Method and device for determining the short-circuit ratio of multiple new energy power plants

Abstract

This application provides a method and apparatus for determining the short-circuit ratio of multiple renewable energy power plants. The method includes: acquiring graph structure data of a target power system, the target power system including multiple devices, including multiple renewable energy power plants; the nodes of the graph structure data represent the characteristic data of the devices, and the edges between the nodes represent the characteristic data of the lines; the characteristic data of the devices include: bus voltage, bus phase angle, active power injection of the bus, and reactive power injection of the bus; the characteristic data of the lines include: line active power and line reactive power; determining the short-circuit ratio of the multiple renewable energy power plants based on the graph structure data and a short-circuit ratio evaluation model; the short-circuit ratio evaluation model is obtained by pre-training a graph convolutional network model based on batch historical graph structure data and the corresponding real short-circuit ratios of the multiple renewable energy power plants. This application can improve the reliability of the short-circuit ratio of multiple renewable energy power plants, thereby improving the stability of power system operation.

Description

Technical Field

[0001] This application relates to the field of power system technology, and in particular to a method and apparatus for determining the short-circuit ratio of multiple new energy power plants. Background Technology

[0002] The short-circuit ratio of multiple renewable energy power plants is an important indicator for measuring the voltage support strength of a power system. With the widespread integration of renewable energy generation, the power grid structure has become more complex, and the short-circuit ratios of different renewable energy power plants vary significantly. Therefore, real-time monitoring and control of the short-circuit ratio of multiple renewable energy power plants is crucial for measuring the voltage strength of a system with multiple renewable energy power plants and for quantitatively assessing the scale of renewable energy integration.

[0003] Currently, the short-circuit ratio is typically evaluated using the short-circuit ratio analysis index. However, this index is an analytical expression of a power flow analysis model and cannot characterize the dynamic control characteristics of the analysis system and renewable energy power generation equipment. In actual operation, the short-circuit ratio constraints of multiple renewable energy power plants are incorporated into the renewable energy operation limit verification through offline calculations. This fails to consider the impact of time-varying factors such as renewable energy output distribution and topology on the short-circuit ratio. Consequently, under certain operating conditions, although the transmission limit of the transmission channel increases, some renewable energy power plants are limited in output due to inaccurate offline short-circuit ratios, resulting in limited renewable energy power generation and an increase in renewable energy curtailment. Summary of the Invention

[0004] To address at least one problem in the prior art, this application proposes a method and apparatus for determining the short-circuit ratio of multiple new energy power plants, thereby improving the reliability of the short-circuit ratio of multiple new energy power plants and thus enhancing the stability of power system operation.

[0005] To address the aforementioned technical problems, this application provides the following technical solution:

[0006] Firstly, this application provides a method for determining the short-circuit ratio of multiple new energy power stations, including:

[0007] Obtain graph structure data of a target power system, which includes multiple devices, including multiple renewable energy power plants. Nodes in the graph structure data represent the characteristic data of the corresponding devices, and edges between nodes represent the characteristic data of the corresponding lines. The characteristic data of the devices includes: bus voltage, bus phase angle, active power injection of the bus, and reactive power injection of the bus. The characteristic data of the lines includes: line active power and line reactive power.

[0008] Based on the graph structure data and the preset short-circuit ratio evaluation model, the short-circuit ratio of the new energy multi-site is determined;

[0009] The preset short-circuit ratio evaluation model is obtained by pre-training a graph convolutional network model based on batch historical graph structure data and their corresponding real short-circuit ratios of multiple new energy power stations.

[0010] In one embodiment, acquiring the graph structure data of the target power system includes:

[0011] Obtain the current steady-state characteristics and topology information of the target power system;

[0012] Based on the steady-state characteristics and topology information, the graph structure data of the target power system is obtained.

[0013] In one embodiment, the method for determining the short-circuit ratio of multiple new energy power stations further includes:

[0014] Collect a training set, which includes: batch historical graph structure data and their respective corresponding real short-circuit ratios of multiple new energy power stations;

[0015] The graph convolutional network model is trained using the training set to obtain the short-circuit ratio evaluation model.

[0016] In one embodiment, prior to acquiring the graph structure data of the target power system, the method further includes:

[0017] Collect a validation set, which includes: batch historical graph structure data and their respective corresponding real short-circuit ratios of multiple new energy power plants;

[0018] Based on each historical graph structure data and the short-circuit ratio evaluation model, the predicted short-circuit ratio of the new energy multi-site corresponding to the historical graph structure data is obtained;

[0019] The mean square error is obtained by processing the predicted short-circuit ratio and the actual short-circuit ratio of the new energy multi-stations corresponding to each historical graph structure data.

