Methods, systems, equipment, and media for recognizing power grid chain failure risk scenarios

Abstract

A power grid chain failure risk scene recognition method, system, device, and medium includes: acquiring power grid historical operation data (S101); performing chain failure simulation on the power grid historical operation data to obtain active power flow entropy (S102); visually displaying the power grid historical operation data and the active power flow entropy to obtain a multi-row heat map (S103); dividing the multi-row heat map into a training set and a test set (S104); training a neural network using the training set to obtain a power grid chain failure risk scene recognition model (S105); and testing the power grid chain failure risk scene recognition model using the test set to obtain a risk scene recognition result (S106).

Description

[Technical Field]

[0001] This application claims priority to the Chinese patent application filed with the China Patent Office on 28 July 2023, application number 202310939393.1, and all contents of the said application are incorporated into this application by reference.

[0002] This application relates to the field of power risk scenarios, specifically to methods, systems, equipment, and media for recognizing power grid chain failure risk scenarios. [Background technology]

[0003] With the application of flexible AC / DC transmission technology in current power grids and the access of new loads such as electric vehicles, the structure of power grids is becoming increasingly complex, and some low-probability events in the power grid can trigger chain reactions, potentially leading to catastrophic consequences. To enhance the reliability and stability of the power grid, avoid chain reactions and large-scale blackouts, reduce the cost and time of recovery, and ensure people's lives and safety, it is possible to take appropriate measures by predicting and identifying potential risks in a timely manner, thereby reducing the occurrence of chain reactions and ensuring the normal operation of the power grid. Therefore, accurately understanding the current operating status of the power grid is of crucial importance in ensuring the normal operation of the power grid.

[0004] With the rise of a new generation of artificial intelligence technologies, exemplified by deep learning, new concepts have emerged for the online identification of power grid chain failure risk scenarios. By using deep learning to uncover the intrinsic link between historical operating conditions and risk scenarios and constructing recognition models, it is possible to increase the speed of online recognition of power grid chain failure risk scenarios, which is extremely important for practical applications. [Overview of the Initiative]

[0005] This application provides a method, system, equipment, and medium for recognizing power grid chain failure risk scenarios that can accurately realize risk recognition in the current operating scenario of the power grid.

[0006] This application is, To acquire power grid operation history data, The aforementioned power grid historical operation data is used to perform a chain failure simulation and obtain the active power flow entropy. The aforementioned power grid history operation data and the aforementioned active power flow entropy are visualized and displayed to obtain a multi-row heatmap. The aforementioned multi-row heatmap is divided into a training set and a test set, The goal is to train a neural network using a training set to obtain a power grid chain failure risk scene recognition model, and The present invention provides a power grid chain failure risk scene recognition method, which includes testing the power grid chain failure risk scene recognition model using a test set and obtaining risk scene recognition results.

[0007] Preferably, a chain failure simulation is performed on the power grid history operation data to obtain the active power flow entropy. Based on the aforementioned power grid historical operation data, a power grid simulation model is constructed, multiple types of system operation scenarios are set, and a chain failure simulation is performed on the power grid lines to obtain the chain failure chain and chain failure load ratio. This includes determining the active power flow entropy based on the chain of faults and the chain of fault load rates.

[0008] Preferably, the power grid history operation data includes power grid chain failure risk scenes, and the power grid history operation data and the active power flow entropy are visualized and displayed to obtain a multi-line heat map. Using the load ratio in the current operating state of the aforementioned power grid chain failure risk scene, the first row of the multi-row heatmap corresponding to the aforementioned power grid chain failure risk scene is drawn. Using the line capacity of the aforementioned power grid chain failure risk scene, the second row of the multi-row heatmap corresponding to the aforementioned power grid chain failure risk scene is drawn, Drawing the third row of a plurality of row heatmaps corresponding to the power grid cascading fault risk scenario by using the node feature vector of the power grid cascading fault risk scenario; Drawing the fourth row of a plurality of row heatmaps corresponding to the power grid cascading fault risk scenario by using the active power flow entropy; Drawing the fifth row of a plurality of row heatmaps corresponding to the power grid cascading fault risk scenario by using the ratio of the occurrence of historical cascading faults in the power grid cascading fault risk scenario to the total fault occurrence, including:

[0009] Preferably, training a neural network by using a training set to obtain a power grid cascading fault risk scenario recognition model specifically includes: Taking a plurality of row heatmaps in a plurality of operation states of the power grid in the training set as the input of a Convolutional Neural Networks (CNN) model, taking the risk scenarios corresponding to various operation states in the training set as the output of the CNN model, and using a cross-entropy loss function as the loss function to train the CNN model to obtain a power grid cascading fault risk scenario recognition model.

