Power grid topology change element identification method, device and equipment and storage medium

Abstract

This application discloses a method, apparatus, device, and storage medium for identifying topology change components in a power grid. Based on the Newton-Raphson method, it calculates a power flow data sample set after a topology change in the power system to simulate power flow data under various topology change scenarios. The power flow data sample set is then preprocessed for visualization to obtain a target power flow data sample set. Based on the target power flow data sample set, a pre-defined convolutional neural network is trained until it reaches a pre-defined convergence condition, resulting in a topology component identification model. Finally, the topology component identification model is used to identify the topology change components in the power system after the topology change. This method establishes a correspondence between power flow data and topology changes through simple visualization preprocessing, and combined with an artificial intelligence model, it can identify topology change components in the power system in a very short time, thus meeting the real-time requirements of online tuning.

Description

Technical Field

[0001] This application relates to the field of power system relay protection technology, and in particular to a method, apparatus, equipment and storage medium for identifying power grid topology change elements. Background Technology

[0002] As power grids expand and their structures become more complex, their topology changes more frequently. Whether altering network operation through switching states or experiencing component outages due to external factors, real-time data collection is essential for understanding the grid topology and ensuring safe and stable operation.

[0003] Currently, topology identification methods only utilize remote signaling data collected by Supervisory Control and Data Acquisition (SCADA) systems. However, remote signaling data is prone to false alarms or omissions, resulting in low accuracy of methods that rely solely on switch status for identification. On the other hand, online setting is an important means to cope with changes in operating modes and improve the performance of relay protection settings in the future, but topology identification is still required before online setting. Topology identification methods are difficult to meet the high real-time requirements of online setting. Summary of the Invention

[0004] This application provides a method, apparatus, device, and storage medium for identifying power grid topology change elements, in order to solve the technical problem that current topology identification methods cannot meet the real-time requirements of online tuning.

[0005] To address the aforementioned technical problems, in a first aspect, this application provides a method for identifying power grid topology change components, comprising:

[0006] Based on the Newton-Raphson method, the power flow data sample set after the topology change of the power system is calculated. The power flow data sample set includes power flow data samples after the power system changes with different topology elements.

[0007] Visual preprocessing is performed on the current data sample set to obtain the target current data sample set;

[0008] Based on the target power flow data sample set, a preset convolutional neural network is trained until the preset convolutional neural network reaches the preset convergence condition, thus obtaining a topological element identification model.

[0009] By using a topology element identification model, the topology-changed elements of the power system after a topology change are identified.

[0010] In some implementations, based on the Newton-Raphson method, a power flow data sample set is calculated after a topology change in the power system, including:

[0011] Obtain first power flow data samples of the power system under actual operating conditions;

[0012] In the simulation environment, each line of the power system is disconnected in sequence, and the power flow distribution of the entire network after each line is disconnected is calculated using the Newton-Raphson method to obtain multiple second power flow data samples.

[0013] The first and second tidal current data samples are used as the tidal current data sample set.

[0014] In some implementations, under a simulation environment, each line of the power system is disconnected sequentially, and the Newton-Raphson method is used to calculate the power flow distribution across the entire network after each line is disconnected, resulting in multiple second power flow data samples, including:

[0015] In the simulation environment, each line of the power system is disconnected in sequence, and the active power of the generator and the active power of the load are randomly changed.

[0016] For each power system after a line disconnection, calculate the nodal admittance matrix of the power system after the line disconnection;

[0017] Calculate the Jacobian matrix based on the nodal admittance matrix;

[0018] Using the modified equation of Newton's method, the third power flow data sample of the power system is calculated based on the Jacobian matrix.

