Wind power plant cluster representative wind power plant selection method and device based on spatial proximity and power clustering, electronic equipment and storage medium

Abstract

The invention discloses a wind power plant cluster representative wind power plant selection method and device based on spatial proximity and power clustering, electronic equipment and a storage medium. The method comprises the following steps: acquiring spatial position information of all wind power plants in a wind power plant group and generated power data at a typical daily sampling moment; dividing all the wind power plants according to the spatial position information, and generating a spatial adjacent structure of a wind power plant group; calculating the neighborhood power entropy of each wind power plant in the spatial adjacent structure of the wind power plant group according to the generated power data; clustering the wind power plants in the spatial adjacent structure of the wind power plant group according to the neighborhood power entropy of the wind power plants to generate a plurality of wind power plant clusters; calculating a cross correlation coefficient between the power of each wind power plant in each wind power plant cluster and the total power of the wind power plants in the cluster, and taking the wind power plant with the maximum cross correlation coefficient as a corresponding cluster representative wind power plant; by implementing the method, the precision and efficiency of wind power plant group power prediction can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of wind power prediction, and in particular to a method, device, electronic equipment and storage medium for selecting wind farm cluster representative wind farms based on spatial proximity and power clustering. Background Art

[0002] With the implementation of my country's "dual carbon" strategic goals and the advancement of the construction of new power systems, wind power, as an important clean energy, has an increasing installed capacity. Accurate prediction of wind power is of vital importance to the optimal configuration of wind power, grid support and power system dispatching. By accurately predicting wind power, not only can the utilization efficiency of wind power resources be improved, but also the grid load fluctuations can be effectively alleviated to ensure the stable operation of the power system. Therefore, improving the accuracy of wind power prediction, especially for the power prediction of large-scale regional wind farm groups, has become an important direction of technical development in the field of wind power.

[0003] Current wind power forecasting methods mainly include methods based on physical models, statistical models and machine learning. These methods have achieved certain results in the power forecasting of a single wind farm, but there are still some significant deficiencies when dealing with large-scale regional wind power forecasting. First, traditional methods often ignore the spatial position relationship between wind farms and fail to make full use of the geographical distribution information of wind farms, resulting in a decrease in prediction accuracy. Secondly, the data processing complexity of large-scale wind farms is relatively high, and directly processing the data of all wind farms may bring redundancy and computational burden. Finally, most of the existing wind farm clustering methods use a single standard and fail to comprehensively consider the spatial heterogeneity and power characteristics of wind farms, resulting in unsatisfactory clustering effects. The root cause of these problems lies in the lack of scientific division of the comprehensive spatial and power characteristics of wind farms, which affects the accuracy and efficiency of regional wind power forecasting. Summary of the invention

[0004] The embodiment of the present invention provides a method, device, electronic device and storage medium for selecting a representative wind farm of a wind farm cluster based on spatial proximity and power clustering. The implementation of the present invention can improve the accuracy and efficiency of wind farm station group power prediction.

[0005] An embodiment of the present invention provides a method for selecting a representative wind farm of a wind farm cluster based on spatial proximity and power clustering, comprising:

[0006] Obtain the spatial location information of all wind farms in the wind farm group and the power generation data at the typical daily sampling time;

[0007] According to the spatial location information, all wind farms are divided to generate a spatial proximity structure of the wind farm group with the goal that the connecting edges between any two wind farms can only intersect at the shared endpoints in the geometric space;

[0008] According to the power generation data, the neighborhood power entropy of each wind farm in the spatial neighborhood structure of the wind farm group is calculated; according to the neighborhood power entropy of the wind farm, the wind farms in the spatial neighborhood structure of the wind farm group are clustered to generate a plurality of wind farm clusters;

[0009] The mutual correlation coefficient between the power of each wind farm in each wind farm cluster and the total power of the wind farms in the cluster is calculated, and the wind farm with the largest mutual correlation coefficient is taken as the corresponding cluster representative wind farm.

[0010] Furthermore, after calculating the correlation coefficient between the power of each wind farm in each wind farm cluster and the total power of the wind farms in the cluster, and taking the wind farm with the largest correlation coefficient as the corresponding cluster representative wind farm, the method further includes:

[0011] The wind power of the wind farm group is predicted according to the cluster representative wind farms.

