Vulnerability risk monitoring method and system based on power transmission and distribution network

Abstract

The invention discloses a vulnerability risk monitoring method and system based on a transmission and distribution power grid, and the method comprises the steps: obtaining power grid data, and carrying out the preprocessing of the power grid data, and obtaining first power grid data; based on the first power grid data, using a neural network model to predict a power grid load demand to obtain prediction data; performing multi-time scale vulnerability index calculation on the prediction data to obtain a calculation result, constructing a Bayesian attack graph according to the calculation result, and identifying a high-risk path; the high-risk path is implemented by formulating a protection strategy, and data generated by risk monitoring is stored, so that the accuracy of risk path identification is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of risk monitoring, and particularly to a vulnerability risk monitoring method and system based on a power transmission and distribution network. Background Art

[0002] With the development of modern power grids, the key position of power transmission and distribution networks in energy distribution and load scheduling has become increasingly prominent. Traditional power transmission and distribution network designs are mostly based on environments with simple structures and relatively stable loads. However, in recent years, due to factors such as the access of renewable energy, the complexity of electricity demand, natural disasters, and cyberattacks, the vulnerability risks of power transmission and distribution network operations have increased significantly.

[0003] In the existing technology for the vulnerability monitoring of power transmission and distribution networks, potential network risks are mainly identified through traditional static assessment methods or prediction technologies based on single indicators. However, these methods often lack dynamic and multi-dimensional comprehensive assessment capabilities, making it difficult to accurately identify high-risk areas. Especially in complex power grid topologies, their risk prediction capabilities and the effectiveness of emergency protection strategy formulation are significantly insufficient. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a vulnerability risk monitoring method and system based on a power transmission and distribution network to solve the problems that traditional methods lack dynamic and multi-dimensional comprehensive assessment capabilities, making it difficult to accurately identify high-risk areas. Especially in complex power grid topologies, their risk prediction capabilities and the effectiveness of emergency protection strategy formulation are significantly insufficient.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a vulnerability risk monitoring method based on a power transmission and distribution network, including:

[0008] Obtain power grid data, preprocess the power grid data to obtain first power grid data;

[0009] Based on the first power grid data, use a neural network model to predict the power grid load demand to obtain prediction data;

[0010] Calculate vulnerability indicators on multiple time scales for the prediction data to obtain a calculation result, and construct a Bayesian attack graph according to the calculation result to identify high-risk paths;

[0011] Implement the high-risk paths by formulating protection strategies and store the data generated by risk monitoring.

[0012] As a preferred solution of the vulnerability risk monitoring method based on the transmission and distribution network according to the present invention, wherein: using a neural network model to predict the grid load demand includes:

[0013] Obtain historical grid data, and use the historical grid data as a training set to input into the neural network model for iterative training;

[0014] Define a loss function and an optimizer to perform iterative optimization of the model parameters. When the loss of the neural network model decreases less than the first threshold during continuous iteration, stop the iteration, output the model parameters, and update the neural network model;

[0015] Obtain real-time grid data, input the real-time grid data into the neural network model, and obtain prediction data.

[0016] As a preferred solution of the vulnerability risk monitoring method based on the transmission and distribution network according to the present invention, wherein: calculating the vulnerability index of multiple time scales for the prediction data includes:

[0017] Conduct a day-ahead vulnerability assessment based on the prediction data, conduct an intraday rolling vulnerability assessment based on the day-ahead assessment results, and calculate the vulnerability of line disconnection based on the rolling assessment results.

[0018] As a preferred solution of the vulnerability risk monitoring method based on the transmission and distribution network according to the present invention, wherein: conducting a day-ahead vulnerability assessment based on the prediction data includes:

[0019] Calculate the average node degree in the transmission and distribution network;

[0020] Extract the closeness centrality and degree of each node, calculate the extended node degree of each node, and store the extended node degree as the node importance assessment result;

[0021] Extract the spare capacity and predicted load demand of each node, and calculate the flexibility margin of each node;

[0022] Adjust the flexibility margin according to the extended node degree;

[0023] Calculate the comprehensive vulnerability index of the node according to the extended node degree and the flexibility margin;

[0024] Use topological data analysis to analyze the importance of the lines between nodes, and calculate the preliminary vulnerability of the lines based on the node vulnerability and the predicted line load values.

