Power information dependent system model construction method based on greedy structure entropy algorithm

The power information dependence system model is constructed through the greedy structural entropy algorithm, which solves the problem of instability in the description of information interaction and interdependence between power system nodes, achieves higher stability and robustness, and provides more effective optimization and management means.

CN119093376APending Publication Date: 2024-12-06BEIHUA UNIV
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411099848.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-12
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

The existing power information-dependent system network model operates unstable, making it difficult to accurately describe the information interaction and interdependence between power system nodes, affecting the robustness of the system.

Method used

The power information dependence system model is constructed based on the greedy structure entropy algorithm, and the power grid and the information network are calculated by calculating the adjacency matrix and weight matrix of the power grid and the information network, and the initial structural entropy is calculated, and the greedy algorithm is used to adjust the dependence relationship between the power grid and the information network to optimize the network structure.

Benefits of technology

It improves the stability and robustness of the power information-dependent system, provides more effective optimization and management methods, and improves the overall performance of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119093376A_ABST
    Figure CN119093376A_ABST
Patent Text Reader

Abstract

The invention discloses an electric power information dependent system model construction method based on a greedy structure entropy algorithm, and belongs to the field of electric power information dependency, and the method comprises the following steps: constructing a power grid model based on a plurality of stations; constructing an information network model based on the communication devices of the plurality of stations and the communication routes of the communication devices; a greedy structure entropy algorithm is adopted to construct a power information dependent system model based on the power grid model and the information network model; wherein the process of constructing the electric power information dependent system model comprises the following steps of: respectively calculating an adjacent matrix and a weight matrix of a power grid model and an information network model; calculating a power grid initial structure entropy based on the adjacency matrix and the weight matrix of the power grid model; calculating an information network initial structure entropy based on the adjacency matrix and the weight matrix of the information network model; based on the power grid initial structure entropy and the information network initial structure entropy, a greedy algorithm is adopted to adjust the dependency relationship between the power network and the information network to obtain a power information dependency system model.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of power information dependency, and in particular relates to a method for constructing a power information dependency system model based on a greedy structural entropy algorithm. Background Art

[0002] As one of the indispensable infrastructures in modern society, the stability and robustness of the operation of the power system plays a vital role in the entire society. With the continuous expansion of the scale and increase in complexity of smart grids, the accurate description of the interaction relationship between nodes within the power system becomes more and more important. The network model of the power information dependency system can better understand the information interaction and interdependence between nodes in the power system. However, the operation of the network model of the power information dependency system in the prior art is not stable. Therefore, in order to better understand and improve the robustness of the power information dependency system, the present invention proposes a method for constructing a power information dependency system model based on a greedy structural entropy algorithm. Summary of the invention

[0003] In order to solve the above technical problems, the present invention proposes a method for constructing a power information dependent system model based on a greedy structural entropy algorithm to solve the problems existing in the above prior art.

[0004] To achieve the above object, the present invention provides a method for constructing a power information dependent system model based on a greedy structural entropy algorithm, comprising:

[0005] Build a power grid model based on several plants and stations;

[0006] Building an information network model based on the communication equipment of the plant and the communication routes of the communication equipment;

[0007] Based on the power grid model and the information network model, a greedy structure entropy algorithm is used to construct a power information dependency system model;

[0008] The process of constructing the power information dependent system model includes:

[0009] Calculate the adjacency matrix and weight matrix of the power grid model and the information network model respectively;

[0010] Calculating the initial structural entropy of the power grid based on the adjacency matrix and the weight matrix of the power grid model;

[0011] Calculating the initial structural entropy of the information network based on the adjacency matrix and weight matrix of the information network model;

[0012] Based on the initial structural entropy of the power grid and the initial structural entropy of the information network, a greedy algorithm is used to adjust the interdependence between the power grid and the information network to obtain a power information interdependence system model.

[0013] Optionally, the power grid model includes a set of power grid nodes and a set of edges connecting power grid nodes;

[0014] Wherein, the grid node set includes a number of plants and stations;

[0015] The grid node connection edge set is the high-voltage transmission lines between various power plants and stations.

[0016] Optionally, the process of building the power grid model includes:

[0017] Define the power injected into the grid node;

[0018] Establish power balance equation for each grid node;

[0019] The power balance equation is iteratively solved using the Newton-Raphson method to obtain the power flow distribution of each grid node;

[0020] Determining node voltage, phase angle, active power, and reactive power based on the power flow distribution of each grid node;

[0021] A power grid model is established based on the active power.

