Vulnerability assessment method considering interaction failure between power information networks under network attack
By establishing a network model and failure model of power information physics, studying the fault propagation of interaction failure between power information networks, evaluating the vulnerability of power information physics systems, solving the problem of vulnerability assessment of power information physics systems under network attacks, and realizing the Analysis of the impact of the coupling interaction between the power network and the information network on fault propagation maximizes the utilization of security defense resources.
Patent Information
- Application Number
- CN202411070241.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-06
- Publication Date
- 2024-11-26
AI Technical Summary
The prior art is difficult to effectively evaluate and analyze the vulnerability of power information physics systems under cyber attacks, especially the impact of the coupling interaction between the power network and the information network on the propagation of faults, making it difficult to maximize protective measures.
By establishing a physically dependent network model of power information, defining the failure rules between power nodes and information nodes, building a physical network failure model of power information networks, studying the fault propagation model of interaction failure between power information networks under network attacks, and proposing a topological structure and power supply capability based on topology and power supply capabilities vulnerability indicators to assess the vulnerability of power information physics systems.
The vulnerability analysis of the power information physics system under cyber attacks is realized, which minimizes the impact of the attacker on the power grid, maximizes the benefits of security defense resources, and provides a criticality indicator and vulnerability assessment method for the power information physics system.
Smart Images

Figure CN119030761A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of power system vulnerability analysis, and in particular to a vulnerability assessment method considering the interaction failure between power information networks under network attacks. Background Art
[0002] With the continuous advancement of smart grid construction, more and more intelligent sensor devices are put into power grid operation, providing effective information support for monitoring and controlling the power grid, making the traditional power system gradually develop into a power cyber-physical system. The power cyber-physical system is a typical cyber-physical coupling system. Compared with the traditional power system, the power cyber-physical system not only needs to deal with damage from the physical level, but also suffers from attacks from the information level. The complex coupling mechanism may cause the power system to suffer serious failures under an information attack.
[0003] Therefore, when studying the power cyber-physical system, it is necessary not only to analyze the fault conditions on the power grid side, but also to conduct a comprehensive vulnerability assessment of the power cyber-physical system in combination with the vulnerability of the information network. Vulnerability assessment refers to the degree and scope of the function and performance loss of the evaluation system when it is subjected to external disturbances or internal failures. Compared with traditional power systems, power cyber-physical systems have higher complexity and uncertainty, and their vulnerability assessment is also more challenging. On the one hand, there are various types, multiple levels, and multiple scales of coupling relationships in the power cyber-physical system, such as information network and physical network, control layer and data layer, local and overall. These coupling relationships make the power cyber-physical system nonlinear, non-stationary, and non-uniform, which increases the difficulty of its dynamic evolution and state change. On the other hand, there are threats and attacks from multiple sources, multiple forms, and multiple targets in the power cyber-physical system. These threats and attacks may target the information network or the physical network or both at the same time, and may occur randomly, intentionally, or mixedly, and may be single-point, multi-point, or coordinated. These threats and attacks make the power cyber-physical system incomplete, unreliable, and unsafe, increasing the difficulty of risk identification and protection measures. Summary of the invention
[0004] The purpose of the present invention is to overcome the shortcomings and deficiencies of the prior art, and to provide a vulnerability assessment method considering the interaction failure between power information networks under network attacks, to achieve coupling modeling of the power grid and the power information network by establishing a power information-physical dependent network model, to establish a power information-physical network failure model, to define the failure rules of power nodes and information nodes, to consider the impact of network failures caused by the coupling interaction between the power grid and the power information network on fault propagation, to establish a fault propagation model for interaction failure between power information networks under network attacks, to propose vulnerability indicators considering the power information-physical system after being attacked based on the vulnerability of the topological structure and the vulnerability of the power supply capacity, to complete the vulnerability analysis of the power information-physical system under network attacks, to minimize the impact of attackers on the power grid, and to maximize the benefits of security defense resources.
