Wind speed determination method and device based on wind power plant, equipment and storage medium

Abstract

The invention provides a wind speed determination method and device based on a wind power plant, equipment and a storage medium, relates to the field of artificial intelligence, in particular to the field of machine learning and statistics, and can be applied to the field of new energy such as wind power generation. According to the scheme, a target time interval is determined according to a wind speed loss moment, and wind speed information of a wind power plant in the target time interval is obtained; the wind speed loss moment is within the target time interval; according to the wind speed information of the wind power plant in the target time interval, determining lagging feature data and future feature data at a wind speed loss moment; the lagging feature data represents wind speed information before a wind speed missing moment in the target time interval, and the future feature data represents wind speed information after the wind speed missing moment in the target time interval; and according to the lagging feature data and the future feature data, determining wind speed information at a wind speed loss moment. The dynamic change of the wind speed is captured by combining past and future characteristics, and the wind speed determination precision is improved.

Description

Technical Field

[0001] The present disclosure relates to the fields of machine learning and statistics in the field of artificial intelligence, and in particular to a method, device, equipment and storage medium for determining wind speed based on a wind farm, which can be used in new energy fields such as wind power generation. Background Art

[0002] As demand for renewable energy continues to grow, developing renewable energy is an inevitable choice for enhancing energy security and achieving energy independence. Wind power generation is an important way to supply renewable energy power, and it is necessary to ensure that wind farms can provide a continuous and stable energy supply.

[0003] However, as wind farms continue to expand in size, wind monitoring data is plagued by a large number of missing values ​​and severe data fragmentation, hindering the analysis of wind farm operating characteristics and the accuracy of power generation forecasts. Therefore, there is an urgent need for an accurate wind speed determination method to address the issue of missing wind speed data in wind farms and provide reliable data support for high-quality scheduling of renewable energy. Summary of the Invention

[0004] The present disclosure provides a method, apparatus, device and storage medium for determining wind speed based on a wind farm.

[0005] According to a first aspect of the present disclosure, a method for determining wind speed based on a wind farm is provided, comprising:

[0006] Determine a target time interval according to the wind speed missing moment, and obtain wind speed information of the wind farm within the target time interval; wherein the wind speed missing moment represents a moment when the wind speed is not monitored, and the wind speed missing moment is within the target time interval;

[0007] Determining, based on the wind speed information of the wind farm in the target time interval, lagged characteristic data and future characteristic data at the moment when the wind speed is missing; wherein the lagged characteristic data represents the wind speed information before the moment when the wind speed is missing in the target time interval, and the future characteristic data represents the wind speed information after the moment when the wind speed is missing in the target time interval;

[0008] The wind speed information at the moment when the wind speed is missing is determined according to the delayed characteristic data and the future characteristic data.

[0009] According to a second aspect of the present disclosure, a wind speed determination device based on a wind farm is provided, comprising:

[0010] an information acquisition unit, configured to determine a target time interval based on a wind speed missing moment, and acquire wind speed information of the wind farm within the target time interval; wherein the wind speed missing moment represents a moment when no wind speed is monitored, and the wind speed missing moment is within the target time interval;

[0011] a feature determination unit, configured to determine, based on the wind speed information of the wind farm in the target time interval, lagged feature data and future feature data at the moment when the wind speed is missing; wherein the lagged feature data represents the wind speed information before the moment when the wind speed is missing in the target time interval, and the future feature data represents the wind speed information after the moment when the wind speed is missing in the target time interval;

[0012] A wind speed determination unit is used to determine the wind speed information at the moment when the wind speed is missing based on the lag characteristic data and the future characteristic data.

[0013] According to a third aspect of the present disclosure, there is provided an electronic device, including:

[0014] at least one processor; and

[0015] a memory communicatively coupled to the at least one processor;

[0016] The memory stores instructions that can be executed by the at least one processor. The instructions are executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect of the present disclosure.

[0017] According to a fourth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause the computer to execute the method described in the first aspect of the present disclosure.

[0018] According to a fifth aspect of the present disclosure, a computer program product is provided, comprising a computer program, which implements the steps of the method described in the first aspect of the present disclosure when executed by a processor.

[0019] According to the technology of the present disclosure, the accuracy of determining the missing wind speed is improved.

[0020] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings are provided to facilitate a better understanding of the present invention and do not constitute a limitation of the present disclosure.

[0022] Figure 1 is a flow chart of a method for determining wind speed based on a wind farm provided in accordance with an embodiment of the present disclosure;

[0023] Figure 2 is a flow chart of a method for determining wind speed based on a wind farm provided in accordance with an embodiment of the present disclosure;

[0024] Figure 3 is a flow chart of a method for determining wind speed based on a wind farm provided in accordance with an embodiment of the present disclosure;

[0025] Figure 4 is a schematic diagram of a process for determining wind speed information according to an embodiment of the present disclosure;

[0026] Figure 5 is a structural block diagram of a wind speed determination device based on a wind farm provided according to an embodiment of the present disclosure;

[0027] Figure 6 is a structural block diagram of a wind speed determination device based on a wind farm provided according to an embodiment of the present disclosure;

[0028] Figure 7 is a block diagram of an electronic device for implementing the wind speed determination method based on a wind farm according to an embodiment of the present disclosure;

[0029] Figure 8 This is a block diagram of an electronic device used to implement the wind speed determination method based on a wind farm according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0030] The following description of exemplary embodiments of the present disclosure is made in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0031] There are a large number of missing values ​​in the wind speed data of wind farms, which makes it difficult to ensure the consistency and integrity of the wind speed data, and brings great obstacles to the refined management and intelligent decision-making of wind farms.

[0032] In the field of wind farm meteorological data management, traditional wind speed data supplementation methods mainly rely on statistical methods and time series analysis techniques. For example, mean filling, linear interpolation, K-nearest neighbor interpolation, and classic time series prediction models. However, with the expansion of wind farm scale and the increase in data complexity, the limitations of traditional methods have gradually become apparent. For example, traditional statistical methods are based on linear assumptions and are difficult to characterize the nonlinear variation characteristics of wind speed under complex meteorological conditions, especially the vertical correlation modeling capabilities of wind speeds at different altitudes; time series prediction models rely on stationarity and fixed period assumptions, and are difficult to adapt to the coupling effects of sudden weather changes, seasonal fluctuations, and long-term trends, and are unable to integrate multi-source heterogeneous data, affecting the accuracy of wind speed data supplementation.

[0033] In recent years, deep learning technology has provided new insights into wind speed data management. Deep learning models such as long-short-term memory networks, temporal convolutional networks, and the Transformer architecture have been used to capture the long-term and short-term dependencies in wind speed data. However, deep learning models typically require a large amount of training data to extract wind speed-related features. High-quality training data is often difficult to obtain, limiting model performance and affecting the model's ability to predict missing values.

[0034] The present disclosure provides a wind speed determination method, device, equipment and storage medium based on a wind farm, which are applied to machine learning and statistics in the field of artificial intelligence, and can be used in new energy fields such as wind power generation to improve the accuracy of wind speed determination.

[0035] It should be noted that the model in this embodiment is not targeted at a specific user and cannot reflect the personal information of a specific user. It should be noted that the data in this embodiment comes from a public data set.

[0036] In the technical solutions disclosed herein, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0037] In order to enable readers to have a deeper understanding of the implementation principle of this disclosure, the following Figures 1-8 The embodiment is further refined.

[0038] Figure 1 FIG1 is a flow chart of a method for determining wind speed based on a wind farm according to an embodiment of the present disclosure. The method can be executed by a wind speed determination device based on a wind farm. Figure 1 As shown, the method includes the following steps:

[0039] S101. Determine a target time interval according to a wind speed missing moment, and obtain wind speed information of the wind farm within the target time interval; wherein the wind speed missing moment represents a moment when the wind speed is not monitored, and the wind speed missing moment is within the target time interval.

