Prediction method and device for wind farm power generation data based on spatiotemporal correlation

Abstract

The application provides a wind farm power generation data prediction method and device based on space-time correlation. According to historical actual power generation data of each wind turbine in a wind farm and historical actual wind speed of each wind turbine in the wind farm, the wind turbines in the wind farm are divided into multiple wind turbine clusters. For any wind turbine cluster, according to historical actual power generation data of each wind turbine in the wind turbine cluster and historical actual total power generation data of the wind turbine cluster, a reference wind turbine in the wind turbine cluster is determined. According to historical actual total power generation data of the wind farm, current actual total power generation data of the wind farm in an upwind direction of the wind farm and current actual power generation data of the reference wind turbine, future expected total power generation data of the wind farm is predicted. In this way, the accuracy of the predicted future expected total power generation data of the wind farm is higher.

Description

Technical Field

[0001] This application relates to the field of wind energy technology, and in particular to a method and apparatus for predicting wind farm power generation data based on spatiotemporal correlation. Background Technology

[0002] As a clean and renewable energy source, wind energy plays an important role in alleviating the shortage of traditional energy sources and reducing environmental pollution. Therefore, the development and utilization of wind power has received increasing attention.

[0003] Developing wind power requires the use of wind power technology, which refers to the technology of generating electricity using wind energy. Due to the cleanliness and renewability of wind energy, wind power accounts for an increasingly larger proportion of the power system.

[0004] In wind power development scenarios, wind farms are typically used, which contain multiple wind turbines that generate electricity.

[0005] However, wind farm power generation data is affected by a variety of factors, exhibiting strong randomness, intermittency, and volatility, posing a significant challenge to the stable operation of the power grid and power dispatch.

[0006] Therefore, in this context, accurate forecasting of wind farm power generation data becomes crucial. For example, using forecasted wind farm power generation data has significant guiding implications for long-term power system dispatch planning, electricity market trading strategies, and wind farm operation and maintenance plans. This can optimize power resource allocation, improve grid efficiency and reliability, reduce economic losses caused by power fluctuations, and further enhance the market competitiveness of wind power. Summary of the Invention

[0007] This application discloses a method and apparatus for predicting wind farm power generation data based on spatiotemporal correlation.

[0008] In a first aspect, this application discloses a prediction method for wind farm power generation data based on spatiotemporal correlation, the method comprising:

[0009] Obtain historical actual power generation data for each wind turbine in the target wind farm, and obtain historical actual wind speed for each wind turbine in the target wind farm;

[0010] Based on the historical actual power generation data and historical actual wind speed of each wind turbine in the target wind farm, the wind turbines in the target wind farm are divided into multiple wind turbine clusters, and each wind turbine cluster includes more than two wind turbines.

[0011] For any wind turbine cluster, a benchmark wind turbine is determined within the wind turbine cluster based on the historical actual power generation data of each wind turbine in the wind turbine cluster and the total historical actual power generation data of the wind turbine cluster.

[0012] Obtain the current actual power generation data of the benchmark wind turbines in each wind turbine cluster, and obtain the current total actual power generation data of the wind farms upwind of the target wind farm; obtain the historical total actual power generation data of the target wind farm based on the historical actual power generation data of each wind turbine in the target wind farm.

[0013] Based on the historical total power generation data of the target wind farm, the current total power generation data of the wind farm upwind of the target wind farm, and the current total power generation data of the benchmark wind turbine, predict the future expected total power generation data of the target wind farm.

[0014] Secondly, this application discloses a method for training a prediction model based on spatiotemporal correlation wind farm power generation data, the method comprising:

[0015] Obtain the training dataset, which includes multiple training data sets, including sample data and labeled data. The sample data includes the total power generation data for the first historical time period of the sample wind farm, the total power generation data for the second historical time period of the wind farm upwind of the sample wind farm, and the sample power generation data for the second historical time period of the benchmark wind turbines within multiple wind turbine clusters in the sample wind farm. The labeled data includes the actual total power generation data for the third historical time period of the sample wind farm. The first historical time period is earlier than the second historical time period, and the second historical time period is earlier than the third historical time period. Each wind turbine cluster includes two or more wind turbines. A wind turbine cluster is obtained by dividing the wind turbines in a sample wind farm into clusters based on the sample power generation data and sample wind speed of each wind turbine in the sample wind farm during the first historical time period. For any wind turbine cluster, the reference wind turbine within the cluster is determined based on the sample power generation data of each wind turbine in the cluster during the first historical time period and the total sample power generation data of the cluster during the first historical time period. The historical actual total power generation data of the target wind farm is obtained based on the historical actual power generation data of each wind turbine in the target wind farm.

[0016] The training model is trained on the training dataset until it converges, thus obtaining a prediction model for wind farm power generation data based on spatiotemporal correlation.

[0017] Thirdly, this application discloses a wind farm power generation data prediction device based on spatiotemporal correlation, the device comprising:

[0018] The first acquisition module is used to acquire historical actual power generation data of each wind turbine in the target wind farm, and to acquire historical actual wind speed of each wind turbine in the target wind farm.

[0019] The partitioning module is used to divide the wind turbines in the target wind farm into multiple wind turbine clusters based on the historical actual power generation data of each wind turbine in the target wind farm and the historical actual wind speed of each wind turbine in the target wind farm. Each wind turbine cluster includes two or more wind turbines.

[0020] The determination module is used to determine a reference wind turbine within any wind turbine cluster based on the historical actual power generation data of each wind turbine in the wind turbine cluster and the total historical actual power generation data of the wind turbine cluster.

[0021] The second acquisition module is used to acquire the current actual power generation data of the reference wind turbines in each wind turbine cluster, acquire the current actual total power generation data of the wind farm upwind of the target wind farm, and acquire the historical actual total power generation data of the target wind farm based on the historical actual power generation data of each wind turbine in the target wind farm.

[0022] The prediction module is used to predict the future expected total power generation of the target wind farm based on the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine.

[0023] Fourthly, this application discloses an apparatus for training a predictive model based on spatiotemporal correlation of wind farm power generation data, the apparatus comprising:

[0024] The third acquisition module is used to acquire the training dataset. The training dataset includes multiple training data sets, including sample data and labeled data. The sample data includes the total sample power generation data for the first historical time period of the sample wind farm, the total sample power generation data for the second historical time period of the wind farm upwind of the sample wind farm, and the sample power generation data for the second historical time period of the benchmark wind turbines within multiple wind turbine clusters in the sample wind farm. The labeled data includes the actual total power generation data for the third historical time period of the sample wind farm. The first historical time period is earlier than the second historical time period, and the second historical time period is earlier than the third historical time period. Each wind turbine cluster includes two or more wind turbines. A wind turbine cluster is obtained by dividing the wind turbines in the sample wind farm into groups based on the sample power generation data of each wind turbine in the sample wind farm during the first historical time period and the sample wind speed of each wind turbine in the sample wind farm during the first historical time period. For any wind turbine cluster, the reference wind turbine within the wind turbine cluster is determined based on the sample power generation data of each wind turbine in the wind turbine cluster during the first historical time period and the total sample power generation data of the wind turbine cluster during the first historical time period. The historical actual total power generation data of the target wind farm is obtained based on the historical actual power generation data of each wind turbine in the target wind farm.

