Power distribution network planning method and device under multi-scene variable, equipment and medium

Abstract

The invention discloses a power distribution network planning method, device, equipment and medium under multiple scene variables, relates to the technical field of power distribution network planning, can capture implied deep relationships among complex nonlinear sequence data, effectively improves prediction precision while considering the multiple scene variables, and has remarkable superiority in the field of source load prediction. According to the scheme, multi-scene variable data are collected; constructing a convolutional recurrent neural network model based on the multi-scene variable data; obtaining initial system topology data of the power distribution network; analyzing the voltage sensitivity of the convolutional recurrent neural network model and the topological data of the initial system of the power distribution network, and obtaining a plurality of potential installation positions of the energy storage equipment and the new energy equipment; and calculating the construction cost and carbon emission of each potential installation position of the energy storage equipment and the new energy equipment, and determining a power distribution network planning scheme, namely the installation positions and capacities of the energy storage equipment and the new energy equipment. The method is used for power distribution network planning.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network planning, and in particular to a distribution network planning method, device, equipment and medium under multiple scenario variables. Background Art

[0002] In the new power system, the source and load sides are highly interactive, and the basic theory of supply and demand balance faces challenges. The access of a large number of distributed new energy sources on the power supply side leads to increased volatility. The electricity consumption behavior on the load side will be affected by the carbon trading market, changing the load curve through electricity costs. At the same time, the behavior of users participating in demand response is also affected by their subjective influence. Therefore, the system balance mechanism changes from "deterministic power generation tracking uncertain load" to "two-way matching of uncertain power generation and uncertain load". These factors will lead to increased source-load uncertainty. When studying the planning problems of new power systems, the influence of the above uncertain factors should be considered, and appropriate algorithms should be adopted to deal with uncertainty, so that the planning scheme has a certain robustness, reduce network losses, and improve voltage quality.

[0003] Traditional statistical methods mainly model the linear relationship between source and load, ignoring the impact of factors such as climate and date type on source and load prediction, and the prediction accuracy is low. Although load forecasting models based on artificial intelligence methods such as support vector machines and feedforward neural networks can achieve relatively ideal prediction accuracy, they often fall into local optimal solutions and gradient disappearance when the source and load data volume is large and the source and load data features are complex. Summary of the invention

[0004] The present invention provides a distribution network planning method, device, equipment and medium under multiple scenario variables, which can capture the implicit deep relationship between complex nonlinear sequence data, effectively improve the prediction accuracy while considering multiple scenario variables, and has significant superiority in the field of source-load prediction.

[0005] In order to achieve the above object, the present invention adopts the following technical scheme: A first aspect of the present invention provides a distribution network planning method under multiple scenario variables, comprising: Collect multi-scenario variable data.

[0006] A convolutional recurrent neural network model is constructed based on the multi-scenario variable data.

[0007] Obtain initial system topology data of the distribution network.

[0008] The voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network is analyzed to obtain multiple potential installation locations for energy storage equipment and new energy equipment.

[0009] The construction cost and carbon emissions of each potential installation location of the energy storage device and the new energy device are calculated, and a distribution network planning scheme is determined, i.e., the installation location and capacity of the energy storage device and the new energy device.

[0010] Furthermore, the distribution network planning method under multiple scenario variables constructs a convolutional recurrent neural network model based on the multiple scenario variable data, including: Extract local spatial features of the multi-scene variable data and generate a feature map.

[0011] The dynamic dependency of the time series in the feature graph is captured to obtain a predicted output sequence.

[0012] For the predicted output sequence, minimizing the mean square error is selected as the loss function and the optimizer updates the model parameters to obtain the convolutional recurrent neural network model after training.

[0013] Furthermore, the distribution network planning method under multiple scenario variables, after calculating the construction cost and carbon emissions of each potential installation location of the energy storage device and the new energy device, and determining the distribution network planning scheme, that is, the installation location and capacity of the energy storage device and the new energy device, further includes: The operation of the distribution network planning scheme in different time periods and different scenarios is simulated, the operating cost and the carbon emissions in each scenario are calculated, and the optimal operating point in each scenario is obtained.

[0014] The optimal operating point is iteratively optimized to obtain an optimization solution for distribution network planning.

[0015] Furthermore, the distribution network planning method under multiple scenario variables, after iteratively optimizing the optimal operation point to obtain the distribution network planning optimization solution, further includes: The distribution network planning optimization scheme is actually tested and verified based on historical data and scenario variables to obtain the verified distribution network planning optimization scheme, that is, the final distribution network planning optimization scheme.

[0016] Furthermore, the distribution network planning method under multiple scenario variables, after collecting the multiple scenario variable data, further includes: The multi-scenario variable data is preprocessed.

[0017] Furthermore, in the distribution network planning method under multiple scenario variables, the preprocessing includes: The preprocessing includes but is not limited to data normalization and missing value filling.