[0010] Preferably, after testing the power grid cascading fault risk scenario recognition model by using a test set to obtain a risk scenario recognition result, further constructing a set of significant risk scenarios by using a clustering algorithm based on the risk scenario recognition result is further included.

[0011] Preferably, the active power flow entropy includes the active power flow entropy of overloaded lines and the active power flow entropy of non-overloaded lines, The expression formula of the active power flow entropy of the non-overloaded line is

Number

[0012] This application is configured with an acquisition module for acquiring power grid historical operation data, a cascading failure simulation module configured to perform cascading failure simulation on the power grid historical operation data to obtain the active power flow entropy, an imaging display module configured to image and display the power grid historical operation data and the active power flow entropy to obtain a multi-row heat map, a classification module configured to classify the multi-row heat map into a training set and a test set, a training module configured to train a neural network using the training set to obtain a power grid cascading failure risk scenario recognition model, The present invention further provides a power grid chain failure risk scene recognition system comprising a test module configured to test the power grid chain failure risk scene recognition model using a test set and to obtain risk scene recognition results.

[0013] Preferably, the chain failure simulation module is A construction and simulation unit is configured to build a power grid simulation model based on the aforementioned power grid historical operation data, set up multiple types of system operation scenarios, perform chain failure simulations on power grid lines, and obtain chain failure chains and chain failure load rates. The system includes an active power flow entropy determination unit configured to determine the active power flow entropy based on the chain of faults and the chain of faults load rate.

[0014] This application is, At least one processor, A memory device containing at least one program, When the at least one program is executed by the at least one processor, the at least one processor further provides an electronic device that realizes the method described above.

[0015] The present invention further provides a computer storage medium that stores a computer program that, when executed by a processor, implements the method described above. [Brief explanation of the drawing]

[0016] The following are the drawings that need to be used in the examples.

[0017] [Figure 1] This is a schematic diagram of a power grid chain failure risk scene recognition method according to an embodiment of the present invention. [Figure 2] This is a multi-line heatmap corresponding to various power grid chain failure risk scenarios according to the embodiment of the present invention. [Figure 3]This is a structural diagram of the power grid chain failure risk scene recognition model according to an embodiment of the present invention. [Figure 4] This is a flowchart of the power grid chain failure risk scene recognition method according to an embodiment of the present invention. [Figure 5] This is a structural block diagram of a power grid chain failure risk scene recognition system according to an embodiment of the present invention. [Modes for carrying out the invention]

[0018] The following describes the technical concepts in the embodiments of this application in conjunction with the drawings of the embodiments. It is clear that the embodiments described are not necessarily all embodiments, but rather only a portion of the embodiments of this application. All other embodiments that can be obtained based on the embodiments of this application without creative work by a person skilled in the art fall within the scope of protection of this application.

[0019] This application provides a method, system, equipment, and medium for recognizing power grid chain failure risk scenarios that can accurately realize risk recognition in the current operating scenario of the power grid.

[0020] As shown in Figures 1 and 4, the power grid chain failure risk scene recognition method according to the present invention includes the following:

[0021] In step 101, power grid operation history data is acquired.

[0022] In step 102, a chain failure simulation is performed on the power grid history operation data to obtain the active power flow entropy.

[0023] Step 102 includes constructing a power grid simulation model based on the power grid history operation data, setting different system operation scenarios, performing a chain failure simulation on the power grid lines to obtain the chain failure chain and chain failure load factor, and determining the active power flow entropy based on the chain failure chain and the chain failure load factor.

[0024] We collect historical power grid operation data, perform chain failure simulations, define active power flow entropy as an index for recognizing power grid chain failure scenes, construct a power grid simulation model, set different system operation scenes, and perform N-1 chain failure simulations on the lines, considering chain failures led by line overload, to obtain chain failure chains and chain failure load rates.

[0025] Entropy represents the degree of uniformity in the distribution of things in space and is used to express the uncertainty of a system. Tidal entropy is a measure of the equilibrium of tidal currents in multiple branch circuits of a system, and is quantified by the distribution of load factors of the branch circuits. The arithmetic progression of load factors is Z = [Z1, Z2, ..., Z n Set ] and the system's tidal entropy F is,

number

[0026] Of these, the number of branch circuits is equivalent to the number of lines in the power grid.

