A method for identifying key nodes of a power grid system

Abstract

The application provides a key node identification method of a power grid system, and belongs to the field of power grid monitoring, and comprises the following steps: step 1, constructing a power grid system model; step 2, calculating the influence of nodes in the power grid system model by using an aggregated local structure information identification key node algorithm ALSI, which comprises the following steps: calculating the contribution of neighbor nodes by using the degree of the neighbor nodes and the layer where the neighbor nodes are located; calculating the influence of the power grid nodes by using the degree of the nodes and the layer where the nodes are located; and calculating the influence of the nodes in the network by comprehensively considering the contribution of the neighbor nodes and the influence of the nodes; and step 3, identifying the key nodes according to the influence of the nodes in the network. The method improves the identification ability of the key nodes; the similar relationship between the nodes and the neighbor nodes, the Ks value and the degree of the neighbor nodes, and the contribution of the neighbor nodes are introduced, so that the detection of the nodes is more effective; the algorithm is simple, has low time complexity, and can be applied to a large network.

Description

Technical Field

[0001] This invention belongs to the field of power grid monitoring technology, and specifically relates to a method for identifying key nodes in a power grid system. Background Technology

[0002] With societal development, the power grid is expanding in scale and increasing in complexity. Power grid failures, caused by natural disasters, human-caused accidents, and cybersecurity issues, occur frequently. Besides large-scale blackouts in a few countries, smaller, localized outages also occur regularly. Analysis shows that blackouts often begin at a specific node and gradually expand, with the final damage related to the node's importance within the overall power grid. Identifying the importance of nodes and implementing corresponding monitoring measures can prevent failures or, when they do occur, allow for rapid location and intervention to minimize losses. Therefore, effective critical node identification methods are crucial for improving the safety and stability of the power grid system.

[0003] Existing methods for identifying critical nodes generally suffer from two problems: simple algorithms produce inaccurate results, while accurate algorithms are overly complex. Algorithms that consider numerous influencing factors, such as the node's global structure, local structure, and intrinsic attributes, can accurately identify critical nodes in the network. However, these algorithms are typically highly complex, time-consuming, and unsuitable for large-scale networks. Conversely, algorithms that consider fewer factors are simpler in concept, have shorter running times, and can produce results more quickly, but their accuracy is lower.

[0004] Therefore, how to accurately and quickly identify key nodes in the power grid is an urgent problem to be solved. Summary of the Invention

[0005] To overcome the shortcomings of the existing technology, the present invention provides a method for identifying key nodes in a power grid system, which solves the problems of inaccurate node influence identification and excessive complexity leading to long running time in the existing methods, thereby improving the node identification efficiency and network security of the power grid system.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A method for identifying critical nodes in a power grid system includes the following steps:

[0008] Construct a power grid system model;

[0009] The Attribution-Integrated Local Structure Information (ALSI) algorithm is used to calculate the influence of nodes in a power grid system model, including:

[0010] The contribution of a neighboring node is calculated by the degree of the neighboring node and the layer in which the neighboring node belongs.

[0011] The influence of a power grid node is calculated by its degree and the layer it belongs to.

[0012] The influence of a node in the network is calculated by combining the contributions of its neighbors and the node's own influence.

[0013] Key nodes are identified based on their influence within the network.

[0014] Preferably, the power grid system model G is:

[0015] G = (V, E)

[0016] Where V = {v1, v2, ... vn} is the set of all nodes in the power grid system, and E = {e1, e2, ... em} is the edge connection between the power grid nodes.

[0017] Preferably, the specific steps for calculating the contribution of neighboring nodes are as follows:

[0018] Let grid node vj be a neighbor node of vi. The initial contribution of vj to vi is Γ(vj), which is calculated based on vj's Ks value and the number of its neighboring nodes DC(vj). The calculation formula is as follows:

[0019] Γ(vj)=Ks(vj)*simjkd(vj,vi)+DC(vj)*Ks(vj)

[0020] Where simjkd(vj,vi) is the correlation degree of the power grid node, and Ks(vj) is the layer value of the power grid node vj in the entire network calculated by the K-shell algorithm;

[0021] Furthermore, depending on whether vj and vi are on the same level, there are three cases: When Ks(vj) is less than Ks(vi), that is, when the Ks value of vj is smaller than the Ks value of vi, in order to reduce the weight of vj in the calculation of vi's influence, the formula for calculating vj's contribution is as follows:

[0022] I(vj)=Γ(vj) / (maxKs+minKs)

[0023] When Ks(vj) is greater than Ks(vi), that is, when the Ks value of vj is greater than the Ks value of vi, in order to increase the weight of vj in the calculation of vi's influence, the formula for calculating vj's contribution is as follows:

[0024] I(vj)=Γ(vj) / (maxKs-minKs)

[0025] When Ks(vj) equals Ks(vi), that is, when vj and vi are located in the same layer, the contribution of vj is calculated as follows:

[0026] I(vj=)Γ(vj) / maxKs

[0027] Where Ks(vi) is the layer value of the power grid node vi in ​​the entire network, calculated by the K-shell algorithm.