[0020] Determine whether the mean square error value is less than the mean square error threshold. If so, the short-circuit ratio evaluation model is deemed to have passed verification.

[0021] In one embodiment, the collection of the training set includes:

[0022] The target power system acquires historical operating data, historical topology information, and the actual short-circuit ratio of multiple new energy power plants under various operating scenarios. The power generation, load, and topology are the same for the same operating scenario, but the power generation, load, and topology are different for each operating scenario.

[0023] Normalize historical operational data under the same operational scenario;

[0024] Based on the normalized historical operation data and the historical topology information for each operation scenario, the historical graph structure data for that operation scenario is obtained.

[0025] The training set is composed of historical graph structure data under various operating scenarios and the actual short-circuit ratios of multiple new energy power stations.

[0026] Secondly, this application provides a device for determining the short-circuit ratio of multiple new energy power stations, comprising:

[0027] The acquisition module is used to acquire graph structure data of a target power system, which includes multiple devices, including multiple renewable energy power plants. Nodes in the graph structure data represent the characteristic data of the corresponding devices, and edges between nodes represent the characteristic data of the corresponding lines. The characteristic data of the devices includes: bus voltage, bus phase angle, active power injection of the bus, and reactive power injection of the bus. The characteristic data of the lines includes: line active power and line reactive power.

[0028] The determination module is used to determine the short-circuit ratio of the new energy multi-site based on the graph structure data and the preset short-circuit ratio evaluation model;

[0029] The preset short-circuit ratio evaluation model is obtained by pre-training a graph convolutional network model based on batch historical graph structure data and their corresponding real short-circuit ratios of multiple new energy power stations.

[0030] In one embodiment, the acquisition module includes:

[0031] The acquisition unit is used to acquire the current steady-state characteristics and topology information of the target power system;

[0032] The unit is used to obtain the graph structure data of the target power system based on the steady-state characteristics and topology information.

[0033] In one embodiment, the short-circuit ratio determination device for multiple new energy power stations further includes:

[0034] The data acquisition module is used to acquire training sets, which include: batch historical graph structure data and their corresponding real short-circuit ratios of multiple new energy power stations;

[0035] The training module is used to train the graph convolutional network model using the training set to obtain the short-circuit ratio evaluation model.

[0036] In one embodiment, the short-circuit ratio determination device for multiple new energy power stations further includes:

[0037] The verification set acquisition module is used to acquire a verification set, which includes: batch historical graph structure data and their respective corresponding real short-circuit ratios of multiple new energy power plants;

[0038] The predicted short-circuit ratio module is used to obtain the predicted short-circuit ratio of the new energy multi-site corresponding to each historical graph structure data and the short-circuit ratio evaluation model.

[0039] The mean square error value module is used to process the mean square error based on the predicted short-circuit ratio and the actual short-circuit ratio of the new energy multi-site corresponding to each historical graph structure data, and obtain the mean square error value.

[0040] The verification module is used to determine whether the mean square error value is less than the mean square error threshold. If so, the short-circuit ratio evaluation model is verified as passed.

[0041] In one embodiment, the acquisition module includes:

[0042] The historical data acquisition unit is used to acquire historical operating data, historical topology information and real short-circuit ratios of multiple new energy power plants under various operating scenarios of the target power system. The power generation, load and topology are the same for the same operating scenario, but the power generation, load and topology are different for each operating scenario.

[0043] The normalization unit is used to normalize historical operational data under the same operational scenario.

[0044] The historical graph structure data unit is obtained and used to obtain the historical graph structure data for each operation scenario based on the normalized historical running data and the historical topology information.

[0045] The combination unit is used to combine historical graph structure data under various operating scenarios and the actual short-circuit ratios of multiple new energy power stations to form the training set.

[0046] Thirdly, this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for determining the short-circuit ratio of new energy multi-site power plants.

[0047] Fourthly, this application provides a computer-readable storage medium storing computer instructions thereon, which, when executed by a processor, implement the method for determining the short-circuit ratio of new energy multi-site power plants.

