CN113722868A  Multiindex power grid node vulnerability assessment method fusing structure hole characteristics  Google Patents
Multiindex power grid node vulnerability assessment method fusing structure hole characteristics Download PDFInfo
 Publication number
 CN113722868A CN113722868A CN202111029535.8A CN202111029535A CN113722868A CN 113722868 A CN113722868 A CN 113722868A CN 202111029535 A CN202111029535 A CN 202111029535A CN 113722868 A CN113722868 A CN 113722868A
 Authority
 CN
 China
 Prior art keywords
 node
 nodes
 power grid
 index
 sorting
 Prior art date
 Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
 Granted
Links
 238000000034 method Methods 0.000 title claims abstract description 22
 230000006870 function Effects 0.000 claims abstract description 31
 238000011156 evaluation Methods 0.000 claims abstract description 21
 238000004422 calculation algorithm Methods 0.000 claims abstract description 20
 238000012163 sequencing technique Methods 0.000 claims abstract description 15
 239000013598 vector Substances 0.000 claims abstract description 12
 238000011478 gradient descent method Methods 0.000 claims abstract description 6
 238000003062 neural network model Methods 0.000 claims abstract description 4
 238000013528 artificial neural network Methods 0.000 claims description 6
 238000010276 construction Methods 0.000 claims description 3
 238000002790 crossvalidation Methods 0.000 claims description 3
 230000004927 fusion Effects 0.000 description 3
 238000004364 calculation method Methods 0.000 description 2
 241001391944 Commicarpus scandens Species 0.000 description 1
 230000002776 aggregation Effects 0.000 description 1
 238000004220 aggregation Methods 0.000 description 1
 230000002146 bilateral effect Effects 0.000 description 1
 230000005540 biological transmission Effects 0.000 description 1
 230000000694 effects Effects 0.000 description 1
 230000005611 electricity Effects 0.000 description 1
 238000002372 labelling Methods 0.000 description 1
 238000004088 simulation Methods 0.000 description 1
Images
Classifications

 G—PHYSICS
 G06—COMPUTING; CALCULATING OR COUNTING
 G06F—ELECTRIC DIGITAL DATA PROCESSING
 G06F30/00—Computeraided design [CAD]
 G06F30/10—Geometric CAD
 G06F30/18—Network design, e.g. design based on topological or interconnect aspects of utility systems, piping, heating ventilation air conditioning [HVAC] or cabling

 G—PHYSICS
 G06—COMPUTING; CALCULATING OR COUNTING
 G06F—ELECTRIC DIGITAL DATA PROCESSING
 G06F30/00—Computeraided design [CAD]
 G06F30/20—Design optimisation, verification or simulation
 G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model

 G—PHYSICS
 G06—COMPUTING; CALCULATING OR COUNTING
 G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
 G06N3/00—Computing arrangements based on biological models
 G06N3/02—Neural networks
 G06N3/08—Learning methods

 G—PHYSICS
 G06—COMPUTING; CALCULATING OR COUNTING
 G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
 G06Q10/00—Administration; Management
 G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
 G06Q10/063—Operations research, analysis or management
 G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
 G06Q10/06393—Scorecarding, benchmarking or key performance indicator [KPI] analysis

 G—PHYSICS
 G06—COMPUTING; CALCULATING OR COUNTING
 G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
 G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
 G06Q50/06—Energy or water supply

 H—ELECTRICITY
 H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
 H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
 H02J3/00—Circuit arrangements for ac mains or ac distribution networks

 G—PHYSICS
 G06—COMPUTING; CALCULATING OR COUNTING
 G06F—ELECTRIC DIGITAL DATA PROCESSING
 G06F2111/00—Details relating to CAD techniques
 G06F2111/02—CAD in a network environment, e.g. collaborative CAD or distributed simulation

 G—PHYSICS
 G06—COMPUTING; CALCULATING OR COUNTING
 G06F—ELECTRIC DIGITAL DATA PROCESSING
 G06F2113/00—Details relating to the application field
 G06F2113/04—Power grid distribution networks