[0040] For example, a wind farm is a wind power plant that generates electricity based on wind power. To efficiently manage the wind farm, data monitoring of the wind farm's operating conditions and surrounding environment is required. In particular, wind speed information, such as wind speed and direction, needs to be monitored. Specifically, wind speed information can include wind speed magnitude and direction. However, as wind farms continue to expand in size, a large number of missing values ​​in wind speed information may exist. Specifically, there are times when wind speed information is not monitored. Therefore, it is necessary to determine the wind speed information at these times to supplement the missing values.

[0041] The moment when no wind speed information was monitored is determined as the wind speed missing moment. For example, the wind speed information monitored by the wind farm over a period of time can be reviewed. Each wind speed information corresponds to its own monitoring moment. The moment when no wind speed information was recorded is determined as the wind speed missing moment. In other words, the wind speed information at the wind speed missing moment needs to be determined. In this embodiment, the wind speed missing moment is a moment in the past.

[0042] A time period is preset, and the time period represents a time length. For example, the preset time period is 24 hours. After determining the moment when the wind speed is missing, the target time interval can be determined based on the moment when the wind speed is missing and the preset time period. The length of the target time interval is the length of the preset time period. For example, the target time interval is from 12 o'clock on a certain day to 12 o'clock on the next day. The moment when the wind speed is missing is within the target time interval, that is, based on the moment when the wind speed is missing, a period of time with a preset time period length containing the moment when the wind speed is missing can be used as the target time interval. For example, the moment when the wind speed is missing is used as the middle moment of the target time interval, and 12 hours are taken before and after the moment when the wind speed is missing to obtain the target time interval.

[0043] Obtain all wind speed information monitored by the wind farm within the target time interval. For example, if the wind speed is missing at 12:00 on May 8, then obtain all wind speed information monitored from 0:00 to 24:00 on May 8.

[0044] S102. Determine, based on the wind speed information of the wind farm within the target time interval, lagged characteristic data and future characteristic data at the moment when the wind speed is missing; wherein the lagged characteristic data represents the wind speed information before the moment when the wind speed is missing within the target time interval, and the future characteristic data represents the wind speed information after the moment when the wind speed is missing within the target time interval.

[0045] For example, wind speed information at multiple moments within a target time interval is obtained, and features are extracted from the wind speed information to obtain lagged feature data and future feature data at the moment when wind speed is missing. Both the lagged feature data and the future feature data are in vector form.

[0046] The delayed characteristic data represents the wind speed information before the wind speed is missing in the target time interval. For example, all wind speed information before the wind speed is missing in the target time interval can be obtained, and feature extraction processing can be performed on these wind speed information to obtain the delayed characteristic data. Alternatively, one or more wind speed information can be extracted from all wind speed information before the wind speed is missing in the target time interval, and the extracted wind speed information can be converted into feature data to obtain the delayed characteristic data. The future characteristic data represents the wind speed information after the wind speed is missing in the target time interval. For example, all wind speed information after the wind speed is missing in the target time interval can be obtained, and feature extraction processing can be performed on these wind speed information to obtain the future characteristic data. Alternatively, one or more wind speed information can be extracted from all wind speed information after the wind speed is missing in the target time interval, and the extracted wind speed information can be converted into feature data to obtain the future characteristic data.

[0047] In this embodiment, the process for generating characteristic data is not specifically limited. For example, all wind speed values ​​prior to the missing wind speed moment can be converted into a single-column matrix, with each row in the matrix representing a moment, and this matrix can be used as the lagged characteristic data. Wind speed information can also include wind direction, and the characteristic data can be a multi-row, two-column matrix, with the first column representing wind speed values ​​and the second column representing wind direction.

[0048] Alternatively, a preset encoding algorithm may be used to convert the format of the determined wind speed information to obtain the delayed characteristic data and the future characteristic data. For example, the preset encoding algorithm may be one-hot encoding.

[0049] S103: Determine the wind speed information at the moment when the wind speed is missing based on the lagged characteristic data and the future characteristic data.

[0050] Exemplarily, the lagged feature data and the future feature data are bidirectional time series features corresponding to the wind speed information at the moment of missing wind speed. The lagged feature data is a forward time series feature, and the future feature data is a backward time series feature. The lagged feature data and the future feature data are combined to determine the characteristics of the wind speed information before and after the moment of missing wind speed. Based on the characteristics of the wind speed information before and after the moment of missing wind speed, the wind speed information at the moment of missing wind speed is determined.

[0051] For example, if the lagged characteristic data indicates that there are four moments before the missing wind speed, with wind speeds of 2 m / s, 2 m / s, 2 m / s, and 1.9 m / s, respectively. Without determining future characteristic data, the wind speed at the missing wind speed moment would most likely be determined to be 2 m / s based solely on the lagged characteristic data. However, the future characteristic data indicates that there are four moments after the missing wind speed moment, with wind speeds of 1.9 m / s, 1.9 m / s, 1.9 m / s, and 1.9 m / s, respectively. By combining the lagged characteristic data with the future characteristic data, we can determine that the wind speed at the missing wind speed moment was 1.9 m / s.

[0052] In this embodiment, various combination strategies can be preset to combine the lagged characteristic data and the future characteristic data. For example, an average value can be calculated based on the lagged characteristic data and the future characteristic data to provide the wind speed information at the time when the wind speed is missing. Alternatively, a weighted calculation can be performed on the lagged characteristic data and the future characteristic data to obtain the wind speed information at the time when the wind speed is missing. In this embodiment, the preset combination strategy is not specifically limited.

[0053] In the disclosed embodiment, for a moment of missing wind speed, a period of time containing the missing wind speed moment can be determined as a target time interval. Wind speed information for the wind farm within the target time interval is obtained. Based on the wind speed information within the target time interval, lagged characteristic data and future characteristic data at the moment of missing wind speed are determined, thereby obtaining a bidirectional time series feature. Based on the lagged characteristic data and future characteristic data, combined with past and future time series features at the moment of missing wind speed, the dynamic changes in wind speed information are captured, thereby improving the accuracy of wind speed information determination.

[0054] Figure 2 A schematic flow chart of a method for determining wind speed based on a wind farm provided in an embodiment of the present disclosure is provided. This embodiment is an optional embodiment based on the above embodiment.

[0055] In this embodiment, the lagged characteristic data and future characteristic data of the moment when the wind speed is missing are determined based on the wind speed information of the wind farm within the target time interval, including: determining the lagged moment and future moment of the moment when the wind speed is missing from the target time interval based on the wind speed information of the wind farm within the target time interval; wherein the lagged moment represents a moment before the wind speed missing moment within the target time interval, and the future moment represents a moment after the wind speed missing moment within the target time interval; determining the lagged characteristic data based on the wind speed information corresponding to the lagged moment, and determining the future characteristic data based on the wind speed information corresponding to the future moment.

[0056] like Figure 2 As shown, the method includes the following steps:

[0057] S201. Determine a target time interval according to a wind speed missing moment, and obtain wind speed information of the wind farm within the target time interval; wherein the wind speed missing moment represents a moment when the wind speed is not monitored, and the wind speed missing moment is within the target time interval.

[0058] For example, this step may refer to the above-mentioned step S101 and will not be described in detail.

[0059] S202. Determine, based on the wind speed information of the wind farm within the target time interval, a lag time and a future time of the wind speed missing time within the target time interval; wherein the lag time represents a time before the wind speed missing time within the target time interval, and the future time represents a time after the wind speed missing time within the target time interval.

[0060] For example, wind speed information of the wind farm at multiple moments within the target time interval is obtained, and each wind speed information obtained corresponds to a moment. A wind speed missing moment is within the target time interval, but has no corresponding wind speed information.