[0025] The training module is used to train the model to be trained based on the training dataset until the model converges, thereby obtaining a prediction model based on the spatiotemporal correlation of wind farm power generation data.

[0026] Fifthly, this application discloses an electronic device comprising: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to perform the method as described in any of the preceding aspects.

[0027] Sixthly, this application discloses a non-transitory computer-readable storage medium that, when the instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to perform the methods described in any of the preceding aspects.

[0028] In a seventh aspect, this application discloses a computer program product in which, when the instructions in the computer program product are executed by a processor of an electronic device, the electronic device is enabled to perform the method described in any of the preceding aspects.

[0029] The technical solution provided in this application may include the following beneficial effects:

[0030] This application obtains historical actual power generation data and historical actual wind speed data for each wind turbine in the target wind farm. Based on the historical actual power generation data and historical actual wind speed data of each wind turbine in the target wind farm, the wind turbines in the target wind farm are divided into multiple wind turbine clusters, each cluster containing two or more wind turbines. For any given wind turbine cluster, a reference wind turbine is determined based on the historical actual power generation data of each wind turbine within that cluster and the total historical actual power generation data of that cluster. The current actual power generation data of the reference wind turbines in each cluster is obtained, as is the current total actual power generation data of the wind farm upwind of the target wind farm; the total historical actual power generation data of the target wind farm is obtained based on the historical actual power generation data of each wind turbine in the target wind farm. Based on the historical total power generation data of the target wind farm, the current total power generation data of the wind farm upwind of the target wind farm, and the current total power generation data of the benchmark wind turbine, predict the future expected total power generation data of the target wind farm.

[0031] In predicting the future expected total power generation of the target wind farm, this application takes into account spatial factors. For example, it considers the impact of the current actual total power generation of the wind farm upwind of the target wind farm on the future expected total power generation of the target wind farm. Since the wind passes through the wind farm upwind of the target wind farm before passing the target wind farm, the wind farm upwind of the target wind farm will absorb some wind energy, for example, by reducing wind speed and / or changing wind direction, which may lead to a decrease in the wind energy that the target wind farm can obtain, and thus a decrease in the total power generation of the target wind farm. Therefore, by considering the impact of the current actual total power generation of the wind farm upwind of the target wind farm on the future expected total power generation of the target wind farm, the accuracy of the predicted future expected total power generation of the target wind farm can be improved.

[0032] In addition, this application also considers the time factor when predicting the future expected total power generation data of the target wind farm. For example, it considers the impact of the historical actual total power generation data of the target wind farm on the future expected total power generation data of the target wind farm. For example, the wind speed and / or wind direction in the area occupied by the wind farm cluster are often stable in adjacent time periods and generally do not change significantly. Therefore, the difference between the historical actual total power generation data of the target wind farm in one time period and the future expected total power generation data of the target wind farm in adjacent time periods is often not large. Therefore, by considering the impact of the historical actual total power generation data of the target wind farm on the future expected total power generation data of the target wind farm, the accuracy of the predicted future expected total power generation data of the target wind farm can be improved.

[0033] Furthermore, when predicting the total expected future power generation of the target wind farm, this application also considers the impact of the current actual power generation data of the benchmark wind turbines within each wind turbine cluster on the total expected future power generation of the target wind farm. For example, by dividing the wind turbines in the target wind farm into multiple wind turbine clusters, and ensuring that the historical actual power generation data of the wind turbines within the same cluster are very similar, and / or that their historical actual wind speeds are very similar, the contribution of the historical actual power generation data of each wind turbine within that cluster to the total historical actual power generation data of that cluster can be determined. The greater the contribution of a wind turbine, the stronger the correlation between its historical actual power generation data and the total historical actual power generation data of that wind turbine cluster. By identifying the wind turbines with the strongest correlation within a wind turbine cluster as the benchmark wind turbines within that cluster, and using the benchmark wind turbines within each wind turbine cluster to process the historical actual power generation data of each wind turbine in the target wind farm, the accuracy of the predicted total future power generation data of the target wind farm can be improved. Attached Figure Description

[0034] Figure 1 This is a flowchart illustrating the steps of a wind farm power generation data prediction method based on spatiotemporal correlation, as described in this application.

[0035] Figure 2 This is a flowchart illustrating the steps of a method for training a predictive model based on spatiotemporal correlation wind farm power generation data, as described in this application.

[0036] Figure 3 This is a schematic diagram of a prediction model for wind farm power generation data based on spatiotemporal correlation, as proposed in this application.

[0037] Figure 4 This is a structural block diagram of a wind farm power generation data prediction device based on spatiotemporal correlation, as described in this application.

[0038] Figure 5 This is a structural block diagram of an apparatus for training a prediction model based on spatiotemporal correlation wind farm power generation data, as described in this application.

[0039] Figure 6 This is a block diagram of an electronic device according to this application.

[0040] Figure 7 This is a block diagram of an electronic device according to this application. Detailed Implementation

[0041] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0042] Reference Figure 1 This paper illustrates a flowchart of a method for predicting wind farm power generation data based on spatiotemporal correlation, which can be applied to electronic devices, including terminals or servers. Specifically, the method includes the following steps:

[0043] In step S101, the historical actual power generation data of each wind turbine in the target wind farm is obtained, and the historical actual wind speed of each wind turbine in the target wind farm is obtained.

[0044] In this application, a wind farm has multiple wind turbines, or multiple wind turbines constitute a wind farm.

[0045] A wind farm occupies an area, and the individual wind turbines in a wind farm are located in different positions within this area.

[0046] A wind farm may be surrounded by other wind farms, and multiple wind farms can form a wind farm cluster.

[0047] The wind speed and / or wind direction in the area occupied by a wind farm cluster tend to be stable in adjacent time periods and generally do not change significantly.

[0048] The target wind farm is one of the wind farms in the wind farm cluster.

[0049] Historical actual power generation data of wind turbines can be understood as: the actual power generation (in kilowatts) or actual power generation (in kilowatt-hours or degrees) of wind turbines during a historical period in the historical process.

[0050] The duration of the historical period can be 6 months, 5 months, 4 months, 3 months, 2 months, 1 month, 2 weeks, 1 week, 48 hours, 36 hours, 24 hours, 12 hours, 6 hours, 2 hours, 1 hour, or half an hour, etc., depending on the actual situation. This application does not limit it in this regard.

[0051] The end point of the historical duration can be the current moment, or it can be an earlier and more recent moment, etc.

[0052] The power generation of wind turbines during historical periods in the historical process includes: the actual power generation of wind turbines during historical periods in the historical process (the actual situation that occurred).

[0053] Alternatively, the power generation of wind turbines during a historical period in the historical process includes: the actual power generation of wind turbines during a historical period in the historical process (the actual situation that occurred).

[0054] The actual power generation or actual electricity generation of a wind turbine during a historical period can be statistically determined and stored in electronic devices in advance, and can be updated in real time.

[0055] Thus, for any wind turbine generator in the target wind farm, the electronic equipment can directly obtain the actual power generation or actual power generation of the wind turbine generator during the historical period of the historical process, and use it as the historical actual power generation data of the wind turbine generator.

[0056] The same applies to each other wind turbine in the target wind farm, thus obtaining the historical actual power generation data of each wind turbine in the target wind farm.

