Topology analysis and verification method based on starting scheme

Abstract

The invention discloses a topology analysis and verification method based on a starting scheme, relates to the technical field of new equipment starting, and solves the problem that a topology analysis method in the prior art cannot fully consider access randomness and intermittency of distributed energy. The method comprises the following steps: collecting distributed power supply historical output data and power grid topological structure data; then, on the basis of historical output data of new energy, modeling is conducted on the uncertainty of wind power and photovoltaic output through a generative adversarial network, typical output scenes with space-time correlation are generated, and the typical scenes are extracted through the scene reduction technology to serve as a new energy output scene set; and finally, on the basis of the new energy output scene set and the constructed topology analysis model, power flow sampling and checking and equipment operation state checking are carried out, so that the problems of randomness and intermittency caused by distributed energy access are effectively solved, and the safety and stability of a power grid under complex operation conditions are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of new device startup, and in particular to a topology analysis and verification method based on a startup scheme. Background Art

[0002] During the startup of new equipment, the startup plan and its verification are crucial. The startup plan is a blueprint for ensuring the safe and efficient connection of new equipment to the grid. It details the steps, sequence, and operating conditions for equipment startup, providing clear operational guidance for grid operators. Verification of the startup plan based on topology analysis is a key step in verifying the feasibility and safety of the startup plan. Topology analysis enables a comprehensive assessment of the structural changes and operating status of the grid after the new equipment is connected, identifying potential electrical connection issues, abnormal power flow distribution, or equipment overload risks. This allows for the optimization of the startup plan in advance, avoiding impacts on the grid or failures caused by equipment startup, and ensuring the safe and stable operation of the grid.

[0003] However, in the current startup plan verification, the existing topology analysis methods have obvious shortcomings for the integration of distributed energy resources (such as photovoltaic and wind power). Distributed energy resources have significant randomness and intermittency, and their output is greatly affected by weather conditions and environmental factors, making it difficult to accurately describe them using traditional deterministic models. Existing topology analysis methods are mostly based on static or deterministic grid structures and operating states, and it is difficult to fully consider the dynamic impact of the volatility of distributed energy output on grid topology and power flow distribution. This may lead to deviations between the verification results and the actual operating conditions, making it impossible to accurately assess the actual operational risks of the grid after the integration of distributed energy resources, thereby affecting the reliability and safety of the startup plan.

[0004] In view of this, a topology analysis and verification method based on the startup scheme is needed. Summary of the Invention

[0005] To address the problem that existing topology analysis methods cannot fully account for the randomness and intermittency of distributed energy access, the present invention provides a topology analysis and verification method based on startup schemes. This method can construct a topology analysis model that adapts to the uncertainty of distributed energy and verify the startup scheme of new equipment based on this model. This method can effectively address the randomness and intermittency problems brought about by distributed energy access, ensuring the security and stability of the power grid under complex operating conditions. The specific technical solution is as follows:

[0006] A topology analysis and verification method based on a startup scheme includes the following steps:

[0007] Collect historical output data of distributed power sources and grid topology data;

[0008] Based on historical output data of renewable energy, a generative adversarial network is used to model the uncertainty of wind power and photovoltaic output, generating typical output scenarios with spatiotemporal correlation. Scenario reduction technology is then used to extract these typical scenarios as the renewable energy output scenario set.

[0009] Construct a topological analysis model, including topological structure identification and topological adjacency matrix modeling;

[0010] Based on the new energy output scenario set and topological adjacency matrix, power flow sampling and verification and equipment operation status verification are carried out.

[0011] Preferably, the process of obtaining the new energy output scenario set is as follows:

[0012] Based on the acquired historical processing data of distributed power sources, data preprocessing is performed on it, including normalization and removal of outliers, and the data is divided into training sets and test sets;

[0013] Define the generator and discriminator, and introduce the self-attention module (SA) or position encoding into the discriminator to enhance the model's ability to capture temporal features;

[0014] Use real historical output data to train the discriminator so that it can distinguish between real data and generated data. Use the generator to generate fake data and train the discriminator to reduce the probability of it being deceived. Alternately train the generator and discriminator until the generator can generate scenes that are difficult to distinguish from real data.

