A post-disaster island division method considering the importance of information-energy coupling nodes

Abstract

The present invention relates to a post-disaster island partitioning method taking into account the importance of information-energy coupling nodes, comprising the following steps: 1) constructing a coupling model of a power communication network and a power network; 2) using a multi-angle, multi-index and subjective and objective evaluation method to obtain the importance of information-energy coupling nodes; 3) constructing an island partitioning model based on the importance of information-energy coupling nodes to perform post-disaster island partitioning of active distribution networks. Compared with the prior art, the present invention has the advantages of effectively avoiding evaluation from a single perspective of information or energy, effectively obtaining system status information, and improving the system's ability to respond to natural disasters.

Description

Technical Field

[0001] The present invention relates to the technical field of improving the elasticity of distribution networks, and in particular to a method for dividing active distribution networks into isolated islands after disasters taking into account the importance of information-energy coupling nodes. Background Art

[0002] With the change of environmental climate, various extreme weather disasters occur frequently, and the characteristics of the distribution network such as wide distribution, many lines, and large number of users determine that it is extremely vulnerable to the threat of natural disasters, thus causing large-scale power outages. Therefore, it is necessary to form a microgrid through island division after a natural disaster occurs to achieve reliable and continuous power supply to key loads and improve the flexibility of the distribution network.

[0003] When an active distribution network fails, the active distribution network is divided into multiple microgrids by means of islanding to achieve continuous and reliable power supply to critical loads. Current research only measures the importance of nodes from a single electrical perspective such as the grid topology and load size when performing islanding, and does not involve the coupling of information and energy. However, the failure of the information system will cause a significant decrease in the reliability of the physical system. Given the deep coupling between the power network and the power communication network, it is necessary to consider the coupling of energy and information in terms of node importance when dividing islands. In addition, existing research on the evaluation of node importance for energy and information coupling is aimed at the power backbone communication network and the transmission network. As the network with the widest coverage, user-oriented, and access to distributed energy, the communication services undertaken by its nodes in the distribution network are different from those in the transmission network. Therefore, it is necessary to analyze the importance of distribution network nodes from the perspective of energy and information coupling, so as to provide a reasonable basis for islanding. Information acquisition is the first step in islanding. Natural disasters put forward higher requirements for the communication of active distribution networks. How to quickly establish information channels after natural disasters and obtain system status information (such as node power, line switch status, etc.) is also a key point in islanding. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a post-disaster island division method taking into account the importance of information-energy coupling nodes.

[0005] The purpose of the present invention can be achieved by the following technical solutions:

[0006] A post-disaster island division method taking into account the importance of information-energy coupling nodes comprises the following steps:

[0007] 1) Construct a coupling model between the power communication network and the power network;

[0008] 2) The importance of information-energy coupling nodes is obtained by adopting a multi-angle, multi-index and subjective-objective evaluation method;

[0009] 3) According to the importance of information-energy coupling nodes, an island partition model is constructed to perform post-disaster island partitioning of the active distribution network.

[0010] The step 1) is specifically as follows:

[0011] The power communication network and the power network are abstracted into a weighted network model. Taking the power network as the basic network, the similarity of the two networks is evaluated through four indicators: node degree, node betweenness, node cohesion and node efficiency value. The coupling node scale value is obtained, and then the coupling network model of the two networks is established.

[0012] The node degree deg(v i ) is:

[0013]

[0014] Among them, deg + (v i )、deg - (v i ) are the nodes v in the two-network coupling network model respectively. i The out-degree and in-degree of

[0015] The node betweenness B(v i ) is:

[0016]

[0017] Among them, δ jk For node v j To node v k The number of shortest paths, δ jk (v i ) is δ jk Passing through node v i The number of nodes; the node cohesion C(v i ) is:

[0018]

[0019] in, is the number of nodes reduced by network contraction after node failure, and N is the total number of nodes;

[0020] The node efficiency value The expression is:

[0021]

[0022] Among them, d ij For node v i To node v jThe number of nodes passed by the shortest path is the number of hops.

[0023] In the step 1), each node v is constructed according to the node degree, node betweenness, node cohesion and node efficiency value. i The topological eigenvalue matrix of According to the dependency relationship between the power communication network and the power network, the correlation matrix K = [k ij ], then the coupling node scale value D r The expression is:

[0024]

[0025]

[0026]

[0027] Among them, r is the coupling node number in the topology of the two-network coupling model, ρ ij is the related node pair based on Spearman correlation coefficient and The mapping ratio, k ij is the node association degree, R is the number of elements in the matrix ψ, the superscript γ represents the element sequence in the matrix ψ, and R is 4, and N is the total number of nodes in the topology of the two-network coupling model.