[0019] In some implementations, the generator's active power and the load's active power are randomly changed, including:

[0020] Based on the power flow relationship of the power system, multiple random numbers are generated within a preset random number generation range;

[0021] Based on multiple random numbers, the generator active power and load active power are randomly changed. The generator active power and load active power are described as follows:

[0022]

[0023] Among them, P g P represents the active power of the generator. gmin P is the minimum threshold for the active power of the generator. gmax P is the maximum active power threshold of the generator. d P represents the active power of the load. dmin P is the minimum active power of the load. dmax k represents the maximum active power of the load. g and k d k is a random number. gmax and k dmaxThis is the maximum value of the preset random number generation range set based on the power flow relationship.

[0024] In some implementations, the current flow data sample set is preprocessed for visualization to obtain the target current flow data sample set, including:

[0025] Based on the power system topology, the power flow data samples in the power flow data sample set are matrixed to obtain the power flow data matrix. The power flow data matrix is ​​the target power flow data sample set, where the power flow data samples include generator active power and load active power. Non-zero elements in the power flow data matrix indicate that adjacent components have electrical or physical connection relationships, and zero elements in the power flow data matrix indicate that adjacent components do not have electrical or physical connection relationships.

[0026] In some implementations, a pre-defined convolutional neural network is trained based on a target power flow data sample set until the pre-defined convolutional neural network reaches a pre-defined convergence condition, thereby obtaining a topological element identification model, including:

[0027] Based on the target current flow data sample set, a pre-set convolutional neural network is trained to obtain the sample identification results for each training session.

[0028] Based on the sample identification results, calculate the loss function value of the pre-defined convolutional neural network;

[0029] By using the control variable method, the hyperparameter variables of the preset convolutional neural network are updated according to the loss function value to obtain a new preset convolutional neural network;

[0030] Based on the target power flow data sample set, continue training the new preset convolutional neural network until the loss function value is less than the preset value or the number of iterations reaches the preset number, and obtain the topology element identification model.

[0031] In some implementations, a topology element identification model is used to identify the topology-changed power system after a topology change, thereby obtaining the topology-changed elements of the power system, including:

[0032] Based on the Newton-Raphson method, calculate the power flow data of the power system after a topology change.

[0033] Visualize and preprocess the trend data to obtain the target trend data;

[0034] Using a topology element identification model, the topology of the power system is identified based on the target power flow data, and the topology change elements of the power system are output.

[0035] Secondly, this application also provides a device for identifying power grid topology change elements, comprising:

[0036] The calculation module is used to calculate the power flow data sample set after a topology change in the power system, based on the Newton-Raphson method. The power flow data sample set includes power flow data samples after changes to different topology components in the power system.

[0037] The visualization module is used to perform visualization preprocessing on the current data sample set to obtain the target current data sample set.

[0038] The training module is used to train a preset convolutional neural network based on the target power flow data sample set until the preset convolutional neural network reaches the preset convergence condition, thereby obtaining a topological element identification model.

[0039] The identification module is used to identify the topology-changed components of the power system after a topology change using a topology component identification model.

[0040] Thirdly, this application also provides a computer device, including a processor and a memory, the memory for storing a computer program, which, when executed by the processor, implements the method for identifying power grid topology change elements as described in the first aspect.

[0041] Fourthly, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method for identifying power grid topology change elements as described in the first aspect.

[0042] Compared with the prior art, this application has at least the following beneficial effects:

[0043] This paper calculates a power flow data sample set after a topology change in a power system based on the Newton-Raphson method. The power flow data sample set includes power flow data samples after changes in different topology components, simulating power flow data under various topology change scenarios. The power flow data sample set is then preprocessed with visualization to obtain a target power flow data sample set, establishing a correspondence between the topological relationships after the topology change and the power flow data samples. This allows the model to learn power flow data under different topological relationships during the model training phase, and subsequently, to identify topological relationships through power flow data during the model application phase. Based on the target power flow data sample set, a pre-defined convolutional neural network is trained until it reaches a pre-defined convergence condition, resulting in a topology component identification model. Finally, the topology component identification model is used to identify the topology-changed components of the power system after the topology change, enabling accurate identification of the topology change location using the power flow data after the topology change. This application establishes a correspondence between power flow data and topology changes through simple visualization preprocessing, and combined with an artificial intelligence model, can identify the topology-changed components of the power system in a very short time, thus meeting the real-time requirements of online tuning. Attached Figure Description