[0018] Furthermore, in the distribution network planning method under multiple scenario variables, the multiple scenario variable data includes: The multi-scenario variable data includes but is not limited to temperature, humidity, wind speed, date type, historical load data and power supply side data.

[0019] A second aspect of the present invention provides a distribution network planning device under multiple scenario variables, comprising: The collection unit is used to collect multi-scenario variable data.

[0020] A construction unit is used to construct a convolutional recurrent neural network model based on the multi-scenario variable data.

[0021] The acquisition unit is used to acquire the initial system topology data of the distribution network.

[0022] An analysis unit is used to analyze the voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network to obtain multiple potential installation locations for energy storage equipment and new energy equipment.

[0023] A calculation unit is used to calculate the construction cost and carbon emissions of each potential installation location of the energy storage device and the new energy device, and determine the distribution network planning scheme, that is, the installation location and capacity of the energy storage device and the new energy device.

[0024] A third aspect of the present invention provides 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 any of the above methods when executing the computer program.

[0025] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of any of the above methods are implemented.

[0026] The present invention provides a distribution network planning method, device, equipment and medium under multi-scenario variables, including: collecting multi-scenario variable data; constructing a convolutional recurrent neural network model based on the multi-scenario variable data; obtaining initial system topology data of the distribution network; analyzing the voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network to obtain multiple potential installation locations of energy storage equipment and new energy equipment; calculating the construction cost and carbon emissions of each potential installation location of the energy storage equipment and the new energy equipment, and determining the distribution network planning scheme, that is, the installation location and capacity of the energy storage equipment and the new energy equipment. Compared with the prior art, the present invention can capture the implicit deep relationship between complex nonlinear sequence data, effectively improve the prediction accuracy while considering multi-scenario variables, and has significant superiority in the field of source-load prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. The drawings are only used to illustrate the implementation methods and are not to be considered as limitations of the present invention.

[0028] Figure 1 It is a schematic diagram of a flow chart of a distribution network planning method under multiple scenario variables in an embodiment of the present invention; Figure 2 It is a flowchart of another distribution network planning method under multiple scenario variables in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of a distribution network planning device under multiple scenario variables in an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of another distribution network planning device under multiple scenario variables in an embodiment of the present invention; Figure 5 The present invention is a schematic diagram of the structure of an electronic device for planning a distribution network under multiple scenario variables in an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all of the embodiments. Based on the embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0030] Unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as commonly understood by technicians in the field of the present invention; the terms used in the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the terms "including" and "having" and any variations thereof in the specification and claims of the present invention and the above-mentioned drawings, and are intended to cover non-exclusive inclusions.

[0031] In the description of the embodiments of the present invention, the technical terms "first", "second", etc. are only used to distinguish different objects, and cannot be understood as indicating or implying relative importance or implicitly indicating the number, specific order or primary and secondary relationship of the indicated technical features. In the description of the embodiments of the present invention, the meaning of "multiple" is more than two, unless otherwise clearly and specifically defined.

[0032] In the description of the embodiments of the present invention, the term "and / or" is only a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0033] In the description of the embodiments of the present invention, the term "multiple" refers to more than two (including two). Similarly, "multiple groups" refers to more than two groups (including two groups), and "multiple pieces" refers to more than two pieces (including two pieces).

[0034] In the description of the embodiments of the present invention, the technical terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc. indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the embodiments of the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the embodiments of the present invention.

[0035] In the description of the embodiments of the present invention, unless otherwise clearly specified and limited, technical terms such as "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the embodiments of the present invention can be understood according to the specific circumstances.

[0036] Embodiment 1 The embodiment of the present invention provides a distribution network planning method under multiple scenario variables, such as Figure 1 As shown, including: S101. Collect multi-scenario variable data.

[0037] The multi-scenario variable data refers to data that requires observation and analysis of multiple variables in different application scenarios. Specifically, the multi-scenario variable data of this embodiment includes but is not limited to temperature, humidity, wind speed, date type, historical load data, and power supply side data.

[0038] S102. Construct a convolutional recurrent neural network model based on multi-scenario variable data.

[0039] Among them, the Convolutional Recurrent Neural Network (CRNN) is a deep learning model that combines the advantages of Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN). CRNN is specifically used to process sequence data with spatiotemporal features, where the CNN part is responsible for feature extraction and the RNN part is responsible for capturing the dynamic dependencies of time series data.

[0040] The structure of the convolutional recurrent neural network model is very suitable for application in the scenario of source load prediction, because the source load data is affected by multiple variables and has temporal and spatial characteristics. The convolutional recurrent neural network model can effectively capture these complex characteristics, especially when considering scene variables, which can effectively improve the robustness and accuracy of the prediction.

[0041] S103: Acquire initial system topology data of the distribution network.

