A power distribution network multi-stage distributed photovoltaic output prediction method and device

Abstract

The present application relates to power distribution network multi-stage distributed photovoltaic output prediction method and device, wherein, the method comprises: establishing power distribution network multi-stage distributed photovoltaic connection distribution relationship data sample library;Based on the historical data of distributed photovoltaic, combine the distributed photovoltaic connection distribution relationship data sample library, establish the power distribution network multi-stage distributed photovoltaic data sample library;Based on the multi-stage distributed photovoltaic data sample library, analyze the influence characteristics of multi-stage distributed photovoltaic output, establish the influence characteristic library for different stage distributed photovoltaic;Based on the influence characteristic library, construct LSTM model, and use the data in the multi-stage distributed photovoltaic data sample library to train, obtain the power distribution network distributed photovoltaic multi-stage output prediction model;The input data of the time period to be predicted is input into the power distribution network distributed photovoltaic multi-stage output prediction model, and the power distribution network multi-stage distributed photovoltaic output prediction result is obtained.The present application can accurately predict the distributed photovoltaic output situation of each stage of power distribution network.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of distributed photovoltaic power station output prediction, in particular to a power distribution network multi-stage distributed photovoltaic output prediction method and device. BACKGROUND

[0002] Under the situation that energy depletion and energy saving and environmental protection problems are becoming increasingly serious, distributed photovoltaic power generation is getting more and more widely used. With the increasing penetration rate of distributed photovoltaic in the power distribution network year by year, while alleviating energy shortage and environmental degradation, the intermittency and volatility of photovoltaic power generation also bring great challenges to the safe, reliable and economic operation of the power distribution network. Accurate distributed photovoltaic output prediction can predict future photovoltaic power generation, provide scientific decision-making basis for power distribution network automatic generation control and power grid dispatching, effectively improve the utilization efficiency of distributed photovoltaic, effectively reduce the impact of high penetration rate of distributed photovoltaic on the power distribution network, improve power supply reliability, and ensure the safe, reliable and economic operation of the power distribution network.

[0003] At present, the distributed photovoltaic output prediction method can be divided into physical method and statistical prediction method according to the prediction principle. The physical method is to use solar radiation transfer equation, solar position model, photovoltaic cell model and inverter efficiency model to establish a prediction algorithm, so as to predict the power generation. The statistical prediction method finds out the internal law between photovoltaic power generation and various influencing factors through statistical analysis of historical operation data, establishes a prediction algorithm to predict the power generation. Artificial intelligence prediction method as a new school of statistical prediction performs outstandingly in distributed photovoltaic prediction, and with the development of various artificial intelligence algorithms, it has been preliminarily applied in the field of distributed photovoltaic prediction.

[0004] Distributed photovoltaic has an impact in the multi-stage of the power distribution network. In order to avoid the impact of high penetration rate of distributed photovoltaic on the power distribution network in many aspects, it is necessary to predict the output of multi-stage distributed photovoltaic in the power distribution network, which is a technical problem to be solved by those skilled in the art. SUMMARY

[0005] The technical problem to be solved by the present application is to provide a power distribution network multi-stage distributed photovoltaic output prediction method and device, which can accurately predict the distributed photovoltaic output of each stage of the power distribution network and predict the output characteristics of the distributed photovoltaic in advance.

[0006] The technical solution adopted by the present application to solve the technical problem is to provide a power distribution network multi-stage distributed photovoltaic output prediction method, comprising the following steps:

[0007] Establishing a power distribution network multi-stage distributed photovoltaic connection distribution relationship data sample library;

[0008] Based on the distributed photovoltaic historical related data, a multi-level distributed photovoltaic data sample library of the power distribution network is established in combination with a multi-level distributed photovoltaic connection distribution relationship data sample library of the power distribution network.

[0009] Based on the multi-level distributed photovoltaic data sample library of the power distribution network, multi-level distributed photovoltaic output influence characteristics are analyzed to establish an influence characteristic library for different levels of distributed photovoltaic.

[0010] Based on the influence characteristic library, an LSTM model is constructed, and data in the multi-level distributed photovoltaic data sample library of the power distribution network is used to train the LSTM model to obtain a multi-level output prediction model of the distributed photovoltaic of the power distribution network.