[0012] Furthermore, according to the spatial location information, with the goal that the connecting edges between any two wind farms can only intersect at the shared endpoints in the geometric space, all wind farms are divided to generate a spatial proximity structure of the wind farm group, including:

[0013] According to the spatial location information of all wind farms in the wind farm group, the Delaunay meshing method is used, with the goal that the connecting edges between any two wind farms can only intersect at the shared endpoints in the geometric space. All wind farms in the wind farm group are divided to generate a Delaunay triangulated network connecting all wind farms.

[0014] According to the spatial location information of all wind farms in the wind farm group, the edge weight threshold of each wind farm is calculated;

[0015] Determine whether the edge weight between each wind farm and the directly adjacent wind farm is greater than the corresponding edge weight threshold. If so, delete the corresponding edge from the Delaunay triangulation to generate an optimized Delaunay triangulation. If two wind farms are directly connected by a connecting edge, the two wind farms are said to be directly adjacent wind farms.

[0016] The wind farms that have no connection edges with other wind farms in the optimized Delaunay triangulation are deleted to generate the spatial proximity structure of the wind farm group.

[0017] Furthermore, the edge weight threshold of each wind farm is calculated according to the spatial location information of all wind farms in the wind farm group, including:

[0018] For each wind farm, according to the spatial location information of all wind farms in the wind farm group, the average value and standard deviation of the edge weights between the current wind farm and the directly adjacent wind farms are calculated to generate a corresponding first average value and first standard deviation;

[0019] The first standard deviations of all wind farms are averaged to generate a second standard deviation;

[0020] The sum of the first average value and the second standard deviation of each wind farm is used as the edge weight threshold of the corresponding wind farm.

[0021] Furthermore, the neighborhood power entropy is calculated by the following formula:

[0022]

[0023] in, For wind farm F i The neighborhood power entropy of ; T is the number of sampling times for sampling the power generation data of each wind farm on a typical day; For the wind farm F in the spatially adjacent structure i The collection of directly adjacent wind farms; P i (t) is the wind farm F i Wind power at time t; P j (t) is the same as F i Wind power of the directly adjacent wind farm at time t.

[0024] Furthermore, the wind farms in the spatial proximity structure of the wind farm group are clustered according to the neighborhood power entropy of the wind farms to generate a number of wind farm clusters, including:

[0025] Add all wind farms in the spatial proximity structure of the wind farm group to the target wind farm set;

[0026] Repeat the clustering operation until the target wind farm set is empty, and generate several wind farm clusters;

[0027] The clustering operation includes:

[0028] The wind farm with the largest neighborhood power entropy in the current target wind farm concentration is selected as the current initial clustering center, and the current clustering cluster is established with it as the core;

[0029] Each wind farm in the current target wind farm set whose distance from the initial cluster center in the spatial proximity structure is not greater than a preset distance is recorded as a target wind farm; wherein the value of the preset distance is greater than 1 and less than the number of wind farms in the current target wind farm set;

[0030] Calculate the similarity between each target wind farm and the current initial cluster center, and add the target wind farms whose similarity is greater than a preset similarity threshold to the current cluster;

[0031] The target wind farm in the current cluster is removed from the target wind farm set, and the target wind farm set is updated;

[0032] The similarity is calculated by the following formula:

[0033]

[0034] Among them, S(F a , F c ) is the target wind farm F a With the current initial cluster center F c The similarity of P is , T is the number of sampling times for the power generation data of each wind farm on a typical day; P a (t) is the target wind farm F a Wind power at time t; The target wind farm F a The average value of the power generation data sampled on a typical day; P c (t) is the initial cluster center F c Wind power at time t; is the initial cluster center F c The average value of the generated power data sampled on a typical day.

[0035] Furthermore, the step of calculating the correlation coefficient between the power of each wind farm in each wind farm cluster and the total power of the wind farms in the cluster, and taking the wind farm with the largest correlation coefficient as the corresponding cluster representative wind farm, includes:

[0036] For each wind farm cluster, a cluster representative calculation operation is performed to generate a cluster representative wind farm of each wind farm cluster;

[0037] The clusters represent computations, including:

[0038] The wind farms in the current wind farm cluster are set as sub-wind farms;

[0039] According to the power generation data of the sub-wind farms at the typical daily sampling time, the power of each sub-wind farm and the current total power of the cluster wind farm are mapped to the corresponding frequency interval through the Sturges rule to generate interval discretization results;