[0048] As can be seen from the above technical solution, this application provides a method and apparatus for determining the short-circuit ratio of a multi-generation renewable energy power station. The method includes: acquiring graph structure data of a target power system, the target power system including multiple devices, the multiple devices including a multi-generation renewable energy power station; nodes in the graph structure data represent the characteristic data of the corresponding devices, and edges between nodes represent the characteristic data of the corresponding lines; the characteristic data of the devices includes: bus voltage, bus phase angle, active power injection of the bus, and reactive power injection of the bus; the characteristic data of the lines includes: line active power and line reactive power; determining the short-circuit ratio of the multi-generation renewable energy power station based on the graph structure data and a preset short-circuit ratio evaluation model; wherein, the preset short-circuit ratio evaluation model is based on... Based on batch historical graph structure data and their corresponding real short-circuit ratios of multiple renewable energy power plants, the proposed method is pre-trained using a graph convolutional network model. This model considers the impact of renewable energy output and topology on the short-circuit ratio of these power plants, improving their reliability and thus enhancing the stability of the power system. Specifically, it allows for real-time monitoring and control of the short-circuit ratio of renewable energy power plants, providing an intuitive and effective measure of the voltage intensity of these plants connected to the power system, and quantitatively assessing the scale of renewable energy integration, thereby improving grid stability. It also considers the characteristics of busbars, lines, and topology connections, resulting in a short-circuit ratio evaluation model to describe changes in the power system topology. The effectiveness of the proposed method is validated by running 40,000 sets of sample data on a 109-node system. The results show that compared to traditional data-driven methods, this approach has significant performance advantages in topology-changing scenarios and can meet the application requirements of complex and variable power grid operation scenarios. Graph network technologies can be applied to large-scale power grids to improve the interpretability of graph networks in power system-related problems. Combining power system physical model-based analysis methods with graph networks can further improve model performance and generalization ability. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 This is a schematic diagram of the first process of the method for determining the short-circuit ratio of multiple new energy power stations in the embodiments of this application;

[0051] Figure 2 This is a second flowchart illustrating the method for determining the short-circuit ratio of multiple new energy power stations in this application embodiment;

[0052] Figure 3 This is the network topology diagram of the 109-node system;

[0053] Figure 4 This is a comparative diagram showing the training and testing losses of the MRSCR1 dataset in the application example of this application as a function of the number of training rounds.

[0054] Figure 5 This is a schematic diagram of the short-circuit ratio determination device for multiple new energy power stations in the embodiments of this application;

[0055] Figure 6 This is a schematic block diagram of the system configuration of an electronic device according to an embodiment of this application. Detailed Implementation

[0056] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0057] In recent years, the proportion of new energy power generation equipment, represented by wind power and photovoltaics, in the power system has been increasing, and the access methods have become more diversified. In some areas, scenarios have emerged where new energy sources are integrated into weak power grids. This process is prone to problems such as broadband oscillations and transient overvoltages, seriously affecting the power quality and supply stability of the power grid. Currently, common methods for determining the short-circuit ratio of multiple renewable energy stations include: online monitoring methods for the aggregated short-circuit ratio (ASCR) and gSCR of wind farms to assess the initial risk of subsynchronous oscillations in the electric field; and evaluation methods for the grid connection strength of a single wind farm, including composite short-circuit ratio (CSCR), weighted short-circuit ratio (WSCR), and equivalent short-circuit ratio (ESCR). These methods only consider the characteristics of the network and not the power electronics equipment itself, possessing good engineering practical value. The generalized short-circuit ratio (gSCR) is also used, but the modal method calculation process is complex, has low engineering practicality, and there is a contradiction between the accuracy and practicality of the short-circuit ratio index. Some scholars have further proposed the multiple renewable energy station short-circuit ratio (MRSCR) and obtained the threshold of the critical short-circuit ratio through extensive simulations.

[0058] However, current offline calculations that incorporate multi-station short-circuit ratio constraints into renewable energy operation limit verification do not consider the impact of renewable energy distribution and topology on the short-circuit ratio. To address the problems of existing technologies and achieve online calculation, this application proposes a novel graph neural network framework and a method and apparatus for determining the short-circuit ratio of multiple renewable energy stations. This method can determine the online short-circuit ratio of multiple renewable energy stations based on message passing GCNConv. While ensuring accuracy, it utilizes power system analysis software for data analysis and considers the impact of renewable energy output and topology changes on the short-circuit ratio of multiple renewable energy stations. It can process the graph structure data of the power system and efficiently aggregate and propagate node and edge information through a message passing mechanism, thereby improving the efficiency of real-time online calculation of the short-circuit ratio of multiple renewable energy stations.