[0011] The input data of the to-be-predicted period is input into the multi-level output prediction model of the distributed photovoltaic of the power distribution network to obtain a multi-level distributed photovoltaic output prediction result of the power distribution network.

[0012] The multi-level distributed photovoltaic connection distribution relationship data sample library of the power distribution network is established in particular as follows.

[0013] The topological structure data of the power distribution network and the distributed photovoltaic user account information are obtained.

[0014] Based on the topological structure data of the power grid and the distributed photovoltaic user account information, the connection relationship among photovoltaic users, transformer areas, feeder lines, busbars, substations and regions is analyzed and verified.

[0015] According to the obtained connection relationship among photovoltaic users, transformer areas, feeder lines, busbars, substations and regions, a multi-level distributed photovoltaic connection distribution relationship data sample library of the power distribution network is established.

[0016] The distributed photovoltaic user account information includes photovoltaic user id, latitude and longitude position, altitude, installed capacity, grid-connected type, measurement table type, installation inclination angle, component area and voltage grade.

[0017] Based on the distributed photovoltaic historical related data, a multi-level distributed photovoltaic data sample library of the power distribution network is established in combination with a multi-level distributed photovoltaic connection distribution relationship data sample library of the power distribution network, in particular as follows.

[0018] The historical active power generation data of the distributed photovoltaic and the corresponding historical micro-meteorological element real-time data are obtained.

[0019] According to the correlation among photovoltaic users in the multi-level distributed photovoltaic connection distribution relationship data sample library of the power distribution network, the historical active power generation data of the distributed photovoltaic and the corresponding historical micro-meteorological element real-time data are completed for missing values.

[0020] According to the completed distributed photovoltaic historical active power generation data and corresponding historical micro-meteorological element real-time data, a multi-level distributed photovoltaic data sample library of a power distribution network is established.

[0021] According to the correlation between photovoltaic users in the distributed photovoltaic connection distribution relationship data sample library of the power distribution network, the missing value completion is performed on the distributed photovoltaic historical active power generation data and corresponding historical micro-meteorological element real-time data, and specifically includes the following steps.

[0022] The missing values of the distributed photovoltaic historical active power generation data and corresponding historical micro-meteorological element real-time data are screened.

[0023] According to the multi-level distributed photovoltaic connection distribution relationship data sample library of the power distribution network, the correlation between photovoltaic users in the same period is calculated.

[0024] The data sample of the photovoltaic user most relevant to the data missing photovoltaic user is selected, and the correlation is used to complete the data of the data missing photovoltaic user.

[0025] According to the multi-level distributed photovoltaic data sample library of the power distribution network, the influence characteristics of multi-level distributed photovoltaic output are analyzed, and an influence characteristic library for different levels of distributed photovoltaic is established, and specifically includes the following steps.

[0026] For different levels of distributed photovoltaic, based on different time scales and different spatial scales in the multi-level distributed photovoltaic data sample library of the power distribution network, the influence characteristics of distributed photovoltaic output are extracted by correlation analysis method, and multiple common influence characteristics are removed.

[0027] The importance of the influence characteristics after removing the multiple common influence characteristics is evaluated by a random forest algorithm, the influence characteristic set is selected for different levels of distributed photovoltaic, and the influence characteristic library for different levels of distributed photovoltaic is constructed.

[0028] The technical solution adopted by the present application to solve the technical problems is to provide a multi-level distributed photovoltaic output prediction device of a power distribution network, which comprises:

[0029] The first establishment module is used for establishing a multi-level distributed photovoltaic connection distribution relationship data sample library of a power distribution network.

[0030] The second establishment module is used for establishing a multi-level distributed photovoltaic data sample library of a power distribution network based on historical correlation data of distributed photovoltaic and in combination with the multi-level distributed photovoltaic connection distribution relationship data sample library of the power distribution network.

[0031] A third establishing module is configured to analyze multi-stage distributed photovoltaic output influence characteristics based on the multi-stage distributed photovoltaic data sample library of the power distribution network, and establish an influence characteristic library for different stages of distributed photovoltaic.