[0040] According to the interval discretization result, calculate the joint probability distribution of the power of each sub-wind farm and the total power of the current clustered wind farm in each frequency interval combination, the marginal probability distribution of the power of each sub-wind farm in each frequency interval, and the marginal probability distribution of the total power of the current clustered wind farm in each frequency interval;

[0041] According to the joint probability distribution of the power of each sub-wind farm and the total power of the current cluster wind farm in each frequency interval combination, the marginal probability distribution of the power of each sub-wind farm in each frequency interval and the marginal probability distribution of the total power of the current cluster wind farm in each frequency interval, the mutual correlation coefficient between the power of each sub-wind farm and the total power of the cluster wind farm is calculated, and the wind farm with the largest mutual correlation coefficient is taken as the cluster representative wind farm of the current cluster.

[0042] Based on the above method embodiment, the present invention provides a corresponding device embodiment.

[0043] An embodiment of the present invention provides a device for selecting wind farm cluster representatives based on spatial proximity and power clustering, comprising: a power generation data acquisition module, a spatial proximity structure generation module, a wind farm cluster generation module and a cluster representative wind farm selection module;

[0044] The power generation data acquisition module is used to acquire the spatial location information of all wind farms in the wind farm group and the power generation data at a typical daily sampling time;

[0045] The spatial proximity structure generation module is used to divide all wind farms according to the spatial position information and generate a spatial proximity structure of the wind farm group with the goal that the connecting edges between any two wind farms can only intersect at the shared endpoints in the geometric space;

[0046] The wind farm clustering generation module is used to calculate the neighborhood power entropy of each wind farm in the spatial neighborhood structure of the wind farm group according to the power generation data; cluster the wind farms in the spatial neighborhood structure of the wind farm group according to the neighborhood power entropy of the wind farms to generate a plurality of wind farm clusters;

[0047] The cluster representative wind farm selection module is used to calculate the correlation coefficient between the power of each wind farm in each wind farm cluster and the total power of the cluster wind farms, and take the wind farm with the largest correlation coefficient as the corresponding cluster representative wind farm.

[0048] Based on the above method embodiment, the present invention provides a corresponding electronic device embodiment.

[0049] An embodiment of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for selecting wind farm cluster representative wind farms based on spatial proximity and power clustering as described in any one of the above method embodiments can be implemented.

[0050] Based on the above method item embodiments, the present invention provides a corresponding storage medium item embodiment.

[0051] An embodiment of the present invention provides a storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for selecting representative wind farms of wind farm clusters based on spatial proximity and power clustering as described in any one of the above method embodiments can be implemented.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] The embodiment of the present invention provides a method, device, electronic device and storage medium for selecting representative wind farms of wind farm clusters based on spatial proximity and power clustering. The method divides all wind farms according to the spatial location information of all wind farms in the wind farm group to generate a spatial proximity structure of the wind farm group; then calculates the neighborhood power entropy of all wind farms in the spatial proximity structure to generate a number of wind farm clusters; finally, according to the correlation coefficient between the power of each wind farm and the total power of the cluster wind farm, selects the representative wind farm of the wind farm cluster to predict the wind power of the wind farm group.

[0054] The present invention divides all wind farms according to the spatial location information of each wind farm, with the goal that the connecting edges between any two wind farms can only intersect at shared endpoints in geometric space, taking into account the spatial heterogeneity of wind farms, and alleviating the problem that the prior art fails to fully utilize the geographical distribution information of wind farms, resulting in a decrease in the accuracy of wind power prediction. In addition, by clustering the wind farms in the spatial proximity structure of the wind farm group according to the neighborhood power entropy of the wind farm, the problem that the prior art lacks consideration of the power characteristics of the wind farm in power prediction is solved. The selection of cluster representatives for the cluster clusters is conducive to reducing the computational burden of power prediction. In summary, the present invention fully considers the spatial proximity relationship between wind farm stations, clusters according to the neighborhood power entropy between wind farm stations, and uses cluster representatives to characterize the information of cluster clusters, thereby improving the accuracy and efficiency of power prediction of wind farm station groups. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a flowchart of a method for selecting representative wind farms of wind farm clusters based on spatial proximity and power clustering provided by an embodiment of the present invention.

[0056] Figure 2It is a flowchart of a method for selecting representative wind farms of wind farm clusters based on spatial proximity and power clustering provided by another embodiment of the present invention.