[0061] The multiple acquired wind speed information is sorted from front to back according to the corresponding time. That is, the time corresponding to the wind speed information is sorted. From the sorted time, the time before the time when the wind speed is missing and the time after the time when the wind speed is missing are determined. The time before the time when the wind speed is missing is determined as the lag time, and the time after the time when the wind speed is missing is determined as the future time. There can be multiple lag times and future times. Both the lag time and the future time are the times when the wind speed information is monitored.

[0062] In this embodiment, based on the wind speed information of the wind farm within the target time interval, the lag moment and future moment of the wind speed missing moment are determined from the target time interval, including: based on the wind speed information of the wind farm within the target time interval, determining the moment when the wind speed is monitored before the wind speed missing moment, and determining the moment when the wind speed is monitored after the wind speed missing moment; determining the moment when the wind speed is monitored before the wind speed missing moment as the lag moment, and determining the moment when the wind speed is monitored after the wind speed missing moment as the future moment.

[0063] Specifically, based on the wind speed information of the wind farm within the target time interval, the time corresponding to each wind speed information is determined. The time corresponding to the wind speed information is compared with the time when the wind speed is missing, and the time when the wind speed information is monitored before the time when the wind speed is missing is determined, as well as the time when the wind speed is monitored after the time when the wind speed is missing.

[0064] The moment when the wind speed information is monitored before the moment when the wind speed is missing is determined as the hysteresis moment, and the moment when the wind speed information is monitored after the moment when the wind speed is missing is determined as the future moment. For example, all moments when the wind speed information is monitored before the moment when the wind speed is missing can be determined as the hysteresis moment, and all moments when the wind speed information is monitored after the moment when the wind speed is missing can be determined as the future moment; or one or more moments can be determined from all moments when the wind speed information is monitored before the moment when the wind speed is missing as the hysteresis moment, and one or more moments can be determined from all moments when the wind speed information is monitored after the moment when the wind speed is missing as the future moment.

[0065] Wind speeds at similar times exhibit strong autocorrelation, with the autocorrelation coefficient exceeding 80%. The closer the times are, the higher the autocorrelation. Therefore, the moments before and after the missing wind speed moment, and the moments closest to the missing wind speed moment, can be used as the lag moment and the future moment, respectively. That is, among all the moments corresponding to wind speed information, the moment before and adjacent to the missing wind speed moment is used as the lag moment; and among all the moments corresponding to wind speed information, the moment after and adjacent to the missing wind speed moment is used as the future moment. In this embodiment, multiple experiments have verified that extracting wind speed information at two moments before and after each other yields the best prediction results. Therefore, among all the moments corresponding to wind speed information, the two moments before and closest to the missing wind speed moment can be used as the lag moment; and the two moments after and closest to the missing wind speed moment can be used as the future moment. For example, if 10 rows of data are collected, each representing a moment in time, and rows 3-7 are missing, then the moments corresponding to rows 3-7 are also missing wind speed moments. For the missing wind speed moment in row 5, the lag moment is the moments corresponding to rows 1 and 2, and the future moment is the moments corresponding to rows 8 and 9. For the missing wind speed moment in row 6, the lag moment is the moments corresponding to rows 1 and 2, and the future moment is the moments corresponding to rows 8 and 9.

[0066] The beneficial effect of such a setting is that it can determine the moments before and after the moment when the wind speed is missing, thereby facilitating the acquisition of bidirectional time series features, capturing the dynamic changes of wind speed information, and improving the prediction accuracy of wind speed information.

[0067] S203: Determine the lag characteristic data according to the wind speed information corresponding to the lag moment, and determine the future characteristic data according to the wind speed information corresponding to the future moment.

[0068] For example, wind speed information corresponding to a delayed time is determined, and format conversion or feature extraction is performed on the wind speed information corresponding to the delayed time to obtain delayed feature data. Wind speed information corresponding to a future time is determined, and format conversion or feature extraction is performed on the wind speed information corresponding to the future time to obtain future feature data. In this embodiment, the calculation method for obtaining the delayed feature data and the future feature data can be the same. For example, a preset encoding algorithm can be used to obtain feature data in vector form.

[0069] In this embodiment, for the moment when wind speed is missing, the moments before and after the moment can be determined, thereby obtaining a bidirectional time series feature, which improves the accuracy of supplementing the missing wind speed, enables the wind farm to more accurately predict power generation, optimize power dispatching, reduce operation and maintenance costs, and significantly improve the economic benefits and market competitiveness of the wind farm.

[0070] S204: Determine the wind speed information at the time when the wind speed is missing based on the lagged characteristic data and the future characteristic data.

[0071] For example, this step may refer to the above-mentioned step S103 and will not be described in detail.

[0072] In the disclosed embodiment, for a moment of missing wind speed, a period of time containing the missing wind speed moment can be determined as a target time interval. Wind speed information for the wind farm within the target time interval is obtained. Based on the wind speed information within the target time interval, lagged characteristic data and future characteristic data at the moment of missing wind speed are determined, thereby obtaining a bidirectional time series feature. Based on the lagged characteristic data and future characteristic data, combined with past and future time series features at the moment of missing wind speed, the dynamic changes in wind speed information are captured, thereby improving the accuracy of wind speed information determination.

[0073] Figure 3 A schematic flow chart of a method for determining wind speed based on a wind farm provided in an embodiment of the present disclosure is provided. This embodiment is an optional embodiment based on the above embodiment.

[0074] In this embodiment, the wind speed information at the moment when the wind speed is missing is determined based on the lagged characteristic data and the future characteristic data, including: determining the first information at the moment when the wind speed is missing based on the lagged characteristic data and the future characteristic data; wherein the first information is the predicted wind speed information; determining the second information at the moment when the wind speed is missing based on the wind speed information at each moment in the target time interval; wherein the second information is the predicted wind speed information; determining the wind speed information at the moment when the wind speed is missing based on the first information and the second information.

[0075] like Figure 3 As shown, the method includes the following steps:

[0076] S301. Determine a target time interval according to a wind speed missing moment, and obtain wind speed information of the wind farm within the target time interval; wherein the wind speed missing moment represents a moment when the wind speed is not monitored, and the wind speed missing moment is within the target time interval.

[0077] For example, this step may refer to the above-mentioned step S101 and will not be described in detail.

[0078] S302. Determine, based on the wind speed information of the wind farm within the target time interval, the lagged characteristic data and the future characteristic data at the moment when the wind speed is missing; wherein the lagged characteristic data represents the wind speed information before the moment when the wind speed is missing within the target time interval, and the future characteristic data represents the wind speed information after the moment when the wind speed is missing within the target time interval.

[0079] For example, this step may refer to the above-mentioned step S102 and will not be described in detail.

[0080] S303: Determine first information at the moment when wind speed is missing based on the lagging characteristic data and the future characteristic data; wherein the first information is the predicted wind speed information.

[0081] Exemplarily, a regression model is pre-constructed and trained. The regression model in this embodiment is not a neural network model, but a machine learning model. This embodiment does not specifically limit the model category and model structure of the regression model. For example, Light GBM can be used as a regression model. The input data of the regression model are lagged feature data and future feature data, and the output data is wind speed information. The wind speed information output by the regression model is determined as the first information. That is, the lagged feature data and future feature data are input into the preset regression model to obtain the first information at the moment when the wind speed is missing. The input data of the regression model is bidirectional feature data. Therefore, the regression model can be a bidirectional short-term time series prediction model.

[0082] That is to say, the first information at the moment when the wind speed is missing can be determined based on the preset regression model according to the lagged characteristic data and the future characteristic data; wherein the preset regression model is a machine learning model, which is used to predict the missing wind speed information, and the first information is the wind speed information predicted by the preset regression model.