[0015] Use the trained GAN model to generate a large number of new energy output scenarios as the new energy output scenario set.

[0016] Preferably, a clustering algorithm is used to divide the generated scenarios into several clusters, and the center point of each cluster is selected as a typical scenario. According to the clustering results, the probability of each typical scenario is calculated, and then typical scenarios are extracted from a large number of generated scenarios, and the typical scenarios replace a large number of generated scenarios as a new energy output scenario set.

[0017] Preferably, the process of topology identification is as follows:

[0018] Preprocess the topological structure data of the power grid and calculate the voltage difference or power difference between nodes;

[0019] Initialize the network structure of R-GCN, and generate an adjacency matrix based on the power grid topology to represent the connection relationship between nodes. Organize the preprocessed node features into a feature matrix as the input of the model.

[0020] The preprocessed data is divided into training set, validation set and test set, and the training set data is further processed to generate triple samples of node connection relationships;

[0021] The adjacency matrix and feature matrix are input, and the relationship between node features is learned through the graph convolution layer and activation function. The cross-entropy loss function is used to calculate the difference between the model output and the true label, and the model parameters are updated through backpropagation. The training is iterated until the model converges or the preset maximum number of iterations is reached.

[0022] The new power grid data is input into the trained R-GCN model, the connection relationship between nodes is output, and the topology of the power grid is identified.

[0023] Preferably, the process of topological adjacency matrix modeling is as follows:

[0024] Obtain node voltage, current, and power and perform data preprocessing;

[0025] Use the node voltage time series data to build the node feature matrix, and build the topological adjacency matrix according to the node association relationship type. The element A in the adjacency matrix A is ij Indicates the type of connection relationship between node i and node j.

[0026] The extracted node feature matrix and adjacency matrix are input into the R-GCN model, and the feature representation of the node is updated through graph convolution operations to generate a more accurate topological structure;

[0027] The topological adjacency matrix is dynamically updated through real-time data input.

[0028] Preferably, the tidal current sampling and verification are specifically as follows:

[0029] Obtain the topological structure data of the power grid and the set of renewable energy output scenarios, and collect line parameters and node electrical parameters;

[0030] Based on the renewable energy output scenario set, the renewable energy power generation in each period is sampled to form a power flow sampling section;

[0031] Combining the topological adjacency matrix with sampled renewable energy output data, a power flow calculation method is used to calculate the power flow distribution of the power grid under different scenarios, including the voltage amplitude and phase angle of each node, as well as the active power and reactive power of each line;

[0032] The calculated node voltage and line transmission power are compared with the preset voltage upper and lower limits and line transmission power upper limit to determine whether there is an over-limit situation. If the node voltage or line transmission power exceeds the constraint range, the operating state is marked as over-limit and the over-limit situation is recorded.

[0033] Preferably, the equipment operation status check is as follows:

[0034] Obtain the grid topology model and renewable energy output scenario set;

[0035] Using topology analysis models and combined with power flow calculation results, the operating status of key equipment such as transmission lines and main transformers in different scenarios is calculated, including transmission power, voltage level, and load factor.

[0036] Based on equipment operating status data, statistics on equipment overload and limit violations in different scenarios are collected, their occurrence probability is calculated, and high-risk equipment and operating time periods are identified;

[0037] Comprehensively analyze equipment overload, over-limit probability, and power flow distribution characteristics, and count dangerous points in power grid operation.

[0038] Preferably, the dangerous points include those containing heavy loads, exceeding limits or high risks, and the judgment conditions are as follows:

[0039] A load rate exceeding 80% of the rated capacity is considered overloaded, and a load rate exceeding 100% is considered overloaded.

[0040] A node voltage lower than 0.95 pu or higher than 1.05 pu is considered out of limit;

[0041] Line transmission power exceeding 90% of the thermal stability limit is considered high risk.