[0057] Figure 3 It is a schematic diagram of the structure of a device for selecting representative wind farms of wind farm clusters based on spatial proximity and power clustering provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0059] like Figure 1 As shown, an embodiment of the present invention provides a method for selecting representative wind farms of wind farm clusters based on spatial proximity and power clustering, which at least includes the following steps:

[0060] Step S1: Acquire the spatial location information of all wind farms in the wind farm group and the power generation data at a typical daily sampling time.

[0061] Specifically, the spatial location information of all wind farms in the wind farm group is obtained, including the geographical coordinates of each wind farm and its relative distribution relationship, as well as the spatial proximity and mutual relationship of each wind farm; at the same time, the power generation data at each time point of a typical day is collected to fully reflect the power output characteristics of the wind farm at different time periods of the day. Preferably, the power generation data of all wind farms in the wind farm group from 0:00 on a typical day to the next 24 hours is collected with a sampling frequency of 15 minutes, and 96 groups of power output are obtained.

[0062] Step S2: based on the spatial location information and the spatial position information, with the goal that the connecting edges between any two wind farms can only intersect at the shared endpoints in the geometric space, all wind farms are divided to generate a spatial proximity structure of the wind farm group.

[0063] Specifically, according to the spatial location information of the wind farm, the wind farm group is divided according to geometric rules to ensure that the connection edges between any two wind farms intersect only at the shared endpoints in the geometric space, thereby avoiding the intersection or overlap of the connection edges. Through this division process, a spatial proximity structure that can accurately reflect the spatial distribution characteristics and proximity relationship of the wind farm group is generated, providing a clear geometric framework for subsequent analysis and processing.

[0064] In a preferred embodiment, the spatial location information is used to divide all wind farms so as to generate a spatial proximity structure of wind farm groups with the goal that the connecting edges between any two wind farms can intersect only at shared endpoints in the geometric space, including:

[0065] According to the spatial location information of all wind farms in the wind farm group, the Delaunay meshing method is used, with the goal that the connecting edges between any two wind farms can only intersect at the shared endpoints in the geometric space. All wind farms in the wind farm group are divided to generate a Delaunay triangulated network connecting all wind farms.

[0066] According to the spatial location information of all wind farms in the wind farm group, the edge weight threshold of each wind farm is calculated;

[0067] Determine whether the edge weight between each wind farm and the directly adjacent wind farm is greater than the corresponding edge weight threshold. If so, delete the corresponding edge from the Delaunay triangulation to generate an optimized Delaunay triangulation. If two wind farms are directly connected by a connecting edge, the two wind farms are said to be directly adjacent wind farms.

[0068] The wind farms that have no connection edges with other wind farms in the optimized Delaunay triangulation are deleted to generate the spatial proximity structure of the wind farm group.

[0069] In actual operation, the edge weight threshold of each wind farm is calculated based on the spatial location information of all wind farms in the wind farm group, including:

[0070] For each wind farm, according to the spatial location information of all wind farms in the wind farm group, the average value and standard deviation of the edge weights between the current wind farm and the directly adjacent wind farms are calculated to generate a corresponding first average value and first standard deviation;

[0071] The first standard deviations of all wind farms are averaged to generate a second standard deviation;

[0072] The sum of the first mean value and the second standard deviation of each wind farm is used as the edge weight threshold of the corresponding wind farm, that is:

[0073] C Global (F)=L mean (F)+σ Global

[0074] Among them, C Global (F) is the edge weight threshold of the wind farm; L mean (F) is the first average value of the wind farm; σ Global is the second standard deviation.

[0075] Step S3: calculating the neighborhood power entropy of each wind farm in the spatial neighborhood structure of the wind farm group according to the power generation data; clustering the wind farms in the spatial neighborhood structure of the wind farm group according to the neighborhood power entropies of the wind farms to generate a plurality of wind farm clusters.

[0076] Specifically, based on the power generation data and combined with the spatial proximity structure of the wind farm group, the power entropy of each wind farm in its neighborhood is calculated to quantify the power distribution characteristics of the surrounding wind farms. Subsequently, with the neighborhood power entropy as the key indicator, cluster analysis is performed on the wind farms in the spatial proximity structure, and the spatial location and power distribution characteristics of the wind farms are comprehensively considered to generate several wind farm clusters, laying the foundation for the characteristic induction and precise analysis of the wind farm group.