[0083] This embodiment can use Light GBM as the regression model. Light GBM is a gradient boosting framework based on decision trees, which is known for its efficient training speed, low memory usage and high accuracy. It is particularly suitable for processing large-scale time series data. The Leaf-wise growth strategy and histogram optimization algorithm in Light GBM can enable the regression model to converge to the optimal solution quickly, while automatically processing category features and missing values, simplifying preprocessing processes such as data encoding. Light GBM, combined with grid search and cross-validation, can find the optimal model parameters for each wind farm. For example, model parameters may include learning rate, tree depth, number of leaf nodes, etc. It ensures the stability and generalization ability of the model in different wind farm scenarios, improves the performance of the model, and provides a solid foundation for subsequent model deployment and application.

[0084] Wind speed information from multiple historical moments can be collected in advance to train the regression model. These historical moments can be long before the wind speed is lost. For these historical moments, wind speed information can be recorded every 15 minutes, generating 96 wind speed information records per day. In real-world wind farms, sensor failures can occur multiple times daily during data recording, and engineers typically respond in one to two hours. Therefore, to simulate this real-world scenario of missing data, five consecutive wind speed records can be discarded, effectively discarding one hour and fifteen minutes of data. Three random discards can occur each day, resulting in a total of 15 wind speed information records lost, representing a daily loss rate of 15.6%.

[0085] In this embodiment, 6 months of historical data from 11 wind farms can be selected as training data, and 1 month of historical data can be selected as test data. At the same time, in order to fully verify the effectiveness of the regression model, 8 months of data from 10 of the wind farms are additionally retrieved as supplementary test data to improve the accuracy of the regression model. During the training of the regression model, the objective function of the model training can be pre-set, and training can be performed based on the specified training data, model parameters, number of iterations, etc. In this embodiment, the training process of the regression model is not specifically limited. For example, for a certain historical moment, the moments when wind speed information is monitored before and after the historical moment are obtained as the lag moment and future moment of the historical moment, respectively. Based on the wind speed information corresponding to the lag moment, the lag feature data of the historical moment is obtained, and based on the wind speed information at the future moment, the future feature data of the historical moment is obtained. The actual wind speed information of the historical moment is used as a label, and the regression model is trained based on the label, the lag feature data, and the future feature data.

[0086] In this embodiment, first information at the moment when wind speed is missing is determined based on lagged characteristic data and future characteristic data, including: obtaining power information and meteorological information of the wind farm at the moment when wind speed is missing; wherein the power information represents the power generation situation of the wind farm, and the meteorological information represents the meteorological situation at a preset height in the area where the wind farm is located; determining power characteristic data based on the power information at the moment when wind speed is missing, and determining meteorological characteristic data based on the meteorological information at the moment when wind speed is missing; wherein the power characteristic data is data in vector form representing the power information, and the meteorological characteristic data is data in vector form representing the meteorological information; determining first information at the moment when wind speed is missing based on the lagged characteristic data, the future characteristic data, the power characteristic data and the meteorological characteristic data.

[0087] Specifically, the power and meteorological information of a wind farm can be monitored in real time. Power information represents the wind farm's power generation. For example, power information may include the wind farm's power ceiling, number of wind turbines, installed capacity, and power generation. Meteorological information represents the weather conditions at preset heights in the area where the wind farm is located. For example, the preset heights can be 10 meters, 30 meters, 50 meters, or 70 meters. Meteorological information can include temperature, humidity, and pressure at different heights. For example, meteorological information may include temperature at 10 meters, humidity at 10 meters, and high pressure at 10 meters.

[0088] After determining the moment when the wind speed is missing, the power information and meteorological information of the wind farm at the moment when the wind speed is missing can be obtained. The power information at the moment when the wind speed is missing is converted into a feature vector as power characteristic data; the meteorological information at the moment when the wind speed is missing is converted into a feature vector as meteorological characteristic data. In this embodiment, the method for obtaining the power characteristic data and meteorological characteristic data is not specifically limited and can be the same as the method for obtaining the lagged characteristic data and the future characteristic data. For example, for meteorological data, the temperature, humidity, and pressure at a height of 10 meters can be used as matrix elements, and the resulting matrix can be used as the meteorological characteristic data.

[0089] The lagged characteristic data, future characteristic data, power characteristic data, and meteorological characteristic data are all input into a preset regression model to obtain the first information at the missing wind speed moment. That is, when training the regression model, not only the wind speed information at the historical moment can be obtained, but also the power information and meteorological information at the historical moment can be obtained. If the power characteristic data and meteorological characteristic data corresponding to the historical moment are input when training the regression model, the power information and meteorological information at the missing wind speed moment must be obtained to determine the first information at the missing wind speed moment.

[0090] The beneficial effect of this setting is that by obtaining power information and meteorological information, multi-source heterogeneous data, such as power generation power, temperature, air pressure, etc., can be integrated to improve the completion accuracy of wind speed information.

[0091] In this embodiment, the power information includes the generated power and the preset power upper limit; the power characteristic data is determined based on the power information at the moment when the wind speed is missing, including: determining the power proportion based on the generated power at the moment when the wind speed is missing and the preset power upper limit; wherein the power proportion represents the ratio between the generated power and the preset power upper limit; the power proportion is binned to obtain the power characteristic data.

[0092] Specifically, the power information may include a preset power upper limit and real-time monitored generated power. The power characteristic data is determined based on the power information, that is, the power characteristic data may characterize relevant characteristics such as the generated power and the power upper limit.

[0093] In order to effectively quantify the impact of wind speed on power, the present embodiment can also determine the power proportion of the generated power, and perform binning on the power proportion to expand the effective features related to power. The power proportion represents the ratio between the generated power and the preset power upper limit, that is, the generated power at the moment when the wind speed is missing is divided by the preset power upper limit to obtain the power proportion at the moment when the wind speed is missing. The power proportion is binned to determine the discretization interval in which the power proportion is located, and the complete power characteristic data is determined based on the binning results and the power information. For example, the binning result of the power proportion can be used as a matrix element in the power characteristic data, so that the power characteristic data can characterize the relevant features of the power proportion.

[0094] The beneficial effect of such an arrangement is that, by determining the power proportion, the power characteristic data is expanded, and the accuracy of wind speed determination is further improved.

[0095] In this embodiment, the first information of the moment when the wind speed is missing is determined based on the lagging characteristic data, future characteristic data, power characteristic data and meteorological characteristic data, including: determining the category characteristic data corresponding to the moment when the wind speed is missing; wherein the category characteristic data represents the time interval in which the wind speed is missing; determining the first information of the moment when the wind speed is missing based on the lagging characteristic data, future characteristic data, power characteristic data, meteorological characteristic data, and category characteristic data.

[0096] Specifically, after determining the moment when the wind speed is missing, the time interval in which the wind speed is missing can be determined. For example, the time interval can be a dimension such as day, hour, minute, morning, noon, evening, early, mid-late, or day of the week. According to the time interval in which the wind speed is missing, the category feature data corresponding to the moment when the wind speed is missing is determined. That is, the time interval in which the wind speed is missing can be represented in the form of a matrix. In this embodiment, there is no specific limitation on the method of determining the category feature data. For example, one-hot encoding can be used to determine the category feature data to achieve feature expansion. Exemplarily, the time interval corresponding to the moment when the wind speed is missing is morning, and the corresponding category feature data can be [0].

[0097] The lagged characteristic data, future characteristic data, power characteristic data, meteorological characteristic data, and category characteristic data are all input into a preset regression model to obtain the first information at the moment when the wind speed is missing. That is, when training the regression model, not only the wind speed information at the historical moment can be obtained, but also the relevant information of the time interval in which the historical moment occurred. If the category characteristic data corresponding to the historical moment is input when training the regression model, it is necessary to determine the category characteristic data corresponding to the moment when the wind speed is missing to determine the first information at the moment when the wind speed is missing.