[0005] To achieve the above object, the technical solution provided by the present invention is: a vulnerability assessment method considering the failure of interaction between power information networks under network attacks, comprising the following steps:
[0006] S1: Based on the real power system structure and the establishment of power information network, considering the mutual coupling dependency between power grid and power information network, combined with the dependent network theory, the power information physical system is modeled as a "partial one-to-one correspondence" power information physical dependent network model;
[0007] S2: By formulating failure rules for physical nodes and information nodes in the power information-physical interdependent network model, a power information-physical network failure model consisting of failure models of the power grid and the power information network is established; wherein, in the power grid, failure rules for power nodes are established by considering the connectivity and functionality of generators, transformers and load nodes, i.e., the failure model of the power grid; in the power information network, failure rules for information nodes are established by distinguishing between routing relay nodes and dispatching control nodes, i.e., the failure model of the power information network; and the fault propagation form and path are established through the failure of various types of nodes and branches, and finally, a power information-physical network failure model is formed;
[0008] S3: Based on the failure model of power information-physical network, a fault propagation model of interaction failure between power information networks is established, and the propagation form and end node of chain failures caused by network attacks are determined, laying a foundation for studying the impact of coupling failure between power grid and information network on fault propagation;
[0009] S4: Criticality indicators are proposed for the weak links of the power information-physical system. Among them, the improved electrical betweenness of the power branch is obtained by combining the electrical betweenness of the power branch and the line load rate as the criticality indicator of the power branch. The weight of the power node transmitting electric energy and the electrical betweenness representing the number of power branches connected to the node are used as the criticality indicator of the power node. Based on the vulnerability of the topological structure and the vulnerability of the power supply capacity, the vulnerability indicators considering the power information-physical system after being attacked are proposed, completing the vulnerability analysis of the power information-physical system under network attacks.
[0010] Further, in step S1, a power grid topology diagram is established according to the actual power system structure; the number of power nodes and the number of power branches are determined, and the interactive interface nodes in the power information network are connected to the power nodes one by one, and the connection relationship between the interactive interface nodes corresponds one by one according to the connection relationship of the power nodes; according to the topological relationship between the power nodes and the interactive interface nodes, appropriate area divisions are selected, and the routing relay nodes are respectively divided into various areas and connected to each interactive interface node; the dispatching control center node will be connected to each routing relay node, and the dispatching control center nodes will also be connected to each other, eventually forming a "partial one-to-one correspondence" power information-physical interdependent network model.
[0011] Further, in step S2, the specific situation of the power information physical network failure model is as follows:
[0012] a. Failure model of power grid: In the power system, through the modeling and analysis of the power grid, the nodes in the power grid can be abstracted into generator nodes, transformer nodes and load nodes; for a generator node, at least one load node or one transformer node must be connected to it in order to play the function of generating power for the power system, so if it is not connected to a load node or a transformer node, it is considered to be failed; for a transformer node, it is a transfer station for communication and transmission of electric energy, it must be connected to at least one power node and one load node, otherwise it will fail; for a load node, if it is not connected to a generator node, it will fail. In reality, a small number of generator nodes have load demands. When they are not connected to other nodes in the power system, if their own capacity can meet their own load demands, they are still considered valid nodes even if they lose the connected branches. After a network attack, the topology of the power system will change. Since we only focus on the active power imbalance of the power system caused by the failure of nodes or branches after the attack, the power flow calculation of the power system can be completed by DC power flow calculation. After an attack, if there is a power branch with power overload in the new equilibrium state, it should also be considered as a failure.
[0013] b. Failure model of the electric power information network: Different from the generator nodes and load nodes in the electric power grid that are physically connected through transmission lines, in the electric power information network, the information transmission between information interaction nodes is independent of each other. There is no direct channel transmission between them. Instead, the information upload and download are executed through routing relay nodes. Therefore, the routing relay node is considered to be a special dispatching control node, and the dispatching control node will master all the information in the electric power information network and will be connected to the routing relay node and the interactive interface node. Therefore, in the electric power information network, as long as each node is connected to another node, it is considered to be working normally, otherwise it is considered to be failed.