[0044] Figure 1 This is a schematic flowchart illustrating the method for identifying power grid topology change components according to an embodiment of this application;

[0045] Figure 2 This is a schematic diagram of the power grid topology shown in the embodiments of this application;

[0046] Figure 3 This is a schematic diagram of a visually preprocessed target power flow data sample shown in an embodiment of this application;

[0047] Figure 4 This is a schematic diagram of the structure of the identification device for power grid topology change elements shown in the embodiments of this application;

[0048] Figure 5 This is a schematic diagram of the structure of a computer device shown in an embodiment of this application. Detailed Implementation

[0049] The technical solutions of 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 of 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.

[0050] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating a method for identifying power grid topology change components according to an embodiment of this application. The method for identifying power grid topology change components according to this application can be applied to computer devices, including but not limited to smartphones, laptops, tablets, desktop computers, physical servers, and cloud servers. Figure 1 As shown, the method for identifying power grid topology change components in this embodiment includes steps S101 to S104, which are detailed below:

[0051] Step S101: Based on the Newton-Raphson method, calculate the power flow data sample set after the topology change of the power system. The power flow data sample set includes power flow data samples after the power system changes with different topology elements.

[0052] In this step, the power flow data samples include power flow data samples under normal operating conditions and under various topology change conditions.

[0053] In some embodiments, step S101 includes:

[0054] Obtain first power flow data samples of the power system under actual operating conditions;

[0055] In the simulation environment, each line of the power system is disconnected in sequence, and the power flow distribution of the entire network after each line is disconnected is calculated using the Newton-Raphson method to obtain multiple second power flow data samples.

[0056] The first tidal data sample and the second tidal data sample are used as the tidal data sample set.

[0057] In this embodiment, in order to more fully simulate the actual operating state of the system, in addition to the power flow samples obtained in actual operation, it is also necessary to simulate power flow samples under various topology change conditions in an offline simulation environment to ensure the performance of the training model.

[0058] In some embodiments, in a simulation environment, each line of the power system is disconnected sequentially, and the power flow distribution of the entire network after each line is disconnected is calculated sequentially using the Newton-Raphson method, resulting in multiple second power flow data samples, including:

[0059] In a simulation environment, each line of the power system is disconnected sequentially, and the active power of the generator and the active power of the load are randomly changed.

[0060] For each disconnection of the power system, calculate the node admittance matrix of the power system after the disconnection;

[0061] Calculate the Jacobian matrix based on the node admittance matrix;

[0062] Using the modified equation of Newton's method, the third power flow data sample of the power system is calculated based on the Jacobian matrix.

[0063] In this embodiment, each line is disconnected sequentially, and the power flow distribution of the entire network after the line disconnection is calculated using the Newton-Raphson method. The calculation process is as follows:

[0064] a. Initialization: Calculate the nodal admittance matrix, given an initial value x(0);

[0065] b. Calculate the Jacobian matrix J;

[0066] c. Calculate the correction amount Δx (k) =-(J) -1 f(x (k) );

[0067] d. Calculate variable x (k+1) =x ( k ) +Δx (k) ;

[0068] b. Calculate the function value f(x) (k+1) ), determine ||f(x) (k+1))-f(x k Check if ||≤ε is true. If it is true, end the iteration; otherwise, go to b.