[0032] A training module is configured to construct an LSTM model based on the influence characteristic library, and train the LSTM model by using data in the multi-stage distributed photovoltaic data sample library of the power distribution network, so as to obtain a multi-stage distributed photovoltaic output prediction model of the power distribution network.

[0033] A prediction module is configured to input input data of a to-be-predicted period into the multi-stage distributed photovoltaic output prediction model of the power distribution network, so as to obtain a multi-stage distributed photovoltaic output prediction result of the power distribution network.

[0034] The technical solution adopted by the present application to solve its technical problems is to provide an electronic device, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the power distribution network multi-stage distributed photovoltaic output prediction method when executing the computer program.

[0035] The technical solution adopted by the present application to solve its technical problems is to provide a computer readable storage medium, which stores a computer program, and the computer program implements the steps of the power distribution network multi-stage distributed photovoltaic output prediction method when executed by a processor.

[0036] Advantages

[0037] Compared with the prior art, the present application has the following advantages and positive effects: the present application constructs multi-stage distributed photovoltaic influence characteristics based on multi-stage distributed photovoltaic connection distribution relationships and multi-stage distributed photovoltaic data, and constructs a multi-stage distributed photovoltaic output prediction model of the power distribution network based on the multi-stage distributed photovoltaic influence characteristics, so that the prediction model can accurately predict the distributed photovoltaic output of each stage of the power distribution network, thereby predicting the output characteristics of the distributed photovoltaic in advance to support the optimization scheduling of the multi-stage operation of the power distribution network and improve the intelligent operation level of the power distribution network. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 is a flowchart of the power distribution network multi-stage distributed photovoltaic output prediction method of the first embodiment of the present application;

[0039] Figure 2 is a schematic diagram of the LSTM model structure in the first embodiment of the present application. DETAILED DESCRIPTION

[0040] The application will be further described in connection with the following specific embodiments. It should be understood that these embodiments are only used to illustrate but not to limit the scope of the application. Furthermore, it should be understood that those skilled in the art can make various modifications or changes to the application after reading the content of the application, and these equivalent forms also fall within the scope defined by the appended claims.

[0041] The first embodiment of the application relates to a power distribution network multi-level distributed photovoltaic output prediction method, wherein the multi-level in the method includes user level, transformer area level, feeder level, bus level, substation level and regional level, wherein the regional level can be county, city, province or specific identified region. The method first establishes a power distribution network multi-level distributed photovoltaic connection distribution relationship data sample library; then cleans and screens the missing values of the historical power generation data of all distributed photovoltaic users in the power distribution network and the corresponding micro-meteorological element historical data, and processes the missing values to establish a photovoltaic data sample library; then, based on a large number of sample data in the photovoltaic data sample library, analyzes the influence characteristics of multi-level distributed photovoltaic output, and establishes an influence characteristic library for different levels of distributed photovoltaic; then, based on a deep learning algorithm, for the influence characteristics of multi-level distributed photovoltaic output, a large number of sample data are trained to establish a multi-level distributed photovoltaic output prediction model; finally, input the input characteristic data of the demand prediction period to obtain the multi-level distributed photovoltaic output prediction results of the user-transformer area-feeder-bus-substation-region in the power distribution network. As shown in the figure, it specifically includes the following steps: Figure 1

[0042] Step 1, establish a power distribution network multi-level distributed photovoltaic connection distribution relationship data sample library.

[0043] This step specifically includes: obtaining power grid topology structure data and distributed photovoltaic user account information; wherein the distributed photovoltaic user account information includes photovoltaic user id, latitude and longitude position, altitude, installed capacity, grid-connected type, measurement table type, installation inclination, component area and voltage grade, etc. Based on the power grid topology structure data and the distributed photovoltaic user account information, analyze and verify the connection relationship between photovoltaic users, transformer areas, feeders, buses, substations and regions. According to the obtained connection relationship between photovoltaic users, transformer areas, feeders, buses, substations and regions, establish a power distribution network multi-level distributed photovoltaic connection distribution relationship data sample library.

[0044] Step 2, based on the historical related data of distributed photovoltaic, combined with the power distribution network multi-level distributed photovoltaic connection distribution relationship data sample library, establish a power distribution network multi-level distributed photovoltaic data sample library.