[0028] The step 2) is specifically as follows:

[0029] Based on the established two-network coupling network model, on the information flow side, the importance of nodes taking into account information flow is analyzed from the two perspectives of communication service and communication reliability; on the energy flow side, the importance of nodes taking into account energy flow is analyzed from three perspectives: topological structure, load properties, and electrical characteristics. Then, a combined weighted evaluation method combining the R-G1 method and the entropy weight method is used to combine the coupling node scale value to obtain the information-energy coupling node importance.

[0030] The expression of the communication reliability rate is:

[0031]

[0032]

[0033] Among them, V i C is the communication reliability rate of node i, is the communication reliability rate of the power communication network without lost nodes, is the communication reliability rate of the power communication network after node i is lost, and M is the total number of nodes in the power communication network.

[0034] The importance of nodes in energy flow is analyzed from three perspectives: topology, load properties, and electrical characteristics. The specific indicators involved include connectivity, network cohesion, load importance, electrical betweenness, and transmission efficiency change rate. The expression of connectivity is:

[0035]

[0036] Among them, L i is the connectivity of node i, is the node efficiency value of node j;

[0037] The expression of the network cohesion is:

[0038]

[0039]

[0040] Among them, α i is the network cohesion of node i, L WAPL,i is the average path length after node i is contracted, n is the number of network nodes after contraction, s ij is the length of the shortest path from node i to node j;

[0041] The expression of load importance is:

[0042]

[0043] Among them, T i is the load importance of node i, H is the load connected to node i, is the primary load connected to node i, is the secondary load connected to node i;

[0044] The expression of the electrical intermediate number is:

[0045]

[0046] Among them, I EBC,i is the electrical betweenness of node i, Ω SE With Ω SK are the sets of power node g and load node l respectively, Ω i is the set of nodes j directly connected to node i, is the active power transmitted from the power source node g to the load node l through the line ij;

[0047] The expression of the transmission efficiency change rate is:

[0048]

[0049] Among them, V i P is the transmission efficiency change rate of node i, For power network transmission efficiency, is the transmission efficiency of the power network after the node i is lost.

[0050] The step 3) specifically comprises the following steps:

[0051] 31) A global information discovery model is used to obtain the system global status information of the entire distribution network. The distributed global information discovery model consists of multiple local agents LA and multiple regional agents. The communication links between the agents are dynamically formed through wireless connections. Each LA node contains four vectors, namely active power, reactive power, line switch status, and node switch status.

[0052] 32) Determine whether the communication network topology has changed. If it has changed, obtain the communication network topology, return to step 2), update the coupling node importance, and if it has not changed, construct an island partitioning model with the equivalent recovery load maximization as the objective function to perform island partitioning. The calculation expression of the coupling node importance is:

[0053] W(r)=D r ×I C (r)+I P (r)

[0054] Among them, D r is the node scale value of the coupling node r, I C (r), I P (r) are the importance of the information flow side and the energy flow side of the coupling node respectively.

[0055] In the step 31), the information exchange step length ε in two scenarios of static communication network and dynamic communication network is considered. mr , then:

[0056]

[0057] Among them, n m 、n r are the degrees of LA nodes m and r in the communication network, is the set of neighbor nodes of LA node m, λ2(L), λ M (L) are the second and M largest eigenvalues ​​in the Laplace matrix L respectively.

[0058] The island partitioning model takes the maximization of equivalent restoration load as the objective function, and then:

[0059]

[0060] Among them, w i is the node importance of the two-network coupling system, γ ik is a 0-1 variable, indicating whether node i has restored power supply, p i is the active power of the load at node i, s i is the switch state variable of node i, c ij is the switch state variable of line ij, v ik is the island partition variable of node i, v ik =1, node i belongs to microgrid k, P i k , V i k are the active output, reactive output and voltage amplitude of node i in microgrid k, Assemble for microgrids;

[0061] The constraints of the island partitioning model include node constraints, microgrid connectivity constraints, microgrid branch node constraints, microgrid load absorption constraints, microgrid safe operation constraints and distribution network environment constraints.