[0069] Where for variable x (k+1) Including the real part e of the node voltage in the (k+1)th iteration i and the imaginary part f i The specific explanation is as follows:

[0070] Let the node voltage For each node, there are two nonlinear equations. Without considering the equations for the equilibrium nodes, there are a total of 2(n-1) equations, as shown below:

[0071]

[0072] The corrected equation for Newton's method is:

[0073]

[0074] in, P represents the node active power, Q represents the node reactive power, G represents the real part of the node admittance matrix element, B represents the imaginary part of the node admittance matrix element, e represents the real part of the node voltage, and f represents the imaginary part of the node voltage. Optionally, the random change of generator active power and load active power includes:

[0075] Based on the power flow relationship of the power system, multiple random numbers are generated within a preset random number generation range;

[0076] Based on a plurality of the aforementioned random numbers, the active power of the generator and the active power of the load are randomly changed, and the active power of the generator and the active power of the load are described as follows:

[0077]

[0078] Among them, P g P represents the active power of the generator. gmin P is the minimum threshold for the active power of the generator. gmax P is the maximum active power threshold of the generator. d P represents the active power of the load. dmin P is the minimum active power of the load. dmax k represents the maximum active power of the load. g and k d k is a random number. gmax and k dmax This is the maximum value of the preset random number generation range set based on the power flow relationship.

[0079] In this embodiment, P gmin and P gmaxP represents the actual minimum and maximum power thresholds of generators in the power system. dmin and P dmax k represents the actual minimum and maximum power thresholds of the load in the power system. gmax and k dmax It can be set based on the topology relationship when the power system experiences surplus output or overload due to topology changes, in order to describe the situation where the power system operates with surplus output or overload due to topology changes.

[0080] Step S102: Perform visualization preprocessing on the current flow data sample set to obtain the target current flow data sample set.

[0081] In this step, visual preprocessing is used to characterize target power flow data samples with the topological relationships of the power system after topology changes, so that the model can learn the correspondence between different topological relationships and target power flow data samples, thereby realizing topology change identification.

[0082] In some embodiments, step S102 includes:

[0083] Based on the topology of the power system, the power flow data samples in the power flow data sample set are matrixed to obtain a power flow data matrix. The power flow data matrix is ​​the target power flow data sample set. The power flow data samples include generator active power and load active power. Non-zero elements in the power flow data matrix indicate that adjacent components have electrical or physical connection relationships, and zero elements in the power flow data matrix indicate that adjacent components do not have electrical or physical connection relationships.

[0084] In this embodiment, the generator active power and load active power are represented by squares of four elements of a matrix; the branch transmission active power is represented by two adjacent elements of the matrix; other elements of the matrix are set to 0. The placement of elements in the matrix is ​​similar to the actual topology diagram. Adjacent non-zero elements in the matrix indicate that there is a physical or electrical connection between the two elements, and zero elements in the matrix indicate that there is no electrical or physical connection between adjacent elements. For example, using... Figure 2 Taking the IEEE 30-node system shown as an example, the visualization results obtained from one sample are as follows: Figure 3 As shown ( Figure 3 The square formed by the four elements represents the generator's active power and the load's active power, while the rectangle formed by the two elements represents the active power transmitted through the branch. The connection relationship of the components corresponds to the actual topology. Figure 1 To).

[0085] Step S103: Based on the target power flow data sample set, train a preset convolutional neural network until the preset convolutional neural network reaches the preset convergence condition to obtain a topology element identification model.

[0086] In this step, before training the CNN, the data sample set is divided into three parts: a training sample set, a validation sample set, and a test sample set. The training sample set is used to train the CNN topology change element recognition model; the validation sample set is used to adjust the CNN network structure parameters and training hyperparameters; and the test sample set is used to test the accuracy of the trained CNN model. Network structure parameters include the number of neural network layers, convolutional kernel size, and pooling kernel size. Training hyperparameters include the learning rate, batch size, number of iterations, and dropout rate.