[0028] Preferably, the calculation process for the maximum and minimum Ks values ​​in the power grid system model G is as follows:

[0029] Given a network G = (V, E), after decomposition using the Ks algorithm, the Ks values ​​of all network nodes are obtained, and the Ks value of node vi is Ks(vi). Find the maximum value maxKs and the minimum value minKs, and the calculation formulas are as follows:

[0030] maxKs=max(Ks(vi))

[0031] minKs = min(Ks(vi)).

[0032] Preferably, the calculation process for the correlation degree of power grid nodes in the power grid system model G is as follows:

[0033] The correlation between power grid nodes vi and vj is defined as their node similarity. The set of neighboring nodes of vi is Ω(vi), and the set of neighboring nodes of vj is Ω(vj). The Jaccard similarity coefficient is used as the correlation of the nodes, and its value is the ratio of the size of the intersection of Ω(vi) and Ω(vj) to the size of the union of Ω(vi) and Ω(vj).

[0034]

[0035] Preferably, the specific steps for calculating the influence of a power grid node itself are as follows:

[0036] The influence of a power grid node vi is calculated based on the number of its neighbors, DC(vi), and the layer number it belongs to. The calculation formula is as follows:

[0037] K(vi)=DC(vi)+Ks(vi) / maxKs.

[0038] Preferably, the influence of the computing node in the network is determined by the following steps:

[0039] The influence of a power grid node is obtained by combining its own influence with the contributions of its neighboring nodes. The calculation formula is as follows:

[0040]

[0041] In the above formula, N(vi) is the set of neighboring nodes of vi, and maxDC is the maximum value of the degree of all power grid nodes in the power grid system.

[0042] Preferably, the identification of key nodes based on the influence of nodes in the network specifically involves: identifying key nodes in the network based on the influence value of each node in the network, where nodes with larger influence values ​​have greater influence than nodes with smaller influence values.

[0043] The method for identifying key nodes in a power grid system provided by this invention has the following advantages:

[0044] This invention proposes an ALSI method for detecting the influence of nodes in a power grid system. It introduces the degree of a node in addition to the Ks value, overcoming the coarseness of the K-shell algorithm and improving the identification ability of key nodes. The method incorporates the similarity relationship between a node and its neighbors, as well as the Ks value and degree of neighboring nodes. The inclusion of neighbor contributions makes node detection more effective. The algorithm is simple, has low time complexity, and is applicable to large-scale networks. Attached Figure Description

[0045] To more clearly illustrate the embodiments and design schemes of the present invention, the accompanying drawings required for this embodiment will be briefly described below. The drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a flowchart of the key node identification method for a power grid system according to Embodiment 1 of the present invention;

[0047] Figure 2 For power system network models;

[0048] Figure 3 The Kendall τ values ​​are obtained by comparing each method with the SIR model on the power grid network. Detailed Implementation

[0049] To enable those skilled in the art to better understand and implement the technical solutions of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and should not be construed as limiting the scope of protection of the present invention.

[0050] Example 1

[0051] This invention provides a method for identifying critical nodes in a power grid system, specifically as follows: Figure 1 As shown, it includes the following steps:

[0052] Step 1: Construct a power grid system model:

[0053] Let G = (V, E) be a power system network, i.e., a power grid system model.

[0054] Where V = {v1, v2, ... vn} is the set of all nodes in the power grid system, and E = {e1, e2, ... em} is the edge connection between the power grid nodes.

[0055] Step 2: The Attribution-Based Local Structure Information (ALSI) algorithm is used to calculate the influence of nodes in the power grid system model. This specifically includes:

[0056] Step 2.1: Calculate the contribution of neighboring nodes by using the degree of neighboring nodes and the layer in which the neighboring nodes are located;

[0057] Let grid node vj be a neighbor node of vi. The initial contribution of vj to vi is Γ(vj), which is calculated based on vj's Ks value and the number of its neighboring nodes DC(vj). The calculation formula is as follows:

[0058] Γ(vj)=Ks(vj)*simjkd(vj,vi)+DC(vj)*Ks(vj) (1)

[0059] Where simjkd(vj,vi) represents the correlation degree of the power grid node, and Ks(vj) represents the layer value of the power grid node vj in the entire network, calculated by the K-shell algorithm.