[0077] Preferably, the neighborhood power entropy is calculated by the following formula:

[0078]

[0079] in, For wind farm F i The neighborhood power entropy of ; T is the number of sampling times for sampling the power generation data of each wind farm on a typical day; For the wind farm F in the spatially adjacent structure i The collection of directly adjacent wind farms; P i (t) is the wind farm F i Wind power at time t; P j (t) is the same as F i Wind power of the directly adjacent wind farm at time t.

[0080] In an optional embodiment, clustering the wind farms in the spatial proximity structure of the wind farm group according to the neighborhood power entropy of the wind farms to generate a plurality of wind farm clusters includes:

[0081] Add all wind farms in the spatial proximity structure of the wind farm group to the target wind farm set;

[0082] Repeat the clustering operation until the target wind farm set is empty, and generate several wind farm clusters;

[0083] The clustering operation includes:

[0084] The wind farm with the largest neighborhood power entropy in the current target wind farm concentration is selected as the current initial clustering center, and the current clustering cluster is established with it as the core;

[0085] Each wind farm in the current target wind farm set whose distance from the initial cluster center in the spatial proximity structure is not greater than a preset distance is recorded as a target wind farm; wherein the value of the preset distance is greater than 1 and less than the number of wind farms in the current target wind farm set; the preset distance here can be 2, and can also be flexibly adjusted according to actual conditions.

[0086] Calculate the similarity between each target wind farm and the current initial cluster center, and add the target wind farms whose similarity is greater than a preset similarity threshold to the current cluster;

[0087] The target wind farm in the current cluster is removed from the target wind farm set, and the target wind farm set is updated;

[0088] The similarity is calculated by the following formula:

[0089]

[0090] Among them, S(F a , F c ) is the target wind farm F a With the current initial cluster center F c The similarity of P is , T is the number of sampling times for the power generation data of each wind farm on a typical day; P a (t) is the target wind farm F a Wind power at time t; The target wind farm F a The average value of the power generation data sampled on a typical day; P c (t) is the initial cluster center F c Wind power at time t; is the initial cluster center F c The average value of the generated power data sampled on a typical day.

[0091] It should be noted here that the preset similarity threshold can be calculated by the following formula:

[0092] S threshold =α·log(n)

[0093] Among them, S threshold is the similarity threshold; α is the weight coefficient, usually α∈[0.5, 0.8]; n is the number of wind farms in the current target wind farm concentration.

[0094] Step S4: Calculate the correlation coefficient between the power of each wind farm in each wind farm cluster and the total power of the wind farms in the cluster, and take the wind farm with the largest correlation coefficient as the corresponding cluster representative wind farm.

[0095] Specifically, the wind farms in each wind farm cluster are analyzed one by one, and based on their power generation data, the correlation coefficient between the power of each wind farm and the total power of the cluster is calculated to quantify the representativeness of the individual wind farm to the overall power characteristics of the cluster. The power output of a wind farm is usually affected by many factors, such as wind speed, weather, etc. These factors may cause the power output between wind farms to be nonlinear. Mutual information can capture more complex dependencies. Subsequently, the wind farm with the highest correlation coefficient is selected as the representative wind farm of the cluster, so as to ensure that the representative wind farm selected in each cluster can accurately reflect the overall power characteristics of the cluster, providing a reliable basis for subsequent analysis and application.

[0096] In a preferred embodiment, the step of calculating the correlation coefficient between the power of each wind farm in each wind farm cluster and the total power of the wind farms in the cluster, and taking the wind farm with the largest correlation coefficient as the corresponding cluster representative wind farm, comprises:

[0097] For each wind farm cluster, a cluster representative calculation operation is performed to generate a cluster representative wind farm of each wind farm cluster;

[0098] The clusters represent computations, including:

[0099] The wind farms in the current wind farm cluster are set as sub-wind farms;

[0100] According to the power generation data of the sub-wind farms at the typical daily sampling time, the power of each sub-wind farm and the current total power of the cluster wind farm are mapped to the corresponding frequency interval through the Sturges rule to generate interval discretization results;

[0101] According to the interval discretization result, calculate the joint probability distribution of the power of each sub-wind farm and the total power of the current clustered wind farm in each frequency interval combination, the marginal probability distribution of the power of each sub-wind farm in each frequency interval, and the marginal probability distribution of the total power of the current clustered wind farm in each frequency interval;