[0014] Further, in step S3, in the electric power information-physical system, the interaction between the power grid and the electric power information network is the power grid's energy supply to the electric power information network and the power information network's control over the power grid; according to the "partial one-to-one correspondence" electric power information-physical dependent network model, a power node will only be connected to one interactive interface node through a dependency edge, and the number of nodes in the power grid is the same as the number of interactive interface nodes in the electric power information network, that is, the direct coupling of the power grid and the power information network only needs to consider the mutual coupling dependency relationship between the power node and the interactive interface node; in the actual electric power information-physical system, since the communication and control devices of the interactive interface will be equipped with a backup power supply, in the event of a power node failure, the failure of the node in the power grid will not cause the interactive interface device and its connected communication branch to fail, but from the perspective of information transmission. From the perspective of control, once a power node fails, its coupled interactive interface node loses the monitoring and control function of the power node and can no longer transmit effective information in the power information network. Therefore, it is assumed that when a power node fails, its coupled interactive interface node will also fail; and for information nodes, when an information node fails, it is impossible to control its dependent power nodes. For the "partial one-to-one correspondence" power information physical dependent network model, if an interactive interface node fails, it will not be able to monitor and control the coupled power nodes, which will cause the power node to be in a blind state. If the power node is in an unsafe or unstable state at this time, the interactive interface node cannot be controlled, which will cause the power node to fail. Therefore, it is assumed that when the interactive interface fails, the coupled power node will fail according to the set probability β;
[0015] Due to the deep coupling between the power grid and the power information network of the power cyber-physical system, the initial fault will not only exist in the power grid or the power information network, but may propagate alternately between the power grid and the power information network according to the coupling interaction relationship between the networks, causing chain failures between the networks; based on the power cyber-physical dependent network model and the power cyber-physical network failure model, a fault propagation model for interactive failures between power information networks caused by network attacks is established;
[0016] The attack of network attackers on the power information-physical system is divided into three parts, namely, obtaining information about the attack target, selecting the attack target and the attack method, and executing the attack; first, a topological model of the power information-physical system is established according to the actual structure of the power information-physical system, the modeling rules of the power grid and the power information network, and the coupling rules between networks. Then, the attacker selects the attack target based on the information of the power information-physical system he has mastered. The attacked power information-physical system will experience a chain failure according to the power information-physical network failure model until the attack no longer generates new failed branches and nodes, and the attack ends. Finally, the attacker launches another attack until all his target attack targets fail.
[0017] Further, the specific operation steps of step S4 are as follows:
[0018] S41: Based on the law of power flow distribution and the characteristics of power transmission in the power system, the improved electrical intermediate value of the power branch is obtained by combining the electrical intermediate value of the power branch and the line load rate as the key index S of the power branch. l :
[0019]
[0020] Where l represents the power branch mn, m represents the starting node of the branch, and n represents the arrival node of the branch; G and L represent the collection of generator and load nodes respectively; P mn (i, j) represents the active power transmitted on line mn when generator node i sends unit active power to load node j, setting the output of other generator nodes and the load demand of load nodes to zero; ω ij Refers to the weight of the amount of electric energy transmitted from generator node i to load node j, ω ij =min(R i ,R j ), where R i represents the rated generating capacity of the generator node i, R j represents the maximum load demand of load node j; f l is the line load factor;
[0021] S42: The criticality index of the power node is defined by the improved electrical betweenness of the power node, which combines the weight of the power node transmitting power and the electrical betweenness that represents the number of power branches connected to the node. The criticality index of the power node is r i It is expressed as follows:
[0022]
[0023] In the formula, S lis the criticality index of the power branch connected to node i, i.e., the electrical betweenness; F(i) refers to the set directly connected to power node i, ω ik Refers to the weight of the amount of electric energy transmitted from node i to node k; the criticality index of a power node combines the ability and weight of the node and its directly connected branch to transmit electric energy in the power network, and also takes into account the conditions of different types of nodes, characterizing the criticality of the node in the power network topology and the criticality in the power transmission process; the larger the criticality index of a power node, the more electric energy the power grid transmits through the node, the more power resources it controls, and the higher its criticality in the power grid;
