Distributed new energy power feature screening method and device, equipment and medium

Abstract

The embodiment of the invention discloses a distributed new energy power feature screening method and device, equipment and a medium. The method comprises the following steps: determining a meteorological factor data set and distributed new energy power matched with each meteorological factor in the meteorological factor set, and determining at least one target cluster according to the meteorological factor data set and each distributed new energy power; determining a target weather process according to the target cluster, and collecting target output data of the distributed new energy in the target weather process; determining each composite feature according to each target weather process and each target output data; screening the composite features to obtain a target feature library; the target feature library is used for training the distributed new energy power prediction model. According to the scheme provided by the embodiment of the invention, the characteristics influencing the distributed new energy power can be accurately determined, and a basis is provided for accurately predicting the distributed new energy power.

Description

Technical Field

[0001] The present invention relates to the field of distributed new energy technology, and in particular to a method, device, equipment and medium for feature screening of distributed new energy power. Background Art

[0002] Distributed Energy Resources (DER) is an energy comprehensive utilization system based on renewable clean energy and distributed at the user end. It is an efficient power supply system installed at the user end. Since the system is close to the consumption site, the energy generated at the power generation site can be dispatched and utilized in a timely manner. Due to the small scale of power generation, it has the advantages of flexibility and controllability, saving land resources and construction costs. At the same time, due to the fit and utilization of new energy, it has the advantages of less pollution and excellent ecological effects. Distributed new energy has become a power generation and energy utilization method with broad development prospects.

[0003] However, due to the distributed characteristics, there are more uncertainties in the distributed renewable energy power generation and power supply process; photovoltaic, wind power and other new energy sources are greatly affected by meteorological factors and have great uncertainty, which leads to unstable power generation.

[0004] How to accurately determine the characteristics that affect the power of distributed renewable energy, and thus accurately predict the power of distributed renewable energy, is a major research issue in the industry. Summary of the invention

[0005] The present invention provides a method, device, equipment and medium for screening characteristics of distributed renewable energy power, so as to accurately determine the characteristics affecting the distributed renewable energy power and provide a basis for accurately predicting the distributed renewable energy power.

[0006] According to one aspect of the present invention, a method for screening characteristics of distributed renewable energy power is provided, the method comprising:

[0007] Determine a meteorological factor data set, and distributed renewable energy power matching each meteorological factor in the meteorological factor set, and determine at least one target cluster according to the meteorological factor data set and each distributed renewable energy power;

[0008] Determine a target weather process according to the target cluster, and collect target output data of the distributed new energy in the target weather process;

[0009] Determining each composite feature according to each of the target weather processes and each of the target output data;

[0010] The composite features are screened to obtain a target feature library; the target feature library is used for training a distributed new energy power prediction model.

[0011] In an optional implementation of this embodiment, the meteorological factor includes at least one of the following: wind speed, wind direction, irradiance, and temperature;

[0012] The meteorological factor data set contains specific values ​​of each meteorological factor at different time nodes;

[0013] Accordingly, the determining of the meteorological factor data set and the distributed new energy power matching each meteorological factor in the meteorological factor set, and determining at least one target cluster according to the meteorological factor data set and each of the distributed new energy power, includes:

[0014] Acquire specific values ​​of each of the meteorological factors at the target historical moment and the power of each of the distributed renewable energy sources at the target historical moment from the target database to obtain each meteorological factor-power data pair;

[0015] Screening each of the meteorological factor-power data pairs to obtain each target data pair;

[0016] Clustering is performed on each of the target data pairs to obtain each of the target clusters.

[0017] In an optional implementation of this embodiment, the screening of each meteorological factor-power data pair to obtain each target data pair includes:

[0018] Calculating the correlation coefficient of each meteorological factor-power data pair respectively, and sorting each data pair according to the set correlation coefficient order to obtain a reference data pair queue;

[0019] The reference data pair queue is traversed to determine a target correlation coefficient, and all data pairs after or before the data pair corresponding to the target correlation coefficient are deleted to obtain the target data pairs.

[0020] In an optional implementation of this embodiment, determining a target weather process according to a target cluster and collecting target output data of the distributed new energy in the target weather process include:

[0021] Determining target meteorological factors in the target cluster, and determining target collection times of the target meteorological factors;

[0022] Determining the target weather process according to each of the target collection moments;

[0023] Collect target output data of the distributed renewable energy at different times during the target weather process.