[0042] The initial system topology data of the distribution network refers to the basic data describing the structure and connection relationship of the distribution network, including information such as nodes, edges, equipment and loads. The initial system topology data is the basis for power system analysis. These data are crucial for the operation, planning and optimization of the power system. Usually, the initial system topology data of the distribution network can be obtained through the power company database, geographic information system or public data set.

[0043] S104. Analyze the voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network to obtain multiple potential installation locations for energy storage equipment and new energy equipment.

[0044] Among them, voltage sensitivity analysis is used to evaluate the sensitivity of voltage at different locations in the distribution network to various factors (such as load changes, power generation changes, etc.). Through voltage sensitivity analysis, it is possible to identify which locations are susceptible to external conditions. At the same time, through voltage sensitivity analysis, it is possible to find locations that have a greater impact on voltage stability and quality, thereby determining multiple potential installation locations for energy storage equipment (such as battery energy storage systems) and new energy equipment (such as solar photovoltaic panels, wind turbines, etc.).

[0045] S105. Calculate the construction cost and carbon emissions of each potential installation location of the energy storage equipment and new energy equipment, and determine the distribution network planning scheme, i.e., the installation location and capacity of the energy storage equipment and new energy equipment.

[0046] Among them, the construction cost of each potential installation location of energy storage equipment and new energy equipment includes but is not limited to equipment cost, installation cost and construction project cost. The construction cost of each potential installation location of energy storage equipment and new energy equipment is the sum of the above-mentioned equipment cost, installation cost and construction project cost.

[0047] The calculation of carbon emissions at each potential installation location of energy storage equipment and new energy equipment can be based on the following methods: (1) Emission factor method: a method for calculating carbon dioxide emissions, which calculates the carbon dioxide emissions from coal, oil, gas, and electricity transfer in and out. The formula is: carbon dioxide emissions = coal emissions + oil emissions + gas emissions - carbon dioxide emissions contained in electricity transfer out of the region, where the emission calculation formula for each fuel is: fuel emissions = fuel consumption in the current year × comprehensive fuel emission factor.

[0048] (2) Mass balance method: Calculate the emissions of various greenhouse gases based on the mass conservation relationship between input and output substances. The formula is: Carbon emissions = Carbon mass in input substances - Carbon mass in output substances.

[0049] (3) Field measurement method: A method based on field measurement and laboratory analysis used to calculate carbon emissions.

[0050] The installation location and capacity, that is, the distribution network planning scheme, can be determined through optimization algorithms (such as linear programming, mixed integer programming, etc.).

[0051] The embodiment of the present invention provides a distribution network planning method under multi-scenario variables, including: collecting multi-scenario variable data; constructing a convolutional recurrent neural network model based on the multi-scenario variable data; obtaining initial system topology data of the distribution network; analyzing the voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network to obtain multiple potential installation locations of energy storage equipment and new energy equipment; calculating the construction cost and carbon emissions of each potential installation location of the energy storage equipment and the new energy equipment, and determining the distribution network planning scheme, that is, the installation location and capacity of the energy storage equipment and the new energy equipment. Compared with the prior art, the embodiment of the present invention can capture the implicit deep relationship between complex nonlinear sequence data, effectively improve the prediction accuracy while considering multi-scenario variables, and has significant superiority in the field of source-load prediction.

[0052] Embodiment 2 The embodiment of the present invention provides a distribution network planning method under multiple scenario variables, such as Figure 2 As shown, including: S201. Collect multi-scenario variable data.

[0053] The multi-scenario variable data of this embodiment includes temperature, humidity, wind speed, date type, historical load data and power supply side data. Specifically, temperature, humidity and wind speed data can be obtained through weather stations, satellite data or online weather services; date types include weekdays, weekends, holidays, which can be obtained according to the current specific date; historical load data can be collected from power companies or smart meters; power supply side data includes the output power, fuel consumption, and equipment status of the power plant, which can be obtained through the power plant monitoring system, fuel supply system and power dispatching center.

[0054] S202: Preprocess multi-scenario variable data.

[0055] The multi-scenario variable data is preprocessed, including but not limited to data normalization and missing value filling. Normalization is a data preprocessing technology used to convert data of different dimensions or value ranges into a relatively uniform range, usually [0, 1] or [-1, 1]. The purpose of normalization is to eliminate the impact caused by different data value ranges, so that different features have the same scale, thereby improving the training effect and convergence speed of the model. Missing Value Imputation is an important step in data preprocessing, which aims to replace the missing parts in the data set with reasonable values ​​to ensure the integrity of the data and the effectiveness of the model.

[0056] S203. Construct a convolutional recurrent neural network model based on multi-scenario variable data.

[0057] S2031. Extract local spatial features of multi-scene variable data and generate a feature map.

[0058] In the construction of the convolutional recurrent neural network model, the convolutional neural network is first used to extract the local spatial features of multi-scene variable data: Assume that the multi-scenario variable data is , the convolution operation can be expressed as:

[0059]

[0060] in, represents the number of time steps, represents the number of features contained in each time step, represents the convolution operation, represents the convolution kernel, Indicates bias is the activation function, Represents a feature map.