[0025] As a preferred solution of the vulnerability risk monitoring method based on the transmission and distribution network according to the present invention, wherein: conducting an intraday rolling vulnerability assessment based on the day-ahead assessment results and calculating the vulnerability of line disconnection based on the rolling assessment results includes:

[0026] Extract the daily flexible margin, real-time load, predicted load of the node, and calculate the real-time flexible margin of the node;

[0027] Extract the real-time load, rated capacity of each line, and the extended node degrees of the two end nodes of the line, and adjust the load distribution factor of the line;

[0028] Calculate the load impact vulnerability of the line based on the load distribution factor and the rated capacity;

[0029] The calculation of the line disconnection vulnerability based on the rolling evaluation results includes:

[0030] Obtain the extended node degrees of the two end nodes of the line and the historical fault impact factor to calculate the disconnection vulnerability weight factor;

[0031] Use the real-time line load data and the power grid topology to calculate the load transfer impact of the line based on the power flow equation;

[0032] Calculate the disconnection vulnerability based on the load transfer impact and the disconnection vulnerability weight factor;

[0033] Comprehensively calculate the comprehensive vulnerability of the node by combining the node flexible margin and the disconnection vulnerability of the surrounding lines;

[0034] Calculate the comprehensive vulnerability of the line according to the line load impact vulnerability and the disconnection vulnerability;

[0035] Integrate the comprehensive vulnerabilities of all nodes and lines to obtain the calculation result of the final comprehensive vulnerability index.

[0036] As a preferred solution of the vulnerability risk monitoring method based on the transmission and distribution power grid described in the present invention, wherein: constructing a Bayesian attack graph includes:

[0037] Define each power grid node as a resource node of the Bayesian network, and use the node comprehensive vulnerability as the node state probability;

[0038] Define an edge for each line in the Bayesian network, use the line comprehensive vulnerability as the edge weight, define the parent node set and the child node set of each node according to the adjacency relationship in the power grid topology, and define the conditional state probability of the current node by combining the parent node state probability and the edge weight;

[0039] Calculate the final state probability of each node to obtain the conditional probability table of each node;

[0040] Use the depth-first search algorithm to enumerate all paths from the source node to the target node in the Bayesian attack graph, and calculate the total risk probability of each path according to the conditional probability table;

[0041] Sort the paths according to the path risk probability, set a second threshold. If the total risk probability of a path is greater than or equal to the second threshold, the path is a high-risk path, and calculate the contribution degree of each node in the high-risk path to the total risk of the path.

[0042] In the high-risk path, select the node with the highest contribution degree as the key node and mark it as the priority protection object.

[0043] As a preferred solution of the vulnerability risk monitoring method based on the transmission and distribution network of the present invention, wherein: the protection strategy includes:

[0044] For the lines in the high-risk path, reallocate the probabilistic power flow, isolate the key nodes in the high-risk path, and allocate reserve capacity in the high-risk area.

[0045] Real-time monitor the change of node status, evaluate the effect of protection measures, and update the node status isolation and edge weights in the Bayesian attack graph according to the effect data.

[0046] In a second aspect, the present invention provides a vulnerability risk monitoring system based on a transmission and distribution network, including:

[0047] A preprocessing module for obtaining grid data and preprocessing the grid data to obtain first grid data;

[0048] A prediction module for predicting the grid load demand based on the first grid data by using a neural network model to obtain prediction data;

[0049] A calculation module for calculating multi-time scale vulnerability indicators for the prediction data to obtain a calculation result, constructing a Bayesian attack graph according to the calculation result, and identifying high-risk paths;

[0050] A storage module for implementing the high-risk path by formulating a protection strategy and storing the data generated by risk monitoring.

[0051] In a third aspect, the present invention provides a computing device, including:

[0052] A memory and a processor;

[0053] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the vulnerability risk monitoring method based on the transmission and distribution network are implemented.

[0054] In a fourth aspect, the present invention provides a computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are executed by a processor, the steps of the vulnerability risk monitoring method based on the transmission and distribution network are implemented.

[0055] Compared with the prior art, the beneficial effects of the present invention are as follows: Through a multi-time scale method of daily evaluation, intra-day rolling evaluation, and line disconnection vulnerability calculation, the present invention comprehensively considers the topological characteristics, flexible margins, and line load distributions of network nodes, can comprehensively capture potential risk changes of the transmission and distribution network at different times, significantly improves the comprehensiveness and accuracy of vulnerability assessment, uses a Bayesian attack graph to identify high-risk paths, accurately locates key nodes and weak links, and improves the accuracy of risk path identification. Description of the Drawings

[0056] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0057] Figure 1 It is a schematic diagram of the overall process logic of the vulnerability risk monitoring method based on the transmission and distribution network according to an embodiment of the present invention. Detailed Embodiments

[0058] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe the detailed embodiments of the present invention with reference to the drawings of the specification. Obviously, the described embodiments are some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0059] Embodiment 1