[0022] Optionally, the power balance equation is expressed as:

[0023]

[0024] Where P i Indicates active power, Q i represents reactive power, i represents the plant node, j represents the imaginary unit, e i represents the real part of the voltage at node i, f i Represents the imaginary part of voltage, G i represents the real part of admittance, B i represents the imaginary part of the admittance, a i represents the current injection amount at node i, b i Represents the amount of current injected at node i.

[0025] Optionally, the expression of the power grid model is:

[0026] P physical =P+diag(L)

[0027] In the formula, P represents the active power between each branch, diag(L) is a vector L with the elements L i is a diagonal matrix with diagonal elements, P physical represents the power flow matrix.

[0028] Optionally, the information network model includes an information network node set and an information network connection edge set;

[0029] Wherein, the information network node set is a communication device;

[0030] The information network connection edge set is a set of communication lines between nodes.

[0031] Optionally, the information network model is constructed using a BA scale-free network.

[0032] Optionally, the expression for calculating the initial structural entropy of the power grid is:

[0033]

[0034] In the formula, u i is a node in the power grid; L p is the threshold of electrical distance; |Z| ij is node u i With node u j The electrical distance between Indicates that at the threshold L p Internal node u i The set of nodes that can be reached, Indicates that at the threshold L p Node u in range i The number of reachable nodes; α is defined as the inverse of the minimum electrical distance between nodes.

[0035] Optionally, the expression for calculating the initial structural entropy of the information network is:

[0036]

[0037] Where, L c is the shortest path threshold, d ij For node v i With node v j The shortest path distance between For node v i At the threshold L c The set of nodes reachable within For node v i At the threshold L c The number of nodes that can be reached within α is the reciprocal of the minimum value of the shortest path between nodes.

[0038] Compared with the prior art, the present invention has the following advantages and technical effects:

[0039] The present invention discloses a method based on greedy structural entropy to construct a power information dependency system, so as to better understand the relationship between nodes within the power system and provide a more effective means for optimizing and managing the power system. The power information dependency system model constructed by the present invention exhibits higher stability and excellent robustness. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0041] Figure 1 An equivalent diagram of power network components and lines based on a greedy structural entropy algorithm according to an embodiment of the present invention;

[0042] Figure 2 IEEE30 node topology diagram of an embodiment of the present invention;

[0043] Figure 3 A distribution diagram of the degree of each node in a scale-free network according to an embodiment of the present invention;

[0044] Figure 4 A probability distribution diagram of node degrees of a scale-free network according to an embodiment of the present invention;

[0045] Figure 5 is the structural entropy of the IEEE30 system node in the embodiment of the present invention;

[0046] Figure 6 The IEEE30 system dependency network visualization model of the embodiment of the present invention;

[0047] Figure 7 The IEEE30 dependent system network efficiency of the embodiment of the present invention;

[0048] Figure 8 is the remaining size of the IEEE30 dependent system network according to an embodiment of the present invention;

[0049] Fig. 9 is the IEEE30 dependent system network residual load of an embodiment of the present invention;

[0050] Fig.10 The network efficiency of the IEEE118 dependent system according to the embodiment of the present invention;

[0051] Fig.11 is the remaining size of the IEEE118 dependent system network according to an embodiment of the present invention;

[0052] Fig.12 It is the network residual load of the IEEE118 dependent system according to the embodiment of the present invention. DETAILED DESCRIPTION

[0053] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0054] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0055] Embodiment 1

[0056] In the present embodiment, a method for constructing a power information dependency system model based on a greedy structural entropy algorithm is provided, comprising the following steps: constructing a power grid model based on a number of plants and stations; constructing an information network model based on the communication equipment of the number of plants and stations and the communication routes of the communication equipment; constructing a power information dependency system model based on the power grid model and the information network model using a greedy structural entropy algorithm; wherein the process of constructing the power information dependency system model comprises: respectively calculating the adjacency matrix and weight matrix of the power grid model and the information network model; calculating the initial structural entropy of the power grid based on the adjacency matrix and weight matrix of the power grid model; calculating the initial structural entropy of the information network based on the adjacency matrix and weight matrix of the information network model; and adjusting the dependency relationship between the power grid and the information network using a greedy algorithm based on the initial structural entropy of the power grid and the initial structural entropy of the information network to obtain the power information dependency system model.

[0057] The specific implementation process is as follows:

[0058] The process of building a power grid model includes:

[0059] The power grid model includes a power grid node set and a power grid node connection edge set; wherein the power grid node set includes a number of power plants; the power grid node connection edge set is the high-voltage transmission lines between the power plants.

[0060] The research focus of the power network is on various types of power plants, including key facilities such as power plants and substations, and these power plants with different functions are uniformly regarded as homogeneous nodes in the network. The power grid has the same physical characteristics and operates around the distribution of power flow. In the power network, the high-voltage transmission lines between various power plants can be regarded as the connecting edges of the power grid. The present invention can simplify the power network into a node set V P and edge set E P The figure G P =(V P ,E P ), and the active power flow value of the power line is used as the weight reference of the initial network.