[0061] The convolution operation extracts local features in the time dimension by sliding the convolution kernel, thereby capturing the correlation between different scene variables and generating a feature map. .

[0062] S2032. Capture the dynamic dependency of the time series in the feature graph to obtain a predicted output sequence.

[0063] The above feature map Input into the recurrent neural network to capture the temporal dynamic relationship of source-load data. Commonly used recurrent units include long short-term memory networks and gated recurrent units, which can effectively solve the gradient vanishing and gradient exploding problems, making them suitable for processing long time series prediction tasks. Among them, the recursive calculation formula of the long short-term memory network is as follows:

[0064]

[0065]

[0066]

[0067]

[0068] in, represents the activation value of the input gate, represents the activation value of the forget gate, represents the activation value of the output gate, Indicates the unit status, represents the hidden layer output, represents the Sigmoid function, represents the hyperbolic tangent function, represents an element-wise multiplication operation, W represents the weight matrix from the input to the hidden layer, U represents the weight matrix from the hidden state to the hidden layer, and b represents the bias vector.

[0069] Through these gating mechanisms, the LSTM network can selectively memorize or forget historical information according to the changes in time steps, thereby capturing the dynamic dependencies of time series in the feature graph and obtaining the predicted output sequence.

[0070] S2033. For the predicted output sequence, minimizing the mean square error is selected as the loss function and the optimizer updates the model parameters to obtain the trained convolutional recurrent neural network model.

[0071] For the predicted output sequence, the loss function is chosen to minimize the mean square error, which is used to measure the error between the predicted value and the true value:

[0072]

[0073] in, Indicates the real source charge value, represents the predicted value of the model, Indicates the sample size.

[0074] The loss function is minimized through an optimization algorithm, such as the Adam optimizer, to update the model parameters and obtain a trained convolutional recurrent neural network model.

[0075] S204: Acquire initial system topology data of the distribution network.

[0076] The initial system topology data of the distribution network refers to the basic data describing the structure and connection relationship of the distribution network, including information such as nodes, edges, equipment and loads. The initial system topology data is the basis for power system analysis. These data are crucial for the operation, planning and optimization of the power system. Usually, the initial system topology data of the distribution network can be obtained through the power company database, geographic information system or public data set.

[0077] S205. Analyze the voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network to obtain multiple potential installation locations for energy storage equipment and new energy equipment.

[0078] The voltage sensitivity of each node is predicted through the convolutional recurrent neural network model. According to the voltage sensitivity distribution, it is determined which nodes are suitable for installing energy storage equipment and new energy equipment. Generally, energy storage equipment is usually installed at nodes with large voltage fluctuations to smooth voltage fluctuations, and new energy equipment is usually installed at nodes with stable voltage and high load to maximize energy utilization efficiency.

[0079] S206. Calculate the construction cost and carbon emissions of each potential installation location of the energy storage equipment and the new energy equipment, and determine the distribution network planning scheme, i.e., the installation location and capacity of the energy storage equipment and the new energy equipment.

[0080] Among them, the construction cost of each potential installation location of energy storage equipment and new energy equipment includes but is not limited to equipment cost, installation cost and construction project cost. The construction cost of each potential installation location of energy storage equipment and new energy equipment is the sum of the above-mentioned equipment cost, installation cost and construction project cost.

[0081] The calculation of carbon emissions at each potential installation location of energy storage equipment and new energy equipment can be based on the following methods: (1) Emission factor method: a method for calculating carbon dioxide emissions, which calculates the carbon dioxide emissions from coal, oil, gas, and electricity transfer in and out. The formula is: carbon dioxide emissions = coal emissions + oil emissions + gas emissions - carbon dioxide emissions contained in electricity transfer out of the region, where the emission calculation formula for each fuel is: fuel emissions = fuel consumption in the current year × comprehensive fuel emission factor.

[0082] (2) Mass balance method: Calculate the emissions of various greenhouse gases based on the mass conservation relationship between input and output substances. The formula is: Carbon emissions = Carbon mass in input substances - Carbon mass in output substances.

[0083] (3) Field measurement method: A method based on field measurement and laboratory analysis used to calculate carbon emissions.

[0084] The installation location and capacity, that is, the distribution network planning scheme, can be determined through optimization algorithms (such as linear programming, mixed integer programming, etc.).

[0085] S207, simulating the operation of the distribution network planning scheme in different time periods and different scenarios, calculating the operating cost and the carbon emissions in each scenario, and obtaining the optimal operating point in each scenario.

[0086] In order to improve the robustness of the model, especially considering the uncertainty of source load prediction, multi-scenario variable simulation is introduced in the training process to increase the diversity of training data through data enhancement or scene sampling methods.