[0042] A computer-readable storage medium includes a stored program, wherein when the program is run, the device where the computer-readable storage medium is located is controlled to execute the above-mentioned topology analysis and verification method based on the startup scheme.

[0043] A processor is used to run a program, wherein the program executes the above-mentioned topology analysis and verification method based on the startup scheme when running.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] The present invention collects historical output data of distributed power sources and grid topology data; then, based on the historical output data of new energy, uses a generative adversarial network to model the uncertainty of wind power and photovoltaic output, generates typical output scenarios with spatiotemporal correlation, and extracts typical scenarios through scenario reduction technology as a new energy output scenario set; moreover, constructs a topological analysis model, including topological structure identification and topological adjacency matrix modeling; finally, based on the new energy output scenario set and the topological adjacency matrix, performs flow sampling and verification and equipment operation status verification. Based on this, the present invention can effectively deal with the randomness and intermittent problems brought about by distributed energy access, and ensure the safety and stability of the power grid under complex operating conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly describes the drawings required for the specific embodiments or the description of the prior art. Similar elements or parts are generally identified by similar reference numerals throughout the drawings. Elements or parts in the drawings are not necessarily drawn to scale.

[0047] Figure 1 is a flow chart of the overall method of the present invention;

[0048] Figure 2 A schematic diagram of the process of obtaining a typical output scenario set;

[0049] Figure 3 Schematic diagram of the process of topology identification;

[0050] Figure 4 Schematic diagram of the process of obtaining a topological adjacency matrix. DETAILED DESCRIPTION

[0051] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0052] It will be understood that when used in this specification and the appended claims, the terms “comprises” and “comprising” indicate the presence of described features, integers, steps, operations, elements and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0053] It should also be understood that the terms used in the present specification are only for the purpose of describing particular embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0054] It should be further understood that the term "and / or" used in the present description and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0055] In one embodiment of the present invention, a topology analysis and verification method based on a startup scheme is provided, such as Figure 1 As shown, the following steps are included:

[0056] Step 1: Collect historical output data of distributed power sources and grid topology data. In addition, also collect load forecasts, tie line plans, power generation plans, and maintenance plans.

[0057] Among them, distributed power sources include photovoltaic and wind power distributed power sources, and historical output data include power and time series; the topological structure data of the power grid includes the connection relationships of nodes, lines, and equipment; load forecasts, interconnection line plans, power generation plans, and maintenance plans are also for subsequent verification and analysis.

[0058] Step 2: Model the uncertainty of new energy output and output the probability distribution model of new energy output and a set of typical output scenarios.

[0059] This step uses a generative adversarial network (GAN) to model the uncertainty of wind power and photovoltaic output based on historical output data of renewable energy, generating typical output scenarios with spatiotemporal correlation. GAN is used to generate a set of typical scenarios for renewable energy output, and scenario simplification techniques (such as synchronous back-substitution elimination) are used to extract typical scenarios to reduce the computational burden.

[0060] like Figure 2 As shown, the details are as follows:

[0061] 1. Based on the acquired historical processing data of distributed power sources, perform data preprocessing, including normalization and removal of outliers, and divide the data into training and test sets;

[0062] 2. Select a GAN architecture (here you can use the modified Wasserstein GAN (WGAN) or its variants (such as WGAN-GP) to improve training stability and generated data quality), define the generator and discriminator, and introduce a self-attention module (SA) or position encoding in the discriminator to enhance the model's ability to capture temporal features.

[0063] The definition generator mentioned above: input a random noise vector and output a new energy output scenario that is close to the real data distribution.

[0064] The definition of the discriminator mentioned above: determines whether the input data is real data or generated data, and outputs a probability value.

[0065] 3. Use real historical output data to train the discriminator so that it can distinguish between real data and generated data. Use the generator to generate fake data and train the discriminator to reduce the probability of it being deceived. Alternately train the generator and discriminator until the generator can generate scenes that are difficult to distinguish from real data.

[0066] In addition, the Wasserstein distance and gradient penalty mechanism can be used to optimize the training process and set appropriate hyperparameters (such as learning rate, batch size, etc.) to improve training efficiency.