[0062] Compared with the prior art, the present invention has the following advantages:

[0063] The present invention takes into account the deep coupling of information and energy, and proposes a microgrid island partitioning method that takes into account the importance of information-energy coupled nodes. Through the distributed information-physical system network architecture, the power communication network is mapped to the power network, and the importance of each node is evaluated by a combined weighted evaluation method combining the R-G1 method and the entropy weight method, which can effectively avoid the irrationality caused by evaluation from a single perspective of information or energy, and respectively selects appropriate information exchange steps for static communication networks and dynamic communication networks for global information discovery, which can effectively obtain system status information. Based on the node importance and system status information of energy and information coupling, when dividing the island, the node recovery set can be reasonably optimized to cover nodes with higher importance and maintain their continuous power supply, thereby improving the system's ability to cope with natural disasters. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 The figure is a flow chart of the method of the present invention.

[0065] Figure 2 Diagram of the coupled network model of a distributed cyber-physical system.

[0066] Figure 3 This is the step for analyzing the importance of coupling nodes.

[0067] Figure 4 This is an improved forward back-step power flow algorithm process.

[0068] Figure 5 A model for distributed global information discovery. DETAILED DESCRIPTION

[0069] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0070] Example

[0071] The present invention provides a method for post-disaster islanding of an active distribution network taking into account the importance of information-energy coupling nodes. The method is oriented to the distribution network, takes the power network as the basic network, and constructs a distributed information-physical system network architecture. The different business characteristics of information flow and energy flow, the node similarity of the power network and the power communication network, and the dependency relationship between the two networks are considered. The importance of each coupling node is obtained by a combined weighted evaluation method combining the R-G1 method and the entropy weight method. Then, on this basis, a method for islanding a microgrid taking into account the importance of nodes is proposed. Firstly, based on a global information discovery model, an information path is quickly established after a disaster with the help of a wireless network, the optimized input parameters of the islanding model and the communication network topology are dynamically obtained, and the importance of the coupling nodes is updated. Furthermore, the active distribution network after the disaster is islanded with the maximization of the equivalent recovery load as the objective function to maintain continuous power supply to important nodes.

[0072] like Figure 1 As shown, the present invention comprises the following steps:

[0073] Step 1: Construct a coupling model of the power communication network and the power network. The power communication network and the power network are abstracted into a weighted network model. The power network is used as the basic network. The two networks are analyzed and similarity evaluated through four indicators: node degree, node betweenness, node cohesion, and node efficiency value. The coupling node scale value is obtained, and then the coupling network model of the two networks is established.

[0074] Step 2: Use a multi-angle, multi-index and subjective and objective evaluation method to obtain the importance of information-energy coupling nodes. Based on the two-network coupling network model established in step 1, on the information flow side, the importance of nodes taking into account information flow is analyzed from the two perspectives of communication service and communication reliability; on the energy flow side, the importance of nodes taking into account energy flow is analyzed from the three perspectives of topological structure, load properties and electrical characteristics. Then, a combined weighted evaluation method combining the R-G1 method and the entropy weight method is used to combine the coupling node scale value obtained in step 1 to obtain the importance of information-energy coupling nodes.

[0075] Step 3: Propose a method for post-disaster island division of active distribution network taking into account the importance of information-energy coupling nodes. First, a global information discovery model is used to obtain the system global state information of the entire distribution network, and then determine whether the communication network topology has changed. If it has changed, the communication network topology is obtained, and the process returns to step 2 to update the importance of coupling nodes. If it has not changed, the island division is performed with the maximization of equivalent recovery load as the objective function.

[0076] The details of each step are as follows

[0077] 1. Analysis of Node Importance in Distributed Cyber-Physical Systems

[0078] Distributed cyber-physical systems are based on sensors and rely on wireless networks (such as Wi-Fi, ZigBee, and 5G) to achieve a close integration of information networks and physical systems through the organic integration of computing, communication, and control as well as deep collaboration. The development of distributed cyber-physical systems has made energy flow increasingly dependent on information flow. Information flow failures may penetrate the boundaries of power communication networks and power networks, directly causing disturbances in the power network, serious failures, system instability, and even system collapse. Therefore, when analyzing the importance of nodes, it is necessary to comprehensively consider information flow and energy flow.

[0079] 1.1 Analysis of the coupling between two networks in distributed cyber-physical systems

[0080] The present invention is based on the IEEE34 node system, takes the power network as the basic network and the power network nodes as the coupling nodes to establish a distributed information-physical system in which the power network and the power communication network are deeply coupled. Figure 2 For the sake of convenience, in this example, the two network nodes are collectively referred to as v i , the node set of the power communication network is The node set of the power network is

[0081] The node characteristics of the two networks are similar. Based on the dual-network coupling model, starting from the nodes, the interdependence of the two networks is analyzed. Through the four similarity evaluation indicators of node degree, node betweenness, node cohesion and node efficiency value, the node scale value of the two-network coupling model can be obtained, which reflects the dependency of the two networks on the nodes and provides a basis for the weight allocation between different indicators of information flow and energy flow for the subsequent node importance evaluation.