 H—ELECTRICITY
 H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
 H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
 H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
 H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]

 Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSSSECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSSREFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
 Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
 Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
 Y04S10/00—Systems supporting electrical power generation, transmission or distribution
 Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the loadside end user applications
Landscapes
 Engineering & Computer Science (AREA)
 Business, Economics & Management (AREA)
 Physics & Mathematics (AREA)
 Theoretical Computer Science (AREA)
 Human Resources & Organizations (AREA)
 General Physics & Mathematics (AREA)
 Evolutionary Computation (AREA)
 Economics (AREA)
 Geometry (AREA)
 Health & Medical Sciences (AREA)
 General Engineering & Computer Science (AREA)
 Strategic Management (AREA)
 Entrepreneurship & Innovation (AREA)
 General Health & Medical Sciences (AREA)
 General Business, Economics & Management (AREA)
 Marketing (AREA)
 Tourism & Hospitality (AREA)
 Development Economics (AREA)
 Artificial Intelligence (AREA)
 Software Systems (AREA)
 Educational Administration (AREA)
 Computer Hardware Design (AREA)
 Data Mining & Analysis (AREA)
 Operations Research (AREA)
 Molecular Biology (AREA)
 Computing Systems (AREA)
 Biophysics (AREA)
 Mathematical Physics (AREA)
 Biomedical Technology (AREA)
 Life Sciences & Earth Sciences (AREA)
 Primary Health Care (AREA)
 Water Supply & Treatment (AREA)
 Game Theory and Decision Science (AREA)
 Computational Linguistics (AREA)
 Quality & Reliability (AREA)
 Computer Networks & Wireless Communication (AREA)
 Computational Mathematics (AREA)
 Public Health (AREA)
 Mathematical Analysis (AREA)
 Mathematical Optimization (AREA)
Abstract
A multiindex power grid node vulnerability assessment method fusing structural hole characteristics is characterized in that node degree and clustering coefficient indexes are selected by considering topological characteristics; selecting lowpower risk factors, node voltage outoflimit risks and node electrical betweenness indexes by considering electrical performance; selecting a loss load index in consideration of the cascading failure result; and selecting PageRank, LeaderRank and HITs indexes based on the sorting of the feature vectors, and comprehensively sorting the power grid nodes by 9 evaluation indexes from the four aspects. Constructing a power grid key node sorting algorithm considering the structural hole characteristics based on ListNet sorting learning, firstly randomly generating initial weight coefficient values of all node indexes, and obtaining initial scores of all nodes by using a scoring function h (x); and then substituting the index weight coefficient into a linear neural network model, updating the index weight coefficient by a gradient descent method, controlling circulation by using iteration times until the obtained index weight coefficient value reaches a stable state, outputting the weight coefficient value at the moment, and finally obtaining the scores of all nodes of the power grid according to a scoring function h (x) and sequencing.
Description
Technical Field
The invention relates to the field of safe and stable operation of a power system, in particular to a multiindex power grid node vulnerability assessment method fusing structure hole characteristics.
Background
With the increasing demand of electricity, the largescale interconnection of power grids is continuously developed, and the power grids become one of the most complex networks in the world. In recent years, major power failure accidents frequently occurring around the world bring huge economic loss and serious influence to society. Often, a power failure accident is caused by a failure of one element in the grid system, which results in the power flowing through the element being transferred to other places, so that the rest of the series of elements in the system are overloaded and quit working, and finally, the grid is crashed. Therefore, the identification of the vulnerable node, namely the node which is easy to break down, in the power grid has important significance for ensuring the safe and stable operation of the power system.
At present, identification of some key nodes in a power grid is carried out mainly on the basis of a complex network theory, and the importance of the power grid nodes is evaluated by methods such as a smallworld model, electric betweenness and degree in combination with characteristics of a power system. Meanwhile, the importance degree of the nodes is measured by introducing internet webpage algorithms such as PageRank, LeaderRank and HITs into a power grid. The comprehensive evaluation method overcomes the limitation of single index evaluation to a certain extent, but the selection of the indexes in the method lacks factors with structural hole characteristics and cannot completely reflect the actual vulnerability of the power grid.