[0024] S43: The vulnerability of the power cyber-physical system is defined as the integrity of the power cyber-physical system and its ability to maintain system functions after a cyber attack. The coupling network node loss rate of the power cyber-physical system after a cyber attack is used to evaluate the topological structure vulnerability of the power cyber-physical system after the attack, and the power system load loss rate after the attack is used to evaluate the power supply capacity vulnerability of the power cyber-physical system after the attack. n and p d It means that the topological structure vulnerability p after the kth attack nk and power supply vulnerability dk As follows:
[0025]
[0026]
[0027]
[0028] Where N0 represents the number of valid nodes in the power cyber-physical system before the attack, N k represents the number of valid nodes remaining in the power cyber-physical system after the kth attack, D0 represents the total load demand of the power system during normal operation, s is the set of subsystems of the power system after the kth attack, z represents the number of the subsystem, and ΔL z represents the load shedding of subsystem z, D z represents the load demand of subsystem z, C z represents the remaining generator capacity of subsystem z, iso represents the set of load nodes that become isolated and fail after the attack, and L b represents the load of the island load node numbered b;
[0029] S44: After the fault occurs, all nodes in the power grid and their connection edge conditions are traversed, and the optimal load reduction model based on DC power flow calculation is used to evaluate the power system loss caused by the cascading fault. The function of the optimal load reduction model is to adjust the generator output and load demand of the power system after cutting off the power limit route, and minimize the load reduction when the power balance of the power system is achieved based on the DC power flow equation calculation. The optimal load reduction model is expressed by the following formula:
[0030]
[0031] StF=AF F-P P
[0032]
[0033] |F l |≤|F lmaX
[0034] -P Gi ≤ΔP Gi ≤P Gimax
[0035] 0≤ΔP Fi ≤P Fi
[0036] Where M is the number of power nodes, P Gi and P Fi are the generator output and load controlled by power node i; ΔP Gi is the output adjustment of the generator at power node i; ΔP Fi is the load reduction of power node i; F is the branch power flow vector, A F-P is the node admittance matrix, P is the node injection power vector; F1 is the branch l power flow, F lmax is the power flow limit of branch l, P Gimax is the power output limit of the generator node i; when an information node i fails, its corresponding power node is uncontrollable, ΔP Gi , ΔP Fi Both are 0.
[0037] Compared with the prior art, the present invention has the following advantages and beneficial effects:
[0038] 1. The present invention develops a power information network failure model, and based on it, establishes a fault propagation model of interaction failure between power information networks under network attacks, which can study the impact of network failure caused by the coupling interaction between the power grid and the power information network on fault propagation.
[0039] 2. The present invention proposes criticality indicators for the weak links of the power information-physical system, and proposes vulnerability indicators considering the power information-physical system after being attacked based on the topological structure vulnerability and power supply capacity vulnerability. It completes the vulnerability analysis of the power information-physical system under network attacks, minimizes the impact of attackers on the power grid, and maximizes the benefits of security defense resources.
[0040] In summary, the existing research on the vulnerability of cyber-physically coupled power systems lacks analysis of the impact of inter-grid coupling on fault propagation. However, inter-grid coupling not only increases the impact range and loss degree of network attacks, but is also crucial to the recovery and protection of cyber-physically coupled power systems. Therefore, it is necessary to consider the impact of inter-grid coupling on fault propagation to solve this problem. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] Figure 1 This is the architecture diagram of the power information-physical interdependent network model.
[0042] Figure 2 Flowchart of the power cyber-physical network failure model.
[0043] Figure 3 This is the architecture diagram of the fault propagation model for interaction failure between power information networks under network attacks.
[0044] Figure 4 Schematic diagram of vulnerability assessment indicators for power cyber-physical systems.
[0045] Figure 5 Flowchart for vulnerability assessment of power cyber-physical systems under cyber attacks. DETAILED DESCRIPTION
[0046] The present invention is further described in detail below in conjunction with embodiments and drawings, but the embodiments of the present invention are not limited thereto.
[0047] This embodiment discloses a vulnerability assessment method considering the failure of interaction between power information networks under network attacks, and the specific situation is as follows:
[0048] Based on the actual power system structure and the establishment of power information network, the mutual coupling dependency between power grid and power information network is considered, and combined with the dependent network theory, a "partial one-to-one correspondence" power information physical dependent network model is built, such as Figure 1As shown. Among them, the connection relationship between the interactive interface nodes is the same as the connection relationship between the power grid nodes they depend on. If the interactive interface nodes and the power grid nodes they are connected to are regarded as a coupling body connected in pairs through dependency edges, then the dependent network model of the entire power information-physical system can be regarded as a coupling system composed of coupling bodies, routing relay nodes, dispatching center nodes, power transmission branches and information transmission branches. Through the basic functions and mutual coupling of these units, the normal operation of the information flow and system can be achieved, ensuring the reliability of the power grid's transmission of electric energy and the reliability of the power information network and its normal monitoring and control functions of the power grid. It specifically includes the following steps:
[0049] 1) Establish a power grid topology diagram based on the actual power system structure.