[0067] 4. Use the trained GAN model to generate a large number of renewable energy output scenarios, ensuring that the generated scenarios have temporal and spatial correlation and randomness. By comparing with real data, the quality of the generated scenarios is evaluated to ensure that they can reflect the uncertainty of renewable energy output.

[0068] 5. Use a clustering algorithm (such as k-means) to divide the generated scenes into several clusters, select the center point of each cluster as a typical scene, calculate the probability of each typical scene based on the clustering results, and then extract the typical scene from a large number of generated scenes.

[0069] Step 3: Construct a topological analysis model, including topological structure identification and topological adjacency matrix modeling.

[0070] Among them, topology recognition specifically involves using a relational graph convolutional neural network (R-GCN) to identify the topology of the power grid after distributed energy is connected. Figure 3 As shown, the details are as follows:

[0071] 1. Based on the grid topology data obtained in step 1, data preprocessing is performed, including data cleaning and normalization, and node features are calculated, including the voltage difference or power difference between nodes, as input to the R-GCN.

[0072] 2. Initialize the R-GCN network structure, including the number of graph convolutional layers, hidden layer dimensions, and activation functions; (The R-GCN model updates the feature representation of nodes by aggregating information from neighboring nodes and can handle relational information in graph structures).

[0073] 3. Based on the grid topology, an adjacency matrix is generated to represent the connection relationship between nodes, and the preprocessed node features are organized into a feature matrix as the input of the model.

[0074] 4. The preprocessed data is divided into training set, validation set and test set. The training set data is further processed to generate triple samples of node connection relationships (such as node pairs and their relationship types) for training the R-GCN model.

[0075] 5. Input the adjacency matrix and feature matrix, and learn the relationship between node features through the graph convolution layer and activation function; use the cross-entropy loss function to calculate the difference between the model output and the true label, and update the model parameters through backpropagation; iterate the training until the model converges or reaches the preset maximum number of iterations.

[0076] 6. Input new grid data into the trained R-GCN model, output the connection relationships between nodes, and identify the topology of the grid. For grids with distributed energy resources connected, the model can identify node types and the connection relationships between nodes, providing detailed topology information.

[0077] The topological adjacency matrix modeling is specifically as follows: establishing a topological adjacency matrix model that adapts to distributed energy access and considering the electrical coupling relationship between nodes.

[0078] After distributed energy is connected, the node voltage amplitude may change, causing the traditional topology identification method to fail. Therefore, it is necessary to analyze the impact of distributed energy access on node voltage correlation.

[0079] The relationship between nodes is classified into separation relationship, up-down relationship, parallel relationship and acceptance relationship.

[0080] For example:

[0081] Separation relationship (electrical distance is far): x ij =x ji =0;

[0082] Upper and lower relationship (electrical distance is closer): x ij =-x ji =1;

[0083] Parallel relationship (electrical distance is very close): x ij =x ji =-2;

[0084] Acceptance relationship (distributed energy access point): x ij =-x ji =0.5;

[0085] like Figure 4 As shown, the topological adjacency matrix modeling is as follows:

[0086] 1. Obtain the node's voltage, current, power and other electrical quantity data and perform data preprocessing. These data will be used to analyze the electrical coupling relationship between nodes.

[0087] 2. Use the node voltage time series data to build a node feature matrix, and build a topological adjacency matrix based on the node association relationship type. The element A in the adjacency matrix A is ij Indicates the type of connection relationship between node i and node j.

[0088] 3. The extracted node feature matrix and adjacency matrix are input into the R-GCN model. Graph convolution operations are used to update the node feature representation. The R-GCN model aggregates information from neighboring nodes, explores potential relationships between nodes, and generates a more accurate topological structure. Through the link prediction function of the R-GCN model, the complete topological structure of the distribution area is gradually generated.

[0089] 4. After distributed energy is connected, the grid topology may change. By inputting real-time data and dynamically updating the topological adjacency matrix, the model can adapt to the dynamic changes of the grid.