[0060] Referring to Figure 1 , an embodiment of the present invention provides a vulnerability risk monitoring method based on a transmission and distribution network, including:

[0061] S100: Obtain power grid data, preprocess the power grid data to obtain first power grid data;

[0062] Specifically, collecting power grid data and transmitting it to the cloud for data preprocessing means extracting topological data, operating state data, and historical dispatching data of the transmission and distribution network from the GIS system, and performing data cleaning and standardization on the collected data;

[0063] The topological data includes a node set, a line set, 110 kV and 35 kV substations, and line section information;

[0064] The operating state data includes voltage, current, power, and load rate;

[0065] Historical scheduling data includes historical load data;

[0066] It should be noted that through the comprehensive collection and cloud transmission of power transmission and distribution network data, the centralized management of power grid operation data is realized, and the data integration efficiency is significantly improved. In the data preprocessing stage, by extracting topological data, operation status data, and historical scheduling data from the GIS system, all-dimensional and high-quality data support is provided for power grid vulnerability assessment. In particular, the extraction of topological data enhances the understanding of the power grid structure, the operation status data provides the basis for real-time monitoring, and the historical scheduling data mines the changing rules of load demand. Data cleaning and standardization processing significantly improve the quality and consistency of data, laying a foundation for subsequent machine learning model training.

[0067] S200: Based on the first power grid data, use a neural network model to predict the power grid load demand and obtain prediction data;

[0068] S300: Calculate multi-time scale vulnerability indicators for the prediction data to obtain calculation results, construct a Bayesian attack graph based on the calculation results, and identify high-risk paths;

[0069] S400: Implement the high-risk paths by formulating protection strategies and store the data generated by risk monitoring.

[0070] It should be noted that a dynamic and multi-dimensional comprehensive evaluation ability is provided, which can accurately identify high-risk areas. Especially in complex power grid topologies, its risk prediction ability and the formulation effect of emergency protection strategies have obvious advantages. By combining real-time load and prediction data to dynamically adjust the flexible margin and load distribution factor, it can accurately reflect the real-time impact of load changes on nodes and lines, improve the real-time and flexibility of power grid vulnerability monitoring, reallocate the probabilistic power flow to relieve the pressure on overloaded lines, isolate key nodes in high-risk paths to prevent risk propagation, and allocate spare capacity in high-risk areas, enhancing the operation stability of the power grid.

[0071] In the embodiment of the present application, the above step S200 includes the following sub-steps A1 - A3;

[0072] In A1: Obtain historical power grid data and input the historical power grid data into the neural network model as a training set for iterative training;

[0073] In A2: Define a loss function and an optimizer to iteratively optimize the model parameters. When the loss decrease of the neural network model is less than the first threshold in consecutive iterations, stop the iteration, output the model parameters, and update the neural network model;

[0074] In A3: Obtain real-time power grid data and input the real-time power grid data into the neural network model to obtain prediction data.

[0075] Specifically, predicting the grid load demand based on the preprocessed data includes using an LSTM model to predict the grid load demand, with ReLU used as the activation function for each layer;

[0076] Taking the historical grid load data as the training set and inputting it into the LSTM model for iterative training, defining a loss function and an Adam optimizer for iterative optimization of the model parameters. When the loss of the LSTM model no longer decreases significantly during consecutive iterations, that is, it is less than the first threshold, and the first threshold is set according to the percentile method, then stop the iteration and output the model parameters to update the LSTM model;

[0077] Inputting the real-time grid load data of the node into the LSTM model to obtain the future grid load demand of the node.

[0078] It should be noted that using the LSTM model for grid load demand prediction, through key steps in stages, solves the problem that traditional load prediction techniques cannot adapt to multi-dimensional time series changes and non-linear load fluctuations in complex grid environments. The training of historical load data provides rich time series information for the model. The combination of the Adam optimizer and the ReLU activation function ensures the efficiency and stability of the training process. The input of real-time data enables the prediction results to dynamically reflect the grid operation status, providing accurate data support for the vulnerability assessment and emergency dispatch of the grid. It can not only significantly improve the accuracy of grid load prediction, but also provide early warnings and optimization bases for weak links in grid operation through forward-looking prediction data, ensuring the safety and stability of the grid in complex environments. By defining a scientific loss function and iteration stop conditions, the model training process is optimized, reducing the consumption of computing resources; it has significant technical advantages in terms of the accuracy, real-time performance, and adaptability of grid load prediction, providing reliable support for the safe operation and intelligent management of the transmission and distribution grid.