[0061] The present invention uses the IEEE standard node system to simulate the power grid model. The IEEE node model is equivalent to a standard test object, which is used to test the performance of different algorithms in different research directions. The data contains various steady-state data, and the results obtained are comparable. These data models are generally simplified models of a certain actual power system. Figure 1An equivalent schematic diagram showing the system's power network components and wiring.

[0062] Therefore, the present invention selects the IEEE 30-node system as the topological model of the power network, which includes 30 nodes, 6 generators, 18 power loads, 6 substations and 41 connection edges. Figure 2 This is the IEEE30 node topology diagram.

[0063] The process of building a power grid model includes: defining the power injected into the grid node; establishing a power balance equation for each grid node; using the Newton-Raphson method to iteratively solve the power balance equation to obtain the power flow distribution of each grid node; determining the node voltage, phase angle, active power, and reactive power based on the power flow distribution of each grid node; and establishing a grid model based on active power. The specific implementation process includes:

[0064] First, a power balance equation is established for each node to calculate the power flow distribution of the power grid. Assume that the injected power of node i is Active power is P i , the reactive power is Q i , by defining the power injected into the node:

[0065]

[0066] In the formula, j represents the imaginary unit, Indicates I i The complex conjugate of .

[0067] The power balance equation in rectangular coordinate form is:

[0068]

[0069] In the formula, e i and e j denote the real part of the voltage at nodes i and j, respectively, and f i Represents the imaginary part of voltage, G i represents the real part of admittance, B i represents the imaginary part of the admittance, G ij and B ij are the conductance and susceptance between nodes i and j, respectively, and a i represents the current injection amount at node i, b i Represents the amount of current injected at node i.

[0070] By using the Newton-Raphson method to solve formula (2), after the iteration is completed, the present invention can determine the power flow distribution of each node. According to the following formula, the power flow distribution of the transmission line can be calculated. After the iteration is completed, the power flow distribution of each node can be obtained.

[0071]

[0072] In the formula, and is the node voltage obtained in the last iteration, y i0 and j0 is the node self-guidance, y ij and ji is the node admittance, P ij and P ji represents the active power between nodes i and j, Q ij and Q ji represents the reactive power between nodes i and j, For U i The complex conjugate of For U j The complex conjugate of represents the transmission power from node i to node j.

[0073] The node voltage, phase angle, active power, reactive power, reactive power, and transmission power between transmission lines are determined by solving the above equations. In the architecture of the power information system, each node of the power grid is equipped with corresponding sensors and controllers, and each transmission line is equipped with switchgear, so that the information network can monitor and control all nodes and transmission lines of the power grid. This configuration allows real-time collection of power grid status data, analysis, and control decisions to ensure stable operation and optimized performance of the power system. Therefore, if active power data is selected to describe the power grid, formula (6) can be used to describe the power grid, thereby establishing a power grid model.

[0074] P physical =P+diag(L) (6)

[0075] In formula (6), the active power between each branch is P∈R n*n Indicates that diag(L) is a vector L with elements L in L i is a diagonal matrix of diagonal elements. physical The non-diagonal elements in represent the injected power of the node, and the diagonal elements represent the injected power of the node. If its active power P i >0 is the generator node, P i =0 is the contact node, P i <0 is a load node. Through the adjacency matrix Characterizes the grid structure of the power grid, where |N P | is the number of nodes in the power grid, and the off-diagonal element A ij ∈{0,1}, represents the connectivity status of the transmission line ij. If A ij =A ji =1, line ij is connected, if Aij =A ji = 0, line ij is disconnected, diagonal element A ii =0.

[0076]

[0077] For the sensors at each node of the power grid and the circuit breakers on each transmission line, define the matrix Indicates the information network control command received by the power grid. ij =M ji =1 means the circuit breaker is closed, i.e. the line is connected; M ij =M ji =0 means the circuit breaker is disconnected, that is, the line is in a disconnected state.

[0078] The process of building an information network model includes:

[0079] The information network model includes an information network node set and an information network connection edge set; wherein the information network node set is the communication equipment; and the information network connection edge set is the set of communication lines between each node.