[0087] After preparing the data sample set, the model is built and its parameters are adjusted using the controlled variable method. First, the number of neural network layers is determined based on the scale and complexity of the sample data features. Generally, more layers lead to a deeper level of abstraction of the input data features and a higher accuracy of understanding. However, the computational complexity also increases with the number of layers, and excessive layers can easily lead to overfitting. Optionally, in this embodiment, the sample data consists only of power grid flow data, and its feature complexity is not high. Therefore, a network structure with moderate complexity, such as "double convolutional layer double pooling layer" or "triple convolutional layer triple pooling layer," is adopted, with the specific choice determined based on specific experiments. Next, the size and number of convolutional kernels are determined. Generally, using multiple small convolutional kernels stacked together is more effective than using a single large convolutional kernel. Therefore, the classic "3×3" convolutional kernel size is initially adopted, and subsequent fine-tuning is performed based on the model's recognition performance.

[0088] In some embodiments, step S103 includes:

[0089] Based on the target power flow data sample set, the preset convolutional neural network is trained to obtain the sample identification results for each training session.

[0090] Based on the sample identification results, the loss function value of the preset convolutional neural network is calculated;

[0091] Using the control variable method, the hyperparameter variables of the preset convolutional neural network are updated according to the loss function value to obtain a new preset convolutional neural network;

[0092] Based on the target power flow data sample set, the new preset convolutional neural network is trained until the loss function value is less than the preset value or the number of iterations reaches the preset number, thus obtaining the topology element identification model.

[0093] In this embodiment, after the CNN network structure parameters are determined, the training hyperparameters are further determined. The learning rate refers to the magnitude of updating the network weights, which is generally between [0.01, 1]. The batch size refers to the number of samples in each epoch of the CNN, which should generally not be too large and is between [8, 128]. The number of iterations refers to the number of times the CNN is trained. The dropout rate refers to the proportion of randomly unactivated neurons in each training session, which is 0.5 by default. By using the control variable method, only one hyperparameter variable is changed each time. The values ​​of the hyperparameters are determined based on the test results of the validation sample set. At this point, the CNN model training is complete.

[0094] Step S104: Using the topology element identification model, the power system after the topology change is identified to obtain the topology change element of the power system.

[0095] In this embodiment, optionally, power flow data of the power system after topology change is calculated based on the Newton-Raphson method; the power flow data is preprocessed for visualization to obtain target power flow data; the topology element identification model is used to identify the topology of the power system based on the target power flow data, and the topology change elements of the power system are output.

[0096] As an example, and not a limitation, the IEEE 30-node power system is used as an example. The topology diagram of this power system is as follows: Figure 2 As shown. The offline simulation uses the N-1 principle commonly used in power grid analysis to generate samples, that is, it considers the operating modes after 41 lines are out of service (in actual applications, the number of lines that are out of service simultaneously can be increased or decreased). Each time, only the active power output of one of the 6 generators is changed, and only the load size of one of the 10 loads is considered to be changed. The total number of samples generated is... There are a total of 2460 sample data points. To simulate noise interference, random noise is added to the disconnected branches during sample generation. The active power flow calculation for the disconnected branches after adding random noise is as follows:

[0097] P = [-P] rand ,P rand ]; where P rand These are predefined parameters.

[0098] The parameter settings for the convolutional neural network are shown in the table below:

[0099]

[0100] The trained convolutional neural network model was tested using a pre-prepared test sample set. The accuracy rate for identifying topology change components was 99.17%, with an average single computation time of 0.002 seconds, achieving the goal of identifying power grid topology change components in a short time. Taking one set of samples as an example, with the disconnected line L1-2, the test results are shown in the table below, with the top ten results presented. The table shows that the proposed method can provide topology change component identification results in a very short time, and the results are sorted in order of probability for on-site dispatchers to refer to.