Disclosure of Invention
At present, sequencing research on key nodes is almost focused on core nodes of a power grid, the nodes at the positions of structural holes are often ignored, and meanwhile, the selection of a single index for sequencing the power grid nodes has great limitation, so that the invention provides a multiindex power grid node vulnerability assessment method fusing structural hole characteristics.
The invention selects 9 indexes from four aspects of topological characteristics, electrical performance, cascading failure consequences and sorting based on characteristic vectors, and sorts the nodes of a power grid comprehensively by using index weight coefficients of all nodes obtained by neural network training and learning and based on a ListNet sorting learning algorithm, and the method comprises the following steps:
step 1: selecting a power grid node index; and respectively selecting 9 index values from the four aspects of topological characteristics, electrical performance, cascading failure consequences and feature vectorbased sequencing as learning features.
1) Aspect of topological characteristics
Node degree: the degree is the total number of all edges on the node, can be used for measuring the connectivity of the node in the network, and can describe partial characteristics of the node of the structural hole;
clustering coefficient: one of the important weighting coefficients used for representing the clustering degree of the network nodes, and the positions of the structural hole nodes in the network are more special, so that the nodes with small clustering coefficients are more likely to become the structural hole nodes.
2) Aspect of electrical performance
Low power risk factor: reflecting the probability of power factor reduction in the power grid and the severity of risk consequences, and quantitatively measuring the operation state of the microgrid at a certain future moment;
node voltage outoflimit risk index: reflecting the possibility and harm of overhigh bus voltage in the system after the power system fails;
node electrical permittivity: the ratio of the number of paths passing through the node to the total number of paths between the two nodes that are the most efficient or the shortest, among the most efficient or shortest paths between the two nodes in the network.
3) Consequences of cascading failures
Load loss: and defining the difference value between the initial value and the load after the load is cut off as the loss load quantity of the power system fault, wherein the loss load can be used for measuring the severity of the node after the fault occurs.
4) Ordering aspect based on feature vectors
PageRank index: the method comprises the steps that a PageRank algorithm is applied to carry out key node sequencing on a power grid on the basis of constructing a power grid directed graph, and nodes with higher PR scores can be presumed to be structure hole nodes more probably according to a PR algorithm principle;
LeaderRank index: the method is an improved algorithm of the PageRank algorithm, a background node is added in a network, the node and all nodes in the network are in bilateral contact to obtain a new network of nodes in strong connection, and then node sequencing is obtained according to PageRank calculation;
HITs index: the authority of the node and the pivot in the network information flow transmission are considered, and more electrical characteristics of the power grid are considered conveniently.
Step 2: learning the index values of the 9 nodes by a ListNet sorting learning algorithm to obtain weight coefficients, and performing comprehensive sorting on the final weight coefficient values obtained by the weight coefficients through cross validation.
In the above method for evaluating vulnerability of multiindex power grid nodes with fusion structure hole characteristics, the concrete model in step 2 is established:
step 21: inputting training data, randomly generating a weight coefficient of each index, and constructing a scoring function h (x), wherein the scoring function of the ith node is h (x)_{i})＝<θ_{i},x_{i}>1,2,3,. eta., m, vector x_{i}＝{x_{1},x_{2},...,x_{9}Denotes 9 evaluation index values corresponding to one node, vector θ_{i}＝{θ_{1},θ_{2},...,θ_{9}The weighting coefficients corresponding to the evaluation indexes of the nodes are obtained, m is the total node number of the power grid, and the score of the node i is h (x)_{i}) Represents;
h(x_{i})＝θ_{i}·x_{i},i＝1,2,...m (1)
in the formula, theta_{i}Is a set of weight coefficients, x, for node i_{i}9 evaluation index value sets corresponding to the node i;
step 22: measuring each of power grids by introducing Luce model through ListNet algorithmProbability P of node ordering mode_{h(x)}Expressing any sorting mode of the power grid nodes as a probability value, and expressing the probability of the sequence by using the sorting probability of the top k items of the scoring function, as shown in formula (2):