[0059] The following examples illustrate this in detail.

[0060] To improve the reliability of the short-circuit ratio of multiple renewable energy power plants and thus enhance the stability of power system operation, this embodiment provides a method for determining the short-circuit ratio of multiple renewable energy power plants, with the execution subject being a short-circuit ratio determination device. This short-circuit ratio determination device includes, but is not limited to, a server, such as... Figure 1 As shown, this method specifically includes the following:

[0061] Step 100: Obtain the graph structure data of the target power system, which includes multiple devices, including multiple new energy power plants; the nodes in the graph structure data represent the characteristic data of the corresponding devices, and the edges between the nodes represent the characteristic data of the corresponding lines; the characteristic data of the devices include: bus voltage, bus phase angle, active power injection of the bus, and reactive power injection of the bus; the characteristic data of the lines include: line active power and line reactive power.

[0062] Specifically, the target power system may include multiple devices, such as generators, transformers, circuit breakers, and new energy power stations.

[0063] Step 200: Determine the short-circuit ratio of the new energy multi-site based on the graph structure data and the preset short-circuit ratio evaluation model; wherein, the preset short-circuit ratio evaluation model is obtained by pre-training a graph convolutional network model based on batch historical graph structure data and the actual short-circuit ratios of their respective new energy multi-sites.

[0064] Specifically, the graph structure data can be input into a preset short-circuit ratio evaluation model, and the output of the preset short-circuit ratio evaluation model can be determined as the short-circuit ratio of the new energy multi-site power plant; the graph convolutional network model can be the GCNConv model. The short-circuit ratio of the new energy multi-site power plant can represent the individual short-circuit ratio of each new energy site within the new energy multi-site power plant, and the true short-circuit ratio of the new energy multi-site power plant can represent the true short-circuit ratio of each new energy site within the new energy multi-site power plant.

[0065] As described above, the short-circuit ratio determination method for multiple new energy power plants provided in this embodiment can reflect the new energy distribution and topology of the target power system based on the graph structure data of the target power system. It can take into account the impact of new energy distribution and topology on the short-circuit ratio, thereby improving the reliability of the short-circuit ratio of multiple new energy power plants and thus improving the stability of power system operation.

[0066] To improve the reliability of graph structure data, in one embodiment, step 100 includes:

[0067] Step 101: Obtain the current steady-state characteristics and topology information of the target power system.

[0068] Specifically, the topology information of a power system can represent the connection methods and organizational structure between various devices in the power system.

[0069] Step 102: Based on the steady-state characteristics and topology information, obtain the graph structure data of the target power system.

[0070] To improve the reliability of the short-circuit ratio evaluation model training, and further apply a reliable short-circuit ratio evaluation model to improve the accuracy of short-circuit ratio determination, such as... Figure 2 As shown, in one embodiment, the method further includes the following step before step 100:

[0071] Step 001: Collect training set, which includes: batch historical graph structure data and their corresponding real short-circuit ratios of multiple new energy power stations.

[0072] Step 002: Use the training set to train the graph convolutional network model to obtain the short-circuit ratio evaluation model.

[0073] To further improve the reliability of the short-circuit ratio evaluation model, in one embodiment, before acquiring the graphical structure data of the target power system, the following steps are also included:

[0074] Step 003: Collect the verification set, which includes: batch historical graph structure data and their corresponding real short-circuit ratios of multiple new energy power stations.

[0075] Step 004: Based on each historical graph structure data and the short-circuit ratio evaluation model, obtain the predicted short-circuit ratio of the new energy multi-site corresponding to the historical graph structure data.

[0076] Step 005: Perform mean square error processing on the predicted short-circuit ratio and the actual short-circuit ratio of the new energy multi-site corresponding to each historical graph structure data to obtain the mean square error value.

[0077] Specifically, the mean square error value can be determined using the following formula:

[0078]

[0079] Where n is the number of samples, i.e., the total number of historical graph structure data, and y i It is the true short-circuit ratio corresponding to the i-th sample. The predicted short-circuit ratio corresponding to the i-th sample.

[0080] Step 006: Determine whether the mean square error value is less than the mean square error threshold. If so, the short-circuit ratio evaluation model is verified.

[0081] Specifically, if the mean square error value is greater than the mean square error threshold, the short-circuit ratio evaluation model can continue to be trained until the short-circuit ratio evaluation model is verified.

