be corrected according to the actual wind speed data. The specific processing steps are as follows:

[0135] Step 1: For each high wind speed period dataset B i , determine the features used for clustering. The power prediction model uses predicted meteorological data as input, so the clustering features must be the predicted meteorological data. Calculate the Pearson correlation coefficient between each predicted meteorological data and the actual power data, and sort them from large to small. The meteorological elements that have the greatest impact on the actual power can be determined as clustering features. Usually, the predicted wind speed and predicted wind direction can be determined as clustering features;

[0136] Step 2: For the clustering features determined in step 1, use the K-means algorithm, density-based clustering algorithm, Gaussian mixture model, etc. to classify the high wind speed period data to obtain sub-datasets and clustering models of different high wind speed modes ( Figure 3 Step S304 in the above method);

[0137] Step 3: For each sub-dataset determined in step 2, establish a correction model for the predicted meteorological data. Take the predicted wind speed data in the sub-dataset as input and the actual wind speed data as output, and use machine learning algorithms such as multi-layer perceptron and support vector machine to establish a wind speed correction model to obtain the corrected predicted wind speed during high wind speed periods ( Figure 3 Step S305 in the above process);

[0138] Step 4: For each normal wind speed period dataset C i , taking the predicted wind speed data as input and the actual wind speed data as output, a wind speed correction model is established using machine learning algorithms such as multi-layer perceptron, support vector machine, and k-nearest neighbor algorithm to obtain the corrected predicted wind speed during the normal wind speed period.

[0139] 3. Establishment of power prediction model Figure 3 Step S306 in the above process);

[0140] The short-term wind power prediction model takes the predicted meteorological data as input. The present invention establishes prediction models under different wind speed modes in high wind speed periods and prediction models in normal wind speed periods respectively. The specific steps are as follows:

[0141] Step 1: For the high wind speed period, follow steps 1-3 in step 2 to obtain sub-datasets of different high wind speed modes and the corrected forecast wind speed for the high wind speed period;

[0142] Step 2: For each sub-dataset determined in step 1, the corrected predicted wind speed during high wind speed periods and other meteorological elements predicted during high wind speed periods are used as inputs, and the actual power in the corresponding sub-dataset is used as output. A statistical learning model based on multiple linear regression, kernel function, neural network, and deep learning methods is established;

[0143] Step 3: For the normal wind speed period, follow step 4 in step 2 to obtain the corrected forecast wind speed for the normal wind speed period;

[0144] Step 4: For the normal wind speed period, the corrected predicted wind speed during the normal wind speed period and other meteorological elements predicted during the normal wind speed period are used as input, and the actual power corresponding to the normal wind speed period is used as output. A statistical learning model based on multivariate linear regression, kernel function, neural network and deep learning methods is established.

[0145] 4. Operation of the power prediction model Figure 3 Step S307 in the above method);

[0146] The above has completed the establishment of wind speed correction models and power prediction models for each data set. In actual operation, it is necessary to obtain weather forecast data for future periods of time. The steps to obtain the predicted wind power based on the weather forecast data are as follows:

[0147] Step 1: Select the time window length. This length should be the same as the time series length of one actual wind speed data in step 3, usually 4 hours;

[0148] Step 2: Determine the high wind speed period and the normal wind speed period. According to the time window length determined in step 1, slide and select the data sequence in the meteorological forecast data, and use the clustering model obtained in step 2 (i.e., the above-mentioned classification model) to obtain the high wind speed category corresponding to the data sequence and the classification accuracy of the clustering model. Continuously slide the time window to obtain the category and classification accuracy of the predicted wind speed data for different time periods. If the classification accuracy of the data sequence is higher than the set threshold (such as 80%), the sequence is regarded as a high wind speed period. If the classification accuracy of the data sequence is lower than the threshold, it means that there is no high wind speed period in the predicted wind speed, and all are normal wind speed periods;

[0149] Step 3: Obtain the corrected forecast wind speed for the high wind speed period. For the high wind speed period determined in step 2, according to its high wind speed category, use the wind speed correction model for the high wind speed period corresponding to the category in step 3 to obtain the corrected forecast wind speed for the high wind speed period;