[0024] In an optional implementation of this embodiment, determining each composite feature according to each target weather process and each target output data includes:

[0025] Performing cross-correlation characteristic analysis on each of the target weather processes and each of the target output data to obtain a cross-correlation characteristic analysis result;

[0026] Feature mining is performed on each of the cross-correlation characteristic analysis results to obtain each of the composite features.

[0027] In an optional implementation of this embodiment, performing feature mining on each of the cross-correlation characteristic analysis results to obtain each of the composite features includes:

[0028] Feature mining is performed on each of the cross-correlation characteristic analysis results through Featurewiz, and the mined features are screened through SGBoost to obtain each of the composite features.

[0029] In an optional implementation of this embodiment, the screening of each of the composite features to obtain a target feature library includes:

[0030] Inputting each of the composite features into a pre-trained feature screening model and a feature elimination model for processing respectively, to obtain the target feature library;

[0031] Correspondingly, after obtaining the target feature library, it also includes:

[0032] Input each target feature in the target feature library into the target machine learning model in turn for iterative training, and obtain the distributed renewable energy power prediction model when the iteration stop condition is met; the distributed renewable energy power prediction model is used to predict the power of distributed renewable energy under different meteorological factors.

[0033] According to another aspect of the present invention, a device for screening characteristics of distributed renewable energy power is provided, the device comprising:

[0034] A target cluster determination module, used to determine a meteorological factor data set and distributed new energy power matching each meteorological factor in the meteorological factor set, and determine at least one target cluster according to the meteorological factor data set and each distributed new energy power;

[0035] A target weather process determination module is used to determine the target weather process according to the target clustering cluster, and collect the target output data of the distributed new energy in the target weather process;

[0036] A composite feature determination module, used to determine each composite feature according to each target weather process and each target output data;

[0037] The target feature library determination module is used to screen each of the composite features to obtain a target feature library; the target feature library is used for training a distributed new energy power prediction model.

[0038] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0039] at least one processor; and

[0040] a memory communicatively connected to the at least one processor; wherein,

[0041] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the characteristic screening method of distributed renewable energy power described in any embodiment of the present invention.

[0042] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the characteristic screening method of distributed renewable energy power described in any embodiment of the present invention when executed.

[0043] According to another aspect of the present invention, a computer program product is provided, including a computer program, which, when executed by a processor, implements the method for screening characteristics of distributed renewable energy power according to any embodiment of the present invention.

[0044] The technical solution of the embodiment of the present invention is to determine a meteorological factor data set and distributed renewable energy power matching each meteorological factor in the meteorological factor set, and determine at least one target clustering cluster based on the meteorological factor data set and each distributed renewable energy power; determine a target weather process based on the target clustering cluster, and collect the target output data of the distributed renewable energy in the target weather process; determine each composite feature based on each target weather process and each target output data; screen each composite feature to obtain a target feature library; the target feature library is used for training a distributed renewable energy power prediction model, which can accurately determine the features that affect the distributed renewable energy power, and provide a basis for accurately predicting the distributed renewable energy power.

[0045] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0047] Figure 1 is a flow chart of a method for screening characteristics of distributed renewable energy power provided according to Embodiment 1 of the present invention;

[0048] Figure 2 is a flow chart of a method for screening characteristics of distributed renewable energy power provided according to Embodiment 2 of the present invention;

[0049] Figure 3 is a structural schematic diagram of a distributed new energy power feature screening device provided according to Embodiment 3 of the present invention;

[0050] Figure 4 It is a structural schematic diagram of an electronic device for implementing the characteristic screening method of distributed renewable energy power according to an embodiment of the present invention. DETAILED DESCRIPTION

[0051] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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 should fall within the scope of protection of the present invention.