[0080] The information network is responsible for collecting power flow data from various stations and equipment in the power grid to monitor the real-time operation status of the power grid. Based on this data and established control standards, the information network will generate various operating instructions to achieve effective control and management of the power grid. This mechanism ensures the stable operation of the power grid while optimizing its performance. By analyzing the power flow data, the information network can predict and respond to potential power grid problems, thereby improving the robustness of the entire system. The information network provides basic services such as relay protection, safety and stability control, so it must meet the following three requirements:

[0081] (1) Meet the reliability requirements of various power services;

[0082] (2) Under normal circumstances, any two network nodes in the information network must ensure that there are at least two independent physical paths to maintain connection;

[0083] (3) On the premise of satisfying the N-1 principle, the N-2 principle should be satisfied as much as possible, that is, the failure of any node in the network should be able to maintain the stable operation of the power business. When two nodes in the network are separated, it should not affect the normal communication of other nodes, and the communication quality degradation and channel congestion of other nodes should be avoided as much as possible.

[0084] Simplify the information network into a graph G consisting of a set of nodes and a set of edges C =(V C ,E C ), V CIt represents the set of nodes in the information network that provide control functions to the power grid, including communication devices in the information network, such as sensors, actuators, control centers, etc. C Represents the set of communication lines between nodes in the information network. The communication link connects the control nodes between them and the control center node (CC), which contains the data transmission medium in the information network. C The number of nodes in the control grid is limited by its size and topology. The minimum control network of the smart grid requires C to have at least one n i ∈V C To control each i∈V P A one-to-one interdependence model makes |V C |=|V P |. Due to the topological restrictions of the power grid, the topology of the control information network usually follows the topology of the power grid. The control packet depends on the control center (CC) and is sent to the control node n i ∈V C . In case of grid failure, the link The status of the CC determines the performance of the corrective actions taken by the CC.

[0085] Connected to the control node can be through the adjacency matrix Characterizes the grid structure of the information network, the non-diagonal element Q ij ∈{0,1}, indicating the connectivity status of the communication line ij. If Q ij =Q ji =1, line ij is connected, if Q ij =Q ji = 0, line ij is disconnected, diagonal element Q ii =0.

[0086] When studying the structure of power information-dependent systems, researchers pointed out that these networks mainly exhibit two types of topological forms. One of them is the binary star topology, which is a standard scale-free network characterized by an intuitive structure and all nodes directly connected to the central node, making it easy to troubleshoot problems at the connection points. However, if the central node encounters a problem, the entire network will face the risk of collapse. The scale-free feature is not only prevalent in many real-world networks, but also applies to information networks. The steps to generate a scale-free information network with N nodes are as follows:

[0087] Initial state: At the beginning, the network consists of a fully connected three-node network.

[0088] Growth process: gradually add new nodes to the network, one at a time. Each new node must establish two edges with any two nodes in the original network to maintain the connectivity of the network.

[0089] Selection of connection edges: The probability that a new node is connected to a node i in the original network is p i :

[0090]

[0091] In formula (8), k i is the degree of node i, ∑ j k j Represents the degree of the total nodes in the network.

[0092] By repeating this process, a scale-free information network with N nodes can be generated. In this network, the node with the highest degree is usually selected as the control center of the power information network. In view of the hub characteristics and importance of the control center, it should be equipped with an independent power supply and uninterruptible power supply system to ensure its continued normal operation. Therefore, in the information network, the control center node is usually not directly connected to any node in the power grid to maintain the stability and security of its operation.

[0093] The second type is the mesh topology. Unlike the binary star topology, this topology network is a typical small-world network with arbitrary nodes, complex structure, high system reliability, and each node is linked to multiple nodes. The degree and betweenness are evenly distributed. In real life, most of them are mesh-type. In the power information-dependent system, the stable operation of the power system depends on the management of the dispatching center. Usually, there are two dispatching centers, the main and the backup, to ensure the real-time transmission of instructions and communication. In the information network, the two nodes with the highest degree are selected to serve as the main dispatching center and the deputy dispatching center to ensure the efficiency of information transmission and the timely response of system instructions. Based on the relevant theories and construction methods of small-world networks and scale-free networks mentioned above, an information network is constructed based on a part of the data in northern Nanjing in China. The average degree <k> shows the average number of connections between nodes and their adjacent nodes in the network. A higher value indicates that the network is richer in information transmission paths; the clustering coefficient C measures the degree of concentration of network nodes to key nodes, that is, the closeness of the connection between nodes and the dispatching center. The larger the value, the closer the connection between nodes and central nodes; the average path L reflects the average distance between nodes in the network. A shorter path means higher information transmission efficiency. In the mesh information network and the dual-star information network, a higher clustering coefficient is conducive to strengthening the control of the control center node over the system and improving the efficiency of information transmission. The dual-star information network has a shorter average path length, which makes the transmission of information between each station and the dispatch center more efficient. Both the mesh information network and the dual-star information network meet the clustering coefficient C>>C random , L ≥ L random . C random and L random It represents the average path and clustering coefficient of a random network when the number of nodes and edges in the network is the same, where Lrandom ≈ <k> / N,C random ≈lnN / ln<K>≈lnN.