[0060] Specifically, the calculation process for the grid node correlation degree simjkd(vj,vi) is as follows:

[0061] The correlation between power grid nodes vi and vj is defined as their node similarity. The set of neighboring nodes of vi is Ω(vi), and the set of neighboring nodes of vj is Ω(vj). The algorithm uses the Jaccard similarity coefficient as the correlation of nodes, which is the ratio of the size of the intersection of Ω(vi) and Ω(vj) to the size of the union of Ω(vi) and Ω(vj).

[0062]

[0063] Furthermore, depending on whether vj and vi are on the same layer, there are three cases. When Ks(vj) is less than Ks(vi), that is, when the Ks value of vj is smaller than the Ks value of vi, in order to reduce the weight of vj in the calculation of vi's influence, the contribution calculation formula for vj is as follows:

[0064] I(vj)=Γ(vj) / (maxKs+minKs) (3)

[0065] When Ks(vj) is greater than Ks(vi), that is, when the Ks value of vj is greater than the Ks value of vi, in order to increase the weight of vj in the calculation of vi's influence, the formula for calculating vj's contribution is as follows:

[0066] I(vj)=Γ(vj) / (maxKs-minKs) (4)

[0067] When Ks(vj) equals Ks(vi), that is, when vj and vi are located in the same layer, the contribution of vj is calculated as follows:

[0068] I(vj=)Γ(vj) / maxKs (5)

[0069] Where Ks(vi) is the layer value of the power grid node vi in ​​the entire network, calculated by the K-shell algorithm.

[0070] Specifically, the calculation process for the maximum and minimum Ks values ​​in the power grid system model G is as follows:

[0071] Given a network G = (V, E), after decomposition using the Ks algorithm, the Ks values ​​of all network nodes are obtained, and the Ks value of node vi is Ks(vi). Find the maximum value maxKs and the minimum value minKs, and the calculation formulas are as follows:

[0072] maxKs=max(Ks(vi)) (6)

[0073] minKs=min(Ks(vi)) (7)

[0074] Step 2.2: Calculate the influence of a power grid node by its degree and the layer it belongs to. The specific steps are as follows:

[0075] The influence of a power grid node vi is calculated based on the number of its neighbors, DC(vi), and the layer number it belongs to. The calculation formula is as follows:

[0076] K(vi)=DC(vi)+Ks(vi) / maxKs (8)

[0077] Step 2.3: Calculate the node's influence in the network by combining the contributions of neighboring nodes and the node's own influence. The specific steps are as follows:

[0078] The influence of a power grid node is obtained by combining its own influence with the contributions of its neighboring nodes. The calculation formula is as follows:

[0079]

[0080] In the above formula, N(vi) is the set of neighboring nodes of vi, and maxDC is the maximum value of the degree of all power grid nodes in the power grid system.

[0081] Step 3: Identify key nodes based on their influence in the network. Specifically, identify key nodes in the network based on the influence value of each node. Nodes with higher influence values ​​have greater influence than nodes with lower influence values.

[0082] The influence of a node is determined by its own influence and the contribution of its neighboring nodes. Therefore, formula (9) is used to calculate the influence value of each node in the network. The nodes are sorted according to their influence. The nodes with larger values ​​have greater influence than the nodes with smaller values. Therefore, key nodes in the network can be identified based on their influence values.

[0083] Experimental verification:

[0084] right Figure 2 In the power system network model constructed in the above steps, all nodes are calculated. After obtaining the weight of each node by running different algorithms in Python, they are ranked. Table 1 shows the results of ranking the ALSI algorithm, eight existing algorithms including BC and CC, and SIR.

[0085] Table 1 shows the sorting results of the algorithm; the last two columns are the SIR sorting and values.

[0086]

[0087] As shown in Table 1, the first eight nodes of the ALSI algorithm are completely consistent with the SIR model. Among them, nodes v4, v2, v1, and v3 are relatively important. According to the algorithm's calculation rules, firstly, based on the Ks algorithm, these four nodes are located in the innermost layer, and their neighboring nodes are all located in the innermost or second-innermost layer. However, the degree of nodes v4 and v2 is greater than that of nodes v1 and v3, and the self-influence of nodes v4 and v2 is greater than that of nodes v1 and v3. Secondly, the contribution of the neighboring nodes of node v4 is greater than that of node v2, therefore node v4 has the greatest influence. Thus, it can be concluded that the node influence calculated using the ALSI method in this power system network model is relatively accurate and can effectively identify key nodes.

[0088] Simultaneously, verification was conducted in the Powergrid network, which contains power grid information for the western states of the United States. The network has 4,941 nodes and 6,594 edges. Nodes are generators, substations, or transformers, and the edges between nodes are power supply lines.

[0089] In the process of method performance verification, the Kendall τ coefficient is used to detect the effectiveness of the method, and its calculation formula is as follows:

[0090]

[0091] In formula (10), n c With n d τ represents the number of consistent and inconsistent elements in two sequences, where n represents the number of elements in the sequence. This coefficient is used to judge the similarity of the sorting results of two sequences. The larger the value of τ, the greater the similarity and the more accurate the sorting results generated by the method.