[0082] (1) Node degree: The node degrees of the power grid and the power communication network are shown in formula (1):

[0083]

[0084] In the formula, deg + (v i) and deg - (v i ) are nodes v i The out-degree and in-degree of .

[0085] (2) Node betweenness: By introducing the node betweenness B(v i ) is a global geometric quantity that characterizes the node v i The influence of the entire network and its impact on (v j ,v k ) is shown in formula (2):

[0086]

[0087] In the formula, δ jk For node v j To node v k The number of shortest paths, δ jk (v i ) is δ jk Passing through node v i The number of

[0088] (3) Node cohesion: Node cohesion reflects the importance of a node by the number of nodes reduced by the network contraction after a node fails. The more nodes are reduced, the more important the node is, as shown in formula (3):

[0089]

[0090] In the formula, is the number of nodes reduced by network contraction after node failure, and N is the total number of nodes;

[0091] (4) Node efficiency value: The node v is measured by the node efficiency value i The closeness of the connection with other nodes is shown in formula (4):

[0092]

[0093] Where, d ij v i →v j (i≠j) The number of nodes passed by the shortest path, that is, the number of hops.

[0094] Through equations (1)-(4), we can get the v of each node in the power grid and power communication network: i The topological eigenvalues ​​of pass Figure 2 The interdependence of nodes between the two networks gives us the correlation matrix K, whose node correlation degree k ij As shown in formula (5):

[0095]

[0096] In order to comprehensively reflect the coupling relationship between the two network topologies and node services, the node scale value D is used. r As shown in formula (6):

[0097]

[0098] Where r is the node number in the topology of the two-network coupling model; ρ ij is the related node pair based on Spearman correlation coefficient ( and ), the calculation formula is shown in formula (7).

[0099]

[0100] Where R is the number of elements in the matrix ψ, that is, R=4.

[0101] Through the similarity evaluation index, the node scale value of the two-network coupling model can be obtained, as shown in Table 1, which reflects the dependency of the two networks on the nodes and provides a basis for the weight distribution between different indicators of information flow and energy flow for the subsequent node importance evaluation.

[0102] Table 1 Node scale values ​​of the two-network coupling model

[0103]

[0104] 1.2. Node importance analysis considering information flow and energy flow

[0105] Coupling nodes have both energy flow and information flow services, which are completely different but based on each other. Energy flow is operated, controlled and managed based on information flow, and energy flow provides power supply for information flow. Therefore, the importance of coupling nodes needs to be analyzed from the perspectives of information and energy.

[0106] 1.2.1 Node importance analysis considering information flow

[0107] Information flow includes various communication services, which support the safe, stable and economical operation of energy systems. Therefore, the present invention analyzes the importance of nodes taking into account information flow from the perspectives of communication services and communication reliability.

[0108] For communication services, existing research mainly focuses on the main network or backbone network, and there are fewer studies involving distribution networks. Starting from the perspective of distribution networks, the present invention takes into account the access of smart distribution networks, distributed new energy and smart users, analyzes the communication service needs of smart distribution networks, and combines the existing distribution network communication services to obtain the importance values ​​of typical distribution network services, as shown in Table 2.

[0109] Table 2 Importance values ​​of typical services in distribution communication network

[0110] Serial number Power communication network business Business Importance 1 Stable system 0.9432 2 Line relay protection 0.9428 3 Advanced Distribution Automation 0.9047 4 Pilot network protection 0.8947 5 Scheduling Automation 0.8653 6 Dispatch Phone 0.8270 7 Wide Area Phase Measurement System 0.8264 8 Electric energy metering system 0.7335 9 Load demand control management 0.6055 10 Distributed power monitoring 0.6045 11 Smart energy meter (area concentration point) 0.5718 12 Equipment operation status monitoring information 0.5583 13 Video conferencing system 0.5073 14 Lightning monitoring and positioning system 0.4809 15 Administrative telephone 0.4557 16 Office Automation 0.3093

[0111] When the power communication network is hit by a natural disaster, it may cause the loss of a node in the power communication network. Therefore, it is necessary to define the node communication reliability rate to measure the impact of the loss of a node on network communication. The node i communication reliability rate The calculation formula is shown in (8).