[0102] According to the joint probability distribution of the power of each sub-wind farm and the total power of the current cluster wind farm in each frequency interval combination, the marginal probability distribution of the power of each sub-wind farm in each frequency interval, and the marginal probability distribution of the total power of the current cluster wind farm in each frequency interval, the mutual correlation coefficient between the power of each sub-wind farm and the total power of the cluster wind farm is calculated, and the wind farm with the largest mutual correlation coefficient is taken as the cluster representative wind farm of the current cluster, that is:

[0103]

[0104] Among them, I(F i , F total) is the sub-wind farm F i The mutual correlation coefficient with the total power of the cluster wind farm, p(p i p total ) is the sub-wind farm F i The joint probability distribution of the power and the total power of the current cluster wind farm; p(p i ) is the sub-wind farm F i The marginal probability distribution of the power; p(p total ) is the marginal probability distribution of the total power of the clustered wind farms.

[0105] Optionally, assume that the power data points of the wind farm are P = [P(1), P(2), ..., P(96)], and each data point represents the power output every 15 minutes within 24 hours in a day.

[0106] According to the minimum value P of the data min and the maximum value P maz , define the interval boundaries.

[0107] Based on the sample size (96 data points), Sturges' Rule is used to determine the reasonable number of intervals k. Sturges' Rule is usually used for small sample data, and the estimation formula is:

[0108] k = [log2(n) + 1]

[0109] Where n = 96, which is the number of data points. k = [log2(96)+1]≈8, 8 intervals are selected.

[0110] If the wind farm power data is continuously distributed and has no significant central trend, use the equal width binning method. If the data is obviously skewed or concentrated, use the equal frequency binning or quantile binning method to ensure that the number of data points in each interval is roughly equal.

[0111] Assign data points to intervals, count the number of data points contained in each interval, and determine the frequency of each interval.

[0112] like Figure 2 As shown, another embodiment of the present invention provides a method for selecting a representative wind farm of a wind farm cluster based on spatial proximity and power clustering. In addition to steps S1 to S4 of the above embodiment, the method further includes the following steps after step S4:

[0113] Step S5: predicting the wind power of the wind farm group according to the cluster representative wind farm.

[0114] Specifically, based on the determined cluster representative wind farm, its power generation data and historical operation characteristics are extracted, and combined with the overall characteristics of the corresponding wind farm cluster, a model for predicting the power of the wind farm group is constructed. By analyzing the power change law of the cluster representative wind farm and its association with external environmental factors (such as meteorological conditions), the total power generation of the entire wind farm group is further calculated and predicted, thus providing a scientific basis for wind power optimization scheduling and stable operation of the power system.

[0115] like Figure 3 As shown, an embodiment of the present invention provides a device for selecting wind farm cluster representatives based on spatial proximity and power clustering, including: a power generation data acquisition module, a spatial proximity structure generation module, a wind farm cluster generation module and a cluster representative wind farm selection module.

[0116] The power generation data acquisition module is used to acquire the spatial location information of all wind farms in the wind farm group and the power generation data at a typical daily sampling time;

[0117] The spatial proximity structure generation module is used to divide all wind farms according to the spatial position information and generate a spatial proximity structure of the wind farm group with the goal that the connecting edges between any two wind farms can only intersect at the shared endpoints in the geometric space;

[0118] The wind farm clustering generation module is used to calculate the neighborhood power entropy of each wind farm in the spatial neighborhood structure of the wind farm group according to the power generation data; cluster the wind farms in the spatial neighborhood structure of the wind farm group according to the neighborhood power entropy of the wind farms to generate a plurality of wind farm clusters;

[0119] The cluster representative wind farm selection module is used to calculate the correlation coefficient between the power of each wind farm in each wind farm cluster and the total power of the cluster wind farms, and take the wind farm with the largest correlation coefficient as the corresponding cluster representative wind farm.

[0120] It should be noted that the embodiment of the device described above corresponds to the above-mentioned embodiment of the present invention, and it can implement any of the above-mentioned methods for selecting wind farm clusters representing wind farms based on spatial proximity and power clustering of the present invention. In addition, the above-mentioned embodiment of the device is merely schematic, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the drawings of the embodiment of the device provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which can be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art can understand and implement it without paying any creative work.

[0121] Based on the above method embodiment of the present invention, a corresponding electronic device embodiment is provided.

[0122] An embodiment of the present invention provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the method for selecting wind farm clusters representing wind farms based on spatial proximity and power clustering as described in any one of the present invention is implemented, or, when the processor executes the computer program, the functions of the modules in the above-mentioned device embodiments are implemented.