[0052] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0053] Embodiment 1

[0054] Figure 1This is a flow chart of a method for screening characteristics of distributed renewable energy power according to the first embodiment of the present invention. This embodiment is applicable to the case of screening characteristics that affect distributed renewable energy power. The method can be performed by a distributed renewable energy power characteristic screening device. The distributed renewable energy power characteristic screening device can be implemented in the form of hardware and / or software. The distributed renewable energy power characteristic screening device can be configured in electronic devices such as computers, servers or tablet computers. Figure 1 As shown, the method includes:

[0055] Step 110: determine a meteorological factor data set and distributed renewable energy powers matching each meteorological factor in the meteorological factor set, and determine at least one target cluster according to the meteorological factor data set and each distributed renewable energy power.

[0056] Among them, meteorological factors may include wind speed, wind direction, irradiance or temperature, etc., which are not limited in this embodiment; the meteorological factor data set contains specific values ​​of each meteorological factor at different time nodes, and the specific value of each meteorological factor is called a meteorological factor data; for example, the temperature at 6 o'clock in the morning on January 1, 2024 is 1 degree Celsius, the wind speed is level 5, and the wind direction is east wind, which is called a meteorological factor data in this embodiment; the temperature at 6 o'clock in the morning on June 1, 2024 is 30 degrees Celsius, the wind speed is level 2, and the wind direction is south wind, which is called a meteorological factor data in this embodiment. It can be understood that in this embodiment, the meteorological factors of the location or area where the distributed new energy is located are obtained, and the locations of the two are consistent.

[0057] In this embodiment, the distributed new energy power that matches each meteorological factor in the meteorological factor set can be the power generation power of the distributed new energy system at the moment corresponding to the target meteorological factor, for example, the power generation power of the distributed new energy system at 6 am on January 1, 2024 in the above-mentioned meteorological factors; or the power generation power of the distributed new energy system at 6 am on June 1, 2024, which is not limited in this embodiment.

[0058] In an optional implementation of this embodiment, determining a meteorological factor data set, and a distributed new energy power matching each meteorological factor in the meteorological factor set, may include: obtaining from a target database the specific values ​​of each of the meteorological factors at the target historical moment, and the power of each of the distributed new energy sources at the target historical moment, to obtain each meteorological factor-power data pair.

[0059] The target database may be any local database storing meteorological factor data and distributed renewable energy power, or a cloud database, which is not limited in the present embodiment.

[0060] Optionally, in this embodiment, the meteorological factors at the target time and the power of the distributed new energy at the target time can be obtained from the target database in sequence, and the meteorological factors at the target time and the power of the distributed new energy can be combined into a meteorological factor-power data pair.

[0061] For example, if the target time is 12:00 on September 1, 2024, then the meteorological factors in the location area of ​​the distributed renewable energy at 12:00 on September 1, 2024, and the power generation power of the distributed renewable energy at 12:00 on September 1, 2024 can be obtained, and the meteorological factors and power generation power at that moment can be combined into a meteorological factor-power data pair.

[0062] Correspondingly, in this embodiment, determining at least one target cluster based on the meteorological factor data set and each of the distributed new energy powers may include: screening each of the meteorological factor-power data pairs to obtain each target data pair; and clustering each of the target data pairs to obtain each of the target clusters.

[0063] In an optional implementation of the present embodiment, after obtaining each meteorological factor-power data pair, each meteorological factor-power data pair can be further screened to obtain each screened target data pair; further, each target data pair can be clustered to obtain multiple target cluster clusters.

[0064] Optionally, in this embodiment, after obtaining each meteorological factor-power data pair, a correlation analysis can be performed on each meteorological factor-power data pair, and each meteorological factor-power data pair with a low correlation can be deleted to obtain a target data pair; further, each target data pair can be subjected to unsupervised clustering to obtain multiple target clustering clusters.

[0065] Step 120: determine a target weather process according to the target cluster, and collect target output data of the distributed new energy in the target weather process.

[0066] The target weather process may be the change of meteorological factors within a time period, which may include the change of one meteorological factor or the change of multiple meteorological factors; for example, it may be the temperature change from 6 am to 8 pm, or the temperature change and wind speed change from 6 am to 8 pm, which are not limited in this embodiment. Correspondingly, the target output data of the distributed new energy may be the power change of the distributed new energy within the time period, or the maximum power of the distributed new energy within the time period, which are not limited in this embodiment.