[0057] The actual power generation of a wind turbine during a historical period can be considered as the actual average power of the wind turbine during that historical period.

[0058] The actual power generation of a wind turbine during a historical period can be considered as the actual total power generation of the wind turbine during that historical period.

[0059] In addition, each wind turbine in each wind farm has a wind speed measuring device, such as a wind measurement tower. The wind speed measuring device on each wind turbine can measure the wind speed. The wind speed actually measured by a wind speed measuring device can be regarded as the actual wind speed of the wind turbine where the wind speed measuring device is located.

[0060] The historical actual wind speed of a wind turbine can be understood as the wind speed measured by the wind speed measuring equipment on the wind turbine during a historical period in the historical process. For example, it could be the average wind speed during the historical period.

[0061] In step S102, based on the historical actual power generation data of each wind turbine in the target wind farm and the historical actual wind speed of each wind turbine in the target wind farm, the wind turbines in the target wind farm are divided into multiple wind turbine clusters, and each wind turbine cluster includes two or more wind turbines.

[0062] In one embodiment of this application, the wind turbines in the target wind farm can be divided into multiple wind turbine clusters based on the historical actual power generation data of each wind turbine in the target wind farm and the historical actual wind speed of each wind turbine in the target wind farm, using the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm.

[0063] Of course, it is understandable that other clustering algorithms can also be used to divide the wind turbines in the target wind farm into multiple wind turbine clusters, such as the k-means algorithm or the Clara algorithm. This application does not limit the specific clustering algorithm.

[0064] Different wind turbine clusters contain different wind turbines.

[0065] The historical actual power generation data of wind turbines in the same wind turbine cluster are very similar, and / or the historical actual wind speeds are very similar.

[0066] Historical actual power generation data of wind turbines in different wind turbine clusters vary to some extent, and may even vary considerably, and / or, historical actual wind speeds of wind turbines in different wind turbine clusters vary to some extent, and may even vary considerably.

[0067] In step S103, for any wind turbine cluster, the reference wind turbine in the wind turbine cluster is determined based on the historical actual power generation data of each wind turbine in the wind turbine cluster and the historical actual total power generation data of the wind turbine cluster.

[0068] Historical actual power generation data of wind turbines can be found in the foregoing description. For example, it can be the actual power generation of the wind turbine during a historical period in the historical process (the actual situation that actually occurred), or the actual power generation of the wind turbine during a historical period in the historical process (the actual situation that actually occurred), which will not be detailed here.

[0069] The historical actual power generation data of a wind turbine cluster can be understood as the sum of the historical actual power generation data of each generator in the wind turbine cluster.

[0070] In one example, assuming that the historical actual power generation data of a wind turbine is the true average power generation of the wind turbine during a historical period in the historical process, then the total historical actual power generation data of a wind turbine cluster can be understood as the sum of the true average power generation of each wind turbine in the wind turbine cluster during a historical period in the historical process.

[0071] Alternatively, in another example, assuming that the historical actual power generation data of wind turbines is the actual power generation of wind turbines within a historical period in the historical process, then the total historical actual power generation data of a wind turbine cluster can be understood as: the sum of the actual power generation of each individual wind turbine in the wind turbine cluster within a historical period in the historical process.

[0072] In one embodiment of this application, step S103 can be implemented through the following process, including:

[0073] 1031. Based on the historical actual power generation data of each wind turbine in the wind turbine cluster and the historical actual power generation data of the wind turbine cluster, determine the correlation between the historical actual power generation data of each wind turbine in the wind turbine cluster and the historical actual power generation data of the wind turbine cluster.

[0074] In this way, the contribution of the historical actual power generation data of each wind turbine in the wind turbine cluster to the total historical actual power generation data of the wind turbine cluster can be determined.

[0075] The greater the contribution of a wind turbine, the greater the correlation between its historical actual power generation data and the total historical actual power generation data of the wind turbine cluster; conversely, the smaller the contribution of a wind turbine, the smaller the correlation between its historical actual power generation data and the total historical actual power generation data of the wind turbine cluster.

[0076] This application supports the use of various methods to determine the correlation between the historical actual power generation data of each wind turbine in the wind turbine cluster and the total historical actual power generation data of the wind turbine cluster. This application does not limit the specific method for determining the correlation between the historical actual power generation data of each wind turbine in the wind turbine cluster and the total historical actual power generation data of the wind turbine cluster.

[0077] In an optional embodiment, the correlation between the historical actual power generation data of a wind turbine within the wind turbine cluster and the total historical actual power generation data of the wind turbine cluster can be determined in the following manner:

[0078] For example, the historical actual power generation data of a single wind turbine and the total historical actual power generation data of the wind turbine cluster can be represented as time series vectors, denoted as A and B respectively:

[0079] The MI value between A and B is defined as:

[0080] I(A,B) = H(A) + H(B) - H(A,B)

[0081] H(A) and H(B) represent the entropy values ​​of variables A and B, respectively.

[0082] H(A, B) represents the joint entropy between variables A and B.

[0083] They are defined as follows:

[0084]

[0085]

[0086] p is the probability density of a single variable.

[0087] p joint Let be the joint probability density between the two variables.

[0088] As can be seen from the above formula, the calculation process of I(A, B) is related to the probability density p and the joint probability density p joint Closely related.

[0089] However, in practice, it is often difficult to obtain an accurate probability density function.

[0090] To address this problem, a nonparametric density estimation method can be used to approximate its value, which is more computationally efficient and more practical.

[0091] Taking a continuous random variable A as an example, the histogram estimation method is used to divide the range of values ​​of this variable into K. F There are three segments of equal length, each segment's length is defined as Δ. F .

[0092] To reduce the error of the estimation method, the number of segments can be determined according to the following formula, where the value of n represents the total number of data points contained in A:

[0093]

[0094]

[0095] Based on this, the entropy value of variable A can be expressed as:

[0096]

[0097]

[0098] p i Let be the probability that sample A belongs to the i-th segment.

[0099] n i Let be the number of samples A that belong to the i-th segment.

[0100] Then H(A) can be rewritten in the following form:

[0101]

[0102] Similarly, the entropy of variable B can be expressed as:

[0103]

[0104] n j Let B be the number of samples that belong to the j-th segment.

[0105] After determining the entropy values ​​of variables A and B, it is necessary to estimate the joint entropy value of variables A and B.

[0106] The histogram estimation method is used to divide the extent of its joint space into K. F *K V There are 3 cells of equal size, and the cell size is Δ. F *Δ V Each cell position is defined by (i, j). This is determined by the number of samples n belonging to cell (i, j). ij Statistically, the joint entropy value can be expressed in the following form:

[0107]

[0108] The MI values ​​of variables A and B can be calculated as follows:

[0109]

[0110] 1032. Based on the correlation between the historical actual power generation data of each wind turbine in the wind turbine cluster and the historical actual power generation data of the wind turbine cluster, determine the benchmark wind turbine in the wind turbine cluster.

[0111] The historical actual power generation data of the benchmark wind turbine in a wind turbine cluster often contributes significantly to the total historical actual power generation data of the wind turbine cluster, and is often the highest one or two, or several, etc.