[0150] Step 4: Obtain the corrected forecast wind speed for the normal wind speed period. For the normal wind speed period determined in step 2, use the wind speed correction model for the normal wind speed period in step 4 to obtain the corrected forecast wind speed for the normal wind speed period;

[0151] Step 5: Obtain the predicted power for high wind speed periods. Using the corrected predicted wind speed for each category of high wind speed period obtained in step 3 as input, use the statistical learning model for each category of high wind speed period established in step 2 of the third section to obtain the power prediction value for each category of high wind speed period;

[0152] Step 6: Obtain the predicted power during the normal wind speed period. Using the corrected predicted wind speed during the normal wind speed period obtained in step 4 as input, use the statistical learning model for the normal wind speed period established in step 4 to obtain the power prediction value during the normal wind speed period;

[0153] Step 7: Joining high wind speed periods with normal wind speed periods. According to the power forecast values ​​of high wind speed periods and normal wind speed periods obtained in steps 5 and 6 of this section, join them in chronological order to obtain the predicted power sequence of various wind turbines in the period to be predicted;

[0154] Step 8: Get the predicted power of the wind turbine farm ( Figure 3 Step S309 in the above). For each type of wind farm, perform the operations of steps 1 to 7 in this section, add the obtained results, and obtain the overall predicted power of the wind turbine farm;

[0155] Through the above steps, the power prediction of the wind turbine considering the effect of cut-out wind speed is completed.

[0156] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods of each embodiment of the present application.

[0157] In this embodiment, a prediction device for the output power of a wind turbine is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0158] Figure 4 : is a structural block diagram of a device for predicting the output power of a wind turbine according to an embodiment of the present application, the device comprising:

[0159] The acquisition module 402 is used to acquire the forecasted meteorological data of the target time period after the current moment, wherein the forecasted meteorological data of the target time period includes the forecasted wind speed data of the target time period; the forecasted wind speed data of the target time period includes the forecasted wind speeds at multiple moments in the target time period;

[0160] A division module 404 is used to divide the target time period into a first time period and a second time period based on the forecast meteorological data of the target time period, wherein there is at least one first moment in the first time period, and the probability that the actual wind speed at the first moment is greater than the cut-out wind speed of the wind turbine of the target model is greater than a preset threshold;

[0161] A correction module 406 is used to correct the predicted wind speed data corresponding to the first time period by using a first wind speed correction model corresponding to the first time period to obtain first corrected wind speed data; and to correct the predicted wind speed data corresponding to the second time period by using a second wind speed correction model corresponding to the second time period to obtain second corrected wind speed data, wherein both the first wind speed correction model and the second wind speed correction model are correction models of the wind turbine of the target model;

[0162] A prediction module 408, configured to predict first power data corresponding to the wind turbine of the target model in the first time period based on the first corrected wind speed data; and predict second power data corresponding to the wind turbine of the target model in the second time period based on the second corrected wind speed data, wherein the first power data includes the predicted output power of the wind turbine of the target model at multiple moments in the first time period, and the second power data includes the predicted output power of the wind turbine of the target model at multiple moments in the second time period;

[0163] The determination module 410 is configured to determine target power data according to the first power data and the second power data, wherein the target power data includes output power of the wind turbine of the target model at multiple moments in the target time period.

[0164] The above device first obtains the predicted meteorological data of the target time period, and then divides the target time period into a first time period and a second time period based on the predicted meteorological data, wherein there is a moment in the first time period when the predicted wind speed exceeds the wind turbine cut-out wind speed, and then applies the first wind speed correction model and the second wind speed correction model to correct the predicted wind speed data for different time periods, and predicts the output power of the wind turbine in different time periods based on the corrected wind speed data, and finally determines the target power data of the wind turbine according to the predicted first power data and the second power data. Since the situation that the wind speed exceeds the wind turbine cut-out wind speed is taken into account in the process of predicting the output power of the wind turbine, that is, the cut-out condition of the wind turbine is effectively identified and processed, thereby improving the accuracy of predicting the output power of the wind turbine, and solving the problem that the traditional method for predicting the output power of the wind turbine does not take the cut-out condition of the wind turbine into account when predicting the output power of the wind turbine, resulting in a low accuracy rate of predicting the output power.