[0027] To more concretely reflect the impact of system indicators on the power grid, we define active power flow entropy in relation to the characteristics of active power flow. As can be seen from major power outages, even small disturbances can cause high-load-rate lines to overload and shut down, potentially leading to chain failures, and the load factor of a line is positively correlated with the probability of its shutdown. Therefore, when the load factor of all lines is less than 1, the active power flow entropy of unoverloaded lines in the system is:

number

number

[0028] In step 103, the power grid history operation data and the active power flow entropy are visualized and displayed to obtain a multi-row heatmap. Of these, the power grid history operation data includes power grid chain failure risk scenes.

[0029] Step 103 is, Using the load ratio at the current operating status of the power grid chain failure risk scene in the aforementioned power grid history operation data, the first row of the multi-row heatmap corresponding to the power grid chain failure risk scene is drawn. Using the line capacity of the power grid chain failure risk scene in the aforementioned power grid history operation data, the second row of the multi-row heatmap corresponding to the power grid chain failure risk scene is drawn. Using the node feature vectors of the power grid chain failure risk scenes in the aforementioned power grid history operation data, the third row of the multi-row heatmap corresponding to the power grid chain failure risk scenes is drawn. Using the aforementioned active power flow entropy, the fourth row of the multi-row heatmap corresponding to the power grid chain failure risk scene is drawn, This includes drawing the fifth row of a multi-row heatmap corresponding to the power grid chain failure risk scene, using the proportion of historical chain failure occurrences in the power grid chain failure risk scene in the power grid history operation data to the total number of failures.

[0030] Visualizing various power grid chain failure risk scenarios and constructing a CNN-based power grid chain failure risk scenario recognition model involves using multi-row heatmaps to visualize various power grid chain failure risk scenarios, generating multi-row heatmaps corresponding to each power grid chain failure risk scenario, inputting the multi-row heatmaps corresponding to each power grid chain failure risk scenario into a pre-trained machine learning model, and obtaining classification results output by the machine learning model for each power grid chain failure risk scenario. As shown in Figure 3, the basic structure of a CNN consists of an input layer, a convolutional layer, a pooling layer (also called a sampling layer), a fully connected layer, and an output layer.

[0031] The CNN-based power grid chain failure risk scene recognition model consists of an input layer, a convolutional layer, a pooling layer, a fully connected layer, and an output layer. The input layer is fed with multi-row heatmap data.

[0032] Referring to Figure 2, for each of the aforementioned power grid chain failure risk scenarios, the first row of the multi-row heatmap corresponding to the power grid chain failure risk scenario is drawn using the load ratio in the current operating state of the power grid chain failure risk scenario, and the load ratio is set to Z k This is shown.

[0033] Using the line capacity of the power grid chain failure risk scene, the second row of the multi-line heatmap corresponding to the power grid chain failure risk scene is drawn, and the line capacity is C m This is shown.

[0034] Using the node feature vector of the power grid chain failure risk scene, the third row of the multi-row heatmap corresponding to the power grid chain failure risk scene is plotted, and in this example, the node feature vector is X Input This is shown.

[0035] Using the active power flow entropy, the fourth row of the multi-row heatmap corresponding to the power grid chain failure risk scene is drawn, and the active power flow entropy is set to H and Hm This is shown.

[0036] Using the proportion of historical chain failures in the power grid chain failure risk scene to the total number of failures, the fifth row of the multi-row heatmap corresponding to the power grid chain failure risk scene is drawn, and the proportion of historical chain failures to the total number of failures is Q m As shown above, a multi-line heatmap corresponding to the power grid chain failure risk scenario has been generated.

[0037] In step 104, the multi-row heatmap is divided into a training set and a test set.

[0038] In step 105, a neural network is trained using a training set to obtain a power grid chain failure risk scene recognition model.

[0039] Step 105 involves taking multi-row heatmaps of different operating conditions of the power grid in the training set as input to a CNN model, taking power grid chain failure risk scenes corresponding to the operating conditions in the training set as output to the CNN model, and training the CNN model using a cross-entropy loss function as the loss function to obtain a power grid chain failure risk scene recognition model.

[0040] When a CNN-based power grid chain failure risk scene recognition model is constructed and multi-row heatmap samples representing different operating conditions of the power grid are used as input to the model, the model's output will be a power grid chain failure risk scene for those operating conditions.

number

[0041] Loss function F in the model loss It employs a cross-entropy loss function, and its formula is:

number

[0042] The constructed CNN-based power grid chain failure risk scene recognition model can analyze chain failures with uncertainties that may occur in the power grid under different operating conditions and output whether or not the operating condition is a risk scene.

[0043] In step 106, the power grid chain failure risk scene recognition model is tested using a test set, and the risk scene recognition results are obtained.

[0044] In actual applications, after testing the power grid chain failure risk scene recognition model using a test set and obtaining risk scene recognition results, it is further included to construct a set of prominent risk scenes using a clustering algorithm based on the risk scene recognition results.