[0087] Use simulation software or programming languages ​​(such as MATLAB, Python, etc.) to establish a simulation model of the distribution network, including the operating characteristics of the equipment, load demand, electricity price and carbon emission factors, etc., run the simulation model in different time periods and different scenarios, and record the operating status of each device (such as charging and discharging status, power generation, etc.). Then, based on the simulation results, calculate the operating cost and carbon emissions in each scenario to obtain the optimal operating point in each scenario.

[0088] S208. Iteratively optimize the optimal operation point to obtain a distribution network planning optimization solution.

[0089] Due to the complexity of distribution network planning, which involves large-scale mixed integer optimization, there is no direct and effective solution. This embodiment proposes a two-layer progressive structure of power network configuration optimization model. This structure handles different constraints and objective functions in the upper and lower layers respectively.

[0090] At the upper level, the discrete planning stage of the distribution network is considered, and the decision variables are the potential installation locations and capacities of the distribution network. At the lower level, the distribution network operates in different scenarios, and the decision variables are the capacities of each device. The upper and lower decision variables affect each other, so it is necessary to iterate the solution. The specific model is:

[0091] This expression can well explain the secondary optimization problem. The upper layer is based on Optimization for the goal, subject to Condition, while the lower layer is is optimized for the target and is subject to conditions.

[0092] The upper model involves mixed integer programming, where the decision variables include the planned location and capacity of the power grid. The objective function is related to the construction cost of the new energy units and energy storage systems:

[0093] in, represents the discount rate, m represents the loan ratio, represents the inflation rate, q represents the service life of the equipment, C swt represents the construction cost of wind power unit capacity, C ypv represents the construction cost of photovoltaic unit capacity, C soss represents the construction cost per unit capacity of the energy storage system, E wt represents the construction capacity of wind power, E pv Represents the construction capacity of photovoltaic power generation, E ess Represents the construction capacity of the energy storage system.

[0094] The upper model constraints include the following: (1) System power balance constraints:

[0095] in, Indicates whether the node The decision variable for installing a new energy unit or energy storage system, 1 means installation, 0 means no installation, and L represents the total load of the region; Indicates the generator power, An indicator variable indicating whether wind turbine i is running, represents the power output of the wind turbine, represents the indicator variable of photovoltaic system i, represents the power output of the photovoltaic system, represents the power contribution of the generator or the mains, represents the operating status indicator variable of energy storage system i, Represents the power output of the energy storage system.

[0096] (2) Constraints on the number of new energy installations:

[0097] in, Indicates the maximum number of new energy installations; represents the indicator variable of the i-th wind turbine, Represents the indicator variable of the i-th photovoltaic power generation system.

[0098] (3) Constraints on new energy installation capacity:

[0099] in, Indicates the maximum installed capacity of renewable energy; Indicates the power of a single wind turbine; Indicates the power of a single photovoltaic system.

[0100] (4) Constraints on the number of energy storage installations:

[0101] in, represents the indicator variable of the i-th energy storage system, Indicates the maximum number of energy storage devices.

[0102] (5) Energy storage capacity constraints:

[0103] in, represents the indicator variable of the i-th energy storage system, Represents the power of a single energy storage system, Indicates the maximum capacity of the energy storage system.

[0104] The lower model considers the operation of the power system in multiple periods, tests the operating costs by simulating load demand and renewable energy output under different scenarios, and feeds this information back to the upper layer to further optimize the network and location capacity. Therefore, the lower model can be described as an optimal power flow problem with multiple periods of coupled continuous variables.

[0105] The lower objective function includes various operating costs, including the electricity purchase cost, power generation cost, carbon emission cost and operation and maintenance cost of wind turbines, photovoltaics, generators and energy storage.

[0106]

[0107] Among them, C p represents the power generation cost, C orepresents the operation and maintenance cost, Represents the sum of carbon emission costs and operation and maintenance costs.

[0108] The lower-level model constraints include the following: (1) Multi-period power balance constraints:

[0109] in, represents the load forecast of node i at time t, represents the active output of the generator at node i at time t, represents the active power output of wind power, Represents the active power output of the photovoltaic unit, Represents the active power output of energy storage, V i (t) represents the voltage value, G ij represents the equivalent conductance of branch ij, B ij represents the susceptance of branch ij, θ ij Represents the voltage phase angle difference across branch ij.

[0110] (2) Wind power active power model: According to the real-time wind speed v(t), the wind power output is modeled using a piecewise linear function:

[0111] Among them, P wt Indicates the rated power of the fan, v r Indicates rated wind speed, v in represents the cut-in wind speed, v out Indicates the cut-out wind speed.

[0112] (3) Photovoltaic unit active power model: The output model of the photovoltaic unit is related to the solar radiation Gc(t) and the temperature Tc(t):

[0113] Among them, T STC Indicates rated temperature, P STC Indicates rated power, G STC represents the rated light intensity; k represents the power temperature coefficient.