[0067] In an optional implementation of this embodiment, the meteorological factor data in the target cluster can be analyzed to determine the time of each meteorological factor data, and the times can be pieced together to obtain the time corresponding to the target weather process, for example, 7 to 8 o'clock or 6 to 12 o'clock, etc.; further, the specific content of each meteorological factor data can be analyzed to obtain the specific weather corresponding to the target weather process, for example, low temperature, or strong wind, etc. After determining the target weather process, the output data of the distributed renewable energy in the target weather process can be further determined, for example, the real-time power of the distributed renewable energy in the target weather process can be determined.

[0068] Step 130: Determine each composite feature according to each target weather process and each target output data.

[0069] In an optional implementation of this embodiment, after determining the target weather process and the target output data of the distributed new energy in the target weather process, multiple composite features can be further determined based on the target weather process and the target output data.

[0070] Optionally, in this embodiment, after determining the target weather process and the target output data of the distributed new energy in the target weather process, relevant characteristic analysis can be performed on each target weather process and the target output data of the distributed new energy, and further, feature mining can be performed on the relevant characteristic analysis results, and various composite features can be constructed based on the mining results.

[0071] It can be understood that in this embodiment, the composite feature can be a combination of different meteorological factors, such as the superposition of temperature and wind speed or the superposition of irradiance and temperature, etc.; different forms of the same meteorological factor, such as the cube of irradiance, are not limited in this embodiment.

[0072] Step 140: Screen the composite features to obtain a target feature library.

[0073] Among them, the target feature library is used for training the distributed new energy power prediction model.

[0074] In an optional implementation of this embodiment, after determining each composite feature, each composite feature can be further screened, for example, by deleting duplicate features or useless features in the composite features, so as to obtain a target feature library; further, the distributed new energy power prediction model can be trained through the target feature library.

[0075] The technical solution of this embodiment is to determine a meteorological factor data set and distributed renewable energy power matching each meteorological factor in the meteorological factor set, and determine at least one target clustering cluster based on the meteorological factor data set and each distributed renewable energy power; determine a target weather process based on the target clustering cluster, and collect the target output data of the distributed renewable energy in the target weather process; determine each composite feature based on each target weather process and each target output data; screen each composite feature to obtain a target feature library; the target feature library is used for training a distributed renewable energy power prediction model, and can accurately determine the features that affect the distributed renewable energy power, providing a basis for accurately predicting the distributed renewable energy power.

[0076] Embodiment 2

[0077] Figure 2 1 is a flow chart of a method for screening the characteristics of distributed renewable energy power according to Embodiment 2 of the present invention. This embodiment is a further refinement of the above technical solution. The technical solution in this embodiment can be combined with each optional solution in one or more of the above embodiments. Figure 2 As shown, the method includes:

[0078] Step 210: determine a meteorological factor data set and distributed renewable energy powers matching each meteorological factor in the meteorological factor set, and determine at least one target cluster according to the meteorological factor data set and each distributed renewable energy power.

[0079] In an optional implementation method of the present embodiment, the specific values ​​of each of the meteorological factors at the target historical moment and the power of each of the distributed new energy sources at the target historical moment can be obtained from the target database to obtain each meteorological factor-power data pair; each of the meteorological factor-power data pairs is screened to obtain each target data pair; each of the target data pairs is clustered to obtain each target cluster cluster.

[0080] The target database may be any local server or cloud server storing data of various meteorological factors and power of distributed new energy, which is not limited in this embodiment. The target historical moment may be any historical moment, which is not limited in this embodiment.

[0081] In this embodiment, data of various meteorological factors, that is, specific values ​​of various meteorological factors at the target historical moment and the power of distributed new energy at the target moment, can be obtained from the target database, so as to obtain meteorological factor-power data pairs at each historical moment; further, each meteorological factor-power data pair can be screened to obtain each target data pair; further, each target data pair can be clustered to obtain each target cluster cluster.

[0082] In an optional implementation of the present embodiment, each of the meteorological factor-power data pairs is screened to obtain each target data pair, which may include: calculating the correlation coefficient of each of the meteorological factor-power data pairs respectively, and sorting each of the data pairs in the order of the set correlation coefficient to obtain a reference data pair queue; traversing the reference data pair queue, determining the target correlation coefficient, and deleting all data pairs behind or in front of the data pair corresponding to the target correlation coefficient to obtain each of the target data pairs.