[0112]

[0113] In the formula, is the communication reliability rate of the power communication network without lost nodes; is the communication reliability rate of the power communication network after node i is lost, as shown in formula (9).

[0114]

[0115] Where M is the total number of nodes in the power communication network.

[0116] 1.2.2 Node importance analysis considering energy flow

[0117] Unlike the power communication network, the power network does not have such a wide variety of communication services. However, when a natural disaster occurs, not only the topology and load distribution must be considered, but also the power concentration and power transmission. Therefore, the importance of nodes taking into account energy flow is analyzed from three perspectives: topology, load properties, and electrical characteristics. The specific indicators are as follows:

[0118] (1) Connectivity

[0119] The average efficiency of each node after a node fails in the system is used to measure the degree of connectivity between the node and the remaining nodes, as shown in formula (10):

[0120]

[0121] In the formula, is the node efficiency value of node j.

[0122] (2) Network cohesion

[0123] When hit by natural disasters, the vulnerability of the power network is inversely correlated with the average path length of the network. When the average path length decreases, the vulnerability of the power network increases significantly.

[0124]

[0125]

[0126] Where, L WAPL,i is the average path length after node i is contracted, n is the number of network nodes after contraction, N is the total number of power network nodes, s ij is the length of the shortest path from node i to node j.

[0127] (3) Load importance

[0128] The load importance indicator is introduced to measure the load importance of the node by the ratio of the primary and secondary loads (i.e., the power load that will have a significant impact if the power supply is interrupted) connected to the node, as shown in the formula.

[0129]

[0130] Where, H is the load connected to node i; is the primary load connected to node i; is the secondary load connected to node i.

[0131] (4) Electrical intermediacy

[0132] Electric energy is different from information flow and does not transmit energy through the shortest path defined in the complex network. Therefore, the node electrical coefficient is defined based on the power transmission distribution coefficient, as shown in Equations (14)-(16).

[0133]

[0134]

[0135]

[0136] In the formula, Ω SE With Ω SK are the collections of power nodes and load nodes respectively; Ω i is the set of nodes directly connected to node i; is the active power transmitted from the power source node g to the load node l through the line ij; are the equivalent impedances between nodes i and g, i and l, j and g, and j and l respectively; z ii and z gg are the input impedances between nodes i and g respectively; z ig is the transfer impedance from node i to node g; x ij is the reactance of line ij.

[0137] (5) Transmission efficiency change rate

[0138] Natural disasters may cause the loss of some key nodes in the power network, resulting in a significant decrease in the connectivity of the network and the efficiency of power transmission. Therefore, the rate of change of the power network transmission efficiency needs to be considered, as shown in formula (17).

[0139]

[0140] In the formula, is the power network transmission efficiency, as shown in formula (18); is the transmission efficiency of the power network after the node i is lost, as shown in formula (19).

[0141]

[0142]

[0143] In the formula, Ω N is the node set of the power network; d min,jk is the shortest electrical distance between node j and node k, that is, the sum of the impedances of the transmission path is the smallest.

[0144] 1.2.3 Analysis of coupling node importance based on the two-network coupling scale

[0145] In order to comprehensively analyze the node importance of each coupling node in the distributed information-physical system, the present invention adopts a subjective and objective evaluation method, that is, a combined weighted evaluation method combining the R-G1 method and the entropy weight method, so as to obtain a reasonable coupling node importance. The analysis steps are as follows: Figure 3 As shown, specifically:

[0146] 1) The importance value of each information flow service is obtained by the combined weighting method, as shown in Table 2;

[0147] 2) According to formulas (8)-(19), the communication reliability rate and communication service value of each node in the power communication network, as well as the connectivity, network cohesion, load importance, electrical betweenness and transmission efficiency change rate of each node in the power network are calculated respectively;

[0148] 3) The importance of each coupling node in the power communication network and the power network is obtained by the combined weighting method;

[0149] 4) Finally, based on the correlation matrix and the scale value of the coupling node, the importance value of the coupling node is obtained by formula (20), and then:

[0150] W(r)=D r ×I C (r)+I P (r) (20)

[0151] Where D r is the node scale value, IC (r), I P (r) are the importance of the information flow side and the energy flow side of the coupling node respectively.