[0112] Alternatively, the correlation between the historical actual power generation data of the benchmark wind turbine in a wind turbine cluster and the total historical actual power generation data of the wind turbine cluster is often high, and it is often the highest one or two, or several, etc.

[0113] Thus, in one embodiment, the wind turbines within the wind turbine cluster can be sorted according to the decreasing correlation between their historical actual power generation data and the total historical actual power generation data of the cluster. For example, the wind turbines with a higher correlation between their historical actual power generation data and the total historical actual power generation data of the cluster are sorted earlier, while those with a lower correlation are sorted later. The correlation can be represented numerically; a larger value indicates a closer correlation, while a smaller value indicates a less close correlation. Then, at least one wind turbine can be selected from the cluster according to the sorting order. For example, the Top N wind turbines can be selected from the cluster according to the sorting order; N is a positive integer greater than or equal to 1, and the specific value of N can be determined according to the actual situation, which is not limited in this application. Then, a reference wind turbine can be determined within the wind turbine cluster based on the selected wind turbine. For example, the selected wind turbine can be used as the reference wind turbine within the wind turbine cluster.

[0114] The number of reference wind turbines in a wind turbine cluster can be 1, 2, or 3, etc.

[0115] In step S104, the current actual power generation data of the reference wind turbines in each wind turbine cluster is obtained, the current actual total power generation data of the wind farm upwind of the target wind farm is obtained, and the historical actual total power generation data of the target wind farm is obtained based on the historical actual power generation data of each wind turbine in the target wind farm.

[0116] The wind speed and / or wind direction of the wind farm cluster where the target wind farm is located are usually stable and do not change significantly.

[0117] For any wind farm in a wind farm cluster other than the target wind farm, if the wind passes through the wind farm first and then the target wind farm, then the wind farm can be the wind farm upwind of the target wind farm, and the target wind farm can be the wind farm downwind of the wind farm.

[0118] In this application, the wind first passes through the wind farm upwind of the target wind farm, and then passes through the target wind farm.

[0119] The current actual power generation data of a wind turbine can be understood as: the actual power generation (in kilowatts) or actual power generation (in kilowatt-hours / kWh) of the wind turbine in the current time period.

[0120] The duration of the current time period can be 5 minutes, 10 minutes, half an hour, 1 hour, 2 hours, 6 hours, 12 hours, 24 hours, 36 hours or 48 hours, etc., depending on the actual situation. This application does not limit it.

[0121] The power generation of a wind turbine in the current time period includes: the actual power generation of the wind turbine in the current time period (the actual situation that occurred), or the power generation of a wind turbine in the current time period includes: the actual power generation of the wind turbine in the current time period (the actual situation that occurred).

[0122] The actual power generation or actual electricity generation of a wind turbine at the current moment within the current time period can be statistically calculated and stored in electronic devices in real time, and can be updated in real time.

[0123] Thus, for any one of the reference wind turbines in each wind turbine cluster, the electronic equipment can directly obtain the actual power generation or actual power generation of that reference wind turbine in the current time period, and use it as the current actual power generation data of that reference wind turbine.

[0124] The same applies to each other reference wind turbine within each wind turbine cluster, thus obtaining the current actual power generation data of the reference wind turbine within each wind turbine cluster.

[0125] The actual power generation of a wind turbine in the current time period can be the actual average power generation of the wind turbine in the current time period.

[0126] The actual power generation of a wind turbine in the current time period can be the actual total power generation of the wind turbine in the current time period.

[0127] The actual power generation or actual electricity generation of the wind turbines in the wind farm upwind of the target wind farm during the current time period can be statistically calculated and stored in electronic devices in real time, and can be updated in real time.

[0128] Thus, for any wind turbine in the wind farm upwind of the target wind farm, the electronic equipment can directly obtain the actual power generation or actual power generation of the wind turbine in the current time period and use it as the current actual power generation data of the wind turbine.

[0129] The same process is repeated for each other wind turbine in the wind farm upwind of the target wind farm, thus obtaining the current actual power generation data of each wind turbine in the wind farm upwind of the target wind farm. Then, the current actual power generation data of each wind turbine in the wind farm upwind of the target wind farm are summed to obtain the current actual power generation data of the wind farm upwind of the target wind farm.

[0130] In addition, the historical actual power generation data of each wind turbine in the target wind farm can be summed to obtain the total historical actual power generation data of the target wind farm.

[0131] In step S105, the expected future total power generation of the target wind farm is predicted based on the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine.

[0132] The total expected future power generation data of the target wind farm can be understood as the sum of the expected future power generation data of each generator in the target wind farm.

[0133] In one example, assuming that the total expected power generation of a wind turbine is the expected average power generation of the wind turbine over a certain period of time in the future, the total expected power generation of the target wind farm can be understood as the sum of the expected average power generation of each wind turbine in the target wind farm over a certain period of time in the future.

[0134] Alternatively, in another example, assuming that the total expected future power generation of a wind turbine is the expected power generation of the wind turbine in a certain period of the future, the total expected future power generation of the target wind farm can be understood as the sum of the expected power generation of each wind turbine in the target wind farm in a certain period of the future.

[0135] The duration of a certain period can be 5 minutes, 10 minutes, half an hour, 1 hour, 2 hours, 6 hours, 12 hours, 24 hours, 36 hours or 48 hours, etc., depending on the actual situation. This application does not limit it.

[0136] This application obtains historical actual power generation data and historical actual wind speed data for each wind turbine in the target wind farm. Based on the historical actual power generation data and historical actual wind speed data of each wind turbine in the target wind farm, the wind turbines in the target wind farm are divided into multiple wind turbine clusters, each cluster containing two or more wind turbines. For any given wind turbine cluster, a reference wind turbine is determined based on the historical actual power generation data of each wind turbine within that cluster and the total historical actual power generation data of that cluster. The current actual power generation data of the reference wind turbines in each cluster is obtained, as is the current total actual power generation data of the wind farm upwind of the target wind farm; the total historical actual power generation data of the target wind farm is obtained based on the historical actual power generation data of each wind turbine in the target wind farm. Based on the historical total power generation data of the target wind farm, the current total power generation data of the wind farm upwind of the target wind farm, and the current total power generation data of the benchmark wind turbine, predict the future expected total power generation data of the target wind farm.

[0137] In predicting the future expected total power generation of the target wind farm, this application takes into account spatial factors. For example, it considers the impact of the current actual total power generation of the wind farm upwind of the target wind farm on the future expected total power generation of the target wind farm. Since the wind passes through the wind farm upwind of the target wind farm before passing the target wind farm, the wind farm upwind of the target wind farm will absorb some wind energy, for example, by reducing wind speed and / or changing wind direction, which may lead to a decrease in the wind energy that the target wind farm can obtain, and thus a decrease in the total power generation of the target wind farm. Therefore, by considering the impact of the current actual total power generation of the wind farm upwind of the target wind farm on the future expected total power generation of the target wind farm, the accuracy of the predicted future expected total power generation of the target wind farm can be improved.

[0138] In addition, this application also considers the time factor when predicting the future expected total power generation data of the target wind farm. For example, it considers the impact of the historical actual total power generation data of the target wind farm on the future expected total power generation data of the target wind farm. For example, the wind speed and / or wind direction in the area occupied by the wind farm cluster are often stable in adjacent time periods and generally do not change significantly. Therefore, the difference between the historical actual total power generation data of the target wind farm in one time period and the future expected total power generation data of the target wind farm in adjacent time periods is often not large. Therefore, by considering the impact of the historical actual total power generation data of the target wind farm on the future expected total power generation data of the target wind farm, the accuracy of the predicted future expected total power generation data of the target wind farm can be improved.