[0094] Table 1

[0095]

[0096]

[0097] The electric power information network has a complex structure, a large scale, and complex and diverse connection relationships, and has obvious scale-free characteristics. Therefore, the present invention constructs a BA scale-free network as the research object of the information network.

[0098] The process of building a power information dependency system includes:

[0099] When analyzing the power information interdependent system, one of the most important steps is to establish a network model of interdependence between the power grid and the information network. In the interdependent network, a single-layer network can be abstracted into a network graph G, that is, G = (V, E, W). V = {v 1 ,v 2 ,...,v n } represents the node set of the network; E = {e 1 ,e 2 ,...,e m } represents the edge set of the network; W represents the weight set of the edge.

[0100] The edges connecting different nodes in a network are called connecting edges. Dependency edges indicate that the function of a node in one network depends on the node of another network. Depending on their directionality, dependency edges are divided into undirected and directed types. The proportion of dependency edges reveals the number of interdependent nodes between different networks, which can be determined by evaluating the dependency strength of the dependent network. The dependency strength q indicates the proportion of nodes with a dependent relationship to the total number of nodes in the dependent network. When q=1, it means that there is a one-to-one dependency relationship between nodes in the network; when q=0, it means that there is no dependency between networks.

[0101] The adjacency matrix is ​​a two-dimensional square matrix used to describe the connection relationship between nodes in a graph. It is usually represented by a square matrix A, where the element a ij Indicates whether the nodes are connected. The adjacency matrix of graph G can be expressed as:

[0102]

[0103] In formula (9): Matrix A is an n-order square matrix, which represents the adjacency matrix. The matrix has symmetry. In such a matrix, the elements on the main diagonal are all 0, and the remaining elements are 1 (indicating that there is a connection between nodes) or 0 (indicating that there is no connection between nodes). Therefore, when storing the adjacency matrix, only the data of the upper triangular or lower triangular matrix needs to be retained. In the adjacency matrix A, each element a ij Defined as:

[0104]

[0105] In formula (10), (v i ,v j ) represents the node v in the graph G i and v j The connecting edges between .

[0106] The weight matrix is ​​similar to the adjacency matrix, which is also a two-dimensional square matrix used to represent the weights of each edge in the graph. This matrix is ​​usually represented by a square matrix W, where the element w ij Represents the weight of the edge. The weight matrix of graph G is expressed as:

[0107]

[0108] In formula (11), W is an n-order square matrix, and the element w in the weight matrix is ij It can be defined as between:

[0109]

[0110] In formula (12): When only the topological structure is considered, the vertex v i and v j The weight of the edge between is 1, so the adjacency matrix and the weight matrix will be the same. In a dependency network, the dependency edges are regarded as a special set of edges, forming a set E inter , which is expressed as:

[0111]

[0112] In formula (13): V A and V B They represent the node sets of two interdependent networks A and B respectively.

[0113] The dependency matrix is ​​a tool used to describe the interdependence between nodes in different networks. Take network A and network B as examples, the dependency matrix E between them is AB It is expressed as:

[0114]

[0115] In formula (14): E AB It is an n×m dimensional matrix, where n represents the number of nodes in network A and m represents the number of nodes in network B. When n=m, that is, the number of nodes in network A and network B is the same, it is a square matrix. Dependency matrix E AB The element e in ij It can be defined as:

[0116]

[0117] In formula (15): (v Ai ,v Bj ) represents the dependency edge between the i-th node in network A and the j-th node in network B.

[0118] The study shows that the power network and the information network show significant similarities in topological structure. By applying the Pearson correlation coefficient method, the degree and betweenness of the IEEE30 nodes in the two networks can be calculated, and then the similarity between the power network and the information network can be evaluated by analyzing the correlation of these parameters.

[0119] The Pearson correlation coefficient measures the linear correlation between two variables, usually represented by the symbol ρ, and its value range is between -1 and 1. The larger the absolute value of ρ, the stronger the linear correlation. ρ>0 means there is a positive correlation between the variables, and ρ<0 means there is a negative correlation between the variables. The calculation formula of the Pearson correlation coefficient ρ between variables X and Y is:

[0120]

[0121] In formula (16): and They represent the mean values ​​of samples of variable X and variable Y respectively.

[0122] According to the data in Table 2, it can be concluded that the topological structure similarity between the IEEE 30-node power network and the information network is very high.