[0165] In an exemplary embodiment, the division module 404 is also used to select M time periods from the target time period according to the target time window length, where M is an integer greater than or equal to 2; determine the specified probability of the time period being a reference time period based on the predicted meteorological data corresponding to each time period in the M time periods through a classification model, and determine the time period as the reference time period when the specified probability is greater than the preset probability, where there is at least one first moment in the reference time period; when there are P reference time periods in the M time periods, determine the union of the P reference time periods as the first time period, and determine the time period in the target time period except the first time period as the second time period, where P is an integer greater than or equal to 1 and less than or equal to M.

[0166] In an exemplary embodiment, the division module 404 is also used to determine the wind speed change trend corresponding to each reference time period in the P reference time periods through the classification model; correct the predicted wind speed data corresponding to the pth reference time period through the wind speed correction model corresponding to the wind speed change trend corresponding to the pth reference time period in the P reference time periods to obtain the pth corrected wind speed data, where p is 1,...,P-1,P, the first wind speed correction model includes wind speed correction models corresponding to different wind speed change trends, and the first corrected wind speed data includes the pth corrected wind speed data.

[0167] In an exemplary embodiment, the prediction module 408 is also used to determine the p-th power data corresponding to the p-th reference time period based on the p-th corrected wind speed data and the predicted target meteorological element data corresponding to the p-th reference time period through a power prediction model corresponding to the wind speed change trend corresponding to the p-th reference time period, wherein the power prediction model is used to predict the output power of the wind turbine of the target model, and the power prediction models corresponding to different wind speed change trends are different. The predicted meteorological data of the target time period also includes the predicted target meteorological element data corresponding to the p-th reference time period, and the p-th power data includes the output power of the wind turbine of the target model at multiple moments in the p-th reference time period; determine the first power data corresponding to the first time period according to the P power data corresponding to the P reference time periods, wherein, when there is a moment with multiple output powers in the P power data, the average of the multiple output powers is determined as the output power of the wind turbine of the target model at that moment.

[0168] In an exemplary embodiment, the above-mentioned device also includes: a processing module, which is used to obtain historical data of a historical time period, wherein the historical data includes: historical meteorological measured data at multiple moments in the historical time period, historical meteorological forecast data and historical output power of the wind turbine of the target model, the historical meteorological measured data includes historical measured wind speed, and the historical meteorological forecast data includes historical predicted wind speed; when it is determined according to the historical data that there is a target historical moment in the historical time period, a historical reference time period is added to the historical time period sequence, wherein the historical measured wind speed at the target historical moment is greater than the cut-out wind speed of the wind turbine of the target model, and the historical reference time period is a time period of a specified time window length before and after the target historical moment in the historical time period, and the specified time window length is half of the target time window length; based on the historical data, a first data set corresponding to the historical time period sequence is determined, wherein the The first data set contains historical meteorological forecast data and historical output power at multiple moments in each historical reference time period in the historical time period sequence; the historical reference time periods in the historical time period sequence are classified by the classification model based on the first data set to obtain N clusters, wherein the wind speed change trend corresponding to the historical reference time period in each cluster of the N clusters is the same, and the wind speed change trends corresponding to the N clusters are different, and N is an integer greater than or equal to 1; the general machine learning model is trained according to the historical forecast wind speed data and the historical measured wind speed data corresponding to the historical reference time period in each cluster to obtain a wind speed correction model corresponding to the wind speed change trend corresponding to each cluster, wherein the historical forecast wind speed data corresponding to the historical reference time period includes the historical forecast wind speeds at multiple moments in the historical reference time period, and the historical measured wind speed data corresponding to the historical reference time period includes the historical measured wind speeds at multiple moments in the historical reference time period.

[0169] In an exemplary embodiment, the correction module 406 is also used to, after obtaining the wind speed correction model corresponding to the wind speed change trend corresponding to each cluster, correct the historical predicted wind speed data corresponding to the historical reference time period in each cluster through the wind speed correction model corresponding to the wind speed change trend corresponding to each cluster to obtain the first historical corrected wind speed data; train the general machine learning model according to the first historical corrected wind speed data, the first historical predicted target meteorological element data and the historical output power data corresponding to the historical reference time period in each cluster to obtain the power prediction model corresponding to the wind speed change trend corresponding to each cluster, wherein the first historical predicted target meteorological element data includes the historical predicted target meteorological element data at multiple moments in the historical reference time period, the historical meteorological prediction data includes the historical predicted target meteorological element data, and the historical output power data corresponding to the historical reference time period includes the historical output power at multiple moments in the historical reference time period.