[0079] In the embodiment of the present application, the above step S300 includes the following sub-steps B1 - B7;

[0080] In B1: Conduct a day-ahead vulnerability assessment based on the prediction data, conduct an intraday rolling vulnerability assessment based on the day-ahead assessment results, and calculate the line disconnection vulnerability based on the rolling assessment results;

[0081] In B2: Calculate the average node degree in the transmission and distribution grid;

[0082] In B3: Extract the closeness centrality and degree of each node, calculate the extended node degree of each node, and store the extended node degree as the node importance assessment result;

[0083] In B4: Extract the spare capacity and predicted load demand of each node, and calculate the flexibility margin of each node;

[0084] In B5: Adjust the flexible margin according to the extended node degree;

[0085] In B6: Calculate the comprehensive vulnerability index of the node according to the extended node degree and the flexible margin;

[0086] In B7: Use topological data analysis to analyze the importance of the lines between nodes, and calculate the preliminary vulnerability of the lines based on the node vulnerability and the predicted line load values.

[0087] Specifically, the average node degree in the transmission and distribution network is calculated as:

[0088]

[0089] where D i is the degree of node i, and V is the total number of nodes;

[0090] Extract the closeness centrality and degree of each node, and calculate the extended node degree of each node as:

[0091]

[0092] where V adj is the set of neighbor nodes of node i, C i and C j are the closeness centralities of nodes i and j, D i and D j are the degrees of nodes i and j;

[0093] Store the calculated extended node degree as the node importance evaluation result;

[0094] Extract the spare capacity and predicted load demand of each node, and calculate the flexible margin of each node as:

[0095]

[0096] where a i is the spare capacity of node i, P i is the predicted load of node i, and μ is the magnification factor;

[0097] Adjust the flexible margin RA′ according to the extended node degree i which is expressed as:

[0098]

[0099] where α is the weight adjustment factor, with a value range of [0, 1], used to control the influence of node importance on the flexible margin, and max(ND) is the maximum value of the extended node degrees of all nodes in the network, used to normalize ND i to the range of [0, 1];

[0100] Calculate the comprehensive vulnerability index C of the node according to the extended node degree and the flexibility margin i Expressed as:

[0101]

[0102] Wherein, β 1 Is a dimensionless adjustment factor, which controls the proportional influence between the flexibility margin and the node degree, and is dynamically adjusted according to the operation scheduling requirements;

[0103] Use topological data analysis to analyze the importance of the lines between nodes, and calculate the preliminary vulnerability G of line ij based on the node vulnerability and the predicted value of the line load ij Expressed as:

[0104]

[0105] Wherein, S ij Is the rated capacity of line ij, g ij Is the original load distribution factor of line ij, calculated through the line historical data or real-time monitoring, β 2 Is a dimensionless adjustment factor, adjusted according to the dimension matching requirements, ND j Is the extended node degree of the other end node j connected to line ij;

[0106] It should be noted that the day-ahead assessment comprehensively analyzes the potential risks of the entire power grid by calculating key indicators such as the average node degree, extended node degree, and flexibility margin, and can identify the nodes with high importance but low flexibility margin in the network, providing a reference basis for the protection strategy in advance.

[0107] In the embodiment of the present application, after completing steps B1-B7 in step S300 above, the following steps B8-B17 are further included;

[0108] In B8: Extract the day-ahead flexibility margin, real-time load, predicted load of the node, and calculate the real-time flexibility margin of the node;

[0109] In B9: Extract the real-time load, rated capacity of each line, and the extended node degrees of the two end nodes, and adjust the load distribution factor of the line;

[0110] In B10: Calculate the load impact vulnerability of the line based on the load distribution factor and the rated capacity;

[0111] In B11: Calculate the line disconnection vulnerability based on the rolling assessment result, including:

[0112] In B12: Obtain the extended node degrees of the nodes at both ends of the line and the historical fault impact factor to calculate the weight factor of line break vulnerability;

[0113] In B13: Use the real-time line load data and the power grid topology to calculate the load transfer impact of the line based on the power flow equation;

[0114] In B14: Calculate the line break vulnerability based on the load transfer impact and the weight factor of line break vulnerability;

[0115] In B15: Calculate the comprehensive vulnerability of the node by integrating the node flexibility margin and the line break vulnerability of the surrounding lines;

[0116] In B16: Calculate the comprehensive vulnerability of the line according to the line load impact vulnerability and the line break vulnerability;

[0117] In B17: Integrate the comprehensive vulnerabilities of all nodes and lines to obtain the calculation result of the final comprehensive vulnerability index.