[0083] The order of setting the correlation coefficients may be from small to large or from large to small, which is not limited in this embodiment. The target correlation coefficient may be a pre-set correlation coefficient, such as 0.8 or 0.7, which is not limited in this embodiment.

[0084] Optionally, in this embodiment, the Pearson correlation coefficient of each meteorological factor-power data pair can be calculated. Furthermore, the calculation results of each Pearson correlation coefficient can be sorted in order from small to large to obtain a reference data pair queue; further, the reference data pair queue can be traversed and all data pairs in front of the target correlation coefficient can be deleted to obtain each target data pair.

[0085] It can be understood that if the correlation coefficient order is set from small to large, then all data pairs before the target correlation coefficient will be deleted subsequently; if the correlation coefficient order is set from large to small, then all data pairs after the target correlation coefficient will be deleted subsequently, that is, the data pairs deleted in this embodiment are data pairs with correlation coefficients smaller than the target correlation coefficient.

[0086] Furthermore, each target data pair may be subjected to unsupervised clustering to obtain each target cluster. Exemplarily, in this embodiment, the OPTICS method may be used to perform cluster analysis on each target data pair to divide the target data pair into several subclusters, ie, different target clusters.

[0087] Step 220, determine the target meteorological factors in the target cluster, and determine the target collection times of the target meteorological factors; determine the target weather process according to the target collection times; collect the target output data of the distributed new energy at different times in the target weather process.

[0088] In an optional implementation of the present embodiment, after determining each target cluster, the target meteorological factor in the target cluster can be further determined, and the target collection time of the target meteorological factor can be determined, that is, the time corresponding to the target meteorological factor; further, the target weather process can be determined according to the target collection time, and further, the target output data of distributed new energy at different times in the target weather process can be collected.

[0089] Step 230: performing cross-correlation characteristic analysis on each of the target weather processes and each of the target output data to obtain a cross-correlation characteristic analysis result; performing feature mining on each of the cross-correlation characteristic analysis results to obtain each of the composite features.

[0090] In an optional implementation of the present embodiment, after determining each target weather process and the target output data of the distributed new energy in each target weather process, the target weather process and the target output data can be further analyzed for cross-correlation characteristics to obtain cross-correlation characteristic analysis results; further, feature mining can be performed on each cross-correlation characteristic analysis result to obtain various composite features.

[0091] Optionally, in this embodiment, performing feature mining on each of the cross-correlation characteristic analysis results to obtain each of the composite features may include: performing feature mining on each of the cross-correlation characteristic analysis results through Featurewiz, and filtering the mined features through SGBoost to obtain each of the composite features.

[0092] The advantage of this setting is that the composite features can be quickly determined, providing a basis for obtaining the target feature library later.

[0093] Step 240: Input each of the composite features into a pre-trained feature screening model and a feature elimination model for processing to obtain the target feature library.

[0094] In an optional implementation of this embodiment, after obtaining each composite feature, each composite feature can be input into a pre-trained feature screening model and a feature elimination model for processing, thereby obtaining a target feature library.

[0095] The feature screening model may be a NWP feature screening model or other feature screening models, which are not limited in this embodiment.

[0096] Optionally, in this embodiment, each composite feature can be input into the NWP feature screening model, and the correlation coefficient of the high-dimensional feature can be calculated and combined with the random forest algorithm for feature screening; further, the screened features can be input into the feature elimination model, and the cross-validation method and recursive model can be used to eliminate low-correlation features, and finally the target feature library can be obtained.

[0097] Step 250: Input each target feature in the target feature library into the target machine learning model in turn for iterative training, and when the iteration stop condition is met, obtain the distributed new energy power prediction model.

[0098] The distributed renewable energy power prediction model is used to predict the power of distributed renewable energy under different meteorological factors.

[0099] In an optional implementation of this embodiment, after obtaining the target feature library, each target feature cluster in the target feature library can be input into the target machine learning model for iterative training, and when the iteration stop condition is met, a distributed new energy power prediction model is trained.

[0100] Optionally, in this embodiment, after the distributed new energy power prediction model is trained, any meteorological data can be input into the distributed new energy for power prediction, so as to obtain the distributed new energy power corresponding to the meteorological data.