[0114] (4) Energy storage model: The charging power limit of the energy storage system under different charging and discharging conditions is considered:

[0115]

[0116] in, Indicates the charging status of the energy storage system. Represents the self-discharge coefficient of the energy storage system, Represents the active power output of energy storage, E max represents the maximum capacity of the energy storage system, represents the charging efficiency of the energy storage system, represents the discharge efficiency of the energy storage system, Indicates the duration of a single period.

[0117] (5) Generator output constraints:

[0118] in, represents the minimum output power of the generator at node i, Represents the maximum output power of the generator at node i.

[0119] S209. Perform actual testing and verification on the distribution network planning optimization plan based on historical data and scenario variables to obtain a verified distribution network planning optimization plan, i.e., a final distribution network planning optimization plan.

[0120] Specifically, the actual test and verification of the distribution network planning optimization scheme based on historical data and scenario variables includes the following steps: (1) Collect historical data: Collect historical load data (peak load, valley load, etc.) in different time periods, meteorological data (temperature, humidity, wind speed, etc.), equipment operation data (charging and discharging status, power generation, etc.), historical electricity prices and carbon emission factors.

[0121] (2) Defining scenario variables: Defining scenario variables can be divided into: defining different load demand scenarios, such as normal operation, peak load, valley load, etc.; defining different meteorological condition scenarios, such as sunny days, cloudy days, wind speed changes, etc.; defining different electricity prices and carbon emission factor scenarios, such as peak electricity prices, valley electricity prices, etc.

[0122] (3) Establish a simulation model: Use simulation software or programming languages ​​(such as MATLAB, Python, etc.) to build a simulation model of the distribution network, including the operating characteristics of the equipment, load demand, electricity price, and carbon emission factors.

[0123] (4) Run the simulation model: Run the simulation model under different historical data and different scenario variables, and record the operating status of each device (such as charging and discharging status, power generation, etc.).

[0124] (5) Verify the optimization scheme: Based on the simulation results, verify the feasibility and effectiveness of the distribution network planning optimization scheme.

[0125] The embodiment of the present invention provides a distribution network planning method under multi-scenario variables, including: collecting multi-scenario variable data; constructing a convolutional recurrent neural network model based on the multi-scenario variable data; obtaining initial system topology data of the distribution network; analyzing the voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network to obtain multiple potential installation locations of energy storage equipment and new energy equipment; calculating the construction cost and carbon emissions of each potential installation location of the energy storage equipment and the new energy equipment, and determining the distribution network planning scheme, that is, the installation location and capacity of the energy storage equipment and the new energy equipment. Compared with the prior art, the embodiment of the present invention can capture the implicit deep relationship between complex nonlinear sequence data, effectively improve the prediction accuracy while considering multi-scenario variables, and has significant superiority in the field of source-load prediction.

[0126] Embodiment 3 The embodiment of the present invention provides a distribution network planning device under multiple scenario variables, such as Figure 3 As shown, including: The collecting unit 31 is used to collect multi-scenario variable data.

[0127] The construction unit 32 is used to construct a convolutional recurrent neural network model based on multi-scenario variable data.

[0128] The acquisition unit 33 is used to acquire initial system topology data of the distribution network.

[0129] The analysis unit 34 is used to analyze the voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network to obtain multiple potential installation locations for energy storage equipment and new energy equipment.

[0130] The calculation unit 35 is used to calculate the construction cost and carbon emissions of each potential installation location of the energy storage device and the new energy device, and determine the distribution network planning scheme, that is, the installation location and capacity of the energy storage device and the new energy device.

[0131] It should be noted here that the detailed description of each component of this embodiment can refer to other embodiments and will not be repeated here.

[0132] The embodiment of the present invention provides a distribution network planning device under multi-scenario variables, including: a collection unit, used to collect multi-scenario variable data; a construction unit, used to construct a convolutional recurrent neural network model based on the multi-scenario variable data; an acquisition unit, used to acquire initial system topology data of the distribution network; an analysis unit, used to analyze the voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network, and obtain multiple potential installation locations of energy storage equipment and new energy equipment; a calculation unit, used to calculate the construction cost and carbon emissions of each potential installation location of the energy storage equipment and the new energy equipment, and determine the distribution network planning scheme, that is, the installation location and capacity of the energy storage equipment and the new energy equipment. Compared with the prior art, the embodiment of the present invention can capture the implicit deep relationship between complex nonlinear sequence data, effectively improve the prediction accuracy while considering multi-scenario variables, and has significant superiority in the field of source-load prediction.

[0133] Embodiment 4 The embodiment of the present invention provides a distribution network planning device under multiple scenario variables, such as Figure 4 As shown, including: The collecting unit 41 is used to collect multi-scenario variable data.

[0134] The preprocessing unit 42 is used to preprocess the multi-scenario variable data.

[0135] The construction unit 43 is used to construct a convolutional recurrent neural network model based on multi-scenario variable data.

[0136] The extraction module 431 is used to extract local spatial features of multi-scene variable data and generate a feature map.