[0139] Furthermore, when predicting the total expected future power generation of the target wind farm, this application also considers the impact of the current actual power generation data of the benchmark wind turbines within each wind turbine cluster on the total expected future power generation of the target wind farm. For example, by dividing the wind turbines in the target wind farm into multiple wind turbine clusters, and ensuring that the historical actual power generation data of the wind turbines within the same cluster are very similar, and / or that their historical actual wind speeds are very similar, the contribution of the historical actual power generation data of each wind turbine within that cluster to the total historical actual power generation data of that cluster can be determined. The greater the contribution of a wind turbine, the stronger the correlation between its historical actual power generation data and the total historical actual power generation data of that wind turbine cluster. By identifying the wind turbines with the strongest correlation within a wind turbine cluster as the benchmark wind turbines within that cluster, and using the benchmark wind turbines within each wind turbine cluster to process the historical actual power generation data of each wind turbine in the target wind farm, the accuracy of the predicted total future power generation data of the target wind farm can be improved.

[0140] In one embodiment of this application, a prediction model based on spatiotemporal correlation of wind farm power generation data can be trained in advance, and then the prediction model based on spatiotemporal correlation of wind farm power generation data can be used to predict the future expected total power generation data of the target wind farm.

[0141] Among them, see Figure 2 The specific training process is as follows:

[0142] In step S201, a training dataset is obtained. The training dataset includes multiple training data sets. These training data sets include sample data and labeled data. The sample data includes the total sample power generation data for the first historical time period of the sample wind farm, the total sample power generation data for the second historical time period of the wind farm upwind of the sample wind farm, and the sample power generation data for the second historical time period of the benchmark wind turbines within multiple wind turbine clusters in the sample wind farm. The labeled data includes the actual total power generation data for the third historical time period of the sample wind farm. The first historical time period is earlier than the second historical time period, and the second historical time period is earlier than the third historical time period.

[0143] Each wind turbine cluster contains two or more wind turbines.

[0144] Multiple wind turbine clusters are obtained by dividing the wind turbines in the sample wind farm based on the sample power generation data of each wind turbine in the first historical time period and the sample wind speed of each wind turbine in the first historical time period.

[0145] For any given wind turbine cluster, the reference wind turbine within that cluster is determined based on sample power generation data from each individual wind turbine within the cluster during the first historical time period, as well as the total sample power generation data from the cluster during the first historical time period.

[0146] The historical actual total power generation data of the target wind farm is obtained based on the historical actual power generation data of each wind turbine in the target wind farm.

[0147] The explanations involved in this step can be found in steps S101 to S104, and will not be elaborated here.

[0148] In step S202, the model to be trained is trained according to the training dataset until the model converges, thereby obtaining a prediction model based on the spatiotemporal correlation of wind farm power generation data.

[0149] Both the sample data and the labeled data are real historical data. Therefore, the prediction model trained on spatiotemporally correlated wind farm power generation data can closely resemble the actual situation, has a high degree of generalization, and the prediction results are more in line with reality.

[0150] In this way, when predicting the future expected total power generation data of the target wind farm in step S105, the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine can be input into the prediction model of wind farm power generation data based on spatiotemporal correlation. This allows the prediction model to process the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine to obtain a processing result. Then, the future expected total power generation data of the target wind farm can be obtained based on the processing result obtained from the prediction model of wind farm power generation data based on spatiotemporal correlation.

[0151] For example, in one embodiment of this application, there is a prediction model for wind farm power generation data based on spatiotemporal correlation. The processing result obtained by the prediction model for wind farm power generation data based on spatiotemporal correlation is the total expected future power generation data of the target wind farm predicted by the prediction model for wind farm power generation data based on spatiotemporal correlation. In this way, the total expected future power generation data of the target wind farm predicted by the prediction model for wind farm power generation data based on spatiotemporal correlation can be determined as the total expected future power generation data of the target wind farm.

[0152] In another embodiment of this application, when training a prediction model based on spatiotemporal correlation of wind farm power generation data, see [reference needed]. Figure 3The bootstrapping sampling method can be used to perform S rounds of sampling on the training dataset (each time 80% of the training dataset is sampled). The training data sampled in each round is used as a subset of the training data. That is, each round of sampling will result in a subset of the training data, resulting in a total of S subsets of the training data.

[0153] Bootstrapping sampling is a statistical method used to handle small datasets. It reconstructs the original distribution by sampling with replacement (where there may be duplicate observations in the sample).

[0154] For any subset of training data, the model to be trained can be trained based on that subset of training data until the model converges, thereby obtaining a prediction model for wind farm power generation data based on spatiotemporal correlation. The same applies to each other subset of training data, thus obtaining a total of S prediction models for wind farm power generation data based on spatiotemporal correlation.

[0155] S is a positive integer greater than or equal to 2; the training data in any two training data subsets are either not all the same or not at all.

[0156] Then, S prediction models based on spatiotemporal correlation of wind farm power generation data can be put online and used, for example, to predict the total expected power generation data of the target wind farm in the future.

[0157] For example, in another embodiment of this application, there are two or more prediction models based on spatiotemporal correlation of wind farm power generation data. Two or more prediction models based on spatiotemporal correlation of wind farm power generation data can be trained in advance, and the training data used when training different prediction models based on spatiotemporal correlation of wind farm power generation data are not all the same or are completely different; the specific training method can be referred to the foregoing description, and will not be detailed here.

[0158] Thus, in step S105, when predicting the future expected total power generation data of the target wind farm, the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine can be input into two or more prediction models based on spatiotemporal correlation of wind farm power generation data. That is, each prediction model based on spatiotemporal correlation of wind farm power generation data will obtain the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine. Each of the above-mentioned prediction models based on spatiotemporal correlation wind farm power generation data processes the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine, and obtains its own processing results. The processing results obtained by each prediction model based on spatiotemporal correlation wind farm power generation data are their respective predictions of the future expected total power generation data of the target wind farm. In this way, the future expected total power generation data of the target wind farm can be obtained based on the processing results obtained by two or more prediction models based on spatiotemporal correlation wind farm power generation data.

[0159] For example, the processing results obtained by the prediction models of wind farm power generation data based on spatiotemporal correlation are: the total expected power generation data of the target wind farm predicted by the prediction models of wind farm power generation data based on spatiotemporal correlation; thus, the average value of the total expected power generation data of the target wind farm predicted by each prediction model of wind farm power generation data based on spatiotemporal correlation can be calculated to obtain the total expected power generation data of the target wind farm, that is, the average value is determined as the total expected power generation data of the target wind farm.

[0160] The models in this application may include Transformer models, and may also include CNN (Convolutional Neural Networks), RNN (Recurrent Neural Network), and LSTM (Long Short-Term Memory). This application does not limit the specific structure of the model.

[0161] It should be noted that, for the sake of simplicity, the method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, because according to this application, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also understand that the embodiments described in the specification are all optional embodiments, and the actions involved are not necessarily required by this application.