[0045] Multi-row heatmaps are divided into training and test sets and input into a recognition model. The model is trained offline, and a clustering method is used to construct a set of prominent risk scenes. Power grid chain failure risk scenes corresponding to active power flow entropy are output. Multi-row heatmaps are divided into training and test sets, samples are input into a CNN-based power grid chain failure risk scene recognition model, feature learning is performed on the multi-row heatmaps using the CNN model, and a clustering method is used to construct a set of prominent risk scenes. The model is trained to output whether or not a scene is a chain failure risk scene under various operating conditions, and the model's recognition ability is verified using the test set. When the model is verified using the test set, given a multi-row heatmap for one operating condition as input, the model can recognize whether or not the scene corresponding to that operating condition belongs to a risk scene and output 0 or 1, where 0 indicates that it does not belong and 1 indicates that it does belong.

[0046] This invention acquires power grid historical operation data, performs a chain failure simulation on the power grid historical operation data, obtains active power flow entropy, visualizes and displays the power grid historical operation data and the active power flow entropy as images, obtains a multi-line heatmap, divides the multi-line heatmap into a training set and a test set, trains a neural network using the training set, obtains a power grid chain failure risk scene recognition model, tests the power grid chain failure risk scene recognition model using the test set, obtains risk scene recognition results, and when applied online, by directly inputting the active power flow entropy into the power grid chain failure risk scene recognition model, risk scene recognition results can be obtained, enabling accurate risk recognition of the current operating scene of the power grid.

[0047] As shown in Figure 5, An acquisition module 10 configured to acquire power grid historical operation data, A chain failure simulation module 20 is configured to perform a chain failure simulation on the aforementioned power grid history operation data and obtain the active power flow entropy, An image display module 30 is configured to visualize and display the aforementioned power grid history operation data and the aforementioned active power flow entropy, and to obtain a multi-line heat map. A partitioning module 40 is configured to divide the aforementioned multi-row heatmap into a training set and a test set, A training module 50 is configured to train a neural network using a training set to obtain a power grid chain failure risk scene recognition model, The present invention further provides a power grid chain failure risk scene recognition system comprising a test module 60 configured to test the power grid chain failure risk scene recognition model using a test set and to obtain risk scene recognition results.

[0048] As one selectable embodiment, the chain failure simulation module 20 is: A construction and simulation unit is configured to construct a power grid simulation model based on the aforementioned power grid historical operation data, set different system operation scenarios, perform chain failure simulations on power grid lines, and obtain chain failure chains and chain failure load rates. The system includes an active power flow entropy determination unit configured to determine the active power flow entropy based on the chain of faults and the chain of fault load ratio.

[0049] The present invention further provides an electronic device comprising one or more processors and a storage device storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in the embodiment of the present invention.

[0050] The present invention further provides a computer storage medium that stores a computer program which, when executed by a processor, implements the method described in the embodiments of the present invention.

[0051] This invention defines active power flow entropy, which reflects the risk of a chain reaction failure in the grid in the current scenario, taking into account the active power flow of the transmission line, and uses the active power flow entropy of the transmission line as an important indicator for identifying power grid chain failure risk scenarios. A deep learning model is applied to uncover the correspondence between the active power flow entropy of the transmission line and power grid chain failure risk scenarios. A multi-row heatmap corresponding to the power grid chain failure risk scenario is drawn by linking the load factor of the power grid, the capacity of the transmission line, and the transmission line nodes. Based on the multi-row heatmap, a power grid chain failure risk scenario recognition model based on a Convolutional Neural Network (CNN) is constructed. The neural network in the model is used to autonomously learn the mapping relationship between the multi-row heatmap and the risk scenario, and a Softmax classifier is used to output the risk scenario recognition result. This method first uses an offline training model to learn the correspondence between the active power flow entropy of historical lines and power grid chain failure risk scenes, thereby obtaining a set of prominent risk scenes. Then, when applied online, the active power flow entropy is directly input into the model to realize risk recognition of the current operating scene of the power grid.