[0118] Specifically, the intra-day rolling vulnerability assessment based on the day-ahead assessment results includes:

[0119] Extract the day-ahead flexibility margin, real-time load, and predicted load of the node, and calculate the real-time flexibility margin FA of node i i Expressed as:

[0120]

[0121] Wherein, P′ i is the real-time load of node i;

[0122] Extract the real-time load, rated capacity, and extended node degrees of the nodes at both ends of each line, and adjust the load distribution factor g′ of the line ij Expressed as:

[0123]

[0124] Calculate the load impact vulnerability G′ of line ij based on the load distribution factor and the rated capacity ij Expressed as:

[0125]

[0126] It should be noted that by combining the day-ahead flexibility margin and the real-time load data, the node flexibility margin and the line load distribution factor are dynamically updated, accurately reflecting the real-time impact of load changes on nodes and lines, and improving the real-time performance and flexibility of power grid vulnerability monitoring;

[0127] Specifically, calculating the line break vulnerability based on the rolling assessment results includes:

[0128] The calculation of the extended node degree of the nodes at both ends of the line and the historical fault impact factor to obtain the weight factor of the line break vulnerability is expressed as:

[0129]

[0130] Among them, H i is the historical fault impact factor, which is obtained by normalizing the number of faults and the impact degree of the line through historical data statistics;

[0131] Using the real-time line load data and the power grid topology, the load transfer impact U of the line is calculated based on the power flow equation kl It is expressed as:

[0132]

[0133] In the formula, ΔP l is the load change of line l, ΔP k is the load change of line k, X lk is the reactance value between line l and line k, X mk is the reactance value between line m and line k, max(ΔP) is the maximum load change value, and L is the set of lines;

[0134] Based on the load transfer impact and the weight factor of the line break vulnerability, the line break vulnerability K is calculated l It is expressed as:

[0135]

[0136] Among them, f kl is the load transfer impact intensity of line k on line l;

[0137] The comprehensive vulnerability V of node i is calculated by integrating the node flexibility margin and the line break vulnerability of the surrounding lines 1 It is expressed as:

[0138]

[0139] Among them, L i is the set of lines directly connected to node i;

[0140] The comprehensive vulnerability V of the line is calculated according to the line load impact vulnerability and the line break vulnerability 2 It is expressed as:

[0141] V 2 =G′ ij +K l

[0142] Integrate the comprehensive vulnerabilities of all nodes and lines to obtain the final calculation result of the comprehensive vulnerability index.

[0143] It should be noted that by calculating the broken-line vulnerability weight factor and the impact of load transfer, the cascading effect of line breakage on the operation of the entire network is quantified. This method effectively solves the problem of insufficient assessment of line breakage risk in the existing technology and significantly improves the stability of power grid operation;

[0144] It should also be noted that by integrating the node flexibility margin, the extended node degree, and the broken-line vulnerability, a comprehensive evaluation index is formed. Compared with a single index, this method provides a more comprehensive perspective for vulnerability analysis, significantly improves the evaluation accuracy, and significantly enhances the accuracy and real-time performance of the vulnerability monitoring of the transmission and distribution network through a calculation method of vulnerability indicators based on multiple time scales. In the day-ahead assessment, the calculation of the average node degree, the extended node degree, and the flexibility margin makes the prediction of potential risks of the power grid more comprehensive, laying a foundation for the early deployment of protection measures. In the intraday rolling assessment, by combining real-time load and prediction data to dynamically adjust the flexibility margin and the load distribution factor, the problem of insufficient real-time performance in traditional methods is effectively solved. In the calculation of broken-line vulnerability, by introducing the historical fault impact factor and the impact of load transfer, the impact of line breakage on the entire network is quantified, providing a more accurate evaluation basis for the risk protection of key lines in the power grid. Finally, through the calculation of the comprehensive vulnerability index, a global and dynamic power grid vulnerability analysis method is provided, providing scientific support for the optimization of intelligent dispatching and protection strategies. The overall design of the present invention provides technical guarantees for emergency protection and operation stability while enhancing the risk perception ability of the transmission and distribution network.

[0145] In the embodiment of the present application, after completing steps B8 - B17 in the above step S300, the following steps B18 - B22 are further included;

[0146] In B18: Define each power grid node as a resource node of the Bayesian network, and use the comprehensive node vulnerability as the node state probability;

[0147] In B18: Define an edge for each line in the Bayesian network, use the comprehensive line vulnerability as the edge weight, define the parent node set and the child node set of each node according to the adjacency relationship in the power grid topology, and define the conditional state probability of the current node by combining the parent node state probability and the edge weight;

[0148] In B19: Calculate the final state probability of each node to obtain the conditional probability table of each node;

[0149] In B20: Use the depth-first search algorithm to enumerate all paths from the source node to the target node in the Bayesian attack graph, and calculate the total risk probability of each path according to the conditional probability table;

[0150] In B21: Sort the paths according to the path risk probability, set a second threshold. If the total risk probability of a path is greater than or equal to the second threshold, the path is a high-risk path, and calculate the contribution degree of each node in the high-risk path to the total risk of the path.