[0162] Reference Figure 4 The diagram shows a structural block diagram of a wind farm power generation data prediction device based on spatiotemporal correlation according to this application. The device includes:

[0163] The first acquisition module 11 is used to acquire historical actual power generation data of each wind turbine in the target wind farm, and to acquire historical actual wind speed of each wind turbine in the target wind farm.

[0164] The division module 12 is used to divide the wind turbines in the target wind farm into multiple wind turbine clusters based on the historical actual power generation data of each wind turbine in the target wind farm and the historical actual wind speed of each wind turbine in the target wind farm. Each wind turbine cluster includes two or more wind turbines.

[0165] The determination module 13 is used to determine a reference wind turbine within any wind turbine cluster based on the historical actual power generation data of each wind turbine in the wind turbine cluster and the total historical actual power generation data of the wind turbine cluster.

[0166] The second acquisition module 14 is used to acquire the current actual power generation data of the reference wind turbines in each wind turbine cluster, acquire the current actual total power generation data of the wind farm upwind of the target wind farm, and acquire the historical actual total power generation data of the target wind farm based on the historical actual power generation data of each wind turbine in the target wind farm.

[0167] The prediction module 15 is used to predict the future expected total power generation of the target wind farm based on the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine.

[0168] In one optional implementation, the prediction module includes:

[0169] The input unit is used to input the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine into the prediction model of wind farm power generation data based on spatiotemporal correlation. This allows the prediction model to process the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine to obtain the processing results.

[0170] The acquisition unit is used to obtain the total expected future power generation data of the target wind farm based on the processing results obtained from the prediction model of wind farm power generation data based on spatiotemporal correlation.

[0171] In one alternative implementation, there are two or more prediction models based on spatiotemporal correlation of wind farm power generation data, and the training data used to train different prediction models based on spatiotemporal correlation of wind farm power generation data are not all the same or are completely different.

[0172] The input unit includes:

[0173] The input subunit is used to input the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine into two or more prediction models based on spatiotemporal correlation of wind farm power generation data. This allows the two or more prediction models based on spatiotemporal correlation of wind farm power generation data to process the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine, respectively, and obtain their respective processing results.

[0174] Accordingly, the acquisition unit includes:

[0175] The acquisition sub-unit is used to obtain the total expected future power generation data of the target wind farm based on the processing results obtained from two or more prediction models of wind farm power generation data based on spatiotemporal correlation.

[0176] In one optional implementation, the processing results obtained by the prediction model based on the spatiotemporal correlation of wind farm power generation data are: the total expected future power generation data of the target wind farm predicted by the prediction model based on the spatiotemporal correlation of wind farm power generation data.

[0177] The acquisition subunit is specifically used to: calculate the average value among the future expected total power generation data of the target wind farm predicted by each prediction model based on spatiotemporal correlation of wind farm power generation data, so as to obtain the future expected total power generation data of the target wind farm.

[0178] In one optional implementation, the partitioning module includes:

[0179] The partitioning unit is used to divide the wind turbines in the target wind farm into multiple wind turbine clusters based on the historical actual power generation data and historical actual wind speed of each wind turbine in the target wind farm, using a representative density-based clustering algorithm, DBSCAN.

[0180] In one optional implementation, the determining module includes:

[0181] The first determining unit is used to determine the correlation between the historical actual power generation data of each wind turbine in the wind turbine cluster and the historical actual power generation data of the wind turbine cluster based on the historical actual power generation data of each wind turbine in the wind turbine cluster and the historical actual power generation data of the wind turbine cluster.

[0182] The second determining unit is used to determine the reference wind turbine in the wind turbine cluster based on the correlation between the historical actual power generation data of each wind turbine in the wind turbine cluster and the historical actual power generation data of the wind turbine cluster.

[0183] In one optional implementation, the second determining unit includes:

[0184] The sorting subunit is used to sort the individual wind turbines in the wind turbine cluster. Within the wind turbine cluster, the wind turbine with the greater correlation between its historical actual power generation data and the total historical actual power generation data of the wind turbine cluster is sorted earlier, and the wind turbine with the smaller correlation between its historical actual power generation data and the total historical actual power generation data of the wind turbine cluster is sorted later.

[0185] The selection subunit is used to select the Top N wind turbines from among the wind turbines in the wind turbine cluster according to the sorting order; N is a positive integer greater than or equal to 1.

[0186] A sub-unit is determined for identifying a reference wind turbine within the wind turbine cluster based on the selected wind turbine.

[0187] Reference Figure 5 The diagram shows a structural block diagram of an apparatus for training a prediction model based on spatiotemporal correlation wind farm power generation data according to this application. The apparatus includes:

[0188] The third acquisition module 21 is used to acquire the training dataset. The training dataset includes multiple training data sets, including sample data and labeled data. The sample data includes the total sample power generation data for the first historical time period of the sample wind farm, the total sample power generation data for the second historical time period of the wind farm upwind of the sample wind farm, and the sample power generation data for the second historical time period of the benchmark wind turbines within multiple wind turbine clusters in the sample wind farm. The labeled data includes the actual total power generation data for the third historical time period of the sample wind farm. The first historical time period is earlier than the second historical time period, and the second historical time period is earlier than the third historical time period. Each wind turbine cluster includes more than two wind turbines. Multiple wind turbine clusters are obtained by dividing the wind turbines in the sample wind farm into groups based on the sample power generation data of each wind turbine in the sample wind farm during the first historical time period and the sample wind speed of each wind turbine in the sample wind farm during the first historical time period. For any wind turbine cluster, the reference wind turbine within the wind turbine cluster is determined based on the sample power generation data of each wind turbine in the wind turbine cluster during the first historical time period and the total sample power generation data of the wind turbine cluster during the first historical time period. The historical actual total power generation data of the target wind farm is obtained based on the historical actual power generation data of each wind turbine in the target wind farm.

[0189] Training module 22 is used to train the model to be trained based on the training dataset until the model to be trained converges, thereby obtaining a prediction model based on the spatiotemporal correlation of wind farm power generation data.

[0190] In one optional implementation, the training module includes:

[0191] The extraction unit is used to extract S rounds of training data from the training dataset using the bootstrapping sampling method, with each round of extracted training data serving as a subset of the training data; S is a positive integer greater than or equal to 2.

[0192] The training unit is used to train the model to be trained on any subset of training data until the model to be trained converges, thereby obtaining a prediction model of wind farm power generation data based on spatiotemporal correlation, resulting in a total of S prediction models of wind farm power generation data based on spatiotemporal correlation; the training data in any two subsets are not all the same or are completely different.

[0193] As the device embodiment is basically similar to the method embodiment, the description is relatively simple, and relevant parts can be found in the description of the method embodiment.

[0194] Optionally, this application also provides an electronic device, including: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the various processes of the above method embodiments and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0195] This application also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the above-described method embodiments and achieves the same technical effects. To avoid repetition, it will not be described again here. The computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc.

[0196] The figure is a block diagram of an electronic device 800 as shown in the six applications. For example, the electronic device 800 can be a mobile phone, computer, digital broadcasting terminal, messaging device, game console, tablet device, medical device, fitness equipment, personal digital assistant, etc.