[0098] The beneficial effect of such a setting is that, by determining the category feature information, the influence of time on wind speed can be taken into account, multi-source data can be integrated, and the accuracy of wind speed determination can be improved.

[0099] In this embodiment, the moment of the wind speed information corresponding to the lag characteristic data is the lag moment, and the moment of the wind speed information corresponding to the future characteristic data is the future moment; the method also includes: obtaining the power information of the wind farm at the lag moment and the power information at the future moment, and obtaining the meteorological information of the wind farm at the lag moment and the meteorological information at the future moment; determining the power characteristic data based on the power information at the lag moment and the power information at the future moment, and determining the meteorological characteristic data based on the meteorological information at the lag moment and the meteorological information at the future moment.

[0100] Specifically, for the moment when wind speed is missing, the power information, meteorological information, etc. at past and future moments will also have a great impact on the current wind speed. Therefore, in this embodiment, the bidirectional time series characteristics of power information and meteorological information can also be determined to determine the wind speed information at the moment when wind speed is missing.

[0101] First, determine the lag time and future time of the missing wind speed. For power information, the power information at the lag time and the power information at the future time can be obtained. Based on the power information at the lag time and the power information at the future time, power characteristic data can be determined. That is, the power characteristic data can represent the two-way power information at the lag time and the future time. For meteorological information, the meteorological information at the lag time and the meteorological information at the future time can be obtained. Based on the meteorological information at the lag time and the meteorological information at the future time, meteorological characteristic data can be determined. That is, the meteorological characteristic data can represent the two-way meteorological information at the lag time and the future time. For example, the meteorological information at the lag time and the meteorological information at the future time can both be used as matrix element values ​​in the meteorological characteristic data.

[0102] Then, the lagged characteristic data, future characteristic data, power characteristic data, and meteorological characteristic data are input into a preset regression model to determine the first information at the time when the wind speed is missing. Alternatively, the lagged characteristic data, future characteristic data, power characteristic data, meteorological characteristic data, and category characteristic data are input into a preset regression model to determine the first information at the time when the wind speed is missing.

[0103] The beneficial effect of this arrangement is that by determining the bidirectional time series characteristics of power information and meteorological information, the influence of power information and meteorological information at past and future times on wind speed can be taken into account, thereby improving the accuracy of wind speed determination.

[0104] S304. Determine second information at the moment when the wind speed is missing based on the wind speed information at each moment within the target time interval; wherein the second information is the predicted wind speed information.

[0105] For example, a statistical algorithm is pre-set, such as a linear interpolation algorithm. Based on the wind speed information at each time point within the target time interval, wind speed information is calculated using a traditional linear interpolation algorithm, and this wind speed information is used as the second information. In this embodiment, the calculation process of the linear interpolation algorithm is not specifically limited.

[0106] That is, according to the wind speed information at each moment in the target time interval, based on a preset statistical algorithm, the second information at the moment when the wind speed is missing is determined; wherein the second information is the wind speed information predicted by the preset statistical algorithm.

[0107] Due to the strong autocorrelation between wind speeds before and after the moment, traditional linear interpolation methods can also achieve accurate predictions. Specifically, a linear function can be used to interpolate the non-missing values ​​before and after the missing wind speed moment. This method is simple and efficient, making it particularly suitable for filling in short-term missing data. Although linear interpolation may be limited in its effectiveness when dealing with long-term missing data, it can provide a relatively reliable baseline and play an important role in subsequent fusion.

[0108] S305: Determine the wind speed information at the time when the wind speed is missing based on the first information and the second information.

[0109] For example, the wind speed information determined by the two algorithms is combined, that is, the first information and the second information are combined to obtain the wind speed information at the moment when the wind speed is missing. For example, the average of the first information and the second information can be calculated as the final wind speed information.

[0110] This embodiment effectively combines the advantages of both algorithms by using linear interpolation and a bidirectional short-term time series regression model, improving prediction accuracy and enhancing the generalization and stability of wind speed determination scenarios. This technical advantage enables wind farms to more accurately predict power generation, optimize power dispatch, reduce operation and maintenance costs, and significantly enhance the economic benefits and market competitiveness of wind farms.

[0111] In this embodiment, the wind speed information at the moment when the wind speed is missing is determined based on the first information and the second information, including: determining a first weight corresponding to the first information and a second weight corresponding to the second information; wherein the first weight represents the prediction accuracy of the first information, and the second weight represents the prediction accuracy of the second information; according to the first weight and the second weight, the first information and the second information are linearly weighted to obtain the wind speed information at the moment when the wind speed is missing.

[0112] Specifically, the regression model and statistical algorithm each have their own corresponding weights, which represent the accuracy of their wind speed prediction. The weight corresponding to the regression model is the first weight, and the weight corresponding to the statistical algorithm is the second weight.

[0113] The first information and the second information are weighted according to the first weight and the second weight to obtain the final wind speed information. For example, the first information and the second information can be linearly weighted fused according to the first weight and the second weight to obtain the wind speed information at the time when the wind speed is missing.

[0114] The beneficial effect of this setup is that it fuses the information obtained from the two algorithms through predetermined weights, improving the stability and accuracy of the forecast. This fusion strategy effectively combines the strengths of both algorithms, and excels particularly when dealing with complex wind speed variations.

[0115] In this embodiment, determining a first weight corresponding to the first information and a second weight corresponding to the second information includes: obtaining a historical data set; wherein the historical data set includes wind speed information at a historical moment, wind speed information before the historical moment, and wind speed information after the historical moment; the historical moment is earlier than the moment when the wind speed is missing; determining a historical lag feature based on the wind speed information before the historical moment, and determining a historical future feature based on the wind speed information after the historical moment; wherein the historical lag feature represents the wind speed information before the historical moment, and the historical future feature represents the wind speed information after the historical moment; according to the historical lag feature, the historical future feature, and the wind speed information at the historical moment, obtaining a first weight corresponding to the first information and a second weight corresponding to the second information.

[0116] Specifically, the first weight and the second weight are pre-determined parameters, which can be directly used when predicting wind speed. The first weight and the second weight can be determined based on wind speed information collected at historical moments.

[0117] When determining the first and second weights, a historical data set can be first obtained. The historical data set may include wind speed information at the historical moment, wind speed information at moments before the historical moment, and wind speed information at moments after the historical moment. There may be multiple moments before the historical moment, and there may also be multiple moments after the historical moment. The historical moment may be much earlier than the moment when the wind speed is missing, so that the moment after the historical moment is also earlier than the moment when the wind speed is missing. The moment before the historical moment is the lag moment of the historical moment, and the moment after the historical moment is the future moment of the historical moment.

[0118] The process for determining wind speed information at the moment of missing wind speed is similar to that for determining lag feature data based on wind speed information at a time after the historical moment, and using this lag feature data as the historical lag feature. Furthermore, future feature data is determined based on wind speed information at a time after the historical moment, and using this future feature data as the historical future feature. Specifically, the historical lag feature represents wind speed information before the historical moment, and the historical future feature represents wind speed information after the historical moment.

[0119] The actual wind speed information collected at a historical moment is used as a label. Based on the historical lag characteristics and the historical future characteristics, a first weight corresponding to the regression model and a second weight corresponding to the statistical algorithm are determined. For example, the first and second weights can be determined to determine when the wind speed information determined using the historical lag characteristics and the historical future characteristics is consistent with the label.

[0120] The beneficial effect of such a setting is that two weights are predetermined based on historical data, which facilitates subsequent calculations directly based on the weights and improves the efficiency of determining wind speed information.