[0170] In an exemplary embodiment, the prediction module 408 is also used to determine the second power data of the second time period based on the second corrected wind speed data and the predicted target meteorological element data corresponding to the second time period through the power prediction model of the wind speed level corresponding to the second time period; wherein the wind speed level corresponding to the second time period is different from the wind speed level corresponding to the first time period, the power prediction model is used to predict the output power of the wind turbine of the target model, different wind speed levels correspond to different power prediction models, and the predicted meteorological data of the target time period also includes the predicted target meteorological element data corresponding to the second time period.

[0171] In an exemplary embodiment, the processing module is further used to obtain historical data of a historical time period, wherein the historical data includes: historical meteorological measured data, historical meteorological forecast data and historical output power of the target model of wind turbine generator at multiple moments in the historical time period, the historical meteorological measured data includes historical measured wind speed, and the historical meteorological forecast data includes historical predicted wind speed; in the case where it is determined according to the historical data that there is a target historical moment in the historical time period, a historical reference time period is determined according to the target historical moment, wherein the historical measured wind speed at the target historical moment is greater than the cut-out wind speed of the target model of wind turbine generator, and the historical reference time period is a time period of a specified time window length before and after the target historical moment in the historical time period; a time period in the historical time period other than the historical reference time period is determined to obtain a historical designated time period; a general machine learning model is trained according to the historical predicted wind speed data and the historical measured wind speed data corresponding to the historical designated time period to obtain the second wind speed correction model, wherein the historical predicted wind speed data corresponding to the historical designated time period includes the historical predicted wind speeds at multiple moments in the historical designated time period, and the historical measured wind speed data corresponding to the historical designated time period includes the historical measured wind speeds at multiple moments in the historical designated time period.

[0172] In an exemplary embodiment, the correction module 406 is also used to, after obtaining the second wind speed correction model, correct the historical predicted wind speed data corresponding to the historical specified time period through the second wind speed correction model to obtain second historical corrected wind speed data; train the general machine learning model according to the second historical corrected wind speed data, the second historical predicted target meteorological element data and the historical output power data corresponding to the historical specified time period to obtain a power prediction model corresponding to the wind speed level corresponding to the second time period, wherein the second historical predicted target meteorological element data includes historical predicted target meteorological element data at multiple moments in the historical specified time period, the historical meteorological forecast data includes the historical predicted target meteorological element data, and the historical output power data corresponding to the historical specified time period includes the historical output power at multiple moments in the historical specified time period.

[0173] An embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when running.

[0174] Optionally, in this embodiment, the storage medium may be configured to store a computer program for performing the following steps:

[0175] S1, obtaining forecasted meteorological data for a target time period after the current moment, wherein the forecasted meteorological data for the target time period includes forecasted wind speed data for the target time period; the forecasted wind speed data for the target time period includes forecasted wind speeds at multiple moments in the target time period;

[0176] S2, dividing the target time period into a first time period and a second time period based on the forecast meteorological data of the target time period, wherein there is at least one first moment in the first time period, and the probability that the actual wind speed at the first moment is greater than the cut-out wind speed of the wind turbine of the target model is greater than a preset threshold;

[0177] S3, correcting the predicted wind speed data corresponding to the first time period by using the first wind speed correction model corresponding to the first time period to obtain first corrected wind speed data; and correcting the predicted wind speed data corresponding to the second time period by using the second wind speed correction model corresponding to the second time period to obtain second corrected wind speed data, wherein the first wind speed correction model and the second wind speed correction model are both correction models of the wind turbine of the target model;

[0178] S4, predicting first power data corresponding to the wind turbine of the target model in the first time period based on the first corrected wind speed data; and predicting second power data corresponding to the wind turbine of the target model in the second time period based on the second corrected wind speed data, wherein the first power data includes the predicted output power of the wind turbine of the target model at multiple moments in the first time period, and the second power data includes the predicted output power of the wind turbine of the target model at multiple moments in the second time period;

[0179] S5. Determine target power data according to the first power data and the second power data, wherein the target power data includes output power of the wind turbine of the target model at multiple moments in the target time period.