[0092] Because the power grid network has a large number of nodes, the infection probability is set to 0.06 in order to analyze the effectiveness of the method in the SIR model. Figure 3 This represents the Kendall values ​​for the ALSI method and the other eight methods in a power grid network. Figure 3 It can be seen that the Kendall value of the ALSI method is better than that of other methods, so the method is more effective and can effectively identify key nodes in the power grid network.

[0093] The above-described embodiments are merely preferred embodiments of the present invention, and the scope of protection of the present invention is not limited thereto. Any simple changes or equivalent substitutions of the technical solutions that can be obviously obtained by those skilled in the art within the scope of the technology disclosed in the present invention shall fall within the scope of protection of the present invention.

Claims

1. A method for identifying key nodes in a power grid system, characterized in that, Includes the following steps: Construct a power grid system model; The Attribution-Integrated Local Structure Information (ALSI) algorithm is used to calculate the influence of nodes in a power grid system model, including: The contribution of a neighboring node is calculated by the degree of the neighboring node and the layer in which the neighboring node belongs. The influence of a power grid node is calculated by its degree and the layer it belongs to. The influence of a node in the network is calculated by combining the contributions of its neighbors and the node's own influence. Identify key nodes based on their influence in the network; In this context, nodes are generators, substations, or transformers, and the edges between nodes are power supply lines. The power grid system model G is: G = (V, E) Where V = {v1, v2, ... vn} is the set of all nodes in the power grid system, and E = {e1, e2, ... em} is the edge connection between the power grid nodes; The specific steps for calculating the contribution of neighboring nodes are as follows: Let grid node vj be a neighbor node of vi. The initial contribution of vj to vi is Γ(vj), which is calculated based on vj's Ks value and the number of its neighboring nodes DC(vj). The calculation formula is as follows: Γ(vj)=Ks(vj)*simjkd(vj,vi)+DC(vj)*Ks(vj) Where simjkd(vj,vi) is the correlation degree of the power grid node, and Ks(vj) is the layer value of the power grid node vj in the entire network calculated by the K-shell algorithm; Furthermore, depending on whether vj and vi are on the same level, there are three cases: When Ks(vj) is less than Ks(vi), that is, when the Ks value of vj is smaller than the Ks value of vi, in order to reduce the weight of vj in the calculation of vi's influence, the formula for calculating vj's contribution is as follows: I(vj)=Γ(vj) / (maxKs+minKs) When Ks(vj) is greater than Ks(vi), that is, when the Ks value of vj is greater than the Ks value of vi, in order to increase the weight of vj in the calculation of vi's influence, the formula for calculating vj's contribution is as follows: I(vj)=Γ(vj) / (maxKs-minKs) When Ks(vj) equals Ks(vi), that is, when vj and vi are located in the same layer, the contribution of vj is calculated as follows: I(vj=)Γ(vj) / maxKs Where Ks(vi) is the layer value of the power grid node vi in ​​the entire network, calculated by the K-shell algorithm; The calculation process for the maximum and minimum Ks values ​​in the power grid system model G is as follows: Given a network G = (V, E), after decomposition using the Ks algorithm, the Ks values ​​of all network nodes are obtained, and the Ks value of node vi is Ks(vi). Find the maximum value maxKs and the minimum value minKs, and the calculation formulas are as follows: maxKs=max(Ks(vi)) minKs=min(Ks(vi)) The calculation process for the correlation degree of power grid nodes in the power grid system model G is as follows: The correlation between power grid nodes vi and vj is defined as their node similarity. The set of neighboring nodes of vi is Ω(vi), and the set of neighboring nodes of vj is Ω(vj). The Jaccard similarity coefficient is used as the correlation of the nodes, and its value is the ratio of the size of the intersection of Ω(vi) and Ω(vj) to the size of the union of Ω(vi) and Ω(vj). The specific steps for calculating the influence of a power grid node itself are as follows: The influence of a power grid node vi is calculated based on the number of its neighbors, DC(vi), and the layer number it belongs to. The calculation formula is as follows: K(vi) = DC(vi) + Ks(vi) / maxKs The influence of the computing node on the network is determined by the following steps: The influence of a power grid node is obtained by combining its own influence with the contributions of its neighboring nodes. The calculation formula is as follows: In the above formula, N(vi) is the set of neighboring nodes of vi, and maxDC is the maximum value of the degree of all power grid nodes in the power grid system.

2. The method for identifying key nodes in a power grid system according to claim 1, characterized in that, The identification of key nodes based on their influence in the network is specifically as follows: key nodes in the network are identified based on the influence value of each node in the network, with nodes having a larger influence value having a greater influence than nodes having a smaller influence value.

Images