in the formula, m represents the total node number of the power grid, and j represents the serial number of the node;
the nodes in the training data are ranked, labeled and scored, and a Luce model is applied to obtain the ranking probability P of the ranked and labeled scores_{y}：
Where m represents the total number of nodes of the grid, y_{j}And y_{k}The ranking label scores of the nodes j and k respectively;
step 23: respectively obtaining probability distribution of the sequences by scoring a scoring function h (x) and a sequencing annotation through a Luce model, and applying cross entropy to measure similarity construction loss function L (y, h (x)) of the probability distribution of the two types of sequences as follows:
step 24: optimizing index weight coefficients by a gradient descent method, and calculating the gradient delta theta of the loss function L (y, h (x)) to theta:
step 25: updating each index weight coefficient until the obtained weight coefficient value reaches a stable state, outputting the weight coefficient value theta' at the moment, and obtaining the final score of the node through a scoring function:
θ'＝θη·Δθ (6)
in the formula, eta is the learning step length in the neural network; updating the weighting coefficients, the final scoring function is obtained by equation (1):
h(x_{i})＝θ'_{i}·x_{i},i＝1,2,...,m (7)
and obtaining the final score of the nodes through the finally updated weight coefficient to realize the sequencing of the power grid nodes.
Drawings
FIG. 1 is a basic flow chart of a multiindex power grid node vulnerability assessment method fusing structure hole characteristics;
FIG. 2 is a power grid key node sorting algorithm based on ListNet sorting learning considering structural hole characteristics;
fig. 3 is a linear neural network model.
Detailed Description
The invention is further described below with reference to the figures and examples.
Referring to fig. 1, a multiindex power grid node vulnerability assessment method fusing structure hole characteristics includes the following steps:
step 1: selecting a power grid node vulnerability assessment index fused with the structural hole characteristics;
in order to enable the ordering of the power grid nodes to be more reasonable and incorporate indexes of structural hole characteristics, node degree and clustering coefficient indexes are selected by considering topological characteristics; selecting lowpower risk factors, node voltage outoflimit risks and node electrical betweenness indexes by considering electrical performance; selecting a loss load index in consideration of the cascading failure result; and selecting PageRank, LeaderRank and HITs indexes based on the sorting of the feature vectors, and comprehensively sorting the power grid nodes by 9 evaluation indexes from the four aspects.
Step 2: referring to fig. 2, the weight coefficient of each index is randomly given, and the score function is used to obtain the initial score of each node; then the weight coefficients are brought into a neural network model and updated by a gradient descent method; the scores of all nodes at the moment are obtained through cross validation until the output index weight coefficient reaches a stable state, so that a power grid key node sorting algorithm considering the structural hole characteristics based on ListNet sorting learning is constructed;
step 21: referring to fig. 2 and 3, training data is input, a weight coefficient of each index is randomly generated, a scoring function h (x) is constructed, and the scoring function of the ith node is h (x)_{i})＝<θ_{i},x_{i}>1,2,3,. eta., m, vector x_{i}＝{x_{1},x_{2},...,x_{9}Denotes 9 evaluation index values corresponding to one node, vector θ_{i}＝{θ_{1},θ_{2},...,θ_{9}The weighting coefficients corresponding to the evaluation indexes of the nodes are obtained, and m is the total number of the nodes of the power grid; the score of node i is determined by the value of its scoring function, which is h (x)_{i}) Represents:
h(x_{i})＝θ_{i}·x_{i},i＝1,2,...m (1)
in the formula, theta_{i}Is a set of weight coefficients, x, for node i_{i}9 evaluation index value sets corresponding to the node i;
step 22: measuring probability P of each node sequencing mode in power grid by introducing ListNet algorithm into Luce model_{h(x)}Expressing any sorting mode of the power grid nodes as a probability value, and expressing the probability of the sequence by using the sorting probability of the top k items of the scoring function, as shown in formula (2):
in the formula, m represents the total node number of the power grid, and j represents the serial number of the node;
the nodes in the training data are ranked, labeled and scored, and a Luce model is applied to obtain the ranking probability P of the ranked and labeled scores_{y}：
Where m represents the total number of nodes of the grid, y_{j}And y_{k}The ranking label scores of the nodes j and k respectively;
step 23: respectively obtaining probability distribution of the sequences by scoring a scoring function h (x) and a sequencing annotation through a Luce model, and applying cross entropy to measure similarity construction loss function L (y, h (x)) of the probability distribution of the two types of sequences as follows:
step 24: optimizing index weight coefficients by a gradient descent method, and calculating the gradient delta theta of the loss function L (y, h (x)) to theta:
step 25: updating each index weight coefficient until the obtained weight coefficient value reaches a stable state, outputting the weight coefficient value theta' at the moment, and obtaining the final score of the node through a scoring function:
θ'＝θη·Δθ (6)
in the formula, eta is the learning step length in the neural network; updating the weighting coefficients, the final scoring function is obtained by equation (1):
h(x_{i})＝θ'_{i}·x_{i},i＝1,2,...,m (7)
the scores of the nodes are obtained through the scoring function, the power grid nodes are sorted, and the specific flow of the power grid key node sorting algorithm considering the structural hole characteristics based on ListNet sorting learning is shown in Table 1.
Table 1 power grid key node sorting algorithm considering structure hole characteristics based on ListNet sorting learning
As can be seen from fig. 3 and table 1, the linear neural network in the algorithm has 9 index values of the grid nodes, a node sorting and labeling score value y, an iteration number T, and a learning step length η as inputs, and has a model weight coefficient θ obtained based on neural network training and learning as an output.
Examples
In order to verify the effectiveness of the proposed multiindex power grid node vulnerability assessment method fusing the structure hole characteristics, the embodiment selects an IEEE118 node system as an example to complete simulation calculation on MATLAB software.
1) Selection of key nodes
Obtaining 9 index specific numerical values of each node of the IEEE118 node system based on MATLAB, evaluating and scoring each node of the IEEE118 node system by the multiindex power grid node vulnerability evaluation method of the fusion structure hole characteristics, selecting nodes with the top ten points of scoring as key nodes of the IEEE118 node system, wherein the key node data of the IEEE118 node system is shown in table 2.
TABLE 2 IEEE118 node System Key node data
As can be seen from table 2, the highest score is 49 nodes by the total evaluation of 9 indexes, that is, 49 nodes are also the most critical nodes. In order to verify the effect of the comprehensive evaluation result and the single index ordering, the evaluation result is compared with a power grid node importance evaluation method of an electric leader rank (elr) algorithm and a power grid node importance evaluation method of an improved HITs algorithm, and the result is shown in table 3.
TABLE 3 IEEE118 System first 10 key nodes three methods contrast value
As can be seen from table 3, the first ten key nodes of the IEEE118 node system obtained by the three different methods are different, and the obtained key node sequences are different because the factors considered by the three methods are different, but the first ten key nodes obtained by the three methods all have nodes No. 12, 49, 80, 92 and 100, which indicates that the proposed multiindex power grid node vulnerability assessment method for the fusion structure hole feature has general applicability to the identification of the power grid key nodes.
2) Key node and common node comparison
For the obtained key nodes of the IEEE118 node system, each index of the key node is analyzed in combination with table 2, and the comparison between the key node and each index mean of all nodes is shown in table 4.
TABLE 4 comparison of key nodes with the mean values of all nodes
As can be seen from table 4, the mean values of the node degree mean value, the electrical medium mean value, the loss load mean value, the LR value, the PR value, and the HITs value of the key nodes are larger than the mean values of all the nodes; the clustering coefficients of the key nodes are lower than the overall clustering coefficient, and the clustering coefficients can represent the aggregation degree of the nodes in the power grid, which shows that the clustering coefficient indexes have little influence on the result of the sequencing model; the lowpower risk factor of the key nodes is lower than the average value of all the nodes, and the lower the index is, the more reactive loads exist in the key nodes of the power grid, so that the operation state and the power supply quality of the power grid are influenced, and therefore the nodes are more prone to faults in the power grid, and further cascading faults occur. The voltage outoflimit risk index of the key node is lower than the average value of all nodes, the smaller the index value is, the more stable the node is, the less fault is easy to occur, namely the node is less important, and therefore the voltage outoflimit risk index has little influence on the comprehensive score of the sequencing model.