[0121] In this embodiment, based on the historical lag characteristics, historical future characteristics, and wind speed information at the historical moment, a first weight corresponding to the first information and a second weight corresponding to the second information are obtained, including: determining the third information of the historical moment based on the historical lag characteristics and the historical future characteristics; wherein the third information is the determined wind speed information of the historical moment; determining the fourth information of the historical moment based on the wind speed information at each moment in the historical data set; wherein the fourth information is the determined wind speed information at the historical moment; obtaining the first weight corresponding to the first information and the second weight corresponding to the second information based on the third information, the fourth information, and the wind speed information at the historical moment.

[0122] Specifically, after obtaining the historical lag features and the historical future features, the historical lag features and the historical future features can be input into a preset regression model to obtain wind speed information output by the regression model. The wind speed information output by the regression model is determined as the third information.

[0123] Obtain wind speed information at each time in the historical data set, input the wind speed information other than the wind speed information at the historical time into a preset statistical algorithm, obtain wind speed information output by the statistical algorithm, and determine the wind speed information as the fourth information. For example, the fourth information corresponding to the historical time can be calculated using a linear interpolation algorithm based on wind speed information before and after the historical time.

[0124] The third and fourth information are fused, using the actual wind speed information at the historical moment as a label to determine the first and second weights for the fusion process. Specifically, the first and second weights are determined to determine when the third and fourth information can be fused into the actual wind speed information.

[0125] That is to say, according to the historical lag characteristics and historical future characteristics, based on the preset regression model, the third information of the historical moment is determined; wherein, the third information is the wind speed information of the historical moment determined by the preset regression model; according to the wind speed information at each moment in the historical data set, based on the preset statistical algorithm, the fourth information of the historical moment is determined; wherein, the fourth information is the wind speed information of the historical moment determined by the preset statistical algorithm; according to the third information, the fourth information, and the wind speed information at the historical moment, the first weight corresponding to the preset regression model and the second weight corresponding to the preset statistical algorithm are obtained.

[0126] The beneficial effect of this setting is that the weight parameters can be continuously adjusted through a large amount of historical data to find the optimal fusion ratio, effectively combining the advantages of the two algorithms and improving the stability and accuracy of the prediction.

[0127] In this embodiment, based on the third information, the fourth information, and the wind speed information at the historical moment, the first weight corresponding to the first information and the second weight corresponding to the second information are obtained, including: based on the wind speed information at the historical moment, linear fitting training is performed on the third information and the fourth information to obtain the first weight corresponding to the first information and the second weight corresponding to the second information.

[0128] Specifically, a Stacking fusion model can be used to perform linear weighted fusion on the calculation results of the regression model and the statistical algorithm to obtain a first weight and a second weight. That is, the actual wind speed information at a historical moment can be used as a label, and the Stacking fusion algorithm can be used to perform linear fitting training on multiple groups of third information and fourth information to obtain the optimal fusion weights, namely the first weight and the second weight. In subsequent predictions, the weights obtained by fitting are directly used to perform linear weighted fusion to obtain the final wind speed information. The Stacking fusion model is an integrated learning method used to improve the prediction performance of wind speed. In this embodiment, the linear fitting training process of the Stacking model is not specifically limited.

[0129] The beneficial effect of this setting is that it adopts the Stacking fusion model, adjusts the weight parameters, finds the optimal fusion ratio, effectively combines the advantages of the two algorithms, and improves the prediction accuracy of wind speed information.

[0130] Figure 4 Schematic diagram of the wind speed information determination process. Figure 4 After determining the missing wind speed moment, basic wind farm information, power information for the target time interval, meteorological information for the target time interval, and wind speed information before and after the missing wind speed moment can be obtained. Basic wind farm information may include site name, longitude, and latitude; power information may include power ceiling, number of wind turbines, installed capacity, and generated power; meteorological information may include temperature, humidity, and pressure at a height of 10 meters; and wind speed information may include wind speed and direction at a height of 10 meters, wind speed and direction at a height of 30 meters, wind speed and direction at a height of 50 meters, and wind speed and direction at a height of 70 meters.

[0131] Feature extraction is performed on the acquired information to obtain time-related category feature data, power-related power feature data, and meteorological feature data related to weather. Category feature data can represent dimensions such as the day, hour, minute, morning, noon, evening, early, mid-to-late month, and day of the week corresponding to the moment when wind speed is missing. Power feature data mainly includes information such as the power upper limit, number of wind turbines, installed capacity, and power generation. To effectively quantify the impact of wind speed on power, the proportion of power generation can also be calculated and binned to expand the effective power-related features. Meteorological feature data can include temperature, humidity, and pressure at a height of 10 meters.

[0132] For the remaining wind speed information before and after the missing wind speed moment, the lagged characteristic data and future characteristic data at the missing wind speed moment can be determined. For power characteristic data, the lagged power characteristic data and future power characteristic data can be obtained based on the power information before and after the missing wind speed moment. For meteorological characteristic data, the lagged meteorological characteristic data and future meteorological characteristic data can be obtained based on the meteorological information before and after the missing wind speed moment.

[0133] All acquired feature data is fed into a bidirectional time series prediction model, typically a pre-defined regression model, to obtain the first information. The remaining wind speed information before and after the missing wind speed moment is fed into a linear interpolation model, where it is used to obtain the second information. A stacking model is used to perform a linear weighted fusion of the first and second information to obtain the wind speed information at the missing moment, thus completing the missing wind speed.

[0134] In the disclosed embodiment, for a moment of missing wind speed, a period of time containing the missing wind speed moment can be determined as a target time interval. Wind speed information for the wind farm within the target time interval is obtained. Based on the wind speed information within the target time interval, lagged characteristic data and future characteristic data at the moment of missing wind speed are determined, thereby obtaining a bidirectional time series feature. Based on the lagged characteristic data and future characteristic data, combined with past and future time series features at the moment of missing wind speed, the dynamic changes in wind speed information are captured, thereby improving the accuracy of wind speed information determination.

[0135] Figure 5 This is a structural block diagram of a wind speed determination device based on a wind farm provided by an embodiment of the present disclosure. For ease of explanation, only the parts related to the embodiment of the present disclosure are shown. Figure 5 The wind speed determination device 500 based on a wind farm includes: an information acquisition unit 501 , a feature determination unit 502 and a wind speed determination unit 503 .

[0136] An information acquisition unit 501 is configured to determine a target time interval based on a wind speed missing moment, and acquire wind speed information of the wind farm within the target time interval; wherein the wind speed missing moment represents a moment when no wind speed is monitored, and the wind speed missing moment is within the target time interval;

[0137] The feature determination unit 502 is configured to determine, based on the wind speed information of the wind farm in the target time interval, lagged feature data and future feature data at the moment when the wind speed is missing; wherein the lagged feature data represents the wind speed information before the moment when the wind speed is missing in the target time interval, and the future feature data represents the wind speed information after the moment when the wind speed is missing in the target time interval;

[0138] The wind speed determining unit 503 is configured to determine the wind speed information at the moment when the wind speed is missing based on the delayed characteristic data and the future characteristic data.

[0139] Figure 6 A structural block diagram of a wind speed determination device based on a wind farm provided in an embodiment of the present disclosure is shown in FIG. Figure 6 As shown, the wind speed determination device 600 based on a wind farm includes an information acquisition unit 601 , a feature determination unit 602 and a wind speed determination unit 603 , wherein the feature determination unit 602 includes a time determination module 6021 and a feature determination module 6022 .

[0140] a time determination module 6021 configured to determine, based on the wind speed information of the wind farm within the target time interval, a lag time and a future time of the wind speed missing time within the target time interval; wherein the lag time represents a time before the wind speed missing time within the target time interval, and the future time represents a time after the wind speed missing time within the target time interval;

[0141] The feature determination module 6022 is configured to determine the delayed feature data based on the wind speed information corresponding to the delayed moment, and to determine the future feature data based on the wind speed information corresponding to the future moment.