[0137] The capture module 432 is used to capture the dynamic dependency of the time series in the feature graph to obtain a predicted output sequence.

[0138] The updating module 433 is used to select the minimization of mean square error as the loss function for predicting the output sequence and the optimizer to update the model parameters to obtain the trained convolutional recurrent neural network model.

[0139] The acquisition unit 44 is used to acquire initial system topology data of the distribution network.

[0140] The analysis unit 45 is used to analyze the voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network to obtain multiple potential installation locations of the energy storage device and the new energy device.

[0141] The calculation unit 46 is used to calculate the construction cost and carbon emissions of each potential installation location of the energy storage device and the new energy device, and determine the distribution network planning scheme, that is, the installation location and capacity of the energy storage device and the new energy device.

[0142] The simulation unit 47 is used to simulate the operation of the distribution network planning scheme in different time periods and different scenarios, calculate the operating cost and the carbon emissions in each scenario, and obtain the optimal operating point in each scenario.

[0143] The iteration unit 48 is used to iteratively optimize the optimal operation point to obtain a distribution network planning optimization solution.

[0144] The testing unit 49 is used to perform actual testing and verification on the distribution network planning optimization scheme based on historical data and scenario variables to obtain a verified distribution network planning optimization scheme, that is, a final distribution network planning optimization scheme.

[0145] It should be noted here that the detailed description of each component of this embodiment can refer to other embodiments and will not be repeated here.

[0146] The embodiment of the present invention provides a distribution network planning device under multi-scenario variables, including: a collection unit, used to collect multi-scenario variable data; a construction unit, used to construct a convolutional recurrent neural network model based on the multi-scenario variable data; an acquisition unit, used to acquire initial system topology data of the distribution network; an analysis unit, used to analyze the voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network, and obtain multiple potential installation locations of energy storage equipment and new energy equipment; a calculation unit, used to calculate the construction cost and carbon emissions of each potential installation location of the energy storage equipment and the new energy equipment, and determine the distribution network planning scheme, that is, the installation location and capacity of the energy storage equipment and the new energy equipment. Compared with the prior art, the embodiment of the present invention can capture the implicit deep relationship between complex nonlinear sequence data, effectively improve the prediction accuracy while considering multi-scenario variables, and has significant superiority in the field of source-load prediction.

[0147] Embodiment 5 An embodiment of the present invention provides an electronic device for planning a distribution network under multiple scenario variables, such as Figure 5 As shown, it includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the distribution network planning method under the above-mentioned multiple scenario variables are implemented.

[0148] A processing device (e.g., a central processing unit, a graphics processing unit, etc.) 51 can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 52 or a program loaded from a storage device 58 into a random access memory (RAM) 53. Various programs and data required for the operation of the electronic device are also stored in the RAM 53. The processing device 51, the ROM 52, and the RAM 53 are connected to each other via a bus 54. An input / output (I / O) interface 55 is also connected to the bus 54.

[0149] Typically, the following devices may be connected to the I / O interface 55: an input device 56 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 57 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 58 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 59. The communication device 59 may allow the electronic device to communicate with other devices wirelessly or by wire to exchange data. Although Figure 5 An electronic device having various devices is shown, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead. Figure 5 Each block shown in the figure may represent one device, or may represent multiple devices as required.

[0150] In particular, according to some embodiments of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, some embodiments of the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In some such embodiments, the computer program can be downloaded and installed from the network through the communication device 59, or installed from the storage device 58, or installed from the ROM 52. When the computer program is executed by the processing device 51, the above-mentioned functions defined in the method of some embodiments of the present invention are executed.

[0151] The embodiment of the present invention provides an electronic device for planning a distribution network under multi-scenario variables, including: collecting multi-scenario variable data; constructing a convolutional recurrent neural network model based on the multi-scenario variable data; obtaining initial system topology data of the distribution network; analyzing the voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network to obtain multiple potential installation locations of energy storage equipment and new energy equipment; calculating the construction cost and carbon emissions of each potential installation location of the energy storage equipment and the new energy equipment, and determining a distribution network planning scheme, that is, the installation location and capacity of the energy storage equipment and the new energy equipment. Compared with the prior art, the embodiment of the present invention can capture the implicit deep relationship between complex nonlinear sequence data, effectively improve the prediction accuracy while considering multi-scenario variables, and has significant superiority in the field of source-load prediction.

[0152] Embodiment 6 An embodiment of the present invention provides a computer-readable storage medium for distribution network planning under multiple scenario variables. A computer program is stored on the computer-readable storage medium. When the computer program is executed by one or more processors, the steps of the distribution network planning method under multiple scenario variables are implemented.