[0152] 2. Microgrid island partitioning method considering the importance of information-energy coupling nodes

[0153] The failures caused by natural disasters are different from general single-point failures. They are mostly cascading failures (i.e., multi-point failures), which split the distribution network into multiple regional power grids. Based on the distributed cyber-physical system network architecture, a microgrid island partitioning method considering the importance of information-energy coupling nodes is proposed. The method is divided into two steps. The first step is a distributed global information discovery model based on multi-agents to obtain global state information; the second step is a hybrid linear programming model with the maximization of equivalent restored load as the objective function. Based on the global state information, the post-disaster active distribution network is divided into multiple island microgrids by controlling the output of DG and energy storage, the state of line switches, and node switches. Each microgrid is powered by DG or DG and energy storage, and the power supply of important nodes is quickly restored after the disaster.

[0154] 2.1 Global Information Discovery Based on Average Consensus Algorithm

[0155] The state information of the entire distribution network is obtained through a distributed global information discovery model, such as Figure 5 As shown, each local agent (LA) represents a device (remote switch or node controller on the line) or a group of devices, and establishes a connection between the local agent and the device through wireless connection such as Wi-Fi or wired connection. The regional agent (RA) has strong computing power and information communication capabilities, and is generally installed at the node where the DG is located. The implementation of the distributed global information discovery solution adopted by the present invention does not require the support of communication infrastructure, and the communication links between the agents can be dynamically formed through wireless connections. Therefore, the distributed global information discovery solution used in the present invention is more flexible than the centralized one.

[0156] make For the LA collection, is the number of LAs in the system. RA also has corresponding modules to implement LA functions, so RA is also included in each LA node m. There are four vectors: p m ,q m 、c m 、s m , respectively, to achieve real-time acquisition of active power, reactive power, line switch status, and node switch status, with vector p m For example, its length is N, where the i-th element corresponds to For the i-th node in the cluster, each LAm calculates the active power of the local node based on p m Initialize, expressed as As shown in formula (21), the average consistency algorithm is then used to obtain The average That is, the active power of all nodes, realizing the global information acquisition of active power, as shown in formula (22).

[0157]

[0158]

[0159] In the formula, is the initialization value of active power of LA node m in the communication network, p i is the active power of node i, M is the number of agents in the system, is the total number of LA nodes in the communication network, for The average value refers to the degree of LA node m in the communication network.

[0160] Unlike the distribution network, the communication network does not rely on the physical structure of the distribution network, but establishes information channels through wireless connections, so it can be dynamically formed after a natural disaster occurs. Therefore, the information exchange step length ε is considered in two scenarios: static communication network and dynamic communication network. mr :

[0161] Scenario 1: The communication network is static, and each LA knows the topology information of the entire network.

[0162] Scenario 2: The communication network is dynamic and each LA does not have the topology information of the entire network.

[0163] For the first scenario, the convergence of the distributed average consensus algorithm is related to the eigenvalues ​​of the Laplace matrix of its communication network, when ε=2 / [λ2(L)+λ M (L)], the convergence speed is the fastest. i (L) refers to the i-th largest eigenvalue in the Laplace matrix L of the graph. Due to the impact of natural disasters, the communication network can be formed dynamically. Each LA cannot obtain the communication topology of the entire system and cannot determine a unified ε. That is, the ε of each LA is different. mr According to the local connectivity information, we have:

[0164]

[0165] Where: n m refers to the degree of node (LA)m in the communication network, Refers to the set of neighbor nodes of agentm.

[0166] 2.2 Island Partition Model

[0167] Based on the obtained node importance and system status information taking into account information-energy coupling, the equivalent restoration load of node i is defined as Taking the maximum equivalent restoration load as the objective function, the post-disaster active distribution network is divided into islands:

[0168]

[0169] In the formula, w i is the node importance of the two-network coupling system; γ ik is a 0-1 variable, indicating whether the power supply of node i is restored; p i is the active power of the load at node i; s i is the switch state variable of node i; c ij is the switch state variable of line ij; v ik is the island partition variable of node i, v ik =1, node i belongs to microgrid k; P i k , V i k are the active output, reactive output and voltage amplitude of node i in microgrid k respectively.

[0170] The constraints include:

[0171] 1) Node constraints

[0172] Each node There is only one microgrid to supply power to it. After the distribution network suffers a natural disaster, some lines may fail permanently, so it is necessary to And line collection Preprocessing is performed to remove nodes and lines that cannot be connected to nodes equipped with DG / energy storage. The node set after processing is The line set is

[0173] Define a binary variable v ik ∈{0,1}, indicating whether node i belongs to microgrid k. When node i belongs to microgrid k, v ik =1; otherwise, v ik =0.

[0174]

[0175] In the formula, A collection of nodes for configuring DG.

[0176] When DGk is installed on node i, node i must belong to microgrid k, as shown in formula (26).