[0151] In B22: In the high-risk paths, select the node with the highest contribution degree as the key node and mark it as the priority protection object.

[0152] Specifically, constructing a Bayesian attack graph to identify high-risk paths according to the index calculation results means defining each power grid node (substation, load center) as a resource node of the Bayesian network, and taking the comprehensive vulnerability of the node as the node state probability:

[0153] According to the power grid topology, define an edge for each line ij in the Bayesian network, use the comprehensive vulnerability of the line as the edge weight, define the parent node set and child node set of each node according to the adjacency relationship in the power grid topology, and combine the parent node state probability and the edge weight to define the conditional state probability P(S j ∣Parent(S j ):

[0154]

[0155] Among them, P(S k ) is the state probability of the parent node S k , and E kj is the edge weight from the parent node S k to the node S j ;

[0156] It should be noted that by modeling the power grid nodes (such as substations, load centers) as resource nodes in the Bayesian network and defining the edge weights based on the line vulnerabilities, a Bayesian attack graph that can describe the potential risk propagation in the system is constructed. This graph can capture the correlation between nodes in the power grid and the propagation path of the fault risk, which helps to comprehensively and dynamically analyze the risk distribution of the power grid and provides strong theoretical support for the identification of high-risk paths;

[0157] Specifically, calculate the final state probability of each node to obtain the conditional probability table of each node;

[0158] The conditional probability table includes the parent node state combination and the corresponding conditional state probability;

[0159] Use the depth-first search algorithm to enumerate all paths from the source node to the target node in the Bayesian attack graph, and calculate the total risk probability D of each path according to the conditional probability table:

[0160]

[0161] Wherein, P(S i ) is the state probability of node S i .

[0162] Sort the paths in descending order according to the path risk probability, set the second threshold through statistical analysis. If the total path risk probability is greater than or equal to the second threshold, the path is a high-risk path, and calculate the contribution degree K(S i ) of each node in the high-risk path to the total path risk:

[0163]

[0164] Wherein, P(S i ) is the state probability of node S i , P(S j ) is the state probability of node P(S j ), is the sum of the state probabilities of all nodes in the path;

[0165] It should be noted that by modeling power grid nodes (such as substations, load centers) as resource nodes in the Bayesian network and defining edge weights based on line vulnerability, a Bayesian attack graph that can describe the potential risk propagation in the system is constructed. This graph can capture the correlation between nodes in the power grid and the propagation path of fault risks, which helps to comprehensively and dynamically analyze the risk distribution of the power grid and provides strong theoretical support for the identification of high-risk paths;

[0166] It should also be noted that the intelligent analysis and accurate identification of the overall risk of the power grid are realized. It can quickly lock high-risk paths in complex topologies and save computing resources; among high-risk paths, select the node with the highest contribution degree as the key node and mark it as the priority protection object; combined with the Bayesian network and real-time data analysis, the dynamic assessment of the power grid node status and path risk is realized, which can timely reflect the changes in the operating state and the law of risk propagation. The high-risk path identification method based on the comprehensive vulnerability index and the Bayesian attack graph can accurately locate weak links and key nodes, significantly improving the accuracy of risk identification. By optimizing path enumeration and risk calculation through depth-first search, the efficiency of high-risk path screening is significantly improved, providing an efficient solution for the real-time monitoring of complex power grid topologies. Dynamically formulate protection strategies according to path risk probability and node contribution degree, realizing the accurate protection of key nodes and minimizing the possibility of risk propagation in the power grid.

[0167] In the embodiment of the present application, the above step S400 includes the following sub-steps C1 - C2;

[0168] In C1: For the lines in the high-risk path, reallocate the probabilistic power flow, isolate the critical nodes in the high-risk path, and allocate spare capacity in the high-risk area.

[0169] In C2: Monitor the change of node status in real time, evaluate the effect of protection measures, and update the node status isolation and edge weights in the Bayesian attack graph according to the effect data.

[0170] Specifically, formulating a protection strategy means reallocating the probabilistic power flow for the lines in the high-risk path, reducing the pressure on overloaded lines, isolating the critical nodes in the high-risk path, preventing the spread of risks, and allocating spare capacity in the high-risk area; monitoring the change of node status in real time after the implementation of the protection strategy, evaluating the effect of protection measures, and updating the node status isolation and edge weights in the Bayesian attack graph according to the effect data.