[0045] ​The step specifically comprises: acquiring distributed photovoltaic historical active power generation data and corresponding historical microclimate element real-time data; wherein the microclimate element real-time data comprises total radiation irradiance, air temperature, relative humidity, wind speed, wind direction, air pressure, rainfall, cloud cover, weather type, etc. According to the correlation between photovoltaic users in the distribution network multi-level distributed photovoltaic connection distribution relationship data sample library, the missing value completion is performed on the distributed photovoltaic historical active power generation data and the corresponding historical microclimate element real-time data. The distribution network multi-level distributed photovoltaic data sample library is established according to the completed distributed photovoltaic historical active power generation data and the corresponding historical microclimate element real-time data.

[0046] When the missing values are completed, if the data vacancy period is short, the interpolation method can be directly used for filling, and when the data vacancy period is long, the correlation method can be used for filling. The correlation method specifically comprises:

[0047] Screening the missing values of the distributed photovoltaic historical active power generation data and the corresponding historical microclimate element real-time data;

[0048] According to the distribution network multi-level distributed photovoltaic connection distribution relationship data sample library, the correlation between photovoltaic users in the same period is calculated;

[0049] The data sample of the photovoltaic user most related to the data vacancy photovoltaic user is selected, and the correlation is used to complete the data of the data vacancy photovoltaic user.

[0050] After the missing data is completed, the embodiment can also normalize all the data, so as to unify the basic measurement unit, so that the neural network model can converge faster during subsequent training. After the normalization processing is completed, the distribution network multi-level distributed photovoltaic data sample library is established based on the normalized data.

[0051] Step 3, based on the distribution network multi-level distributed photovoltaic data sample library, the multi-level distributed photovoltaic output influence characteristics are analyzed, and the influence characteristic library for different levels of distributed photovoltaic is established. The step specifically comprises:

[0052] For different levels of distributed photovoltaic, based on different time scales and different spatial scales in the distribution network multi-level distributed photovoltaic data sample library, the influence characteristics of the distributed photovoltaic output are extracted by the correlation analysis method, and the multiple common influence characteristics are eliminated. By eliminating the multiple common influence characteristics, the dimension of the influence characteristics is reduced, so as to accelerate the training speed of the subsequent neural network model.

[0053] The importance of the impact features excluding the multiple common impact features is evaluated by a random forest algorithm, and the impact feature set is selected for different cascade distributed photovoltaics, and the impact feature library for different cascade distributed photovoltaics is constructed.

[0054] In step 4, an LSTM model is constructed based on the impact feature library, and the data in the power distribution network multi-cascade distributed photovoltaic data sample library is used to train the LSTM model, and a power distribution network distributed photovoltaic multi-cascade output prediction model is obtained.

[0055] In this step, the LSTM model is an advanced version of the recurrent neural network (RNN), and unlike the RNN, the LSTM model selectively receives information and selectively stores information.

[0056] As shown in Figure 2 , the LSTM model has three gate units, (1) input gate: decides whether information is input to the memory unit at this time; (2) output gate: decides whether information is output from the memory unit at each time. (3) forget gate: decides whether the value in the memory unit at each time is forgotten. If it is opened, the value in the memory unit will be cleared, that is, forgotten.

[0057] The LSTM model mainly has three stages:

[0058] (1) The forgetting stage: selectively forgets the input transmitted by the previous node. The specific calculation is as follows:

[0059] The forget gate output f t : f t =σ(W hf H t-1 +W xf x t +b f )

[0060] f represents forget, which is a forget gate that controls whether the value of the previous state needs to be forgotten or remembered for the next operation.

[0061] (2) Selective memory stage: selectively memorize the input of this stage.

[0062] The input gate output i t : i t =σ(W hi H t-1 +W xi x t +b i )

[0063] Temporary state

[0064] Update temporary state Get new cell state C t :

[0065] (3) Output stage: determine the current output. The new cell of the last stage is changed through a tanh activation function to obtain the output h t . h t-1 and x t are two inputs of the output gate, and the calculation formula is as follows:

[0066] o t = sigma(W xo H t-1 + W xo x t + b0)

[0067] In the formula, sigma() is an activation function, generally a sigmoid function, and the value obtained after the activation function is between [0, 1], 1 indicates that the gate is completely open, and 0 indicates that it is closed; W f , W i , W C , W o correspond to the weights of the forget gate, the input gate, the memory cell, and the output gate; b f , b i , b C , b o are the corresponding bias vectors, represents the multiplication operation; represents the addition operation.