[0177]

[0178] 2) Microgrid connectivity constraints

[0179] A radial distribution network is equivalent to a tree, and a microgrid is equivalent to a subtree whose root node is a node equipped with a DG. Due to the connectivity of the tree, when node i belongs to microgrid k, its parent node must also belong to microgrid k.

[0180]

[0181] Where: θ k (i) is the parent node set of node i.

[0182] 3) Microgrid branch node constraints

[0183] When both node i and node j belong to microgrid k, that is, v ik =v jk =1, the branch between them also belongs to microgrid k. Combining formula (27), it can be obtained that when the child nodes of nodes i and j belong to microgrid k, branch ij also belongs to microgrid k, and the state of this branch must be closed.

[0184]

[0185] Where h = ζ k (i,j) refers to the child node of branch ij in microgrid k.

[0186] 4) Microgrid load absorption constraints

[0187] When the load of node i is picked up by microgrid k, the following two conditions must be met: 1) Node i belongs to microgrid k, that is, v ik =1; 2) The switch state of node i is in the closed state, that is, s i =1. Combining the two, we can get the constraint condition: ik ·s i = 1. This constraint is a quadratic constraint, so we define a binary variable γ ik ∈{0,1} to linearize it.

[0188]

[0189]

[0190]

[0191] 5) Constraints on safe operation of microgrids

[0192] When dividing the isolated microgrid, it is also necessary to meet safe operation conditions, such as power balance and voltage within the limit. When dividing the microgrid into isolated islands, there may be multiple power sources, that is, the divided isolated microgrid contains DG and energy storage, not just a one-way flow, so the improved forward back-substitution flow algorithm is used as the operation constraint.

[20] .

[0193] The output constraints of DG and energy storage in the distribution network are:

[0194]

[0195] In the formula, Ω B A node set configured with an energy storage device.

[0196] The improved forward back generation algorithm must meet two assumptions: 1) the distribution network must be the receiving end network, that is, the load capacity is greater than the capacity of the distributed generation; 2) the load distributed on the feeder is a constant power load. The algorithm mainly includes the following two parts: flexible node numbering and traditional forward back generation algorithm. First, the forward back generation calculation order is determined by the flexible node numbering, and then the node power flow and voltage are calculated by the traditional forward back generation algorithm. The algorithm flow is as follows: Figure 4 shown.

[0197] The voltage of each node in microgrid k must be within the safe operating range. Taking node i as an example, the following constraints must be met:

[0198]

[0199] 6) Distribution network environmental constraints

[0200] Due to natural disasters, some lines or switches may fail and be in an open state. Let the set of faulty lines in an open state be Its constraints are shown in equation (34). In contrast, the set of lines that are closed due to faults is denoted as Its constraints are shown in equation (35). Similarly, let the set of fault nodes in the disconnected state be The set of faulty nodes in the closed state is The corresponding constraints are shown in equations (36) and (37) respectively.

[0201]

[0202]

[0203]

[0204]

Claims

1. A post-disaster island division method taking into account the importance of information-energy coupling nodes, characterized in that: The following steps are involved: 1) Construct a coupling model between the power communication network and the power network; 2) The importance of information-energy coupling nodes is obtained by adopting a multi-angle, multi-index and subjective-objective evaluation method; 3) According to the importance of information-energy coupling nodes, an island partitioning model is constructed to perform post-disaster island partitioning of active distribution networks; The step 1) is specifically as follows: The power communication network and the power network are abstracted into a weighted network model. With the power network as the basic network, the similarity of the two networks is evaluated by four indicators: node degree, node betweenness, node cohesion and node efficiency value. The coupling node scale value is obtained, and then the coupling network model of the two networks is established. The step 2) is specifically as follows: Based on the established two-network coupling network model, on the information flow side, the importance of nodes taking into account information flow is analyzed from the two perspectives of communication business and communication reliability; on the energy flow side, the importance of nodes taking into account energy flow is analyzed from three perspectives: topological structure, load properties, and electrical characteristics. Then, the importance of information-energy coupling nodes is obtained by combining the combined weighted evaluation method of R-G1 method and entropy weight method with the coupling node scale value. The step 3) specifically comprises the following steps: 31) A global information discovery model is used to obtain the system global status information of the entire distribution network. The distributed global information discovery model consists of multiple local agents LA and multiple regional agents. The communication links between the agents are dynamically formed through wireless connections. Each LA node contains four vectors, namely active power, reactive power, line switch status, and node switch status. 32) Determine whether the communication network topology has changed. If it has changed, obtain the communication network topology, return to step 2), update the coupling node importance, and if it has not changed, construct an island partitioning model with the equivalent recovery load maximization as the objective function to perform island partitioning. The calculation expression of the coupling node importance is: W(r)=D r ×I C (r)+I P (r) Among them, D r is the node scale value of coupling node r1, I C (r), I P (r) are the importance of the information flow side and the energy flow side of the coupling node respectively; The island partitioning model takes the maximization of equivalent restoration load as the objective function, and then: Among them, w i is the node importance of the two-network coupling system, γ ik is a 0-1 variable, indicating whether node i has restored power supply, p i is the active power of the load at node i, s i is the switch state variable of node i, c ij is the switch state variable of line ij, v ik is the island partition variable of node i, v ik =1, node i belongs to microgrid k, P i k , V i k are the active output, reactive output and voltage amplitude of node i in microgrid k, Assemble for microgrids; The constraints of the island partitioning model include node constraints, microgrid connectivity constraints, microgrid branch node constraints, microgrid load absorption constraints, microgrid safe operation constraints and distribution network environment constraints.