[0101] The solution of this embodiment can perform feature mining on each of the cross-correlation characteristic analysis results through Featurewiz, and screen the mined features through SGBoost to obtain each of the composite features; further, each of the composite features is respectively input into a pre-trained feature screening model and a feature elimination model for processing to obtain the target feature library, and the target feature library can be quickly determined; and after obtaining the target feature library, each target feature cluster in the target feature library can be continuously input into the target machine learning model for iterative training, and when the iteration stop condition is met, a distributed new energy power prediction model is trained to provide a strong basis for distributed new energy power prediction.

[0102] In order to better understand the feature screening method of distributed renewable energy power involved in this embodiment, a specific example is used below to illustrate it:

[0103] Step 1: Perform Pearson correlation coefficient analysis on meteorological factors of different terrains such as wind speed, wind direction, irradiance and temperature with wind power and photovoltaic power generation, and quantitatively calculate the correlation results; use the OPTICS method to perform cluster analysis on sample data such as historical power generation data and sensitive meteorological factors.

[0104] Step 1.1: Select n distributed wind farms and photovoltaic farms in different terrains, and m meteorological factors such as wind speed, wind direction, irradiation, and temperature;

[0105] Step 1.2: Use the Pearson coefficient to conduct correlation analysis on the meteorological factors and wind / solar power generation in step 1.1, quantitatively calculate the correlation relationship of new energy data in different terrains, and divide them into characteristics of strong correlation coefficient to low correlation coefficient;

[0106] Step 1.3: Preliminarily screen the feature set in step 1.2 by sorting the correlation coefficients, retain the features with a Pearson coefficient greater than 0.8, filter the features with weak correlation coefficients, and construct the minimum set of sensitive meteorological factors for wind power / photovoltaic power generation;

[0107] Step 1.4: Considering the impact of distributed sensitive meteorological factors on the power generation data of new energy sites, the minimum feature set of sensitive meteorological factors extracted in step 1.3 is combined with the historical power of new energy sites, and the OPTICS method is used for cluster analysis. It is divided into several sub-clusters, thereby paving the way for the construction of new energy composite features.

[0108] Step 2: Analyze the cross-correlation characteristics of the output of distributed renewable energy stations and weather processes such as low temperature and strong wind, and use the Featurewiz method to mine features in combination with the composite meteorological characteristics of weather processes. Use recursive XGBoost to optimize the expanded features and establish composite features that integrate distributed renewable energy output characteristics, climate characteristics and other different indicators.

[0109] Step 2.1: Collect distributed renewable energy station output data and weather process characteristics;

[0110] Step 2.2: Establish different weather processes such as low temperature and strong wind, and conduct cross-correlation analysis with the output characteristics of distributed renewable energy stations;

[0111] Step 2.3: Use the Featurewiz method to mine the features of different weather processes, and use the XGBoost method for automatic feature optimization technology;

[0112] Step 2.4: Finally, the optimized composite features are obtained.

[0113] Step 3: Based on the distributed renewable energy multi-dimensional meteorological feature library, a NWP feature screening model is established based on algorithms such as the Pearson coefficient and random forest. A model for eliminating low-correlation features is constructed based on cross-validation and recursive algorithms, and finally the minimum key feature set is obtained to effectively improve the prediction accuracy.

[0114] Step 3.1: Construct a distributed new energy multi-dimensional meteorological feature library based on the meteorological features obtained in step 2;

[0115] Step 3.2: Input the multi-dimensional meteorological characteristics of distributed renewable energy into the NWP feature screening model, calculate the correlation coefficient of high-dimensional features and combine the random forest algorithm for feature screening.

[0116] Step 3.3: By inputting the above-screened features into the feature elimination model, cross-validation method and recursive model are used to eliminate low-correlation features, and finally the minimum key feature set is obtained.

[0117] Embodiment 3

[0118] Figure 3 Schematic diagram of a distributed new energy power feature screening device provided according to the third embodiment of the present invention. Figure 3 As shown, the device includes: a target cluster determination module 310, a target weather process determination module 320, a composite feature determination module 330 and a target feature library determination module 340.