[0090] Step 4: Verify the startup plan based on topology analysis, including power flow sampling and verification and equipment operating status verification.

[0091] This step verifies the startup plan based on the new energy output scenario set (from step 2), the grid topology model (from step 3), the grid topology, and the topological adjacency matrix (from step 3), so that the verification process can fully consider its randomness and intermittency, and thus make the verification results more adaptable to the access of distributed energy, and will not be affected by the access of distributed power sources and reduce the accuracy of the verification.

[0092] The specific steps of power flow sampling and verification are as follows: based on the set of renewable energy output scenarios, the renewable energy power generation in each period is sampled to form a power flow sampling section; combined with the topological adjacency matrix, the power flow distribution of the power grid under different scenarios is calculated. The details are as follows:

[0093] 1. Data collection: Obtain the grid topology data collected in step 1 and the renewable energy output scenario set obtained in step 2, and collect line parameters and node electrical parameters. Line parameters include impedance and transmission power upper limit, and node electrical parameters include upper and lower voltage limits.

[0094] 2. Based on the renewable energy output scenario set, the renewable energy power generation in each time period is sampled to form a power flow sampling section. Methods such as Latin hypercube sampling are used to generate representative sampling points to ensure that the sampled data can cover the uncertainty of renewable energy output and provide diverse input scenarios for power flow calculations.

[0095] 3. Combining the topological adjacency matrix and the sampled renewable energy output data, use the power flow calculation method (such as the Newton-Raphson method) to calculate the power flow distribution of the power grid under different scenarios, including the voltage amplitude and phase angle of each node and the active power and reactive power of each line, so as to obtain the operating status of the power grid under different renewable energy output scenarios.

[0096] 4. Compare the calculated node voltage and line transmission power with the preset upper and lower voltage limits and upper limit of the line transmission power to determine whether any limit violations exist. If any node voltage or line transmission power exceeds the constraint range, the operating state is marked as out of limit and the violation is recorded to provide a basis for subsequent safety assessments.

[0097] The specific verification of equipment operating status is as follows: Based on the topological analysis model, the operating status of transmission lines, main transformers and other equipment in different scenarios is counted, the probability of equipment overload and over-limit is analyzed, and the dangerous points of power grid operation are counted; the details are as follows:

[0098] 1. The grid topology model and renewable energy output scenario set are used as input, combined with the topological adjacency matrix to provide basic data and structural framework for equipment operation status analysis.

[0099] 2. Use the topology analysis model and combine it with the power flow calculation results to calculate the operating status of key equipment such as transmission lines and main transformers in different scenarios, including transmission power, voltage level, and load rate.

[0100] 3. Based on the equipment operating status data, count the equipment's overload (load rate exceeds the rated value) and over-limit (voltage and power exceed the safe range) situations in different scenarios, calculate their probability of occurrence, and identify high-risk equipment and operating time periods.

[0101] 4. Comprehensively analyze equipment overload, over-limit probability, and power flow distribution characteristics, and count dangerous points in power grid operation. The criteria for determining dangerous points are as follows:

[0102] Equipment load rate: A load rate exceeding 80% of the rated capacity is considered overloaded, and a load rate exceeding 100% is considered out of limit.

[0103] Voltage level: A node voltage lower than 0.95 pu or higher than 1.05 pu is considered out of limit.

[0104] Power flow distribution: Line transmission power exceeding 90% of the thermal stability limit is considered high risk.

[0105] To better illustrate the verification process of step 4, the following is a further explanation based on a specific case:

[0106] The input data is as follows:

[0107] New energy output scenario set:

[0108] Scenario 1: PV output 100MW, wind power output 50MW.

[0109] Scenario 2: PV output 80MW, wind power output 60MW.

[0110] Scenario 3: PV output 120MW, wind power output 40MW.

[0111] The optimized grid topology model includes nodes A, B, C, D, and connecting lines AB, BC, CD, and DA.