[0171] It should be noted that reallocating the probabilistic power flow effectively alleviates the load pressure on the lines in the high-risk path, enhances the operation stability of the power grid, isolating critical nodes blocks the risk propagation path, reduces the possibility of chain reactions, the spare capacity allocation mechanism quickly responds to the load changes in the high-risk area, providing a solid disaster tolerance capacity guarantee for the power grid, and monitoring the node status in real time and dynamically updating the Bayesian attack graph makes the risk monitoring and the adjustment of protection strategies more accurate and scientific. Overall, the present invention not only improves the risk perception ability of the transmission and distribution power grid, but also significantly enhances its protection and recovery capabilities, providing strong technical support for building a more intelligent operation management system for the transmission and distribution power grid.

[0172] Specifically, storing and backing up the data generated during the risk monitoring process means storing the collected power grid data, the power grid load demand prediction results, the vulnerability index calculation results, and the protection strategy in the database, setting security access measures, synchronously backing up the stored data to the cloud, and regularly detecting the integrity of the stored data and the backup data. After the detection, generate an integrity detection record and synchronously store it in the database.

[0173] It should be noted that through standardized database management, systematically storing the load prediction results, vulnerability indicators, and protection strategies forms a highly integrated data center, facilitating efficient query and subsequent analysis. Secondly, using cloud backup technology further improves the data redundancy and disaster recovery capabilities on the basis of local storage, enabling the system to ensure data security even in extreme cases. Moreover, regularly performing integrity detection ensures the authenticity and reliability of the data, and realizes transparent data management through the integrity detection record, providing a solid technical foundation for the long-term operation of the transmission and distribution power grid.

[0174] The above is a schematic solution of a vulnerability risk monitoring method based on the transmission and distribution power grid in this embodiment. It should be noted that the technical solution of the vulnerability risk monitoring system based on the transmission and distribution power grid belongs to the same concept as the technical solution of the above-mentioned vulnerability risk monitoring method based on the transmission and distribution power grid. For the details not described in detail in the technical solution of the vulnerability risk monitoring system based on the transmission and distribution power grid in this embodiment, reference can be made to the description of the technical solution of the above-mentioned vulnerability risk monitoring method based on the transmission and distribution power grid.

[0175] The vulnerability risk monitoring system based on the transmission and distribution power grid in this embodiment includes:

[0176] A preprocessing module, configured to obtain power grid data, preprocess the power grid data, and obtain first power grid data;

[0177] A prediction module, configured to predict the power grid load demand based on the first power grid data by using a neural network model, and obtain prediction data;

[0178] A calculation module, configured to calculate vulnerability indicators on multiple time scales for the prediction data, obtain a calculation result, construct a Bayesian attack graph according to the calculation result, and identify high-risk paths;

[0179] A storage module, configured to implement the high-risk paths by formulating protection strategies and store the data generated by risk monitoring.

[0180] This embodiment also provides a computing device applicable to the vulnerability risk monitoring of the transmission and distribution power grid, including:

[0181] A memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the vulnerability risk monitoring method based on the transmission and distribution power grid as proposed in the above embodiment.

[0182] This embodiment also provides a storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the vulnerability risk monitoring method based on the transmission and distribution power grid as proposed in the above embodiment.

[0183] The storage medium proposed in this embodiment and the vulnerability risk monitoring method based on the transmission and distribution power grid proposed in the above embodiment belong to the same inventive concept. The technical details not described in detail in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0184] From the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software and necessary general-purpose hardware, and of course, it can also be implemented by hardware. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disc of a computer, etc., and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of the various embodiments of the present invention.

Claims

1. A vulnerability risk monitoring method based on a transmission and distribution network, characterized in that: include: Acquire power grid data, and preprocess the power grid data to obtain first power grid data; Based on the first power grid data, using a neural network model to predict the power grid load demand to obtain prediction data; Calculating the multi-time scale vulnerability index on the predicted data to obtain calculation results, constructing a Bayesian attack graph based on the calculation results, and identifying high-risk paths; The high-risk paths are implemented by formulating protection strategies and storing the data generated by risk monitoring.

2. The vulnerability risk monitoring method based on the transmission and distribution network according to claim 1 is characterized in that: The use of neural network models to predict power grid load demand includes: Obtain historical power grid data, and input the historical power grid data into the neural network model as a training set for iterative training; Define a loss function and an optimizer to iteratively optimize the model parameters. When the loss of the neural network model decreases below a first threshold during continuous iterations, stop iterating and output the model parameters, and update the neural network model. Obtain real-time power grid data, input the real-time power grid data into the neural network model, and obtain predicted data.