[0101] L1－2 0.9882 L27－29 0.0054 L2－6 0.0370 L21－22 0.0044 L2－4 0.0278 L23－24 0.0044 L6－6 0.0134 L22－24 0.0031 L5－7 0.0108 L24－25 0.0018

[0102] To implement the power grid topology change element identification method corresponding to the above method embodiments, and to achieve the corresponding functions and technical effects. See also Figure 4 , Figure 4 This diagram illustrates a structural block diagram of a power grid topology change element identification device according to an embodiment of this application. For ease of explanation, only the parts relevant to this embodiment are shown. The power grid topology change element identification device provided in this embodiment includes:

[0103] The calculation module 401 is used to calculate the power flow data sample set after a topology change in the power system based on the Newton-Raphson method. The power flow data sample set includes power flow data samples after the power system changes in different topology elements.

[0104] The visualization module 402 is used to perform visualization preprocessing on the current data sample set to obtain the target current data sample set.

[0105] The training module 403 is used to train a preset convolutional neural network based on the target power flow data sample set until the preset convolutional neural network reaches a preset convergence condition, thereby obtaining a topology element identification model.

[0106] The identification module 404 is used to identify the power system after a topology change using the topology element identification model, and to obtain the topology change element of the power system.

[0107] In some embodiments, the computing module 401 includes:

[0108] The acquisition unit is used to acquire first power flow data samples of the power system under actual operating conditions.

[0109] The calculation unit is used to sequentially disconnect each line of the power system in a simulation environment and use the Newton-Raphson method to sequentially calculate the power flow distribution of the entire network after each line is disconnected, thereby obtaining multiple second power flow data samples.

[0110] The first tidal data sample and the second tidal data sample are used as the tidal data sample set.

[0111] In some embodiments, the computing unit is specifically used for:

[0112] In a simulation environment, each line of the power system is disconnected sequentially, and the active power of the generator and the active power of the load are randomly changed.

[0113] For each disconnection of the power system, calculate the node admittance matrix of the power system after the disconnection;

[0114] Calculate the Jacobian matrix based on the node admittance matrix;

[0115] Using the modified equation of Newton's method, the third power flow data sample of the power system is calculated based on the Jacobian matrix.

[0116] In some embodiments, the random alteration of the generator active power and the load active power includes:

[0117] Based on the power flow relationship of the power system, multiple random numbers are generated within a preset random number generation range;

[0118] Based on a plurality of the aforementioned random numbers, the active power of the generator and the active power of the load are randomly changed, and the active power of the generator and the active power of the load are described as follows:

[0119]

[0120] Among them, P g P represents the active power of the generator. gmin P is the minimum threshold for the active power of the generator. gmax P is the maximum active power threshold of the generator. d P represents the active power of the load. dmin P is the minimum active power of the load. dmax k represents the maximum active power of the load. g and k d k is a random number. gmax and k dmax This is the maximum value of the preset random number generation range set based on the power flow relationship.

[0121] In some embodiments, the visualization module 402 is specifically used for:

[0122] Based on the topology of the power system, the power flow data samples in the power flow data sample set are matrixed to obtain a power flow data matrix. The power flow data matrix is ​​the target power flow data sample set. The power flow data samples include generator active power and load active power. Non-zero elements in the power flow data matrix indicate that adjacent components have electrical or physical connection relationships, and zero elements in the power flow data matrix indicate that adjacent components do not have electrical or physical connection relationships.

[0123] In some embodiments, the training module 403 is specifically used for:

[0124] Based on the target power flow data sample set, the preset convolutional neural network is trained to obtain the sample identification results for each training session.

[0125] Based on the sample identification results, the loss function value of the preset convolutional neural network is calculated;

[0126] Using the control variable method, the hyperparameter variables of the preset convolutional neural network are updated according to the loss function value to obtain a new preset convolutional neural network;

[0127] Based on the target power flow data sample set, the new preset convolutional neural network is trained until the loss function value is less than the preset value or the number of iterations reaches the preset number, thus obtaining the topology element identification model.

[0128] In some embodiments, the identification module 404 is specifically used for:

[0129] Based on the Newton-Raphson method, the power flow data of the power system after the topology change is calculated.

[0130] The current flow data is preprocessed for visualization to obtain the target current flow data;

[0131] Using the topology element identification model, the topology of the power system is identified based on the target power flow data, and the topology change elements of the power system are output.