[0123] Exemplarily, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to implement the present invention. The one or more modules may be a series of computer program instruction segments capable of implementing specific functions, which are used to describe the execution process of the computer program in the terminal device.

[0124] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0125] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the terminal device, and uses various interfaces and lines to connect various parts of the entire terminal device.

[0126] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the terminal device by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created according to the use of the mobile phone, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0127] Based on the above method embodiment, the present invention provides a storage medium embodiment;

[0128] Another embodiment of the present invention provides a storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute any one of the above-mentioned methods for selecting representative wind farms for wind farm clusters based on spatial proximity and power clustering of the present invention.

[0129] The storage medium is a computer-readable storage medium, and the computer program includes computer program code, which may be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0130] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.

[0131] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for selecting representative wind farms in wind farm clusters based on spatial proximity and power clustering, characterized in that: include: Obtain the spatial location information of all wind farms in the wind farm group and the power generation data at the typical daily sampling time; According to the spatial location information, all wind farms are divided to generate a spatial proximity structure of the wind farm group with the goal that the connecting edges between any two wind farms can only intersect at the shared endpoints in the geometric space; According to the power generation data, the neighborhood power entropy of each wind farm in the spatial neighborhood structure of the wind farm group is calculated; according to the neighborhood power entropy of the wind farm, the wind farms in the spatial neighborhood structure of the wind farm group are clustered to generate a plurality of wind farm clusters; The mutual correlation coefficient between the power of each wind farm in each wind farm cluster and the total power of the wind farms in the cluster is calculated, and the wind farm with the largest mutual correlation coefficient is taken as the corresponding cluster representative wind farm.

2. The method for selecting representative wind farms of wind farm clusters based on spatial proximity and power clustering according to claim 1, characterized in that: After calculating the correlation coefficient between the power of each wind farm and the total power of the wind farms in the cluster in each wind farm cluster, and taking the wind farm with the largest correlation coefficient as the corresponding cluster representative wind farm, the method further includes: The wind power of the wind farm group is predicted according to the cluster representative wind farms.

3. The method for selecting representative wind farms of wind farm clusters based on spatial proximity and power clustering according to claim 2, characterized in that: According to the spatial position information, with the goal that the connecting edges between any two wind farms can only intersect at the shared endpoints in the geometric space, all wind farms are divided to generate a spatial proximity structure of the wind farm group, including: According to the spatial location information of all wind farms in the wind farm group, the Delaunay meshing method is used, with the goal that the connecting edges between any two wind farms can only intersect at the shared endpoints in the geometric space. All wind farms in the wind farm group are divided to generate a Delaunay triangulated network connecting all wind farms. According to the spatial location information of all wind farms in the wind farm group, the edge weight threshold of each wind farm is calculated; Determine whether the edge weight between each wind farm and the directly adjacent wind farm is greater than the corresponding edge weight threshold. If so, delete the corresponding edge from the Delaunay triangulation to generate an optimized Delaunay triangulation. If two wind farms are directly connected by a connecting edge, the two wind farms are said to be directly adjacent wind farms. The wind farms that have no connection edges with other wind farms in the optimized Delaunay triangulation are deleted to generate the spatial proximity structure of the wind farm group.

4. The method for selecting representative wind farms of wind farm clusters based on spatial proximity and power clustering according to claim 3, characterized in that: The step of calculating the edge weight threshold of each wind farm according to the spatial location information of all wind farms in the wind farm group includes: For each wind farm, according to the spatial location information of all wind farms in the wind farm group, the average value and standard deviation of the edge weights between the current wind farm and the directly adjacent wind farms are calculated to generate a corresponding first average value and first standard deviation; The first standard deviations of all wind farms are averaged to generate a second standard deviation; The sum of the first average value and the second standard deviation of each wind farm is used as the edge weight threshold of the corresponding wind farm.

5. The method for selecting representative wind farms of wind farm clusters based on spatial proximity and power clustering according to claim 4, characterized in that: The neighborhood power entropy is calculated by the following formula: in, For wind farm F i The neighborhood power entropy of ; T is the number of sampling times for sampling the power generation data of each wind farm on a typical day; For the wind farm F in the spatially adjacent structure i The collection of directly adjacent wind farms; P i (t) is the wind farm F i Wind power at time t; P j (t) is the same as F i Wind power of the directly adjacent wind farm at time t.