3. The vulnerability risk monitoring method based on the transmission and distribution network according to claim 2 is characterized in that: Calculating the multi-time scale vulnerability index for the forecast data includes: A day-ahead vulnerability assessment is conducted based on the forecast data, an intraday rolling vulnerability assessment is conducted based on the day-ahead assessment results, and the line disconnection vulnerability is calculated based on the rolling assessment results.

4. The vulnerability risk monitoring method based on the transmission and distribution network according to claim 3 is characterized in that: The day-ahead vulnerability assessment based on forecast data includes: Calculate the average node degree in the transmission and distribution network; Extract the closeness centrality and degree of each node, calculate the extended node degree of each node, and store the extended node degree as the node importance evaluation result; Extract the spare capacity and predicted load demand of each node and calculate the flexibility margin of each node; Adjust the flexibility margin according to the extended node degree; The comprehensive vulnerability index of the node is calculated based on the extended node degree and flexibility margin; The importance of lines between nodes is analyzed using topological data, and the preliminary vulnerability of lines is calculated based on node vulnerability and line load prediction values.

5. The vulnerability risk monitoring method based on the transmission and distribution network according to claim 3 is characterized in that: The daily rolling vulnerability assessment based on the day-ahead assessment results and the calculation of line disconnection vulnerability based on the rolling assessment results include: Extract the day-ahead flexibility margin, real-time load and forecasted load of the node, and calculate the real-time flexibility margin of the node; Extract the real-time load, rated capacity and extended node degree of nodes at both ends of each line, and adjust the load distribution factor of the line; Calculate the load shock vulnerability of the line based on the load distribution factor and rated capacity; The calculation of line disconnection vulnerability based on the rolling assessment result includes: Obtain the extended node degree and historical fault impact factor of the nodes at both ends of the line to calculate the disconnection vulnerability weight factor; Using real-time line load data and grid topology, calculate the load transfer impact of the line based on power flow equations; Calculate the line break vulnerability based on the load transfer impact and the line break vulnerability weight factor; The comprehensive vulnerability of the node is calculated by integrating the node flexibility margin and the disconnection vulnerability of the surrounding lines; Calculate the comprehensive vulnerability of the line based on the line load impact vulnerability and line break vulnerability; The comprehensive vulnerability of all nodes and lines is integrated to obtain the final calculation result of the comprehensive vulnerability index.

6. The vulnerability risk monitoring method based on the transmission and distribution network according to claim 1 or 5, characterized in that: Building a Bayesian attack graph involves: Define each grid node as a resource node in the Bayesian network, and use the node's comprehensive vulnerability as the node's state probability; In the Bayesian network, an edge is defined for each line, and the comprehensive vulnerability of the line is used as the weight of the edge. The parent node set and child node set of each node are defined according to the adjacency relationship in the power grid topology. The conditional state probability of the current node is defined by combining the parent node state probability and the edge weight. Calculate the final state probability of each node to obtain the conditional probability table of each node; Use the depth-first search algorithm to enumerate all paths from the source node to the target node in the Bayesian attack graph, and calculate the total risk probability of each path according to the conditional probability table; Sort the paths according to the path risk probability, set a second threshold, if the total risk probability of the path is greater than or equal to the second threshold, the path is a high-risk path, and calculate the contribution of each node in the high-risk path to the total risk of the path; In the high-risk path, the node with the highest contribution is selected as the key node and marked as the priority protection object.

7. The vulnerability risk monitoring method based on the transmission and distribution network according to claim 6 is characterized in that: Protection strategies include: For lines in high-risk paths, the probability flow is redistributed, key nodes in high-risk paths are isolated, and spare capacity is deployed in high-risk areas; Monitor node status changes in real time, evaluate the effectiveness of protection measures, and update node status isolation and edge weights in the Bayesian attack graph based on the effect data.

8. A system using the method for monitoring vulnerability risk based on a transmission and distribution network as described in any one of claims 1 to 7, characterized in that: include: A preprocessing module, used for acquiring power grid data, and preprocessing the power grid data to obtain first power grid data; A prediction module, configured to predict the load demand of a power grid using a neural network model based on the first power grid data to obtain prediction data; A calculation module, used to calculate the multi-time scale vulnerability index of the predicted data to obtain a calculation result, construct a Bayesian attack graph according to the calculation result, and identify high-risk paths; The storage module is used to implement the high-risk path by formulating a protection strategy and store data generated by risk monitoring.

9. An electronic device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the vulnerability risk monitoring method based on the transmission and distribution network as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the method for monitoring vulnerability risk based on a transmission and distribution network as described in any one of claims 1 to 7.

Images