[0132] The aforementioned device for identifying power grid topology change components can implement the power grid topology change component identification method of the above method embodiments. The options in the above method embodiments are also applicable to this embodiment, and will not be detailed here. The remaining content of this application embodiment can be referred to the content of the above method embodiments, and will not be repeated in this embodiment.

[0133] Figure 5 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Figure 5 As shown, the computer device 5 of this embodiment includes: at least one processor 50 ( Figure 5 (Only one is shown) a processor, a memory 51, and a computer program 52 stored in the memory 51 and executable on the at least one processor 50, wherein the processor 50 executes the computer program 52 to implement the steps in any of the above method embodiments.

[0134] The computer device 5 may be a smartphone, tablet, desktop computer, or cloud server, among other computing devices. This computer device may include, but is not limited to, a processor 50 and a memory 51. Those skilled in the art will understand that... Figure 5 The computer device 5 is merely an example and does not constitute a limitation on the computer device 5. It may include more or fewer components than shown in the figure, or combine certain components, or different components, such as input / output devices, network access devices, etc.

[0135] The processor 50 may be a Central Processing Unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0136] In some embodiments, the memory 51 may be an internal storage unit of the computer device 5, such as a hard disk or memory of the computer device 5. In other embodiments, the memory 51 may be an external storage device of the computer device 5, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 5. Furthermore, the memory 51 may include both internal and external storage units of the computer device 5. The memory 51 is used to store the operating system, applications, bootloader, data, and other programs, such as the program code of the computer program. The memory 51 can also be used to temporarily store data that has been output or will be output.

[0137] In addition, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps in any of the above method embodiments.

[0138] This application provides a computer program product that, when run on a computer device, enables the computer device to execute the steps described in the various method embodiments above.

[0139] In the several embodiments provided in this application, it will be understood that each block in the flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the figures. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved.

[0140] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0141] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of this application. It should be understood that the above descriptions are merely specific embodiments of this application and are not intended to limit the scope of protection of this application. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application for those skilled in the art.

Claims

1. A method for identifying power grid topology change components, characterized in that, include: Based on the Newton-Raphson method, a power flow data sample set is calculated after a topology change in the power system. This power flow data sample set includes power flow data samples after changes to different topology components. The calculation of the power flow data sample set after a topology change based on the Newton-Raphson method includes: Obtain first power flow data samples of the power system under actual operating conditions; In the simulation environment, each line of the power system is disconnected in sequence, and the power flow distribution of the entire network after each line is disconnected is calculated using the Newton-Raphson method to obtain multiple second power flow data samples. The first tidal data sample and the second tidal data sample are used as the tidal data sample set; The tidal current data sample set is subjected to visualization preprocessing to obtain a target tidal current data sample set; wherein, the visualization preprocessing of the tidal current data sample set to obtain the target tidal current data sample set includes: Based on the topology of the power system, the power flow data samples in the power flow data sample set are matrixed to obtain a power flow data matrix. The power flow data matrix is ​​the target power flow data sample set. The power flow data samples include generator active power and load active power. Non-zero elements in the power flow data matrix indicate that adjacent components have electrical or physical connection relationships, and zero elements in the power flow data matrix indicate that adjacent components do not have electrical or physical connection relationships. Based on the target power flow data sample set, a preset convolutional neural network is trained until the preset convolutional neural network reaches the preset convergence condition, thereby obtaining a topological element identification model. Using the aforementioned topology element identification model, the power system after a topology change is identified, and the topology change elements of the power system are obtained.