[0082] To improve the reliability of the training set, and thus the reliability of the short-circuit ratio evaluation model training, in one embodiment, step 001 includes:

[0083] Step 011: Obtain historical operating data, historical topology information, and actual short-circuit ratios of multiple new energy power plants for the target power system under various operating scenarios. The power generation, load, and topology are the same for the same operating scenario, but the power generation, load, and topology are different for each operating scenario.

[0084] Step 012: Normalize the historical operation data under the same operation scenario.

[0085] Step 013: Based on the normalized historical running data and the historical topology information for each operating scenario, obtain the historical graph structure data for that operating scenario.

[0086] Step 014: Combine the historical graph structure data under each operation scenario with the actual short-circuit ratios of multiple new energy power stations to form the training set.

[0087] To further illustrate this solution, this application provides an application example of a method for determining the short-circuit ratio of multiple new energy power plants, as described in detail below:

[0088] Step 1: Sample Generation Phase. A training sample set is established based on steady-state data from the power system's operational history and derived simulation data. To enhance the model's applicability to diverse operating scenarios, power generation and load levels fluctuate within a certain range in the derived simulation data. The training set includes various power generation-load distributions and topology combinations.

[0089] During sample generation, the input characteristics of the samples need to be determined in advance. This application example uses steady-state data as the sample characteristics for evaluating the transient stability of the power system; the selected steady-state characteristics are shown in Table 1. The power subscripts v and e represent the vertices or edges of the graph containing the features, respectively. The topological information of each sample is represented by an adjacency matrix.

[0090] Table 1

[0091] Input characteristics meaning <![CDATA[P v ]]> Active power injection of the bus <![CDATA[Q v ]]> Reactive power injection from the bus V Bus voltage θ bus phase angle <![CDATA[P e ]]> Line active power <![CDATA[Q e ]]> Line reactive power

[0092] Each node feature in graph-structured data is a four-dimensional vector [P] v Q v In a power system, lines are modeled as edges in graph-structured data, and the feature vector of each edge is a two-dimensional vector [P]. e Q e To avoid convergence difficulties caused by excessively large numerical differences between features, Z-score standardization is used to normalize the bus and line feature data collected during the sample generation stage.

[0093]

[0094] in, It is feature X i The mean, It is X i The standard deviation of the features. After normalization, combined with the topological information represented by the adjacency matrix, they form graphical data representing each operational scenario.

[0095] Step 2: Graph structure data processing stage. Its main function is to reconstruct the long vector data obtained in the sample generation stage into graph structure data, which is used as input to the GCNConv message passing model. Each bus in the power system is modeled as a node in the graph structure data, and the node's features are the characteristics of the corresponding bus. For busbars that are simultaneously connected to generators and loads, power is injected to obtain their superposition value.

[0096] Step 3: Offline training phase, mainly involving building the model structure and training the model parameters. A training model for the short-circuit ratio of new energy multi-stations based on GCNConv is defined. It includes two GCN convolutional layers. It can implement convolution operations on graph-structured data by defining initialization, parameter reset, forward propagation, and message passing methods. It executes the message passing mechanism by calling the propagate method and aggregates the neighbor information of each node to update the node features.

[0097] The processed graph dataset was split into training and test sets in an 8:2 ratio. The training set was used to determine various hyperparameters and weights of GCNConv, while the test set was used to evaluate the performance of the trained GCNConv model.

[0098] After the model is built, the graph structure data training set obtained in the graph structure data processing stage is fed into the GCNConv model for offline model training. During training, the predicted values ​​of the GCNConv model are compared with the sample labels, the loss function is calculated, and the model's internal weights and bias parameters are updated using the backpropagation algorithm. The loss function used in training is the mean squared error (MSE), and these loss values ​​are output in each batch and epoch.

[0099] Taking a 109-node system as an example, 20 different operating modes and topologies were considered. The network topology diagram for the 109 nodes is as follows: Figure 3 As shown. The input feature is the 156 edges of 109 nodes. The output feature is the short-circuit ratio of the 6 new energy multi-site stations, that is, the short-circuit ratio of the 6 wind turbine nodes 96, 97, 104, 105, 106, and 107.