[0119] The target cluster determination module 310 is used to determine a meteorological factor data set and distributed new energy power matching each meteorological factor in the meteorological factor set, and determine at least one target cluster according to the meteorological factor data set and each distributed new energy power;

[0120] A target weather process determination module 320 is used to determine a target weather process according to a target cluster, and collect target output data of the distributed new energy in the target weather process;

[0121] A composite feature determination module 330, for determining each composite feature according to each target weather process and each target output data;

[0122] The target feature library determination module 340 is used to screen each of the composite features to obtain a target feature library; the target feature library is used for training the distributed new energy power prediction model.

[0123] The scheme of this embodiment determines a meteorological factor data set and distributed new energy power matching each meteorological factor in the meteorological factor set through a target cluster determination module, and determines at least one target cluster based on the meteorological factor data set and each distributed new energy power; determines a target weather process based on the target cluster through a target weather process determination module, and collects the target output data of the distributed new energy in the target weather process; determines each composite feature based on each target weather process and each target output data through a composite feature determination module; and screens each composite feature through a target feature library determination module to obtain a target feature library, which can accurately determine the features that affect the distributed new energy power and provide a basis for accurately predicting the distributed new energy power.

[0124] In an optional implementation of this embodiment, the meteorological factor includes at least one of the following: wind speed, wind direction, irradiance, and temperature;

[0125] The meteorological factor data set contains specific values ​​of each meteorological factor at different time nodes;

[0126] Correspondingly, the target cluster determination module 310 is specifically used to obtain the specific values ​​of each of the meteorological factors at the target historical moment and the power of each of the distributed new energy sources at the target historical moment from the target database, and obtain each meteorological factor-power data pair;

[0127] Screening each of the meteorological factor-power data pairs to obtain each target data pair;

[0128] Clustering is performed on each of the target data pairs to obtain each of the target clusters.

[0129] In an optional implementation of this embodiment, the target cluster determination module 310 is further specifically used to calculate the correlation coefficient of each meteorological factor-power data pair respectively, and sort each data pair according to the set correlation coefficient order to obtain a reference data pair queue;

[0130] The reference data pair queue is traversed to determine a target correlation coefficient, and all data pairs after or before the data pair corresponding to the target correlation coefficient are deleted to obtain the target data pairs.

[0131] In an optional implementation of this embodiment, the target weather process determination module 320 is specifically used to determine the target meteorological factors in the target cluster, and determine each target collection time of the target meteorological factors;

[0132] Determining the target weather process according to each of the target collection moments;

[0133] Collect target output data of the distributed renewable energy at different times during the target weather process.

[0134] In an optional implementation of this embodiment, the composite feature determination module 330 is specifically used to perform a cross-correlation characteristic analysis on each of the target weather processes and each of the target output data to obtain a cross-correlation characteristic analysis result;

[0135] Feature mining is performed on each of the cross-correlation characteristic analysis results to obtain each of the composite features.

[0136] In an optional implementation of the present embodiment, the composite feature determination module 330 is further specifically used to perform feature mining on each of the cross-correlation characteristic analysis results through Featurewiz, and to filter the mined features through SGBoost to obtain each of the composite features.

[0137] In an optional implementation of this embodiment, the target feature library determination module 340 is specifically used to input each of the composite features into a pre-trained feature screening model and a feature elimination model for processing, so as to obtain the target feature library;

[0138] Correspondingly, the feature screening device for distributed renewable energy power also includes: a model training module, which is used to input each target feature in the target feature library into the target machine learning model in turn for iterative training, and when the iteration stop condition is met, the distributed renewable energy power prediction model is obtained; the distributed renewable energy power prediction model is used to predict the power of distributed renewable energy under different meteorological factors.

[0139] The characteristic screening device for distributed renewable energy power provided in the embodiment of the present invention can execute the characteristic screening method for distributed renewable energy power provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0140] Embodiment 4

[0141] Figure 4 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0142] like Figure 4 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0143] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0144] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The processor 11 executes the various methods and processes described above, such as a method for screening the characteristics of distributed new energy power.

[0145] In some embodiments, the characteristic screening method of distributed renewable energy power can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the characteristic screening method of the distributed renewable energy power described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to perform the characteristic screening method of the distributed renewable energy power by any other appropriate means (for example, by means of firmware).