[0123] Table 2

[0124]

[0125]

[0126] In the power information dependent network, not only the properties of its own network structure but also the properties of the dependent network structure should be considered. The dependency analysis method based on degree centrality mainly focuses on the connection relationship between nodes and their adjacent nodes. This method is simple and intuitive, but it ignores the influence of non-adjacent nodes and only uses limited network topology information. In contrast, the dependency analysis method based on betweenness centrality calculates the number of all shortest paths passing through a specific node, and more comprehensively evaluates the importance of the node in the network to measure whether the node is in an important position for information transmission. In fact, the influence of each node gradually decreases with the increase of the distance between nodes, so the influence of most nodes only acts on their multi-hop neighbors. In complex networks, the different adjacency relationships of each node will cause differences in the network topology.

[0127] The present invention constructs a greedy structural entropy dependency method. Structural entropy, as an indicator to measure network complexity and uncertainty, reflects the information entropy of the network. It quantifies the structural information of the network by examining the connection mode and mutual relationship between nodes, reveals the diversity and complexity of information flow in the network, and indicates the stability and efficiency of the system. In the power system, the nodes can be power plants, transmission lines or load nodes, and the connections between them represent interactions and dependencies. In a dependent network, structural entropy can help identify key nodes and connections in the network, which are particularly important when establishing dependencies between the power grid and the information network. The greedy algorithm is a method of gradually adding nodes and connections. By continuously optimizing the network structure, a power information dependency network with certain structural characteristics is finally obtained.

[0128] Structural entropy is based on the entropy theory in information theory and is used to measure the degree of influence of network topology on information transmission. Specifically, structural entropy reveals the location of nodes in the entire network architecture and their connection relationships, thereby expressing the overall organization and operation complexity of the network. High structural entropy indicates that the internal connections of the network are diverse and complex, which plays a vital role in whether the entire network is stable and reliable. Next, according to the greedy strategy, the feature with the maximum structural entropy is selected from the entire feature set and added to the feature subset. Then, for the remaining features, the structural entropy is calculated based on the current feature subset and the feature with the maximum structural entropy is selected for addition. Repeat this process until no further features can be selected. By maximizing the structural entropy as the basis for connection selection, you can try to find the connection method that increases the overall structural entropy of the system at each step.

[0129] As shown in formula (17), P ij Represents the probability of information being transmitted from node i to node j. This probability is a key indicator for measuring the strength of interaction between nodes in the network and the possibility of information flow.

[0130]

[0131] Among them, ω ij is the weight of the edge connecting node i and node j, N (i) The set of neighbor nodes of node i. The denominator is the sum of the weights of all edges of node i.

[0132] For each node i, calculate its probability value P ij , calculate the structural entropy of the node:

[0133]

[0134] According to formula (18), in order to describe the connection between grid nodes in terms of electrical characteristics, the present invention constructs the grid electrical distance structure entropy to measure the local influence of nodes in the grid.

[0135]

[0136] u i is a node in the power grid; L p is the threshold of electrical distance; |Z| ij is node u i With node u j The electrical distance between Indicates that at the threshold L p Internal node u i The set of nodes that can be reached, Indicates that at the threshold L p Node u in range i The number of nodes that can be reached. α is defined as the inverse of the minimum electrical distance between nodes.

[0137] In order to describe the local influence of nodes in the information network, the information node structure entropy is defined as:

[0138]

[0139] Where L c is the shortest path threshold, d ij For node v i With node v j The shortest path distance between For node v i At the threshold L c The set of nodes reachable within For node v i At the threshold L c The number of nodes that can be reached within α is the reciprocal of the minimum value of the shortest path between nodes.

[0140] The initial structural entropy of the power grid and information network is calculated using the calculation formulas (19) and (20) of the structural entropy. This process includes a quantitative evaluation of the interaction between nodes in the network. Using the greedy algorithm, the interdependence between the power grid and the information network can be gradually adjusted to achieve the purpose of optimizing the network structure and improving system efficiency.

[0141] The present invention adopts the standard IEEE 30 node system as the power grid, and the actual information network is the BA network, such as Figure 6 As shown, blue represents 30 power grid nodes, green represents 30 information network nodes, the internal connection edges of the network are represented by solid lines, and the dependent edges between the power grid and the information network are represented by dotted lines, constructing a power-information dependent system model.

[0142] For the interdependent network of power grid and information network, maximizing structural entropy helps to identify and establish the optimal interdependence relationship, thereby improving the stability and robustness of the entire system. The power information interdependent system model is constructed by the greedy structural entropy method, with the aim of reducing the uncertainty of the system by optimizing the network structure, so that the system can operate more stably in the face of external disturbances or internal faults, reducing the risk of cascading failures.