[0050] 2) Determine the number of power nodes and the number of power branches, connect the interactive interface nodes in the power information network to the power nodes one by one, and the connection relationship between the interactive interface nodes corresponds one by one to the connection relationship of the power nodes.
[0051] 3) According to the topological relationship between power nodes and interactive interface nodes, appropriate area division is selected, and the routing relay nodes are divided into each area and connected with each interactive interface node.
[0052] 4) The dispatching control center node will be connected to each routing relay node, and the dispatching control center nodes will also be connected to each other.
[0053] According to the constructed power information-physical interdependent network model, the power information-physical network failure model is established, such as Figure 2 As shown. Generally, a network attacker’s attack on the power information-physical system can be divided into three parts, namely obtaining information about the attack target, selecting the attack target and attack method, and executing the attack. First, a topological model of the power information-physical system is established based on the actual structure of the power information-physical system, the modeling rules of the power grid and the power information network defined above, and the network coupling rules. Then, the attacker selects the attack target to attack based on the information about the power information-physical system he has mastered. The attacked system will experience a chain failure according to the network failure model until the attack no longer generates new failed branches and nodes, and the attack ends. Then, the attacker launches another attack until all of his target attack targets fail. The flowchart of the entire attack process can be shown as follows Figure 2 The process is represented by .
[0054] In order to better explain the fault propagation process of the interaction failure between power information networks under network attacks, this embodiment will Figure 3 The represented power cyber-physical system is used as the target system for attack, and an attack is launched against it to show the chain development process of failures within the system. Figure 3 In the figure, the square represents an attack, and the number in the square represents the number of attacks; G, D, and P represent the generator node, substation node, and load node in the power grid, respectively; a represents the interactive interface node connected to the power node, b represents the routing relay node, and c represents the dispatch control node. Assuming that the power grid and the power information network are coupled and interacted, the probability of network failure caused by attacking the power information network node is less than 1, that is, β is less than 1. Then simulate an attack on the system in the figure, and the attack process can be shown as follows:
[0055] 1) First attack: Attack the a1G1 coupling body, causing the failure of a1 and G1 nodes and the branches connected to them, the a2 node loses all information branch connections and fails, and the D1 node loses all generator nodes connected to it and also fails. The attack ends, and the failure range caused by the attack is a1, a2, G1, D1 nodes and the branches connected to them.
[0056] 2) Second attack: Attack the a5T1 coupling body, causing the failure of a5 and T1 nodes and the branches connected to them. The power of G3, G4, and G5 nodes cannot be transmitted to G2 and D2 nodes through T1, and the power grid is divided into two left and right subsets. The attack ends, and the attack causes a small range of a5 and T1 nodes and their connected branches.
[0057] 3) The third attack: Attack the a3G2 coupling body, causing the failure of the a3 and G2 nodes and the roads they are connected to. The D2 node loses all power sources and fails. The a4 node loses all branch connections and also fails. The attack ends, and the failure range caused by the attack is the a3, G2, D2, a4 nodes and the branches they are connected to. After the attack, all load nodes in the power system fail.
[0058] 4) The fourth attack: attacking the a6 node, causing the a6 node and the roads connected to it to fail. According to the rule that the probability of inter-network impact is less than one, it is assumed that this attack did not cause the G3 node to fail. The attack ends, and the failure range caused by the attack is the a6 node and the branches connected to it.
[0059] 5) The fifth attack: attack the a7G4 coupling body, causing the a7 node and the G4 node and the branches connected to them to fail. The attack ends, and the attack only causes the a7 and G4 nodes and the branches connected to them to fail in a small area.
[0060] 6) The sixth attack: attacking the a8 node, causing the a8 node and the roads connected to it to fail. According to the rule that the probability of inter-network impact is less than one, the attack is set to cause the G5 node to fail. The attack ends, and the failure range caused by the attack is a6, G5 nodes and the branches connected to them.
[0061] 7) The seventh attack: Attack on the a9T2 coupling body, causing the a9 node and the T2 node and the roads connected to them to fail. The G3 node lost all load nodes and transmission node connection branches, causing the G3 node to fail. This attack is over. After this attack, all nodes in the power grid fail and the power system completely collapses. All interactive interface nodes in the power information network also fail, and the power information network loses the ability to detect and control the power grid. At this point, all network attacks are over.