[0142] In one example, the time determination module 6021 includes:

[0143] A first determining submodule is configured to determine, based on the wind speed information of the wind farm within the target time interval, a time when the wind speed is monitored before the wind speed missing time, and a time when the wind speed is monitored after the wind speed missing time;

[0144] The second determining submodule is configured to determine the time when the wind speed is monitored before the wind speed missing time as the lag time, and determine the time when the wind speed is monitored after the wind speed missing time as the future time.

[0145] In one example, the wind speed determination unit 603 includes:

[0146] A first determining module is configured to determine first information of the wind speed missing moment based on the lag characteristic data and the future characteristic data; wherein the first information is the predicted wind speed information;

[0147] A second determining module is configured to determine second information of the wind speed missing moment based on the wind speed information at each moment within the target time interval; wherein the second information is the predicted wind speed information;

[0148] A wind speed determination module is used to determine the wind speed information at the moment when the wind speed is missing based on the first information and the second information.

[0149] In one example, the first determining module includes:

[0150] an information acquisition submodule, configured to acquire power information and meteorological information of the wind farm at the moment when the wind speed is missing; wherein the power information represents the power generation of the wind farm, and the meteorological information represents the meteorological conditions at a preset height in the area where the wind farm is located;

[0151] a feature extraction submodule, configured to determine power characteristic data based on the power information at the time when the wind speed is missing, and to determine meteorological characteristic data based on the meteorological information at the time when the wind speed is missing; wherein the power characteristic data is data in the form of a vector representing the power information, and the meteorological characteristic data is data in the form of a vector representing the meteorological information;

[0152] The regression submodule is used to determine the first information of the wind speed missing moment based on the lag characteristic data, the future characteristic data, the power characteristic data and the meteorological characteristic data.

[0153] In one example, the power information includes the generated power and a preset power upper limit; the feature extraction submodule is specifically used to:

[0154] Determine the power proportion according to the generated power at the moment of wind speed loss and the preset power upper limit; wherein the power proportion represents the ratio between the generated power and the preset power upper limit;

[0155] The power proportions are binned to obtain the power characteristic data.

[0156] In one example, the regression submodule is used to:

[0157] Determining the category characteristic data corresponding to the moment when the wind speed is missing; wherein the category characteristic data represents the time interval in which the wind speed is missing;

[0158] The first information of the wind speed missing moment is determined according to the hysteresis characteristic data, the future characteristic data, the power characteristic data, the meteorological characteristic data, and the category characteristic data.

[0159] In one example, the time of the wind speed information corresponding to the hysteresis characteristic data is the hysteresis time, and the time of the wind speed information corresponding to the future characteristic data is the future time; the first determining module further includes:

[0160] a bidirectional acquisition submodule, configured to acquire power information of the wind farm at the lag moment and power information at the future moment, and to acquire meteorological information of the wind farm at the lag moment and meteorological information at the future moment;

[0161] The bidirectional extraction submodule is used to determine power characteristic data based on the power information at the lag moment and the power information at the future moment, and to determine meteorological characteristic data based on the meteorological information at the lag moment and the meteorological information at the future moment.

[0162] In one example, the wind speed determination module includes:

[0163] A weight determination submodule, configured to determine a first weight corresponding to the first information and a second weight corresponding to the second information; wherein the first weight represents the prediction accuracy of the first information, and the second weight represents the prediction accuracy of the second information;

[0164] The weighted processing submodule is used to perform linear weighted processing on the first information and the second information according to the first weight and the second weight to obtain the wind speed information at the moment when the wind speed is missing.

[0165] In one example, the weight determination submodule is specifically used to:

[0166] Acquire a historical data set; wherein the historical data set includes wind speed information at a historical moment, wind speed information before the historical moment, and wind speed information after the historical moment; the historical moment is earlier than the moment when the wind speed is missing;

[0167] Determining a historical lag feature based on wind speed information before the historical moment, and determining a historical future feature based on wind speed information after the historical moment; wherein the historical lag feature represents wind speed information before the historical moment, and the historical future feature represents wind speed information after the historical moment;

[0168] According to the historical lag feature, the historical future feature, and the wind speed information at the historical moment, a first weight corresponding to the first information and a second weight corresponding to the second information are obtained.

[0169] In one example, the weight determination submodule is specifically used to:

[0170] Determining third information of the historical moment according to the historical lag feature and the historical future feature; wherein the third information is the wind speed information of the determined historical moment;

[0171] Determining fourth information at the historical moment based on the wind speed information at each moment in the historical data set; wherein the fourth information is the determined wind speed information at the historical moment;

[0172] According to the third information, the fourth information, and the wind speed information at the historical moment, a first weight corresponding to the first information and a second weight corresponding to the second information are obtained.

[0173] In one example, the weight determination submodule is specifically used to:

[0174] According to the wind speed information at the historical moment, linear fitting training is performed on the third information and the fourth information to obtain a first weight corresponding to the first information and a second weight corresponding to the second information.

[0175] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device.

[0176] Figure 7 A structural block diagram of an electronic device provided in an embodiment of the present disclosure, such as Figure 7 As shown, the electronic device 700 includes: at least one processor 702; and a memory 701 communicatively connected to the at least one processor 702; wherein the memory stores instructions that can be executed by the at least one processor 702, and the instructions are executed by the at least one processor 702 to enable the at least one processor 702 to execute the wind speed determination method based on a wind farm disclosed in the present invention.

[0177] The electronic device 700 further includes a receiver 703 and a transmitter 704. The receiver 703 is used to receive instructions and data sent by other devices, and the transmitter 704 is used to send instructions and data to external devices.

[0178] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0179] According to an embodiment of the present disclosure, the present disclosure also provides a computer program product, which includes: a computer program, the computer program is stored in a readable storage medium, at least one processor of an electronic device can read the computer program from the readable storage medium, and at least one processor executes the computer program so that the electronic device executes the solution provided by any of the above embodiments.

[0180] Figure 8 A schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are provided as examples only and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0181] like Figure 8As shown, the device 800 includes a computing unit 801, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage unit 808 into a random access memory (RAM) 803. Various programs and data required for the operation of the device 800 can also be stored in the RAM 803. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0182] Various components in device 800 are connected to I / O interface 805, including an input unit 806, such as a keyboard, mouse, etc.; an output unit 807, such as various types of displays, speakers, etc.; a storage unit 808, such as a magnetic disk, optical disk, etc.; and a communication unit 809, such as a network card, modem, wireless communication transceiver, etc. The communication unit 809 allows device 800 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0183] The computing unit 801 can be various general-purpose and / or specialized processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as the wind farm-based wind speed determination method. For example, in some embodiments, the wind farm-based wind speed determination method can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed on the device 800 via the ROM 802 and / or the communication unit 809. When the computer program is loaded into the RAM 803 and executed by the computing unit 801, one or more steps of the wind farm-based wind speed determination method described above can be performed. Alternatively, in other embodiments, the computing unit 801 may be configured to execute the wind speed determination method based on a wind farm in any other appropriate manner (for example, by means of firmware).

[0184] Various embodiments of the systems and techniques described above can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0185] The program code for implementing the method of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device so that when the program code is executed by the processor or controller, the functions / operations specified in the flow chart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0186] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in conjunction with an instruction execution system, device or equipment. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium can include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0187] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0188] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0189] A computer system may include a client and a server. The client and server are generally remote from each other and typically interact via a communication network. This client-server relationship is established by computer programs running on the respective computers and establishing a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, a host product within a cloud computing service ecosystem that addresses the management difficulties and limited business scalability of traditional physical hosts and VPS services ("Virtual Private Servers" or simply "VPS"). The server may also be a server in a distributed system or a server integrated with blockchain.