[0153] It should be noted that in some embodiments of the present invention, the computer-readable medium mentioned above may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, 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 above. In some embodiments of the present invention, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In some embodiments of the present invention, the computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than computer readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0154] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with 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"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0155] The computer-readable medium may be included in the device; or it may exist independently without being assembled into the electronic device. The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device: collects multi-scenario variable data; constructs a convolutional recurrent neural network model based on the multi-scenario variable data; obtains the initial system topology data of the distribution network; analyzes the voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network to obtain multiple potential installation locations of energy storage devices and new energy devices; calculates the construction cost and carbon emissions of each potential installation location of the energy storage device and the new energy device, and determines the distribution network planning scheme, that is, the installation location and capacity of the energy storage device and the new energy device. The computer program code for performing the operations of some embodiments of the present invention may be written in one or more programming languages ​​or a combination thereof, and the programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" language or similar programming languages. The program code may execute entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer, or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0156] The embodiment of the present invention provides a computer-readable storage medium for distribution network planning under multi-scenario variables, including: collecting multi-scenario variable data; constructing a convolutional recurrent neural network model based on the multi-scenario variable data; obtaining initial system topology data of the distribution network; analyzing the voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network to obtain multiple potential installation locations of energy storage equipment and new energy equipment; calculating the construction cost and carbon emissions of each potential installation location of the energy storage equipment and the new energy equipment, and determining the distribution network planning scheme, that is, the installation location and capacity of the energy storage equipment and the new energy equipment. Compared with the prior art, the embodiment of the present invention can capture the implicit deep relationship between complex nonlinear sequence data, effectively improve the prediction accuracy while considering multi-scenario variables, and has significant superiority in the field of source-load prediction.

[0157] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, not to limit them; although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention. In particular, as long as there is no structural conflict, the various technical features mentioned in each embodiment can be combined in any way. The present invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions that fall within the scope of the claims.

Claims

1. A distribution network planning method under multiple scenario variables, characterized in that: include: Collect multi-scenario variable data; Constructing a convolutional recurrent neural network model based on the multi-scenario variable data; Obtaining initial system topology data of the distribution network; Analyzing the voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network to obtain multiple potential installation locations of energy storage equipment and new energy equipment; The construction cost and carbon emissions of each of the potential installation locations of the energy storage device and the new energy device are calculated, and a distribution network planning scheme, i.e., the installation location and capacity of the energy storage device and the new energy device, is determined.

2. The distribution network planning method under multiple scenario variables according to claim 1 is characterized in that: Constructing a convolutional recurrent neural network model based on the multi-scenario variable data, including: Extracting local spatial features of the multi-scenario variable data and generating a feature map; Capturing the dynamic dependency of the time series in the feature graph to obtain a predicted output sequence; For the predicted output sequence, minimizing the mean square error is selected as the loss function and the optimizer updates the model parameters to obtain the convolutional recurrent neural network model after training.

3. The distribution network planning method under multiple scenario variables according to claim 1 is characterized in that: After calculating the construction cost and carbon emissions of each potential installation location of the energy storage device and the new energy device, and determining the distribution network planning scheme, i.e., the installation location and capacity of the energy storage device and the new energy device, it also includes: Simulate the operation of the distribution network planning scheme in different time periods and different scenarios, calculate the operating cost and the carbon emissions in each scenario, and obtain the optimal operating point in each scenario; The optimal operating point is iteratively optimized to obtain an optimization solution for distribution network planning.

4. The distribution network planning method under multiple scenario variables according to claim 3 is characterized in that: After iteratively optimizing the optimal operating point to obtain the distribution network planning optimization solution, the method further includes: The distribution network planning optimization scheme is actually tested and verified based on historical data and scenario variables to obtain the verified distribution network planning optimization scheme, that is, the final distribution network planning optimization scheme.

5. The distribution network planning method under multiple scenario variables according to claim 1 is characterized in that: After collecting multi-scenario variable data, it also includes: The multi-scenario variable data is preprocessed.

6. The distribution network planning method under multiple scenario variables according to claim 5 is characterized in that: The pre-processing comprises: The preprocessing includes but is not limited to data normalization and missing value filling.

7. The distribution network planning method under multiple scenario variables according to claim 1 is characterized in that: The multi-scenario variable data includes: The multi-scenario variable data includes but is not limited to temperature, humidity, wind speed, date type, historical load data and power supply side data.

8. A distribution network planning device under multiple scenario variables, characterized in that: include: A collection unit, used for collecting multi-scenario variable data; A construction unit, used to construct a convolutional recurrent neural network model based on the multi-scenario variable data; An acquisition unit, used for acquiring initial system topology data of the distribution network; An analysis unit, configured to analyze the voltage sensitivity of the convolutional recurrent neural network model and the initial system topology data of the distribution network to obtain a plurality of potential installation locations of energy storage equipment and new energy equipment; A calculation unit is used to calculate the construction cost and carbon emissions of each potential installation location of the energy storage device and the new energy device, and determine the distribution network planning scheme, that is, the installation location and capacity of the energy storage device and the new energy device.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images