[0100] To ensure a relatively uniform distribution of total power generation and load levels considered in the samples, the power generation and load levels for each dataset were categorized and randomly varied during sample generation: the total power generation and load levels of the power system varied uniformly between 80% and 120%. 2025 samples were generated for each pattern, providing a total of 40,500 samples to validate the model's performance. The GCNConv model was implemented using the PyTorch Geometric framework. The initial learning rate was set to 0.01 after hyperparameter optimization using the Adam optimizer, and then reduced to 90% after 10 iterations. The batch size was set to 32, and the number of epochs was 200. Figure 4 This is a diagram comparing the training and testing losses of the MRSCR1 dataset as a function of the number of training epochs. Figure 4 Curve A in the figure represents the training loss as a function of the number of training epochs, while curve B represents the test loss as a function of the number of training epochs.

[0101] To evaluate the performance of the GCNConv model in calculating the short-circuit ratio of new energy substations under different topologies and operating modes, the mean square error (MSE) was selected to comprehensively assess the performance of various models.

[0102]

[0103] Where n is the number of samples, y i It is the true value of the i-th sample. The predicted value of the i-th sample.

[0104] First, the short-circuit ratio of new energy multi-stations was calculated and compared using the least squares fitting algorithm. The MSE calculated by the least squares fitting algorithm is smaller than the MSE calculated by the GCNConv model, but it can only be applied to one topology and not to other topologies.

[0105] Next, three common data-driven methods were selected for comparison: Random Forest (RF), XGBoost (XGB), and Deep Neural Network (DNN). RF and XGB each have 100 trees. The DNN is a 4-layer fully connected network with 512, 256, 128, and 64 neurons per layer, respectively. ReLU was chosen as the activation function, and Adam was chosen as the optimizer. The initial learning rate was 0.001, and the learning rate decayed to 90% of the original rate every 10 epochs. The DNN was implemented using the PyTorch framework. The RF, XGB, DNN, and GCNCV models were trained on the dataset. Table 2 shows the performance metrics of different models on different test sets.

[0106] Table 2

[0107]

[0108]

[0109] The overall accuracy of GCNConv is higher than that of RF, XGB, and DNN. This fully demonstrates the advantages of the GCNConv model in considering datasets with topology changes. In this application example, the GCNConv model can effectively process graph-structured power system topology data and can calculate the short-circuit ratio of multiple renewable energy stations online in real time.

[0110] Step 4: In the GCNConv online application stage, a graph structure data is constructed using the current node data, edge data, and topology information. This data is then input into the trained GCNConv model to output the short-circuit ratio of the new energy power station. The trained GCNConv model can be equivalent to the short-circuit ratio evaluation model mentioned above.

[0111] From a software perspective, in order to improve the reliability of the short-circuit ratio of multiple renewable energy power plants and thus enhance the stability of power system operation, this application provides an embodiment of a device for determining the short-circuit ratio of multiple renewable energy power plants, which implements all or part of the aforementioned method for determining the short-circuit ratio of multiple renewable energy power plants. See [link to relevant documentation]. Figure 5 The short-circuit ratio determination device for multiple new energy power stations specifically includes the following components:

[0112] Acquisition module 01 is used to acquire graph structure data of a target power system, which includes multiple devices, including multiple renewable energy power plants. Nodes in the graph structure data represent the characteristic data of the corresponding devices, and edges between nodes represent the characteristic data of the corresponding lines. The characteristic data of the devices includes: bus voltage, bus phase angle, active power injection of the bus, and reactive power injection of the bus. The characteristic data of the lines includes: line active power and line reactive power.

[0113] The determination module 02 is used to determine the short-circuit ratio of the new energy multi-site based on the graph structure data and the preset short-circuit ratio evaluation model; wherein, the preset short-circuit ratio evaluation model is obtained by pre-training a graph convolutional network model based on batch historical graph structure data and the actual short-circuit ratio of each corresponding new energy multi-site.

[0114] In one embodiment, the acquisition module includes:

[0115] The acquisition unit is used to acquire the current steady-state characteristics and topology information of the target power system;

[0116] The unit is used to obtain the graph structure data of the target power system based on the steady-state characteristics and topology information.

[0117] In one embodiment, the short-circuit ratio determination device for multiple new energy power stations further includes:

[0118] The data acquisition module is used to acquire training sets, which include: batch historical graph structure data and their corresponding real short-circuit ratios of multiple new energy power stations;

[0119] The training module is used to train the graph convolutional network model using the training set to obtain the short-circuit ratio evaluation model.

[0120] In one embodiment, the short-circuit ratio determination device for multiple new energy power stations further includes:

[0121] The verification set acquisition module is used to acquire a verification set, which includes: batch historical graph structure data and their respective corresponding real short-circuit ratios of multiple new energy power plants;

[0122] The predicted short-circuit ratio module is used to obtain the predicted short-circuit ratio of the new energy multi-site corresponding to each historical graph structure data and the short-circuit ratio evaluation model.