[0112] Topological adjacency matrix:

[0113] The verification process is as follows:

[0114] Power flow sampling and verification: Flow calculations are performed for each scenario to determine the transmission power and node voltage for each line. For example, in scenario 1, the transmission power of line AB is 150 MW, and the voltage at node A is 225 kV. Equipment operating status verification: Statistics are collected to analyze the operating status of equipment and determine whether there are overload or over-limit risks.

[0115] For example, the safety margin of line AB is 10%, and the safety margin of line CD is 5%.

[0116] Output verification results:

[0117] Safety assessment results of the startup plan:

[0118] Overall Conclusion: The startup plan passed verification, but under Scenario 2, the safety margin of Line CD is low and requires attention. Risk Warning: Line CD has the risk of exceeding the limit in high-load scenarios.

[0119] Statistics of device operating status:

[0120] Device Name Voltage level Transmission power (MW) Safety margin (%) Line AB 220kV 150 10 Line CD 220kV 180 5

[0121] Limit crossing probability, overload probability:

[0122] Device Name Limit crossing probability (%) Overload probability (%) Line AB 0 5 Line CD 10 20

[0123] Analysis of new energy output scenarios:

[0124] Scene Number New Energy Station Active power output (MW) Scenario 1 Photovoltaic and wind power 150 Scenario 2 Photovoltaic and wind power 140 Scenario 3 Photovoltaic and wind power 160

[0125] Scenario risk assessment:

[0126] Scene Number Node voltage exceeding limit probability (%) Branch current exceeding limit probability (%) Scenario 1 0 0 Scenario 2 0 10 Scenario 3 0 5

[0127] In summary, the present invention collects historical output data of distributed power sources and grid topology data; then, based on the historical output data of new energy, uses a generative adversarial network to model the uncertainty of wind power and photovoltaic output, generates typical output scenarios with spatiotemporal correlation, and extracts typical scenarios through scenario reduction technology as a new energy output scenario set; moreover, constructs a topological analysis model, including topological structure identification and topological adjacency matrix modeling; finally, based on the new energy output scenario set and the topological adjacency matrix, performs flow sampling and verification and equipment operating status verification. Based on this, the present invention can effectively deal with the randomness and intermittent problems brought about by distributed energy access, and ensure the safety and stability of the power grid under complex operating conditions.

[0128] Those skilled in the art will appreciate that the units of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition of each example has been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0129] In the embodiments provided by the present invention, it should be understood that the division of units is merely a logical function division, and there may be other division methods in actual implementation, for example, multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored, etc.

[0130] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0131] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-0nly Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc., various media that can store program code.

[0132] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. 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 make equivalent replacements for some or all of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions 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 description of the present invention.

Claims

1. A topology analysis and verification method based on a startup scheme, characterized in that: The following steps are involved: Collect historical output data of distributed power sources and grid topology data; Based on historical output data of renewable energy, a generative adversarial network is used to model the uncertainty of wind power and photovoltaic output, generating typical output scenarios with spatiotemporal correlation. Scenario reduction technology is then used to extract these typical scenarios as the renewable energy output scenario set. Construct a topological analysis model, including topological structure identification and topological adjacency matrix modeling; Based on the new energy output scenario set and topological adjacency matrix, power flow sampling and verification and equipment operation status verification are carried out.

2. A topology analysis and verification method based on a startup scheme according to claim 1, characterized in that: The process of obtaining the new energy output scenario set is as follows: Based on the acquired historical processing data of distributed power sources, data preprocessing is performed on it, including normalization and removal of outliers, and the data is divided into training sets and test sets; Define the generator and discriminator, and introduce the self-attention module (SA) or position encoding into the discriminator to enhance the model's ability to capture temporal features; Use real historical output data to train the discriminator so that it can distinguish between real data and generated data. Use the generator to generate fake data and train the discriminator to reduce the probability of it being deceived. Alternately train the generator and discriminator until the generator can generate scenes that are difficult to distinguish from real data. Use the trained GAN model to generate a large number of new energy output scenarios as the new energy output scenario set.