[0197] Reference Figure 6 The electronic device 800 may include one or more of the following components: a processing component 802, a memory 804, a power supply component 806, a multimedia component 808, an audio component 810, an input / output (I / O) interface 812, a sensor component 814, and a communication component 816.

[0198] Processing component 802 typically controls the overall operation of electronic device 800, such as operations associated with display, telephone calls, data communication, camera operation, and recording operations. Processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the methods described above. Furthermore, processing component 802 may include one or more modules to facilitate interaction between processing component 802 and other components. For example, processing component 802 may include a multimedia module to facilitate interaction between multimedia component 808 and processing component 802.

[0199] Memory 804 is configured to store various types of data to support the operation of device 800. Examples of this data include instructions for any application or method operating on electronic device 800, contact data, phonebook data, messages, images, videos, etc. Memory 804 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0200] Power supply component 806 provides power to various components of electronic device 800. Power supply component 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to electronic device 800.

[0201] Multimedia component 808 includes a screen that provides an output interface between the electronic device 800 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touchscreen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors may sense not only the boundaries of the touch or swipe action but also the duration and pressure associated with the touch or swipe operation. In some embodiments, multimedia component 808 includes a front-facing camera and / or a rear-facing camera. When the device 800 is in an operating mode, such as a shooting mode or a video mode, the front-facing camera and / or the rear-facing camera may receive external multimedia data. Each front-facing camera and rear-facing camera may be a fixed optical lens system or have focal length and optical zoom capabilities.

[0202] Audio component 810 is configured to output and / or input audio signals. For example, audio component 810 includes a microphone (MIC) configured to receive external audio signals when electronic device 800 is in an operating mode, such as call mode, recording mode, and voice recognition mode. The received audio signals may be further stored in memory 804 or transmitted via communication component 816. In some embodiments, audio component 810 also includes a speaker for outputting audio signals.

[0203] I / O interface 812 provides an interface between processing component 802 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to, home buttons, volume buttons, power buttons, and lock buttons.

[0204] Sensor assembly 814 includes one or more sensors for providing state assessments of various aspects of electronic device 800. For example, sensor assembly 814 may detect the on / off state of device 800, the relative positioning of components such as the display and keypad of electronic device 800, changes in position of electronic device 800 or a component of electronic device 800, the presence or absence of user contact with electronic device 800, orientation or acceleration / deceleration of electronic device 800, and temperature changes of electronic device 800. Sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. Sensor assembly 814 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, sensor assembly 814 may also include an accelerometer, gyroscope, magnetometer, pressure sensor, or temperature sensor.

[0205] Communication component 816 is configured to facilitate wired or wireless communication between electronic device 800 and other devices. Electronic device 800 can access wireless networks based on communication standards, such as WiFi, carrier networks (such as 2G, 3G, 4G, or 5G), or combinations thereof. In one exemplary embodiment, communication component 816 receives broadcast signals or broadcast operation information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, communication component 816 also includes a near-field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on radio frequency identification (RFID) technology, Infrared Data Association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

[0206] In an exemplary embodiment, the electronic device 800 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the methods described above.

[0207] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 804 including instructions, which can be executed by a processor 820 of an electronic device 800 to perform the above-described method. For example, the non-transitory computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0208] Figure 7This is a block diagram of an electronic device 1900 shown in this application. For example, the electronic device 1900 can be provided as a server.

[0209] Reference Figure 7 The electronic device 1900 includes a processing component 1922, which further includes one or more processors, and memory resources represented by memory 1932 for storing instructions, such as application programs, that can be executed by the processing component 1922. The application programs stored in memory 1932 may include one or more modules, each corresponding to a set of instructions. Furthermore, the processing component 1922 is configured to execute instructions to perform the methods described above.

[0210] Electronic device 1900 may also include a power supply component 1926 configured to perform power management of electronic device 1900, a wired or wireless network interface 1950 configured to connect electronic device 1900 to a network, and an input / output (I / O) interface 1958. Electronic device 1900 can operate on an operating system stored in memory 1932, such as Windows Server™, Mac OS X™, Unix™, Linux™, FreeBSD™, or similar.

[0211] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0212] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0213] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

[0214] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this application can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0215] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0216] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0217] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0218] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0219] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0220] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for predicting wind farm power generation data based on spatiotemporal correlation, characterized in that, The method includes: Obtain historical actual power generation data for each wind turbine in the target wind farm, and obtain historical actual wind speed for each wind turbine in the target wind farm; Based on the historical actual power generation data and historical actual wind speed of each wind turbine in the target wind farm, the wind turbines in the target wind farm are divided into multiple wind turbine clusters, and each wind turbine cluster includes more than two wind turbines. For any wind turbine cluster, a benchmark wind turbine is determined within the wind turbine cluster based on the historical actual power generation data of each wind turbine in the wind turbine cluster and the total historical actual power generation data of the wind turbine cluster. Obtain the current actual power generation data of the benchmark wind turbines in each wind turbine cluster, and obtain the current total actual power generation data of the wind farms upwind of the target wind farm; obtain the historical total actual power generation data of the target wind farm based on the historical actual power generation data of each wind turbine in the target wind farm. Based on the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine, the data are input into a prediction model based on spatiotemporal correlation of wind farm power generation data to predict the future expected total power generation data of the target wind farm. The step of determining a benchmark wind turbine within the wind turbine cluster based on the historical actual power generation data of each wind turbine within the cluster and the total historical actual power generation data of the wind turbine cluster includes: Based on the historical actual power generation data of each wind turbine in the wind turbine cluster and the historical actual power generation data of the wind turbine cluster, determine the correlation between the historical actual power generation data of each wind turbine in the wind turbine cluster and the historical actual power generation data of the wind turbine cluster. The individual wind turbines within the wind turbine cluster are sorted. Within the wind turbine cluster, the wind turbine with the greater correlation between its historical actual power generation data and the total historical actual power generation data of the wind turbine cluster is sorted earlier, while the wind turbine with the smaller correlation between its historical actual power generation data and the total historical actual power generation data of the wind turbine cluster is sorted later. According to the sorting order, select the Top N wind turbines from each wind turbine in the wind turbine cluster; N is a positive integer greater than or equal to 1. The reference wind turbine within the wind turbine cluster is determined based on the selected wind turbine.

2. The method according to claim 1, characterized in that, The prediction of the target wind farm's future expected total power generation, based on the target wind farm's historical actual total power generation data, the current actual total power generation data of wind farms upwind of the target wind farm, and the current actual power generation data of benchmark wind turbines, includes: The historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine are input into the prediction model of wind farm power generation data based on spatiotemporal correlation. The prediction model of wind farm power generation data based on spatiotemporal correlation processes the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine to obtain the processing results. Based on the processing results obtained from the prediction model of wind farm power generation data based on spatiotemporal correlation, the total expected future power generation data of the target wind farm is obtained.

3. The method according to claim 2, characterized in that, There are two or more prediction models for wind farm power generation data based on spatiotemporal correlation. The training data used to train different prediction models for wind farm power generation data based on spatiotemporal correlation are not all the same or are completely different. The process involves inputting the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine into a prediction model for wind farm power generation data based on spatiotemporal correlation. This allows the prediction model to process the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine, obtaining the processing results, including: The historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine are respectively input into two or more prediction models based on spatiotemporal correlation of wind farm power generation data. This allows the two or more prediction models based on spatiotemporal correlation of wind farm power generation data to process the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine, and obtain their respective processing results. Accordingly, obtaining the future expected total power generation data of the target wind farm based on the processing results obtained from the prediction model of wind farm power generation data based on spatiotemporal correlation includes: Based on the processing results obtained from two or more prediction models of wind farm power generation data based on spatiotemporal correlation, the total expected future power generation data of the target wind farm is obtained.