6. The method for selecting representative wind farms of wind farm clusters based on spatial proximity and power clustering according to claim 5, characterized in that: The wind farms in the spatial proximity structure of the wind farm group are clustered according to the neighborhood power entropy of the wind farms to generate a number of wind farm clusters, including: Add all wind farms in the spatial proximity structure of the wind farm group to the target wind farm set; Repeat the clustering operation until the target wind farm set is empty, and generate several wind farm clusters; The clustering operation includes: The wind farm with the largest neighborhood power entropy in the current target wind farm concentration is selected as the current initial clustering center, and the current clustering cluster is established with it as the core; Each wind farm in the current target wind farm set whose distance from the initial cluster center in the spatial proximity structure is not greater than a preset distance is recorded as a target wind farm; wherein the value of the preset distance is greater than 1 and less than the number of wind farms in the current target wind farm set; Calculate the similarity between each target wind farm and the current initial cluster center, and add the target wind farms whose similarity is greater than a preset similarity threshold to the current cluster cluster; The target wind farm in the current cluster is removed from the target wind farm set, and the target wind farm set is updated; The similarity is calculated by the following formula: Among them, S(F a , F c ) is the target wind farm F a With the current initial cluster center F c The similarity of P is , T is the number of sampling times for the power generation data of each wind farm on a typical day; P a (t) is the target wind farm F a Wind power at time t; The target wind farm F a The average value of the power generation data sampled on a typical day; P c (t) is the initial cluster center F c Wind power at time t; is the initial cluster center F c The average value of the generated power data sampled on a typical day.

7. The method for selecting representative wind farms of wind farm clusters based on spatial proximity and power clustering according to claim 6, characterized in that: The step of calculating the correlation coefficient between the power of each wind farm in each wind farm cluster and the total power of the wind farms in the cluster, and taking the wind farm with the largest correlation coefficient as the corresponding cluster representative wind farm, includes: For each wind farm cluster, a cluster representative calculation operation is performed to generate a cluster representative wind farm of each wind farm cluster; The clusters represent computations, including: The wind farms in the current wind farm cluster are set as sub-wind farms; According to the power generation data of the sub-wind farms at the typical daily sampling time, the power of each sub-wind farm and the current total power of the cluster wind farm are mapped to the corresponding frequency interval through the Sturges rule to generate interval discretization results; According to the interval discretization result, calculate the joint probability distribution of the power of each sub-wind farm and the total power of the current clustered wind farm in each frequency interval combination, the marginal probability distribution of the power of each sub-wind farm in each frequency interval, and the marginal probability distribution of the total power of the current clustered wind farm in each frequency interval; According to the joint probability distribution of the power of each sub-wind farm and the total power of the current cluster wind farm in each frequency interval combination, the marginal probability distribution of the power of each sub-wind farm in each frequency interval and the marginal probability distribution of the total power of the current cluster wind farm in each frequency interval, the mutual correlation coefficient between the power of each sub-wind farm and the total power of the cluster wind farm is calculated, and the wind farm with the largest mutual correlation coefficient is taken as the cluster representative wind farm of the current cluster.

8. A device for selecting wind farm cluster representative wind farms based on spatial proximity and power clustering, characterized in that: include: Power generation data acquisition module, spatial neighbor structure generation module, wind farm cluster generation module and cluster representative wind farm selection module; The power generation data acquisition module is used to acquire the spatial location information of all wind farms in the wind farm group and the power generation data at a typical daily sampling time; The spatial proximity structure generation module is used to divide all wind farms according to the spatial position information and generate a spatial proximity structure of the wind farm group with the goal that the connecting edges between any two wind farms can only intersect at the shared endpoints in the geometric space; The wind farm clustering generation module is used to calculate the neighborhood power entropy of each wind farm in the spatial neighborhood structure of the wind farm group according to the power generation data; cluster the wind farms in the spatial neighborhood structure of the wind farm group according to the neighborhood power entropy of the wind farms to generate a plurality of wind farm clusters; The cluster representative wind farm selection module is used to calculate the correlation coefficient between the power of each wind farm in each wind farm cluster and the total power of the cluster wind farms, and take the wind farm with the largest correlation coefficient as the corresponding cluster representative wind farm.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method for selecting wind farm cluster representative wind farms based on spatial proximity and power clustering as described in any one of claims 1 to 7 can be implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it can implement the method for selecting wind farm cluster representative wind farms based on spatial proximity and power clustering as described in any one of claims 1 to 7.

Images