[0146] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0147] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0148] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0149] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0150] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0151] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0152] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0153] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for screening characteristics of distributed renewable energy power, characterized in that: include: Determine a meteorological factor data set, and distributed new energy power matching each meteorological factor in the meteorological factor set, and determine at least one target cluster according to the meteorological factor data set and each distributed new energy power; Determine a target weather process according to the target cluster, and collect target output data of the distributed new energy in the target weather process; Determining each composite feature according to each of the target weather processes and each of the target output data; The composite features are screened to obtain a target feature library; the target feature library is used for training a distributed new energy power prediction model.

2. The method for screening the characteristics of distributed renewable energy power according to claim 1, characterized in that: The meteorological factors include at least one of the following: wind speed, wind direction, irradiance and temperature; The meteorological factor data set contains specific values ​​of each meteorological factor at different time nodes; Accordingly, the determining of the meteorological factor data set and the distributed new energy power matching each meteorological factor in the meteorological factor set, and determining at least one target cluster according to the meteorological factor data set and each of the distributed new energy power, includes: Acquire specific values ​​of each of the meteorological factors at the target historical moment and the power of each of the distributed renewable energy sources at the target historical moment from the target database to obtain each meteorological factor-power data pair; Screening each of the meteorological factor-power data pairs to obtain each target data pair; Clustering is performed on each of the target data pairs to obtain each of the target clusters.

3. The method for screening the characteristics of distributed renewable energy power according to claim 2, characterized in that: The step of screening each of the meteorological factor-power data pairs to obtain each target data pair includes: Calculating the correlation coefficient of each meteorological factor-power data pair respectively, and sorting each data pair according to the set correlation coefficient order to obtain a reference data pair queue; The reference data pair queue is traversed to determine a target correlation coefficient, and all data pairs after or before the data pair corresponding to the target correlation coefficient are deleted to obtain the target data pairs.

4. The method for screening the characteristics of distributed renewable energy power according to claim 1, characterized in that: The step of determining a target weather process according to a target cluster and collecting target output data of the distributed new energy in the target weather process includes: Determining target meteorological factors in the target cluster, and determining target collection times of the target meteorological factors; Determining the target weather process according to each of the target collection moments; Collect target output data of the distributed renewable energy at different times during the target weather process.

5. The method for screening the characteristics of distributed renewable energy power according to claim 1, characterized in that: Determining each composite feature according to each target weather process and each target output data includes: Performing cross-correlation characteristic analysis on each of the target weather processes and each of the target output data to obtain a cross-correlation characteristic analysis result; Feature mining is performed on each of the cross-correlation characteristic analysis results to obtain each of the composite features.

6. The method for screening the characteristics of distributed renewable energy power according to claim 5, characterized in that: The performing feature mining on each of the cross-correlation characteristic analysis results to obtain each of the composite features includes: Feature mining is performed on each of the cross-correlation characteristic analysis results through Featurewiz, and the mined features are screened through SGBoost to obtain each of the composite features.

7. The method for screening the characteristics of distributed renewable energy power according to claim 1, characterized in that: The step of screening each of the composite features to obtain a target feature library includes: Inputting each of the composite features into a pre-trained feature screening model and a feature elimination model for processing respectively, to obtain the target feature library; Correspondingly, after obtaining the target feature library, it also includes: Input each target feature in the target feature library into the target machine learning model in turn for iterative training, and obtain the distributed renewable energy power prediction model when the iteration stop condition is met; the distributed renewable energy power prediction model is used to predict the power of distributed renewable energy under different meteorological factors.

8. A distributed new energy power feature screening device, characterized in that: include: A target cluster determination module, used to determine a meteorological factor data set and distributed new energy power matching each meteorological factor in the meteorological factor set, and determine at least one target cluster according to the meteorological factor data set and each distributed new energy power; A target weather process determination module is used to determine the target weather process according to the target clustering cluster, and collect the target output data of the distributed new energy in the target weather process; A composite feature determination module, used to determine each composite feature according to each target weather process and each target output data; The target feature library determination module is used to screen each of the composite features to obtain a target feature library; the target feature library is used for training a distributed new energy power prediction model.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the characteristic screening method of distributed renewable energy power described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the characteristic screening method of distributed renewable energy power according to any one of claims 1 to 7 when executed.

Images