[0190] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved. This is not a limitation herein.

[0191] The above specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.

Claims

1. A method for determining wind speed based on a wind farm, comprising: Determine a target time interval according to the wind speed missing moment, and obtain wind speed information of the wind farm within the target time interval; wherein the wind speed missing moment represents a moment when the wind speed is not monitored, and the wind speed missing moment is within the target time interval; Determining, based on the wind speed information of the wind farm in the target time interval, lagged characteristic data and future characteristic data at the moment when the wind speed is missing; wherein the lagged characteristic data represents the wind speed information before the moment when the wind speed is missing in the target time interval, and the future characteristic data represents the wind speed information after the moment when the wind speed is missing in the target time interval; The wind speed information at the moment when the wind speed is missing is determined according to the delayed characteristic data and the future characteristic data.

2. The method according to claim 1, wherein The determining, based on the wind speed information of the wind farm in the target time interval, the lagged characteristic data and the future characteristic data at the moment when the wind speed is missing includes: Determining, based on the wind speed information of the wind farm within the target time interval, a lag time and a future time of the wind speed missing time within the target time interval; wherein the lag time represents a time before the wind speed missing time within the target time interval, and the future time represents a time after the wind speed missing time within the target time interval; The hysteresis characteristic data is determined according to the wind speed information corresponding to the hysteresis moment, and the future characteristic data is determined according to the wind speed information corresponding to the future moment.

3. The method according to claim 2, wherein: The determining, based on the wind speed information of the wind farm in the target time interval, the lag time and the future time of the wind speed missing time in the target time interval includes: determining, based on the wind speed information of the wind farm within the target time interval, a time when the wind speed is monitored before the wind speed missing time, and determining a time when the wind speed is monitored after the wind speed missing time; The time when the wind speed is monitored before the wind speed missing time is determined as the lag time, and the time when the wind speed is monitored after the wind speed missing time is determined as the future time.

4. The method according to any one of claims 1 to 3, wherein The determining, based on the hysteresis characteristic data and the future characteristic data, the wind speed information at the moment when the wind speed is missing includes: Determining first information of the wind speed missing moment according to the hysteresis characteristic data and the future characteristic data; wherein the first information is the predicted wind speed information; Determining second information of the wind speed missing moment according to the wind speed information at each moment within the target time interval; wherein the second information is the predicted wind speed information; The wind speed information at the moment when the wind speed is missing is determined according to the first information and the second information.

5. The method according to claim 4, wherein The determining, based on the hysteresis characteristic data and the future characteristic data, the first information of the wind speed missing moment includes: Obtaining power information and meteorological information of the wind farm at the moment when the wind speed is missing; wherein the power information represents the power generation situation of the wind farm, and the meteorological information represents the meteorological situation at a preset height in the area where the wind farm is located; Determining power characteristic data based on the power information at the time when the wind speed is missing, and determining meteorological characteristic data based on the meteorological information at the time when the wind speed is missing; wherein the power characteristic data is data in a vector form representing the power information, and the meteorological characteristic data is data in a vector form representing the meteorological information; The first information of the wind speed missing moment is determined according to the hysteresis characteristic data, the future characteristic data, the power characteristic data and the meteorological characteristic data.

6. The method according to claim 5, wherein: The power information includes the generated power and the preset power upper limit; The determining of power characteristic data according to the power information at the moment when the wind speed is missing includes: Determine the power proportion according to the generated power at the moment of wind speed loss and the preset power upper limit; wherein the power proportion represents the ratio between the generated power and the preset power upper limit; The power proportions are binned to obtain the power characteristic data.

7. The method according to claim 5 or 6, wherein: The determining, based on the hysteresis characteristic data, the future characteristic data, the power characteristic data, and the meteorological characteristic data, the first information of the wind speed missing moment includes: Determining the category characteristic data corresponding to the moment when the wind speed is missing; wherein the category characteristic data represents the time interval in which the wind speed is missing; The first information of the wind speed missing moment is determined according to the hysteresis characteristic data, the future characteristic data, the power characteristic data, the meteorological characteristic data, and the category characteristic data.

8. The method according to any one of claims 4 to 7, wherein The time of the wind speed information corresponding to the hysteresis characteristic data is the hysteresis time, and the time of the wind speed information corresponding to the future characteristic data is the future time; the method further includes: Acquiring power information of the wind farm at the lag time and power information at the future time, and acquiring meteorological information of the wind farm at the lag time and meteorological information at the future time; Power characteristic data is determined based on the power information at the delayed moment and the power information at the future moment, and meteorological characteristic data is determined based on the meteorological information at the delayed moment and the meteorological information at the future moment.

9. The method according to any one of claims 4 to 8, wherein The determining, based on the first information and the second information, the wind speed information at the moment when the wind speed is missing includes: Determining a first weight corresponding to the first information and a second weight corresponding to the second information; wherein the first weight represents the prediction accuracy of the first information, and the second weight represents the prediction accuracy of the second information; According to the first weight and the second weight, linear weighted processing is performed on the first information and the second information to obtain the wind speed information at the moment when the wind speed is missing.

10. The method according to claim 9, wherein: The determining a first weight corresponding to the first information and a second weight corresponding to the second information includes: Acquire a historical data set; wherein the historical data set includes wind speed information at a historical moment, wind speed information before the historical moment, and wind speed information after the historical moment; the historical moment is earlier than the moment when the wind speed is missing; Determining a historical lag feature based on wind speed information before the historical moment, and determining a historical future feature based on wind speed information after the historical moment; wherein the historical lag feature represents wind speed information before the historical moment, and the historical future feature represents wind speed information after the historical moment; According to the historical lag feature, the historical future feature, and the wind speed information at the historical moment, a first weight corresponding to the first information and a second weight corresponding to the second information are obtained.

11. The method according to claim 10, wherein: The obtaining, according to the historical hysteresis feature, the historical future feature, and the wind speed information at the historical moment, a first weight corresponding to the first information and a second weight corresponding to the second information includes: Determining third information of the historical moment according to the historical lag feature and the historical future feature; wherein the third information is the wind speed information of the determined historical moment; Determining fourth information at the historical moment based on the wind speed information at each moment in the historical data set; wherein the fourth information is the determined wind speed information at the historical moment; A first weight corresponding to the first information and a second weight corresponding to the second information are obtained according to the third information, the fourth information, and the wind speed information at the historical moment.

12. The method according to claim 11, wherein Obtaining, according to the third information, the fourth information, and the wind speed information at the historical moment, a first weight corresponding to the first information and a second weight corresponding to the second information, including: According to the wind speed information at the historical moment, linear fitting training is performed on the third information and the fourth information to obtain a first weight corresponding to the first information and a second weight corresponding to the second information.

13. A wind speed determination device based on a wind farm, comprising: an information acquisition unit, configured to determine a target time interval based on a wind speed missing moment, and acquire wind speed information of the wind farm within the target time interval; wherein the wind speed missing moment represents a moment when no wind speed is monitored, and the wind speed missing moment is within the target time interval; a feature determination unit, configured to determine, based on the wind speed information of the wind farm in the target time interval, lagged feature data and future feature data at the moment when the wind speed is missing; wherein the lagged feature data represents the wind speed information before the moment when the wind speed is missing in the target time interval, and the future feature data represents the wind speed information after the moment when the wind speed is missing in the target time interval; A wind speed determination unit is used to determine the wind speed information at the moment when the wind speed is missing based on the lag characteristic data and the future characteristic data.

14. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 12.

15. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-12.

16. A computer program product, wherein The invention comprises a computer program, which implements the steps of the method according to any one of claims 1 to 12 when the computer program is executed by a processor.

Images