[0123] The mean square error value module is used to process the mean square error based on the predicted short-circuit ratio and the actual short-circuit ratio of the new energy multi-site corresponding to each historical graph structure data, and obtain the mean square error value.

[0124] The verification module is used to determine whether the mean square error value is less than the mean square error threshold. If so, the short-circuit ratio evaluation model is verified as passed.

[0125] In one embodiment, the acquisition module includes:

[0126] The historical data acquisition unit is used to acquire historical operating data, historical topology information and real short-circuit ratios of multiple new energy power plants under various operating scenarios of the target power system. The power generation, load and topology are the same for the same operating scenario, but the power generation, load and topology are different for each operating scenario.

[0127] The normalization unit is used to normalize historical operational data under the same operational scenario.

[0128] The historical graph structure data unit is obtained and used to obtain the historical graph structure data for each operation scenario based on the normalized historical running data and the historical topology information.

[0129] The combination unit is used to combine historical graph structure data under various operating scenarios and the actual short-circuit ratios of multiple new energy power stations to form the training set.

[0130] The embodiments of the short-circuit ratio determination device for new energy multi-stations provided in this specification can be used to execute the processing flow of the embodiments of the above-described method for determining the short-circuit ratio of new energy multi-stations. Its functions will not be repeated here, but can be referred to the detailed description of the embodiments of the above-described method for determining the short-circuit ratio of new energy multi-stations.

[0131] Figure 6 This is a schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention, such as... Figure 6 As shown, the electronic device includes: a memory 601, a processor 602, and a computer program stored in the memory 601 and executable on the processor 602. When the processor 602 executes the computer program, it implements the following method:

[0132] Obtain graph structure data of a target power system, which includes multiple devices, including multiple renewable energy power plants. Nodes in the graph structure data represent the characteristic data of the corresponding devices, and edges between nodes represent the characteristic data of the corresponding lines. The characteristic data of the devices includes: bus voltage, bus phase angle, active power injection of the bus, and reactive power injection of the bus. The characteristic data of the lines includes: line active power and line reactive power.

[0133] Based on the graph structure data and the preset short-circuit ratio evaluation model, the short-circuit ratio of the new energy multi-site is determined;

[0134] The preset short-circuit ratio evaluation model is obtained by pre-training a graph convolutional network model based on batch historical graph structure data and their corresponding real short-circuit ratios of multiple new energy power stations.

[0135] This embodiment discloses a computer program product, which includes a computer program that, when executed by a processor, implements the following method:

[0136] Obtain graph structure data of a target power system, which includes multiple devices, including multiple renewable energy power plants. Nodes in the graph structure data represent the characteristic data of the corresponding devices, and edges between nodes represent the characteristic data of the corresponding lines. The characteristic data of the devices includes: bus voltage, bus phase angle, active power injection of the bus, and reactive power injection of the bus. The characteristic data of the lines includes: line active power and line reactive power.

[0137] Based on the graph structure data and the preset short-circuit ratio evaluation model, the short-circuit ratio of the new energy multi-site is determined;

[0138] The preset short-circuit ratio evaluation model is obtained by pre-training a graph convolutional network model based on batch historical graph structure data and their corresponding real short-circuit ratios of multiple new energy power stations.

[0139] This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the following method:

[0140] Obtain graph structure data of a target power system, which includes multiple devices, including multiple renewable energy power plants. Nodes in the graph structure data represent the characteristic data of the corresponding devices, and edges between nodes represent the characteristic data of the corresponding lines. The characteristic data of the devices includes: bus voltage, bus phase angle, active power injection of the bus, and reactive power injection of the bus. The characteristic data of the lines includes: line active power and line reactive power.

[0141] Based on the graph structure data and the preset short-circuit ratio evaluation model, the short-circuit ratio of the new energy multi-site is determined;

[0142] The preset short-circuit ratio evaluation model is obtained by pre-training a graph convolutional network model based on batch historical graph structure data and their corresponding real short-circuit ratios of multiple new energy power stations.