[0062] Criticality indicators are proposed for the weak links of the system, and vulnerability indicators are proposed based on the vulnerability of the topology structure and the vulnerability of the power supply capacity after the system is attacked, such as Figure 4 shown.
[0063] Based on the law of power flow distribution and the characteristics of power transmission in the power system, the improved electrical betweenness of the power branch is obtained by combining the electrical betweenness of the power branch and the line load rate as the key index S of the power branch. l :
[0064]
[0065] Where l represents the power branch mn, m represents the starting node of the branch, and n represents the arrival node of the branch; G and L represent the collection of generator and load nodes respectively; P mn (i, j) represents the active power transmitted on line mn when generator node i sends unit active power to load node j, setting the output of other generator nodes and the load demand of load nodes to zero; ω ij Refers to the weight of the amount of electric energy transmitted from generator node i to load node j, ω ij =min(R i ,R j ), where R i represents the rated generating capacity of the generator node i, R j represents the maximum load demand of load node j; f l is the line load factor;
[0066] The criticality index of the power node is defined by the improved electrical betweenness of the power node, which combines the weight of the power node's transmitted energy and the electrical betweenness that characterizes the number of power branches connected to the node. The criticality index of the power node is r i It is expressed as follows:
[0067]
[0068] In the formula, S lis the criticality index of the power branch connected to node i, i.e., the electrical betweenness; F(i) refers to the set directly connected to power node i, ω ik Refers to the weight of the amount of electric energy transmitted from node i to node k; the criticality index of a power node combines the ability and weight of the node and its directly connected branch to transmit electric energy in the power network, and also takes into account the conditions of different types of nodes, characterizing the criticality of the node in the power network topology and the criticality in the power transmission process. The larger the criticality index of a power node, the more electric energy the power grid transmits through the node, the more power resources it controls, and the higher its criticality in the power grid;
[0069] The vulnerability of the power cyber-physical system is defined as the integrity of the power cyber-physical system and its ability to maintain system functions after a cyber attack. The coupled network node survival rate of the power cyber-physical system after a cyber attack is used to evaluate the topological structure vulnerability of the power cyber-physical system after the attack, and the load loss rate of the power system after the attack is used to evaluate the power supply capacity vulnerability of the power cyber-physical system after the attack. n and p d It means that the topological structure vulnerability p after the kth attack nk and power supply vulnerability dk As follows:
[0070]
[0071]
[0072]
[0073] Where N0 represents the number of valid nodes in the power cyber-physical system before the attack, N k represents the number of valid nodes remaining in the power cyber-physical system after the kth attack, D0 represents the total load demand of the power system during normal operation, s is the set of subsystems of the power system after the kth attack, z represents the number of the subsystem, and ΔL z represents the load shedding of subsystem z, D z represents the load demand of subsystem z, C z represents the remaining generator capacity of subsystem z, iso represents the set of load nodes that become isolated and fail after the attack, and L b represents the load of the island load node numbered b;
[0074] After integrating the power information-physical interdependent network model, the power information-physical network failure model, the fault propagation model of the interaction failure between power information networks, and the vulnerability index of the power information-physical system, a set of security assessment methods for the power information-physical system under network attacks is obtained. The main process is divided into three steps: first, the power information-physical system is modeled, then the attack method is selected according to the model to produce cascading failures, and finally the vulnerability analysis is performed according to the consequences of the attack. The specific process is as follows: Figure 4 shown.
[0075] After the fault occurs, all nodes in the power grid and their connection edges are traversed, and the optimal load reduction model based on DC power flow calculation is used to evaluate the power system loss caused by the cascading fault. The function of the optimal load reduction model is to adjust the generator output and load demand of the power system after cutting off the power limit route, and minimize the load reduction when the power system power balance is achieved based on the DC power flow equation calculation. The optimal load reduction model is expressed by the following formula:
[0076]
[0077] stF=A F-P P
[0078]
[0079] |F1|≤F lmax
[0080] -P Gi ≤ΔP Gi ≤P Gimax
[0081] 0≤ΔP Fi ≤P Fi
[0082] Where M is the number of power nodes, P Gi and P Fi are the generator output and load controlled by power node i; ΔP Gi is the output adjustment of the generator at power node i; ΔP Fi is the load reduction of power node i; F is the branch power flow vector, A F-P is the node admittance matrix, P is the node injection power vector; F1 is the branch l power flow, F lmax is the power flow limit of branch l, P Gimax is the power output limit of the generator node i; when an information node i fails, its corresponding power node is uncontrollable, ΔP Gi , ΔP Fi Both are 0.