For the key nodes of the IEEE118 node system, data of most indexes of the key nodes are larger than those of other common nodes, so that the important positions of the key nodes in the whole node system are well illustrated, and meanwhile, the key nodes play an important role in a power grid, so that the key nodes are prone to faults in the actual power grid operation process, cascading faults are caused, the power grid is finally broken down, and a major power failure accident is caused. In order to prevent the occurrence of grid cascading failures, key nodes of a power grid are subjected to important protection, which is necessary for preventing a blackout accident of the power grid.
Claims (2)
1. A multiindex power grid node vulnerability assessment method fusing structure hole characteristics comprises the following steps:
step 1: selecting a power grid node vulnerability assessment index fused with the structural hole characteristics;
in order to enable the ordering of the power grid nodes to be more reasonable and incorporate indexes of structural hole characteristics, node degree and clustering coefficient indexes are selected by considering topological characteristics; selecting lowpower risk factors, node voltage outoflimit risks and node electrical betweenness indexes by considering electrical performance; selecting a loss load index in consideration of the cascading failure result; and selecting PageRank, LeaderRank and HITs indexes based on the sorting of the feature vectors, and comprehensively sorting the power grid nodes by 9 evaluation indexes from the four aspects.
Step 2: constructing a power grid key node sorting algorithm based on ListNet sorting learning and considering the characteristics of the structural holes;
randomly giving the weight coefficient of each index, and obtaining the initial score of each node by using a scoring function; substituting the index weight coefficients into a neural network model and updating each index weight coefficient by using a gradient descent method; and performing cross validation until the output index weight coefficient reaches a stable state to obtain the score of each node at the moment, thereby realizing the key node sorting algorithm of the power grid.
2. The method for evaluating the vulnerability of multiindex grid nodes fusing structure hole characteristics according to claim 1, wherein in step 2 of claim 1, a grid key node sorting algorithm considering structure hole characteristics based on ListNet sorting learning is constructed, and the steps are as follows:
step 21: randomly generating initial weight coefficient values of evaluation indexes of all nodes, and constructing a scoring function h (x), wherein the scoring function of the ith node is h (x)_{i})＝<θ_{i},x_{i}>I 1,2,3, …, m, vector x_{i}＝{x_{1},x_{2},...,x_{9}Denotes 9 evaluation index values corresponding to one node, vector θ_{i}＝{θ_{1},θ_{2},...,θ_{9}The weighting coefficients corresponding to all evaluation indexes of the nodes are obtained, m is the total node number of the power grid, the score of the node i is determined by the value of a scoring function of the node i, and h (x) is used_{i}) Represents:
h(x_{i})＝θ_{i}·x_{i},i＝1,2,…m (1)
in the formula, theta_{i}Is a set of weight coefficients, x, for node i_{i}9 evaluation index value sets corresponding to the node i;
step 22: measuring probability P of each node sequencing mode in power grid by introducing ListNet algorithm into Luce model_{h(x)}Expressing any sorting mode of the power grid nodes as a probability value, and expressing the probability of the sequence by using the sorting probability of the top k items of the scoring function, as shown in formula (2):
in the formula, m represents the total node number of the power grid, and j represents the serial number of the node;
the nodes in the training data are ranked, labeled and scored, and a Luce model is applied to obtain the ranking probability P of the ranked and labeled scores_{y}：
Where m represents the total number of nodes of the grid, y_{j}And y_{k}The ranking label scores of the nodes j and k respectively;
step 23: respectively obtaining probability distribution of the sequences by scoring a scoring function h (x) and a sequencing annotation through a Luce model, and applying cross entropy to measure similarity construction loss function L (y, h (x)) of the probability distribution of the two types of sequences as follows:
step 24: optimizing index weight coefficients by a gradient descent method, and calculating the gradient delta theta of the loss function L (y, h (x)) to theta:
step 25: updating each index weight coefficient until the obtained weight coefficient value reaches a stable state, outputting the weight coefficient value theta' at the moment, and obtaining the final score of the node through a scoring function:
θ'＝θη·Δθ (6)
in the formula, eta is the learning step length in the neural network; updating the weighting coefficients, the final scoring function is obtained by equation (1):
h(x_{i})＝θ′_{i}·x_{i},i＝1,2,…,m (7)
and obtaining the final score of the nodes through the finally updated weight coefficient to realize the sequencing of the power grid nodes.
Priority Applications (1)
Application Number  Priority Date  Filing Date  Title 