2. The method for identifying power grid topology change components as described in claim 1, characterized in that, In the simulation environment, each line of the power system is disconnected sequentially, and the power flow distribution of the entire network after each line is disconnected is calculated using the Newton-Raphson method, resulting in multiple second power flow data samples, including: In a simulation environment, each line of the power system is disconnected sequentially, and the active power of the generator and the active power of the load are randomly changed. For each disconnection of the power system, calculate the node admittance matrix of the power system after the disconnection; Calculate the Jacobian matrix based on the node admittance matrix; Using the modified equation of Newton's method, the third power flow data sample of the power system is calculated based on the Jacobian matrix.

3. The method for identifying power grid topology change components as described in claim 2, characterized in that, The random alteration of generator active power and load active power includes: Based on the power flow relationship of the power system, multiple random numbers are generated within a preset random number generation range; Based on a plurality of the aforementioned random numbers, the active power of the generator and the active power of the load are randomly changed, and the active power of the generator and the active power of the load are described as follows: ； in, The active power of the generator. This represents the minimum threshold for the active power of the generator. This represents the maximum active power threshold for the generator. For the active power of the load, This represents the minimum active power of the load. This represents the maximum active power of the load. and It is a random number. and This is the maximum value of the preset random number generation range set based on the power flow relationship.

4. The method for identifying power grid topology change components as described in claim 3, characterized in that, The step of training a preset convolutional neural network based on the target power flow data sample set until the preset convolutional neural network reaches a preset convergence condition to obtain a topology element identification model includes: Based on the target power flow data sample set, the preset convolutional neural network is trained to obtain the sample identification results for each training session. Based on the sample identification results, the loss function value of the preset convolutional neural network is calculated; Using the control variable method, the hyperparameter variables of the preset convolutional neural network are updated according to the loss function value to obtain a new preset convolutional neural network; Based on the target power flow data sample set, the new preset convolutional neural network is trained until the loss function value is less than the preset value or the number of iterations reaches the preset number, thus obtaining the topology element identification model.

5. The method for identifying power grid topology change components as described in claim 4, characterized in that, The process of using the topology element identification model to identify the power system after a topology change, and obtaining the topology change elements of the power system, includes: Based on the Newton-Raphson method, the power flow data of the power system after the topology change is calculated. The current flow data is preprocessed for visualization to obtain the target current flow data; Using the topology element identification model, the topology of the power system is identified based on the target power flow data, and the topology change elements of the power system are output.

6. A device for identifying power grid topology change components, characterized in that, include: The calculation module is used to calculate a power flow data sample set after a topology change in a power system based on the Newton-Raphson method. The power flow data sample set includes power flow data samples after changes in different topology components of the power system. The calculation of the power flow data sample set after a topology change in a power system based on the Newton-Raphson method includes: Obtain first power flow data samples of the power system under actual operating conditions; In the simulation environment, each line of the power system is disconnected in sequence, and the power flow distribution of the entire network after each line is disconnected is calculated using the Newton-Raphson method to obtain multiple second power flow data samples. The first tidal data sample and the second tidal data sample are used as the tidal data sample set; A visualization module is used to perform visualization preprocessing on the current data sample set to obtain a target current data sample set; wherein, the visualization preprocessing on the current data sample set to obtain the target current data sample set includes: Based on the topology of the power system, the power flow data samples in the power flow data sample set are matrixed to obtain a power flow data matrix. The power flow data matrix is ​​the target power flow data sample set. The power flow data samples include generator active power and load active power. Non-zero elements in the power flow data matrix indicate that adjacent components have electrical or physical connection relationships, and zero elements in the power flow data matrix indicate that adjacent components do not have electrical or physical connection relationships. The training module is used to train a preset convolutional neural network based on the target power flow data sample set until the preset convolutional neural network reaches a preset convergence condition, thereby obtaining a topological element identification model. The identification module is used to identify the power system after a topology change using the topology element identification model, and to obtain the topology change element of the power system.

7. A computer device, characterized in that, It includes a processor and a memory, the memory being used to store a computer program, which, when executed by the processor, implements the method for identifying power grid topology change elements as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the method for identifying power grid topology change elements as described in any one of claims 1 to 5.

Images