[0180] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

[0181] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail herein.

[0182] An embodiment of the present application further provides a computer program product, including a computer program, and when the computer program is executed by a processor, the steps in any one of the above method embodiments are performed.

[0183] An embodiment of the present application further provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0184] Optionally, in this embodiment, the processor may be configured to perform the following steps through a computer program:

[0185] S1, obtaining forecasted meteorological data for a target time period after the current moment, wherein the forecasted meteorological data for the target time period includes forecasted wind speed data for the target time period; the forecasted wind speed data for the target time period includes forecasted wind speeds at multiple moments in the target time period;

[0186] S2, dividing the target time period into a first time period and a second time period based on the forecast meteorological data of the target time period, wherein there is at least one first moment in the first time period, and the probability that the actual wind speed at the first moment is greater than the cut-out wind speed of the wind turbine of the target model is greater than a preset threshold;

[0187] S3, correcting the predicted wind speed data corresponding to the first time period by using the first wind speed correction model corresponding to the first time period to obtain first corrected wind speed data; and correcting the predicted wind speed data corresponding to the second time period by using the second wind speed correction model corresponding to the second time period to obtain second corrected wind speed data, wherein the first wind speed correction model and the second wind speed correction model are both correction models of the wind turbine of the target model;

[0188] S4, predicting first power data corresponding to the wind turbine of the target model in the first time period based on the first corrected wind speed data; and predicting second power data corresponding to the wind turbine of the target model in the second time period based on the second corrected wind speed data, wherein the first power data includes the predicted output power of the wind turbine of the target model at multiple moments in the first time period, and the second power data includes the predicted output power of the wind turbine of the target model at multiple moments in the second time period;

[0189] S5. Determine target power data according to the first power data and the second power data, wherein the target power data includes output power of the wind turbine of the target model at multiple moments in the target time period.

[0190] In an exemplary embodiment, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0191] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail herein.

[0192] Obviously, those skilled in the art should understand that the above modules or steps of the present application can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order from that herein, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present application is not limited to any specific combination of hardware and software.

[0193] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for predicting the output power of a wind turbine generator, characterized in that: include: Acquire forecasted meteorological data for a target time period after the current moment, wherein the forecasted meteorological data for the target time period includes forecasted wind speed data for the target time period; the forecasted wind speed data for the target time period includes forecasted wind speeds at multiple moments in the target time period; Dividing the target time period into a first time period and a second time period based on the forecast meteorological data of the target time period, wherein there is at least one first moment in the first time period, and the probability that the actual wind speed at the first moment is greater than the cut-out wind speed of the wind turbine of the target model is greater than a preset threshold; Correcting the predicted wind speed data corresponding to the first time period by using a first wind speed correction model corresponding to the first time period to obtain first corrected wind speed data; and correcting the predicted wind speed data corresponding to the second time period by using a second wind speed correction model corresponding to the second time period to obtain second corrected wind speed data, wherein both the first wind speed correction model and the second wind speed correction model are correction models of the wind turbine of the target model; Predicting first power data corresponding to the wind turbine of the target model in the first time period based on the first corrected wind speed data; and predicting second power data corresponding to the wind turbine of the target model in the second time period based on the second corrected wind speed data, wherein the first power data includes the predicted output power of the wind turbine of the target model at multiple moments in the first time period, and the second power data includes the predicted output power of the wind turbine of the target model at multiple moments in the second time period; Target power data is determined according to the first power data and the second power data, wherein the target power data includes output power of the wind turbine generator of the target model at multiple moments in the target time period.

2. The method according to claim 1, characterized in that Dividing the target time period into a first time period and a second time period based on the forecast meteorological data of the target time period includes: Selecting M time periods from the target time period according to the target time window length, where M is an integer greater than or equal to 2; Determine, by means of a classification model, a specified probability that each time period in the M time periods is a reference time period based on the predicted meteorological data corresponding to the time period, and determine that the time period is the reference time period when the specified probability is greater than a preset probability, wherein at least one first moment exists in the reference time period; In the case where there are P reference time periods in the M time periods, the union of the P reference time periods is determined as the first time period, and the time period in the target time period except the first time period is determined as the second time period, where P is an integer greater than or equal to 1 and less than or equal to M.