CN202111029535.8A CN113722868B (en)  20210903  20210903  Multiindex power grid node vulnerability assessment method integrating structural hole characteristics 
Applications Claiming Priority (1)
Application Number  Priority Date  Filing Date  Title 

CN202111029535.8A CN113722868B (en)  20210903  20210903  Multiindex power grid node vulnerability assessment method integrating structural hole characteristics 
Publications (2)
Publication Number  Publication Date 

CN113722868A true CN113722868A (en)  20211130 
CN113722868B CN113722868B (en)  20240326 
Family
ID=78681350
Family Applications (1)
Application Number  Title  Priority Date  Filing Date 

CN202111029535.8A Active CN113722868B (en)  20210903  20210903  Multiindex power grid node vulnerability assessment method integrating structural hole characteristics 
Country Status (1)
Country  Link 

CN (1)  CN113722868B (en) 
Cited By (2)
Publication number  Priority date  Publication date  Assignee  Title 

CN114598612A (en) *  20220318  20220607  西华大学  Cascading failure model of power communication convergence network and node vulnerability assessment method 
CN117172286A (en) *  20230905  20231205  岭南师范学院  Multilayer fusion network key node identification method based on improved structure hole 
Citations (5)
Publication number  Priority date  Publication date  Assignee  Title 

WO2016026355A1 (en) *  20140818  20160225  国家电网公司  Voltage sag simulation and evaluation method of active power distribution grid 
CN107871216A (en) *  20171220  20180403  广东电网有限责任公司惠州供电局  A kind of recognition methods of power distribution network fragility node 
CN108667005A (en) *  20180413  20181016  湘潭大学  A kind of quiet dynamic bind vulnerability assessment method of power grid counted and new energy influences 
CN110020815A (en) *  20190507  20190716  云南电网有限责任公司  A kind of comprehensive vulnerability inder calculation method of the grid nodes based on analytic network process 
AU2020103195A4 (en) *  20201103  20210114  East China University Of Science And Technology  A Method for Detecting Vulnerability of Largescale Power Grid Based On Complex Network 

2021
 20210903 CN CN202111029535.8A patent/CN113722868B/en active Active
Patent Citations (5)
Publication number  Priority date  Publication date  Assignee  Title 

WO2016026355A1 (en) *  20140818  20160225  国家电网公司  Voltage sag simulation and evaluation method of active power distribution grid 
CN107871216A (en) *  20171220  20180403  广东电网有限责任公司惠州供电局  A kind of recognition methods of power distribution network fragility node 
CN108667005A (en) *  20180413  20181016  湘潭大学  A kind of quiet dynamic bind vulnerability assessment method of power grid counted and new energy influences 
CN110020815A (en) *  20190507  20190716  云南电网有限责任公司  A kind of comprehensive vulnerability inder calculation method of the grid nodes based on analytic network process 
AU2020103195A4 (en) *  20201103  20210114  East China University Of Science And Technology  A Method for Detecting Vulnerability of Largescale Power Grid Based On Complex Network 
NonPatent Citations (4)
Title 