3. A topology analysis and verification method based on a startup scheme according to claim 2, characterized in that: A clustering algorithm is used to divide the generated scenarios into several clusters. The center point of each cluster is selected as a typical scenario. Based on the clustering results, the probability of each typical scenario is calculated. Then, typical scenarios are extracted from a large number of generated scenarios, and the typical scenarios replace a large number of generated scenarios as the new energy output scenario set.

4. A topology analysis and verification method based on a startup scheme according to claim 1, characterized in that: The process of topology identification is as follows: Preprocess the topological structure data of the power grid and calculate the voltage difference or power difference between nodes; Initialize the network structure of R-GCN, and generate an adjacency matrix based on the power grid topology to represent the connection relationship between nodes. Organize the preprocessed node features into a feature matrix as the input of the model. The preprocessed data is divided into training set, validation set and test set, and the training set data is further processed to generate triple samples of node connection relationships; Input the adjacency matrix and feature matrix, and learn the relationship between node features through the graph convolution layer and activation function; The cross-entropy loss function is used to calculate the difference between the model output and the true label, and the model parameters are updated through backpropagation; the training is iterated until the model converges or the preset maximum number of iterations is reached. The new power grid data is input into the trained R-GCN model, the connection relationship between nodes is output, and the topology of the power grid is identified.

5. A topology analysis and verification method based on a startup scheme according to claim 4, characterized in that: The process of topological adjacency matrix modeling is as follows: Obtain node voltage, current, and power and perform data preprocessing; Use the node voltage time series data to build the node feature matrix, and build the topological adjacency matrix according to the node association relationship type. The element A in the adjacency matrix A is ij Indicates the type of connection relationship between node i and node j. The extracted node feature matrix and adjacency matrix are input into the R-GCN model, and the feature representation of the node is updated through graph convolution operations to generate a more accurate topological structure; The topological adjacency matrix is dynamically updated through real-time data input.

6. A topology analysis and verification method based on a startup scheme according to claim 1, characterized in that: The details of tidal current sampling and verification are as follows: Obtain the topological structure data of the power grid and the set of renewable energy output scenarios, and collect line parameters and node electrical parameters; Based on the renewable energy output scenario set, the renewable energy power generation in each period is sampled to form a power flow sampling section; Combining the topological adjacency matrix with sampled renewable energy output data, a power flow calculation method is used to calculate the power flow distribution of the power grid under different scenarios, including the voltage amplitude and phase angle of each node, as well as the active power and reactive power of each line; The calculated node voltage and line transmission power are compared with the preset voltage upper and lower limits and line transmission power upper limit to determine whether there is an over-limit situation. If the node voltage or line transmission power exceeds the constraint range, the operating state is marked as over-limit and the over-limit situation is recorded.

7. A topology analysis and verification method based on a startup scheme according to claim 6, characterized in that: The equipment operation status verification is as follows: Obtain the grid topology model and renewable energy output scenario set; Using topology analysis models and combined with power flow calculation results, the operating status of key equipment such as transmission lines and main transformers in different scenarios is calculated, including transmission power, voltage level, and load factor. Based on equipment operating status data, statistics on equipment overload and limit violations in different scenarios are collected, their occurrence probability is calculated, and high-risk equipment and operating time periods are identified; Comprehensively analyze equipment overload, over-limit probability, and power flow distribution characteristics, and count dangerous points in power grid operation.

8. A topology analysis and verification method based on a startup scheme according to claim 7, characterized in that: Dangerous points include those involving heavy loads, exceeding limits, or high risks. The judgment conditions are as follows: A load rate exceeding 80% of the rated capacity is considered overloaded, and a load rate exceeding 100% is considered overloaded. A node voltage lower than 0.95 pu or higher than 1.05 pu is considered out of limit; Line transmission power exceeding 90% of the thermal stability limit is considered high risk.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the topology analysis and verification method based on the startup scheme according to any one of claims 1 to 8.

10. A processor, characterized in that: The processor is used to run a program, wherein the program, when running, executes the topology analysis and verification method based on the startup scheme according to any one of claims 1 to 8.

Images