3. The method according to claim 2, characterized in that Correcting the predicted wind speed data corresponding to the first time period by using the first wind speed correction model corresponding to the first time period to obtain first corrected wind speed data includes: Determine the wind speed change trend corresponding to each reference time period in the P reference time periods by using the classification model; The predicted wind speed data corresponding to the p-th reference time period is corrected by a wind speed correction model corresponding to the wind speed change trend corresponding to the p-th reference time period among the P reference time periods to obtain the p-th corrected wind speed data, wherein p is 1, ..., P-1, P, the first wind speed correction model includes wind speed correction models corresponding to different wind speed change trends, and the first corrected wind speed data includes the p-th corrected wind speed data.

4. The method according to claim 3, characterized in that Predicting first power data corresponding to the wind turbine generator of the target model in the first time period based on the first corrected wind speed data includes: Determine the p-th power data corresponding to the p-th reference time period based on the p-th corrected wind speed data and the predicted target meteorological element data corresponding to the p-th reference time period through a power prediction model corresponding to the wind speed change trend corresponding to the p-th reference time period, wherein the power prediction model is used to predict the output power of the wind turbine of the target model, and different power prediction models correspond to different wind speed change trends. The predicted meteorological data of the target time period also includes the predicted target meteorological element data corresponding to the p-th reference time period, and the p-th power data includes the output power of the wind turbine of the target model at multiple moments in the p-th reference time period; The first power data corresponding to the first time period is determined based on the P power data corresponding to the P reference time periods, wherein, when there is a plurality of output powers at one moment in the P power data, the average of the plurality of output powers is determined as the output power of the wind turbine of the target model at that moment.

5. The method according to claim 4, characterized in that The method further comprises: Acquire historical data of a historical time period, wherein the historical data includes: historical meteorological measured data, historical meteorological forecast data and historical output power of the wind turbine of the target model at multiple moments in the historical time period, the historical meteorological measured data includes historical measured wind speed, and the historical meteorological forecast data includes historical forecast wind speed; In the case where it is determined according to the historical data that there is a target historical moment in the historical time period, a historical reference time period is added to the historical time period sequence, wherein the historical measured wind speed at the target historical moment is greater than the cut-out wind speed of the wind turbine of the target model, and the historical reference time period is a time period of a specified time window length before and after the target historical moment in the historical time period, and the specified time window length is half of the target time window length; Determine a first data set corresponding to the historical time period sequence based on the historical data, wherein the first data set includes historical meteorological forecast data and historical output power at multiple moments in each historical reference time period in the historical time period sequence; Classifying the historical reference time periods in the historical time period sequence based on the first data set by the classification model to obtain N clusters, wherein the wind speed change trend corresponding to the historical reference time period in each cluster of the N clusters is the same, the wind speed change trends corresponding to the N clusters are different from each other, and N is an integer greater than or equal to 1; A general machine learning model is trained according to the historical predicted wind speed data and the historical measured wind speed data corresponding to the historical reference time period in each cluster to obtain a wind speed correction model corresponding to the wind speed change trend corresponding to each cluster, wherein the historical predicted wind speed data corresponding to the historical reference time period includes the historical predicted wind speeds at multiple moments in the historical reference time period, and the historical measured wind speed data corresponding to the historical reference time period includes the historical measured wind speeds at multiple moments in the historical reference time period.

6. The method according to claim 5, characterized in that After obtaining the wind speed correction model corresponding to the wind speed change trend corresponding to each cluster, the method further includes: Correcting the historical predicted wind speed data corresponding to the historical reference time period in each cluster by using the wind speed correction model corresponding to the wind speed change trend corresponding to each cluster to obtain first historical corrected wind speed data; A general machine learning model is trained according to the first historical corrected wind speed data, the first historical prediction target meteorological element data and the historical output power data corresponding to the historical reference time period in each cluster to obtain a power prediction model corresponding to the wind speed change trend corresponding to each cluster, wherein the first historical prediction target meteorological element data includes the historical prediction target meteorological element data at multiple moments in the historical reference time period, the historical meteorological forecast data includes the historical prediction target meteorological element data, and the historical output power data corresponding to the historical reference time period includes the historical output power at multiple moments in the historical reference time period.