[0143] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0144] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0145] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0146] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0147] In the description of this specification, the references to terms such as "an embodiment," "a specific embodiment," "some embodiments," "for example," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0148] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for determining the short-circuit ratio of multiple new energy power stations, characterized in that, include: Obtain graph structure data of a target power system, which includes multiple devices, including multiple renewable energy power plants. Nodes in the graph structure data represent the characteristic data of the corresponding devices, and edges between nodes represent the characteristic data of the corresponding lines. The characteristic data of the devices includes: bus voltage, bus phase angle, active power injection of the bus, and reactive power injection of the bus. The characteristic data of the lines includes: line active power and line reactive power. Based on the graph structure data and the preset short-circuit ratio evaluation model, the short-circuit ratio of the new energy multi-site is determined; The preset short-circuit ratio evaluation model is obtained by pre-training a graph convolutional network model based on batch historical graph structure data and their corresponding real short-circuit ratios of multiple new energy power stations.

2. The method for determining the short-circuit ratio of multiple new energy power stations according to claim 1, characterized in that, The acquisition of the graph structure data of the target power system includes: Obtain the current steady-state characteristics and topology information of the target power system; Based on the steady-state characteristics and topology information, the graph structure data of the target power system is obtained.

3. The method for determining the short-circuit ratio of multiple new energy power stations according to claim 1, characterized in that, Also includes: Collect a training set, which includes: batch historical graph structure data and their respective corresponding real short-circuit ratios of multiple new energy power stations; The graph convolutional network model is trained using the training set to obtain the short-circuit ratio evaluation model.

4. The method for determining the short-circuit ratio of multiple new energy power stations according to claim 3, characterized in that, Before acquiring the graph structure data of the target power system, the following is also included: Collect a validation set, which includes: batch historical graph structure data and their respective corresponding real short-circuit ratios of multiple new energy power plants; Based on each historical graph structure data and the short-circuit ratio evaluation model, the predicted short-circuit ratio of the new energy multi-site corresponding to the historical graph structure data is obtained; The mean square error is obtained by processing the predicted short-circuit ratio and the actual short-circuit ratio of the new energy multi-stations corresponding to each historical graph structure data. Determine whether the mean square error value is less than the mean square error threshold. If so, the short-circuit ratio evaluation model is deemed to have passed verification.

5. The method for determining the short-circuit ratio of multiple new energy power stations according to claim 3, characterized in that, The training set includes: The target power system acquires historical operating data, historical topology information, and the actual short-circuit ratio of multiple new energy power plants under various operating scenarios. The power generation, load, and topology are the same for the same operating scenario, but the power generation, load, and topology are different for each operating scenario. Normalize historical operational data under the same operational scenario; Based on the normalized historical operation data and the historical topology information for each operation scenario, the historical graph structure data for that operation scenario is obtained. The training set is composed of historical graph structure data under various operating scenarios and the actual short-circuit ratios of multiple new energy power stations.

6. A device for determining the short-circuit ratio of multiple new energy power stations, characterized in that, include: The acquisition module is used to acquire graph structure data of a target power system, which includes multiple devices, including multiple renewable energy power plants. Nodes in the graph structure data represent the characteristic data of the corresponding devices, and edges between nodes represent the characteristic data of the corresponding lines. The characteristic data of the devices includes: bus voltage, bus phase angle, active power injection of the bus, and reactive power injection of the bus. The characteristic data of the lines includes: line active power and line reactive power. The determination module is used to determine the short-circuit ratio of the new energy multi-site based on the graph structure data and the preset short-circuit ratio evaluation model; The preset short-circuit ratio evaluation model is obtained by pre-training a graph convolutional network model based on batch historical graph structure data and their corresponding real short-circuit ratios of multiple new energy power stations.

7. The short-circuit ratio determination device for multiple new energy power stations according to claim 6, characterized in that, The acquisition module includes: The acquisition unit is used to acquire the current steady-state characteristics and topology information of the target power system; The unit is used to obtain the graph structure data of the target power system based on the steady-state characteristics and topology information.

8. The short-circuit ratio determination device for multiple new energy power stations according to claim 6, characterized in that, Also includes: The data acquisition module is used to acquire training sets, which include: batch historical graph structure data and their corresponding real short-circuit ratios of multiple new energy power stations; The training module is used to train the graph convolutional network model using the training set to obtain the short-circuit ratio evaluation model.

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 program, it implements the method for determining the short-circuit ratio of new energy multi-site power plants as described in any one of claims 1 to 5.

10. A computer-readable storage medium storing computer instructions thereon, characterized in that, When the instruction is executed by the processor, it implements the method for determining the short-circuit ratio of new energy multi-site power plants as described in any one of claims 1 to 5.

Images