[0068] The LSTM model is improved on the basis of the RNN, aiming to improve the long-term data dependence problem of the RNN. In view of the serious gradient disappearance problem of the RNN, the long-term memory function unit of the LSTM model enables data and information to be better filtered, and its gate structure enables the network to reasonably output information when a large amount of data flows into the network.

[0069] In this embodiment, the influence feature set selected for different cascade distributed photovoltaics from the influence feature library is used to train and test the LSTM model using a large amount of sample data, and appropriate parameters are obtained. After the training is completed, the distributed photovoltaic multi-stage output prediction algorithm model of the power distribution network can be obtained.

[0070] Step 5: input the input data of the to-be-predicted period into the distributed photovoltaic multi-stage output prediction model of the power distribution network to obtain the distributed photovoltaic multi-stage output prediction result of the power distribution network, that is, the user-subdistrict-feeder-busbar-substation-region multi-stage distributed photovoltaic output prediction result in the power distribution network.

[0071] It can be found that the application is based on the multi-stage distributed photovoltaic connection distribution relationship and the multi-stage distributed photovoltaic data, and the multi-stage distributed photovoltaic influence characteristics are constructed, and then the multi-stage distributed photovoltaic output prediction model of the power distribution network is constructed based on the multi-stage distributed photovoltaic influence characteristics, so that the prediction model can accurately predict the distributed photovoltaic output of each stage of the power distribution network, thereby predicting the output characteristics of the distributed photovoltaic in advance, supporting the optimization scheduling of the multi-stage operation of the power distribution network, and improving the intelligent operation level of the power distribution network.

[0072] The second embodiment of the application relates to a multi-stage distributed photovoltaic output prediction device of a power distribution network, comprising:

[0073] A first establishing module is configured to establish a multi-stage distributed photovoltaic connection distribution relationship data sample library of the power distribution network.

[0074] A second establishing module is configured to establish a multi-stage distributed photovoltaic data sample library of the power distribution network based on the historical related data of the distributed photovoltaic and in combination with the multi-stage distributed photovoltaic connection distribution relationship data sample library of the power distribution network.

[0075] A third establishing module is configured to analyze the influence characteristics of the multi-stage distributed photovoltaic output based on the multi-stage distributed photovoltaic data sample library of the power distribution network, and establish an influence characteristic library for the distributed photovoltaic of different stages.

[0076] A construction training module is configured to construct an LSTM model based on the influence characteristic library, and train the LSTM model by using the data in the multi-stage distributed photovoltaic data sample library of the power distribution network, so as to obtain a multi-stage output prediction model of the distributed photovoltaic of the power distribution network.

[0077] A prediction module is configured to input the input data of a to-be-predicted period into the multi-stage output prediction model of the distributed photovoltaic of the power distribution network, so as to obtain a multi-stage distributed photovoltaic output prediction result of the power distribution network.

[0078] The first establishing module comprises:

[0079] A first obtaining unit is configured to obtain the power grid topology structure data and the distributed photovoltaic user account information.

[0080] An analysis unit is configured to analyze and verify the connection relationship among the photovoltaic users, the transformer areas, the feeder lines, the busbars, the substations and the regions based on the power grid topology structure data and the distributed photovoltaic user account information.

[0081] A first establishing unit is configured to establish a multi-stage distributed photovoltaic connection distribution relationship data sample library of the power distribution network according to the obtained connection relationship among the photovoltaic users, the transformer areas, the feeder lines, the busbars, the substations and the regions.

[0082] The distributed photovoltaic user account information includes a photovoltaic user ID, a latitude and longitude position, an altitude, an installed capacity, a grid-connected type, a meter type, an installation inclination angle, a component area, and a voltage level.

[0083] The second establishing module includes:

[0084] A second acquisition unit is configured to acquire distributed photovoltaic historical active power generation data and corresponding historical micro-meteorological element real-time data.