7. The method according to claim 1, characterized in that Predicting second power data corresponding to the wind turbine generator of the target model in the second time period based on the second corrected wind speed data includes: Determine the second power data for the second time period based on the second corrected wind speed data and the forecast target meteorological element data corresponding to the second time period by using a power prediction model for the wind speed level corresponding to the second time period; Among them, the wind speed level corresponding to the second time period is different from the wind speed level corresponding to the first time period, the power prediction model is used to predict the output power of the wind turbine of the target model, the power prediction models corresponding to different wind speed levels are different, and the predicted meteorological data of the target time period also includes the predicted target meteorological element data corresponding to the second time period.

8. The method according to claim 7, characterized in that The method further comprises: Acquire historical data of a historical time period, wherein the historical data includes: historical meteorological measured data, historical meteorological forecast data and historical output power of the wind turbine of the target model at multiple moments in the historical time period, the historical meteorological measured data includes historical measured wind speed, and the historical meteorological forecast data includes historical forecast wind speed; In the case where it is determined according to the historical data that there is a target historical moment in the historical time period, a historical reference time period is determined according to the target historical moment, wherein the historical measured wind speed at the target historical moment is greater than the cut-out wind speed of the wind turbine of the target model, and the historical reference time period is a time period of the historical time period with a specified time window length before and after the target historical moment; Determine a time period in the historical time period except the historical reference time period to obtain a historical designated time period; The general machine learning model is trained according to the historical predicted wind speed data and the historical measured wind speed data corresponding to the historical specified time period to obtain the second wind speed correction model, wherein the historical predicted wind speed data corresponding to the historical specified time period includes the historical predicted wind speeds at multiple moments in the historical specified time period, and the historical measured wind speed data corresponding to the historical specified time period includes the historical measured wind speeds at multiple moments in the historical specified time period.

9. The method according to claim 8, characterized in that After obtaining the second wind speed correction model, the method further includes: Correcting the historical predicted wind speed data corresponding to the historical specified time period by the second wind speed correction model to obtain second historical corrected wind speed data; The general machine learning model is trained according to the second historical corrected wind speed data, the second historical predicted target meteorological element data and the historical output power data corresponding to the historical specified time period to obtain a power prediction model corresponding to the wind speed level corresponding to the second time period, wherein the second historical predicted target meteorological element data includes the historical predicted target meteorological element data at multiple moments in the historical specified time period, the historical meteorological forecast data includes the historical predicted target meteorological element data, and the historical output power data corresponding to the historical specified time period includes the historical output power at multiple moments in the historical specified time period.

10. A device for predicting the output power of a wind turbine generator, characterized in that: include: An acquisition module, used to acquire the forecasted meteorological data of a target time period after the current moment, wherein the forecasted meteorological data of the target time period includes the forecasted wind speed data of the target time period; the forecasted wind speed data of the target time period includes the forecasted wind speeds at multiple moments in the target time period; a division module, configured to divide the target time period into a first time period and a second time period based on the forecast meteorological data of the target time period, wherein there is at least one first moment in the first time period, and the probability that the actual wind speed at the first moment is greater than the cut-out wind speed of the wind turbine of the target model is greater than a preset threshold; a correction module, configured to correct the predicted wind speed data corresponding to the first time period by using a first wind speed correction model corresponding to the first time period to obtain first corrected wind speed data; and to correct the predicted wind speed data corresponding to the second time period by using a second wind speed correction model corresponding to the second time period to obtain second corrected wind speed data, wherein both the first wind speed correction model and the second wind speed correction model are correction models of the wind turbine of the target model; A prediction module, configured to predict first power data corresponding to the wind turbine of the target model in the first time period based on the first corrected wind speed data; and predict second power data corresponding to the wind turbine of the target model in the second time period based on the second corrected wind speed data, wherein the first power data includes the predicted output power of the wind turbine of the target model at multiple moments in the first time period, and the second power data includes the predicted output power of the wind turbine of the target model at multiple moments in the second time period; A determination module is used to determine target power data according to the first power data and the second power data, wherein the target power data includes the output power of the wind turbine of the target model at multiple moments in the target time period.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein the program executes the method according to any one of claims 1 to 9 when executed.

12. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to execute the method according to any one of claims 1 to 9 through the computer program.

13. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 9 is implemented.

Images