4. The method according to claim 3, characterized in that, The processing results obtained by the prediction model based on spatiotemporal correlation of wind farm power generation data are as follows: the total expected future power generation data of the target wind farm predicted by the prediction model based on spatiotemporal correlation of wind farm power generation data. The process of obtaining the total expected future power generation data of the target wind farm based on the processing results of two or more prediction models for wind farm power generation data based on spatiotemporal correlation includes: The average value of the future expected total power generation data of the target wind farm is calculated among the prediction models of each wind farm power generation data based on spatiotemporal correlation, and the future expected total power generation data of the target wind farm is obtained.

5. The method according to claim 1, characterized in that, Based on the historical actual power generation data and historical actual wind speed of each wind turbine in the target wind farm, the wind turbines in the target wind farm are divided into multiple wind turbine clusters, including: Based on the historical actual power generation data and historical actual wind speed of each wind turbine in the target wind farm, the representative density-based clustering algorithm DBSCAN is used to divide the wind turbines in the target wind farm into multiple wind turbine clusters.

6. The method according to any one of claims 1-4, characterized in that, The training methods for the prediction model include: Obtain the training dataset; The training dataset includes multiple training datasets; The training data includes sample data and labeled data; The sample data includes the total sample power generation data for the first historical time period of the sample wind farm, the total sample power generation data for the second historical time period of the wind farm upwind of the sample wind farm, and the sample power generation data for the second historical time period of the benchmark wind turbines in multiple wind turbine clusters within the sample wind farm. The labeled data includes the actual total power generation data for the third historical period of the sample wind farm; The first historical time period is earlier than the second historical time period, and the second historical time period is earlier than the third historical time period. Each wind turbine cluster includes two or more wind turbines. Multiple wind turbine clusters are obtained by dividing the wind turbines in the sample wind farm based on the sample power generation data of each wind turbine in the sample wind farm during the first historical time period and the sample wind speed of each wind turbine in the sample wind farm during the first historical time period. For any wind turbine cluster, the reference wind turbine in the wind turbine cluster is determined based on the sample power generation data of each wind turbine in the wind turbine cluster during the first historical time period and the total sample power generation data of the wind turbine cluster during the first historical time period. The historical actual total power generation data of the target wind farm is obtained based on the historical actual power generation data of each wind turbine in the target wind farm. The training model is trained on the training dataset until it converges, thus obtaining a prediction model for wind farm power generation data based on spatiotemporal correlation.

7. The method according to claim 6, characterized in that, The process of training the model to be trained based on the training dataset until the model converges, thereby obtaining a prediction model based on spatiotemporal correlation of wind farm power generation data, includes: The bootstrapping sampling method is used to sample the training dataset in S rounds, with each round of sampled training data serving as a subset of the training data; S is a positive integer greater than or equal to 2. For any subset of training data, the model to be trained is trained according to the subset of training data until the model to be trained converges, thereby obtaining a prediction model of wind farm power generation data based on spatiotemporal correlation, resulting in a total of S prediction models of wind farm power generation data based on spatiotemporal correlation; the training data in any two subsets of training data are either not all the same or not at all.

8. A wind farm power generation data prediction device based on spatiotemporal correlation, characterized in that, The device includes: The first acquisition module is used to acquire historical actual power generation data of each wind turbine in the target wind farm, and to acquire historical actual wind speed of each wind turbine in the target wind farm. The partitioning module is used to divide the wind turbines in the target wind farm into multiple wind turbine clusters based on the historical actual power generation data of each wind turbine in the target wind farm and the historical actual wind speed of each wind turbine in the target wind farm. Each wind turbine cluster includes two or more wind turbines. The determination module is used to determine a reference wind turbine within any wind turbine cluster based on the historical actual power generation data of each wind turbine in the wind turbine cluster and the total historical actual power generation data of the wind turbine cluster. The second acquisition module is used to acquire the current actual power generation data of the reference wind turbines in each wind turbine cluster, acquire the current actual total power generation data of the wind farm upwind of the target wind farm, and acquire the historical actual total power generation data of the target wind farm based on the historical actual power generation data of each wind turbine in the target wind farm. The prediction module is used to predict the future expected total power generation of the target wind farm by inputting the historical actual total power generation data of the target wind farm, the current actual total power generation data of the wind farm upwind of the target wind farm, and the current actual power generation data of the benchmark wind turbine into a prediction model based on spatiotemporal correlation of wind farm power generation data. The determining module includes: The first determining unit is used to determine the correlation between the historical actual power generation data of each wind turbine in the wind turbine cluster and the historical actual power generation data of the wind turbine cluster based on the historical actual power generation data of each wind turbine in the wind turbine cluster and the historical actual power generation data of the wind turbine cluster. The second determining unit is used to sort the individual wind turbines within the wind turbine cluster. Within the wind turbine cluster, the wind turbine with the greater correlation between its historical actual power generation data and the total historical actual power generation data of the wind turbine cluster is sorted earlier, while the wind turbine with the smaller correlation between its historical actual power generation data and the total historical actual power generation data of the wind turbine cluster is sorted later. According to the sorting order, the Top N wind turbines are selected from the individual wind turbines within the wind turbine cluster, where N is a positive integer greater than or equal to 1. The base wind turbine within the wind turbine cluster is determined based on the selected wind turbines.

9. The apparatus according to claim 8, characterized in that, The training device for the prediction model includes: The third acquisition module is used to acquire the training dataset. The training dataset includes multiple training data sets, including sample data and labeled data. The sample data includes the total sample power generation data for the first historical time period of the sample wind farm, the total sample power generation data for the second historical time period of the wind farm upwind of the sample wind farm, and the sample power generation data for the second historical time period of the benchmark wind turbines within multiple wind turbine clusters in the sample wind farm. The labeled data includes the actual total power generation data for the third historical time period of the sample wind farm. The first historical time period is earlier than the second historical time period, and the second historical time period is earlier than the third historical time period. Each wind turbine cluster includes two or more wind turbines. A wind turbine cluster is obtained by dividing the wind turbines in the sample wind farm into groups based on the sample power generation data of each wind turbine in the sample wind farm during the first historical time period and the sample wind speed of each wind turbine in the sample wind farm during the first historical time period. For any wind turbine cluster, the reference wind turbine within the wind turbine cluster is determined based on the sample power generation data of each wind turbine in the wind turbine cluster during the first historical time period and the total sample power generation data of the wind turbine cluster during the first historical time period. The historical actual total power generation data of the target wind farm is obtained based on the historical actual power generation data of each wind turbine in the target wind farm. The training module is used to train the model to be trained based on the training dataset until the model converges, thereby obtaining a prediction model based on the spatiotemporal correlation of wind farm power generation data.

10. An electronic device, characterized in that, include: A processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the method as described in any one of claims 1 to 7.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 7.

Images