[0085] A completion unit is configured to complete missing values in the distributed photovoltaic historical active power generation data and the corresponding historical micro-meteorological element real-time data according to correlations between photovoltaic users in the power distribution network multi-level distributed photovoltaic connection distribution relationship data sample library.

[0086] A second establishing unit is configured to establish a power distribution network multi-level distributed photovoltaic data sample library according to the completed distributed photovoltaic historical active power generation data and the corresponding historical micro-meteorological element real-time data.

[0087] The completion unit includes:

[0088] A screening subunit is configured to screen missing values in the distributed photovoltaic historical active power generation data and the corresponding historical micro-meteorological element real-time data.

[0089] A calculation subunit is configured to calculate correlations between photovoltaic users in the same period according to the power distribution network multi-level distributed photovoltaic connection distribution relationship data sample library.

[0090] A completion subunit is configured to select data samples of photovoltaic users most relevant to the photovoltaic user with data missing, and complete data of the photovoltaic user with data missing by using the correlations.

[0091] The third establishing module includes:

[0092] An extraction unit is configured to extract influence characteristics of distributed photovoltaic output by a correlation analysis method based on different time scales and different spatial scales in the power distribution network multi-level distributed photovoltaic data sample library for different levels of distributed photovoltaic, and eliminate multiple common influence characteristics.

[0093] A construction unit is configured to evaluate importance of the influence characteristics from which the multiple common influence characteristics are eliminated by a random forest algorithm, select an influence characteristic set for different levels of distributed photovoltaic, and construct an influence characteristic library for different levels of distributed photovoltaic.

[0094] The third embodiment of the present application relates to an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the multi-stage distributed photovoltaic output prediction method of the power distribution network according to the first embodiment when executing the computer program.

[0095] The fourth embodiment of the present application relates to a computer readable storage medium, which stores a computer program, wherein the computer program implements the steps of the multi-stage distributed photovoltaic output prediction method of the power distribution network according to the first embodiment when executed by a processor.

[0096] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage, etc.) containing computer-usable program code.

[0097] The present application is described with reference to flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus generate a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks Figure 1 The functions specified in one or more flows and / or blocks

[0098] These computer program instructions can also be stored in a computer-readable memory capable of directing the computer or other programmable data processing apparatus to work in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction method, which implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks Figure 1 The functions specified in one or more flows and / or blocks

[0099] These computer program instructions can also be loaded into a computer or other programmable data processing apparatus, so that a series of operation steps are performed on the computer or other programmable data processing apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable data processing apparatus provide a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks Figure 1steps of the functions specified in the block or blocks.

[0100] The above description is merely that of a specific implementation of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, and all such changes or replacements shall be encompassed within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be subject to the scope of protection of the claims.

Claims

1. A method for predicting multi-level distributed photovoltaic output in a distribution network, characterized in that: The following steps are involved: Establish a sample database of distribution relationship data of multi-level distributed photovoltaic connections in distribution networks; Based on the distributed photovoltaic historical data, combined with the distribution network multi-level distributed photovoltaic connection distribution relationship data sample library, a distribution network multi-level distributed photovoltaic data sample library is established; Based on the multi-level distributed photovoltaic data sample library of the distribution network, the impact characteristics of the multi-level distributed photovoltaic output are analyzed, and an impact characteristic library for distributed photovoltaics at different levels is established; An LSTM model is constructed based on the impact feature library, and the LSTM model is trained using data in the distribution network multi-level distributed photovoltaic data sample library to obtain a distribution network distributed photovoltaic multi-level output prediction model; input parameter data of the time period to be predicted is input into the distribution network distributed photovoltaic multi-level output prediction model to obtain a distribution network multi-level distributed photovoltaic output prediction result; Among them, the multiple levels include user level, substation level, feeder level, busbar level, substation level and regional level.

2. The method for predicting multi-level distributed photovoltaic output in a distribution network according to claim 1, characterized in that: The establishment of a data sample library of multi-level distributed photovoltaic connection distribution relationship data of a distribution network specifically includes: Obtain distribution network topology data and distributed photovoltaic user ledger information; Analyze and verify the connection relationship between photovoltaic users, substations, feeders, busbars, substations and regions based on the grid topology data and distributed photovoltaic user ledger information; Based on the obtained connection relationships between photovoltaic users, substations, feeders, busbars, substations and regions, a distribution network multi-level distributed photovoltaic connection distribution relationship data sample library is established.