[0083] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.
Claims
1. A vulnerability assessment method considering the interaction failure between power information networks under network attacks is characterized by: The following steps are involved: S1: Based on the real-world power system structure and the construction of the power information network, and considering the mutual coupling and dependency between the power grid and the power information network, combined with the dependent network theory, the power cyber-physical system is modeled as a "partial one-to-one correspondence" power cyber-physical dependent network model; S2: By formulating failure rules for physical nodes and information nodes in the power cyber-physical interdependent network model, a power cyber-physical network failure model is established, consisting of failure models for the power grid and the power cyber-physical network. Specifically, in the power grid, failure rules for power nodes are established by considering the connectivity and functionality of generators, transformers, and load nodes, i.e., the power grid failure model. In the power cyber-physical network, failure rules for information nodes are established by distinguishing between routing relay nodes and dispatching control nodes, i.e., the power cyber-physical network failure model. The failure of various types of nodes and branches is used to determine the form and path of fault propagation, ultimately forming the power cyber-physical network failure model. S3: Based on the power cyber-physical network failure model, a fault propagation model for interaction failures between power and information networks was established. This model established the propagation form and end point of cascading faults caused by cyber attacks, laying the foundation for studying the impact of coupling failures between the power and information networks on fault propagation. S4: Criticality indicators are proposed for the weak links of the power cyber-physical system. Among them, the improved electrical betweenness of the power branch is obtained by combining the electrical betweenness of the power branch and the line load rate as the criticality indicator of the power branch. The weight of the power node transmitting electric energy and the electrical betweenness representing the number of power branches connected to the node are combined as the criticality indicator of the power node. Based on the vulnerability of the topological structure and the vulnerability of the power supply capacity, the vulnerability index considering the attack on the power cyber-physical system is proposed to complete the vulnerability analysis of the power cyber-physical system under network attacks.
2. The vulnerability assessment method for considering interaction failure between power information networks under network attacks according to claim 1 is characterized in that: In step S1, a power grid topology diagram is established according to the actual power system structure; Determine the number of power nodes and power branches, connect the interactive interface nodes in the power information network to the power nodes one by one, and the connection relationship between the interactive interface nodes corresponds one by one to the connection relationship between the power nodes; Based on the topological relationship between power nodes and interactive interface nodes, appropriate regional divisions are selected, and routing relay nodes are divided into each region and connected to each interactive interface node; the dispatching control center node will be connected to each routing relay node, and the dispatching control center nodes will also be connected to each other, ultimately forming a "partial one-to-one correspondence" power information-physical interdependent network model.
3. The vulnerability assessment method for considering interaction failure between power information networks under network attacks according to claim 1 is characterized in that: In step S2, the details of the power cyber-physical network failure model are as follows: a. Failure model of the power grid: Through modeling and analysis of the power grid in the power system, the nodes in the power grid can be abstracted into generator nodes, transformer nodes, and load nodes; For a generator node, there must be at least one load node or one transformer node connected to it in order to play the role of generating power for the power system. Therefore, if it is not connected to a load node or a transformer node, it is considered to be failed. For a transformer node, it is a transfer station for communicating and transmitting electric energy. It must be connected to at least one power node and one load node, otherwise it will fail. For a load node, if it is not connected to a generator node or a transformer node, it is considered to be failed. In reality, a small number of generator nodes have their own load demands. When they are not connected to other nodes in the power system, if their own capacity can meet their own load demands, they are still considered to be valid nodes even if they lose the connected branches. After a network attack occurs, the topology of the power system will change. Since we only focus on the active power imbalance of the power system caused by the failure of nodes or branches after the attack, the power system flow calculation can be completed using DC flow calculation. After an attack, if there is a power branch power overload in the new equilibrium state, it should also be considered to be failed. b. Failure model of the electric power information network: Unlike the generator nodes and load nodes in the electric power grid, which are physically connected through transmission lines, in the electric power information network, the information transmission between information interaction nodes is independent of each other. There is no direct channel transmission between them. Instead, information upload and download are executed through routing relay nodes. Therefore, routing relay nodes are considered to be a special dispatching control node, and the dispatching control node will master all the information in the electric power information network and will have connections with routing relay nodes and interactive interface nodes. Therefore, in the electric power information network, as long as each node is connected to another node, it is considered to be working normally, otherwise it is considered to be failed.