Claims

1. A method for recognizing a power grid chain failure risk scene, which is performed by an electronic device comprising a processor and a memory device configured to store a program, The aforementioned electronic device acquires power grid historical operation data, The aforementioned electronic device performs a chain failure simulation on the power grid history operation data and obtains the active power flow entropy. The electronic device visualizes and displays the power grid history operation data and the active power flow entropy, and obtains a multi-line heat map. The electronic device divides the multi-row heatmap into a training set and a test set, The aforementioned electronic device trains a neural network using a training set to obtain a power grid chain failure risk scene recognition model. The electronic device tests the power grid chain failure risk scene recognition model using a test set and obtains the risk scene recognition result, The aforementioned electronic device trains a neural network using a training set to obtain a power grid chain failure risk scene recognition model. The electronic device takes multi-row heatmaps of multiple operating conditions of the power grid in a training set as input to a convolutional neural network (CNN) model, risk scenes corresponding to various operating conditions in the training set as output to the CNN model, and trains the CNN model using a cross-entropy loss function as the loss function to obtain a power grid chain failure risk scene recognition model. Power grid chain failure risk scene recognition method.

2. The electronic device performs a chain failure simulation on the power grid history operation data and obtains the active power flow entropy, The aforementioned electronic device constructs a power grid simulation model based on the power grid historical operation data, sets up multiple types of system operation scenarios, performs a chain failure simulation on the power grid lines, and obtains the chain failure chain and chain failure load ratio. The electronic device determines the active power flow entropy based on the chain of faults and the chain of fault load rates, The method according to claim 1.

3. The aforementioned power grid history operation data includes power grid chain failure risk scenarios, The electronic device visualizes and displays the power grid history operation data and the active power flow entropy, and obtains a multi-line heat map. The electronic device draws the first row of a multi-row heatmap corresponding to the power grid chain failure risk scene using the load ratio in the current operating status of the power grid chain failure risk scene. The electronic device uses the capacity of the lines in the power grid chain failure risk scene to draw the second row of the multi-row heatmap corresponding to the power grid chain failure risk scene, The electronic device uses the node feature vector of the power grid chain failure risk scene to draw the third row of the multi-row heatmap corresponding to the power grid chain failure risk scene, The electronic device uses the active power flow entropy to draw the fourth row of the multi-row heatmap corresponding to the power grid chain failure risk scene, The electronic device includes drawing the fifth row of a multi-row heatmap corresponding to the power grid chain failure risk scene using the proportion of historical chain failure occurrences in the power grid chain failure risk scene to the total number of failures, The method according to claim 1.

4. The active power flow entropy includes the active power flow entropy of the overloaded line and the active power flow entropy of the non-overloaded line. The expression for the active power flow entropy of the aforementioned non-overloaded transmission line is: [Math 1] And, H is the active power flow entropy, R is a constant, and Z ｋ is the load factor of the k-th branch circuit, and Z ｋ＋１ P is the load factor of the (k+1)th branch circuit, n is the total number of branch circuits, k is the kth branch circuit, and P ｋ The load factor is (Z ｋ Z ｋ＋１ The number of non-overload branch circuits located in ] is the ratio to the total number of branch circuits. The expression for the active power flow entropy of the overloaded line is: [Math 2] And, H ｍ is the effective power flow entropy of the overload line m, h is the basic entropy value of the overload line, (Z ｗ , Z ｗ＋１ is the load rate interval where the line m is located, n ｗ is the number of branch circuits located in the load rate interval where the line m is located, C ｍ is the capacity of the line m, C ｍａｘ is the maximum capacity of the lines in the system, F ｍａｘ is the maximum allowable load rate of the line, P ｗ is the ratio of the number of overload branch circuits located in the load rate (Z ｗ , Z ｗ＋１ to the total number of branch circuits, The method according to claim 1.

5. An acquisition module configured to acquire power grid historical operation data, A chain failure simulation module configured to perform a chain failure simulation on the aforementioned power grid history operation data and obtain the active power flow entropy, An image display module configured to visualize and display the aforementioned power grid history operation data and the aforementioned active power flow entropy, and to obtain a multi-line heat map, A partitioning module configured to divide the aforementioned multi-row heatmap into a training set and a test set, A training module configured to train a neural network using a training set and obtain a power grid chain failure risk scene recognition model, The system includes a test module configured to test the power grid chain failure risk scene recognition model using a test set and to obtain risk scene recognition results. A power grid cascading failure risk scenario recognition system.

6. The aforementioned chain failure simulation module is A construction and simulation unit is configured to build a power grid simulation model based on the aforementioned power grid historical operation data, set up multiple types of system operation scenarios, perform chain failure simulations on power grid lines, and obtain chain failure chains and chain failure load rates. The system includes an active power flow entropy determination unit configured to determine the active power flow entropy based on the chain of faults and the chain of faults load rate, The system according to claim 5.

7. At least one processor, A memory device having at least one program stored in it, When the at least one program is executed by the at least one processor, the at least one processor performs the method according to any one of claims 1 to 4. electronic equipment.

8. When executed by the processor, a computer program is stored which performs the method according to any one of claims 1 to 4. Computer storage medium.

Images