2. A post-disaster island division method taking into account the importance of information-energy coupling nodes according to claim 1, characterized in that: The node degree deg(v i ) is: Among them, deg + (v i )、deg - (v i ) are the nodes v in the two-network coupling network model respectively. i The out-degree and in-degree of The node betweenness B(v i ) is: Among them, δ jk For node v j To node v k The number of shortest paths, δ jk (v i ) is δ jk Passing through node v i the number of The node cohesion C(v i ) is: Among them, k vi is the number of nodes reduced by network contraction after node failure, and N is the total number of nodes; The node efficiency value The expression is: Among them, d ij For node v i To node v j The number of nodes passed by the shortest path is the number of hops.

3. A post-disaster island division method taking into account the importance of information-energy coupling nodes according to claim 2, characterized in that: In the step 1), each node v is constructed according to the node degree, node betweenness, node cohesion and node efficiency value. i The topological eigenvalue matrix of According to the dependency relationship between the power communication network and the power network, the correlation matrix K = [k ij ], then the coupling node scale value D r The expression is: Among them, r is the coupling node number in the topology of the two-network coupling model, ρ ij is the related node pair based on Spearman correlation coefficient and The mapping ratio, k ij is the node association degree, R is the number of elements in the matrix ψ, the superscript γ represents the element sequence in the matrix ψ, and R is 4, and N is the total number of nodes in the topology of the two-network coupling model.

4. A post-disaster island division method taking into account the importance of information-energy coupling nodes according to claim 1, characterized in that: The expression of the communication reliability rate is: Among them, V i C is the communication reliability rate of node i, is the communication reliability rate of the power communication network without lost nodes, is the communication reliability rate of the power communication network after node i is lost, and M is the total number of nodes in the power communication network.

5. The post-disaster island division method taking into account the importance of information-energy coupling nodes according to claim 1, characterized in that: The importance of nodes in energy flow is analyzed from three perspectives: topology, load properties, and electrical characteristics. The specific indicators involved include connectivity, network cohesion, load importance, electrical betweenness, and transmission efficiency change rate. The expression of connectivity is: Among them, L i is the connectivity of node i, is the node efficiency value of node j; The expression of the network cohesion is: Among them, α i is the network cohesion of node i, L WAPL,i is the average path length after node i is contracted, n is the number of network nodes after contraction, s ij is the length of the shortest path from node i to node j; The expression of load importance is: Among them, T i is the load importance of node i, H is the load connected to node i, is the primary load connected to node i, is the secondary load connected to node i; The expression of the electrical intermediate number is: Among them, I EBC,i is the electrical betweenness of node i, κ and Ω SK are the sets of power node g and load node l respectively, Ω i is the set of nodes j directly connected to node i, is the active power transmitted from the power source node g to the load node l through the line ij; The expression of the transmission efficiency change rate is: in, is the transmission efficiency change rate of node i, For power network transmission efficiency, is the transmission efficiency of the power network after the node i is lost.

6. A post-disaster island division method taking into account the importance of information-energy coupling nodes according to claim 1, characterized in that: In the step 31), the information exchange step length ε in two scenarios of static communication network and dynamic communication network is considered. mr , then: Among them, n m 、n r are the degrees of LA nodes m and r in the communication network, is the set of neighbor nodes of LA node m, λ2(L), λ M (L) are the second and M largest eigenvalues ​​in the Laplace matrix L respectively.

Images