[0143] The specific simulation case analysis includes:

[0144] With the rapid advancement of smart grid technology, the stability and robustness of the power system have become key considerations because they directly affect the continuity of power supply and the overall performance of the system. In order to better understand and improve the robustness of the power information dependent system, the present invention uses the IEEE-30 and IEEE-118 systems in the IEEE standard test system database through the MATLAB simulation platform, compares the network models constructed based on degree, betweenness and the greedy structural entropy method proposed by the present invention, and explores the impact of different network connection methods on the robustness of the power information dependent system by evaluating indicators such as network efficiency, network residual scale, and network residual load. Network efficiency is one of the key indicators for evaluating network performance. It reveals the efficiency of information or power transmission between nodes and indicates the speed and accuracy of the network when transmitting resources or data. High network efficiency means that the average shortest path length between any two points in the network is short, and the interaction between nodes consumes less energy, thereby ensuring the efficient operation of the network. High network residual scale means that even if some nodes or connections fail due to external attacks, internal failures or other reasons, the network can still maintain a large operating scale. Specifically, it indicates the proportion of nodes and connections in the network that can still operate normally, reflecting the resilience and fault tolerance of the network. A large network residual scale indicates that the network can effectively resist external interference, maintain the integrity of its core functions and structure, and ensure the uninterrupted flow of key services and information. A high network residual load refers to the maximum information flow or power flow that the network can still carry after experiencing load or attack. This includes not only the physical carrying capacity of the network, but also the adaptability and adjustment ability of the network under complex conditions. A network with a high residual load can quickly redistribute the load when a node or connection fails, optimize resource utilization, and reduce the decline in overall system performance.

[0145] According to the network model constructed by the present invention, a random attack is carried out on the network model, and the influence of the three evaluation indicators on the network stability is analyzed by comparing different dependency modes. Figure 7-Figure 9 As shown. As can be seen from the figure, the method adopted by the present invention has higher network efficiency, network residual scale and network residual load than the degree and betweenness dependent model when facing the same number of attacks. When carrying out random attacks, the present invention can observe that the construction method of the present invention performs better in maintaining network efficiency, because this method improves the efficiency of information flow by optimizing the topological structure of the network. Compared with the degree and betweenness dependent method, the greedy structure entropy method can better adjust the network structure to reduce the impact of the loss of key nodes or links on network efficiency. When comparing different dependency methods, the present invention can more effectively protect important nodes and links in the network, thereby retaining a larger network scale after being subjected to random attacks. This is because the method reduces the impact of attacks by optimizing the overall structure of the network, while the degree and betweenness dependent method may be more susceptible to random attacks, resulting in a rapid reduction in network scale. The method of the present invention helps to disperse the load after the attack and reduce the pressure on a single node or link by improving the structural adaptability and elasticity of the network, thereby improving the overall load tolerance of the network. In contrast, the degree and betweenness dependent method may not be able to effectively disperse the additional load caused by the attack, causing the network to reach the load limit faster after being attacked.

[0146] The method of the present invention is applied to the IEEE 118 node system for analysis. Figure 10-12 As shown, it can be seen that the dependent method constructed by the present invention can still be applied to a larger system.

[0147] Technological progress has led to a more complex network structure. At the same time, the interdependence between networks has also been strengthened, which shows that the interconnection and interaction between different network systems have become closer and more important. The present invention demonstrates the advantages of the greedy structural entropy method in constructing a power information interdependent system model by comparing the impact of random attacks connected in different interdependent ways. Specifically, the simulation results show that the network constructed by the greedy structural entropy method is superior to the traditional degree-dependent and betweenness-dependent construction methods in terms of network efficiency, network residual scale, and network residual load. Compared with the traditional degree-based or betweenness-based connection method, the greedy structural entropy method can more effectively optimize network connections, improve structural entropy, and improve the overall efficiency and reliability of the network. This shows that in actual power systems, by optimizing the interdependent connection mode between power networks and information networks, the stability and robustness of the system can be effectively enhanced. This optimization helps to ensure the continuity of power supply and the overall performance of the system.

[0148] The core idea of ​​the greedy structural entropy method is to use the structural characteristics of the network to optimize the overall reliability of the network. Unlike the method that simply relies on the degree or betweenness of the node, the greedy structural entropy method takes into account the overall structure of the network and improves the network's anti-attack ability and robustness by optimizing the structural entropy of the network. Specifically, this method fine-tunes the node connection method in the network to improve the overall structural entropy of the network, thereby enhancing the reliability of the network in the face of random attacks. In addition, the greedy structural entropy method not only improves the robustness of the network, but also maintains the high efficiency of the network to a certain extent. This is because while optimizing the network structure, the method also fully considers the needs of network efficiency, ensuring that the network will not significantly sacrifice its transmission efficiency while improving its robustness.