崔力民;李刚;王天军;石江红;宋雨;: "智能电网信息通信系统风险评估的模糊测度方法", 电力建设, no. 05 * 
白加林;刘天琪;曹国云;陈陈;: "电力系统脆弱性评估方法综述", 电网技术, no. 2 * 
贺海磊;周勤勇;张彦涛;赵珊珊;梁才浩;: "考虑自然灾害的电网规划方案暂态风险评估", 陕西电力, no. 06 * 
韩忠明;吴杨;谭旭升;段大高;杨伟杰;: "面向结构洞的复杂网络关键节点排序", 物理学报, no. 05 * 
Cited By (4)
Publication number  Priority date  Publication date  Assignee  Title 

CN114598612A (en) *  20220318  20220607  西华大学  Cascading failure model of power communication convergence network and node vulnerability assessment method 
CN114598612B (en) *  20220318  20221206  西华大学  Cascading failure model of power communication convergence network and node vulnerability assessment method 
CN117172286A (en) *  20230905  20231205  岭南师范学院  Multilayer fusion network key node identification method based on improved structure hole 
CN117172286B (en) *  20230905  20240507  岭南师范学院  Multilayer fusion network key node identification method based on improved structure hole 
Also Published As
Publication number  Publication date 

CN113722868B (en)  20240326 
Similar Documents
Publication  Publication Date  Title 

EP3780307B1 (en)  Online dynamic decisionmaking method and system for unit restoration  
CN110034581B (en)  Interval electrical medium number vulnerability assessment method for power system under wind power gridconnected condition  
Liu et al.  Recognition and vulnerability analysis of key nodes in power grid based on complex network centrality  
CN107895947A (en)  A kind of power distribution network vulnerable line discrimination method  
CN113722868A (en)  Multiindex power grid node vulnerability assessment method fusing structure hole characteristics  
CN114665498A (en)  Active power distribution network fragile node identification method considering new energy influence  
CN110428005B (en)  Umbrella algorithmbased dynamic security misclassification constraint method for power system  
CN104992266A (en)  Method of determining power grid node importance degree and system thereof  
CN108767848B (en)  Electric power system fragile line identification and cascading failure prevention method  
CN110350522B (en)  Electric power system fragile line identification method based on weighted H index  
Shuvro et al.  Predicting cascading failures in power grids using machine learning algorithms  
CN107622360A (en)  A kind of critical circuits recognition methods for considering subjective and objective factor  
CN104156774B (en)  A kind of electric power support method for considering adjacent system  
Zhang et al.  Vulnerable point identification using heterogeneous interdependent node theory for distribution systems  
CN111404163A (en)  Electromagnetic looped network openloop method  
CN116502380A (en)  Power grid transient state instability positioning method and system based on interpretable graph neural network  
CN115659833A (en)  Power network node vulnerability assessment method based on BP neural network  
Qu et al.  Electric power cyberphysical systems vulnerability assessment under cyber attack  
Sachan et al.  Multiple correspondence analysis to study failures in a diverse population of a cable  
CN113725855B (en)  Electric DebtRank algorithmbased electric power system fragile line identification method  
Gu et al.  Identification of vulnerable nodes in power grids based on graph deep learning algorithm  
CN110365006A (en)  A kind of subarea division method based on nwbbo algorithm  
US20240054259A1 (en)  Method for identifying fragile lines in power grids based on electrical betweenness  
Wang et al.  Network attack risk assessment of power grid CPS System based on attacker’s perspective  
Han et al.  A method for searching splitting surface considering network splitting adaptation index 
Legal Events
Date  Code  Title  Description 

PB01  Publication  
PB01  Publication  
SE01  Entry into force of request for substantive examination  
SE01  Entry into force of request for substantive examination  
GR01  Patent grant  
GR01  Patent grant 