3. The method for predicting multi-level distributed photovoltaic output in a distribution network according to claim 2, characterized in that: The distributed photovoltaic user ledger information includes: photovoltaic user ID, longitude and latitude location, altitude, installed capacity, grid connection type, meter type, installation inclination, component area and voltage level.

4. The method for predicting multi-level distributed photovoltaic output in a distribution network according to claim 1, characterized in that: The distributed photovoltaic historical data is combined with the distribution network multi-level distributed photovoltaic connection distribution relationship data sample library to establish a distribution network multi-level distributed photovoltaic data sample library, specifically including: Obtain historical distributed photovoltaic active power data and corresponding historical micro-meteorological element data; According to the correlation between photovoltaic users in the distribution network multi-level distributed photovoltaic connection distribution relationship data sample library, missing values ​​are filled in the distributed photovoltaic historical power generation active power data and the corresponding historical micro-meteorological element real-time data; A multi-level distributed photovoltaic data sample library for distribution network is established based on the completed distributed photovoltaic historical power generation active power data and the corresponding historical micro-meteorological element real-time data.

5. The method for predicting multi-level distributed photovoltaic output in a distribution network according to claim 4, characterized in that: The method of completing missing values ​​of the distributed photovoltaic historical active power generation data and the corresponding historical micro-meteorological element real-time data based on the correlation between photovoltaic users in the distribution network multi-level distributed photovoltaic connection distribution relationship data sample library specifically includes: Screening for missing values ​​in the distributed photovoltaic historical power generation active power data and the corresponding historical micro-meteorological element real-time data; Calculate the correlation between photovoltaic users in the same period based on the distribution network multi-level distributed photovoltaic connection distribution relationship data sample library; The data samples of photovoltaic users that are most relevant to the photovoltaic users with missing data are selected, and the data of the photovoltaic users with missing data are supplemented by using the correlation.

6. The method for predicting multi-level distributed photovoltaic output in a distribution network according to claim 1, characterized in that: The method of analyzing the impact characteristics of multi-level distributed photovoltaic output based on the distribution network multi-level distributed photovoltaic data sample library and establishing an impact characteristic library for distributed photovoltaics at different levels specifically includes: For distributed photovoltaics at different levels, based on the different time scales and different spatial scales in the multi-level distributed photovoltaic data sample library of the distribution network, the influence characteristics of distributed photovoltaic output are extracted through the correlation analysis method, and multiple common influence characteristics are eliminated; The importance of influencing features after eliminating multiple common influencing features is evaluated through the random forest algorithm, and an influencing feature set is selected for different levels of distributed photovoltaics to construct an influencing feature library for different levels of distributed photovoltaics.

7. A multi-level distributed photovoltaic output prediction device for a distribution network, characterized in that: include: The first establishment module is used to establish a data sample library of multi-level distributed photovoltaic connection distribution relationship of the distribution network; The second establishment module is used to establish a distribution network multi-level distributed photovoltaic data sample library based on the distributed photovoltaic historical related data and the distribution network multi-level distributed photovoltaic connection distribution relationship data sample library; The third establishment module is used to analyze the impact characteristics of multi-level distributed photovoltaic output based on the multi-level distributed photovoltaic data sample library of the distribution network, and establish an impact characteristic library for distributed photovoltaics of different levels; Constructing a training module for constructing an LSTM model based on the impact feature library, and training the LSTM model using data from the distribution network multi-level distributed photovoltaic data sample library to obtain a distribution network distributed photovoltaic multi-level output prediction model; A prediction module, configured to input the input parameter data of the time period to be predicted into the distribution network distributed photovoltaic multi-level output prediction model to obtain a distribution network multi-level distributed photovoltaic output prediction result; Among them, the multiple levels include user level, substation level, feeder level, busbar level, substation level and regional level.

8. 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 steps of the method for predicting multi-level distributed photovoltaic output in a distribution network as claimed in any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for predicting multi-level distributed photovoltaic output in a distribution network as claimed in any one of claims 1 to 6 are implemented.

Images