[0149] The present invention constructs a power network model based on the actual power network data of the IEEE30 standard node system, and constructs a scale-free information network model based on the structural characteristics of the power network, which is used as the basis for constructing a power information dependency system model. Furthermore, the present invention proposes an innovative way to define dependency relationships. This method not only takes into account the electrical distance and the shortest path distance of the network, but also adopts the concept of structural entropy on this basis, and recalculates the updated structural entropy value. In order to optimize the constructed power information dependency system model, the present invention adopts a greedy algorithm for corresponding optimization processing. In order to evaluate the performance of the proposed model, the present invention compares the newly constructed model with the traditional model based on node degree and betweenness centrality dependency, especially the performance of system robustness in the face of random attacks. Through a series of simulation analyses in the MATLAB environment, the results show that the power information dependency system model constructed by the present invention exhibits higher stability and excellent robustness.

[0150] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.< / k>

Claims

1. A method for constructing a power information dependent system model based on a greedy structural entropy algorithm, characterized in that: The following steps are involved: Build a power grid model based on several plants and stations; Building an information network model based on the communication equipment of the plant and the communication routes of the communication equipment; Based on the power grid model and the information network model, a greedy structure entropy algorithm is used to construct a power information dependency system model; The process of constructing the power information dependent system model includes: Calculate the adjacency matrix and weight matrix of the power grid model and the information network model respectively; Calculating the initial structural entropy of the power grid based on the adjacency matrix and the weight matrix of the power grid model; Calculating the initial structural entropy of the information network based on the adjacency matrix and weight matrix of the information network model; Based on the initial structural entropy of the power grid and the initial structural entropy of the information network, a greedy algorithm is used to adjust the interdependence between the power grid and the information network to obtain a power information interdependence system model.

2. The method for constructing a power information dependent system model based on a greedy structural entropy algorithm according to claim 1, characterized in that: The power grid model includes a set of power grid nodes and a set of edges connecting power grid nodes; Wherein, the grid node set includes a number of plants and stations; The grid node connection edge set is the high-voltage transmission lines between various power plants and stations.

3. The method for constructing a power information dependent system model based on a greedy structural entropy algorithm according to claim 1, characterized in that: The process of building a power grid model includes; Define the power injected into the grid node; Establish power balance equation for each grid node; The power balance equation is iteratively solved using the Newton-Raphson method to obtain the power flow distribution of each grid node; Determining node voltage, phase angle, active power, and reactive power based on the power flow distribution of each grid node; A power grid model is established based on the active power.

4. The method for constructing a power information dependent system model based on a greedy structural entropy algorithm according to claim 3, characterized in that: The expression of the power balance equation is: Where P i Indicates active power, Q i represents reactive power, i represents the plant node, j represents the imaginary unit, e i represents the real part of the voltage at node i, f i Represents the imaginary part of voltage, G i represents the real part of admittance, B i represents the imaginary part of the admittance, a i represents the current injection amount at node i, b i Represents the amount of current injected at node i.

5. The method for constructing a power information dependent system model based on a greedy structural entropy algorithm according to claim 3, characterized in that: The expression of the power grid model is: P.S physical NP+diag(L) In the formula, P represents the active power between each branch, P∈R n*n , diag(L) is a vector L with elements L i is a diagonal matrix with diagonal elements, P physical represents the power flow matrix.

6. The method for constructing a power information dependent system model based on a greedy structural entropy algorithm according to claim 1, characterized in that: The information network model includes an information network node set and an information network connection edge set; Wherein, the information network node set is a communication device; The information network connection edge set is a set of communication lines between nodes.

7. The method for constructing a power information dependent system model based on a greedy structural entropy algorithm according to claim 6, characterized in that: The information network model is constructed using BA scale-free network.

8. The method for constructing a power information dependent system model based on a greedy structural entropy algorithm according to claim 1, characterized in that: The expression for calculating the initial structural entropy of the power grid is: In the formula, u i is a node in the power grid; L p is the threshold of electrical distance; |Z ij is node u i With node u j The electrical distance between Indicates that at the threshold L p Internal node u i The set of nodes that can be reached, Indicates that at the threshold L p Node u in range i The number of reachable nodes; α is defined as the inverse of the minimum electrical distance between nodes.

9. The method for constructing a power information dependent system model based on a greedy structural entropy algorithm according to claim 1, characterized in that: The expression for calculating the initial structural entropy of the information network is: Where, L c is the shortest path threshold, d ij For node v i With node v j The shortest path distance between For node v i At the threshold L c The set of nodes reachable within For node v i At the threshold L c The number of nodes that can be reached within α is the reciprocal of the minimum value of the shortest path between nodes.

Citation Information

Cited By

  • Intelligent recycling system and method for renewable resources

    CN119784370A