4. The vulnerability assessment method for considering interaction failure between power information networks under network attacks according to claim 1 is characterized in that: In step S3, in the power cyber-physical system, the interaction between the power grid and the power information network is the power grid supplying energy to the power information network and the power information network controlling the power grid; According to the "partial one-to-one correspondence" power cyber-physical dependency network model, a power node is only connected to one interactive interface node via a dependency edge. The number of nodes in the power grid is the same as the number of interactive interface nodes in the power information network. In other words, the direct coupling between the power grid and the power information network only needs to consider the mutual coupling dependency between the power node and the interactive interface node. In actual power cyber-physical systems, since the communication and control devices of the interactive interfaces are equipped with backup power supplies, the failure of a power node in the power grid will not cause the interactive interface devices and their connected communication branches to fail. However, from the perspective of information transmission and control, once a power node fails, its coupled interactive interface nodes lose their monitoring and control functions over the power node and can no longer transmit valid information within the power information network. Therefore, when a power node fails, its coupled interactive interface nodes are also assumed to fail. For information nodes, if an information node fails, it cannot control its dependent power nodes. In the "partial one-to-one correspondence" power cyber-physical interdependent network model, if an interactive interface node fails, it cannot monitor and control the coupled power nodes, which will cause the power node to be in a blind state. If the power node is in an unsafe or unstable state at this time, the interactive interface node cannot control it, which will cause the power node to fail. Therefore, when the interactive interface fails, the coupled power node is assumed to fail with a set probability β. Due to the deep coupling between the power grid and the power information network in the power cyber-physical system, the initial fault will not only exist in the power grid or the power information network, but may propagate alternately between the power grid and the power information network based on the coupling interaction relationship between the networks, triggering a cascading failure between the networks. Based on the power cyber-physical interdependent network model and the power cyber-physical network failure model, a fault propagation model for interactive failures between power information networks caused by cyber attacks is established. The attack of network attackers on power cyber-physical systems is divided into three parts: obtaining information about the attack target, selecting the attack target and attack method, and executing the attack; first, a topological model of the power cyber-physical system is established based on the actual structure of the power cyber-physical system, the modeling rules of the power grid, the power information network, and the coupling rules between networks. Then, the attacker selects the attack target based on the information about the power cyber-physical system he has mastered. After the attack, the power cyber-physical system will suffer a chain failure according to the failure model of the power cyber-physical network until the attack no longer generates new failed branches and nodes. At this time, the attack ends. Finally, the attacker launches another attack until all his target attack targets fail.
5. The vulnerability assessment method considering interaction failure between power information networks under network attacks according to claim 1 is characterized in that: The specific operation steps of step S4 are as follows: S41: Based on the law of power flow distribution and the characteristics of power transmission in the power system, the improved electrical betweenness of the power branch is obtained by combining the electrical betweenness of the power branch and the line load rate as the key index S of the power branch. l : Where l represents the power branch mn, m represents the starting node of the branch, and n represents the arrival node of the branch; G and L represent the collection of generator and load nodes respectively; P mn (i, j) represents the active power transmitted on line mn when generator node i sends unit active power to load node j, setting the output of other generator nodes and the load demand of the load node to zero; ω ij Refers to the weight of the amount of electric energy transmitted from generator node i to load node j, ω ij =min(R i ,R j ), where R i represents the rated generating capacity of generator node i, R j represents the maximum load demand of load node j; f l is the line load factor; S42: The criticality index of the power node is defined by the improved electrical betweenness of the power node, which combines the weight of the power transmitted by the power node and the electrical betweenness that represents the number of power branches connected to the node. The criticality index of the power node r i It is expressed as follows: Where S l is the criticality index of the power branch connected to node i, that is, the electrical betweenness; F(i) refers to the set directly connected to power node i, ω ik Refers to the weight of the amount of power transmitted from node i to node k; The criticality index of a power node combines the power transmission capacity and weight of the node and its directly connected branches in the power network, taking into account the different types of nodes. It represents the criticality of the node in the power network topology and its criticality in the power transmission process. The larger the criticality index of a power node, the more power the power network transmits through the node, the more power resources it controls, and the higher its criticality in the power network. S43: The vulnerability of the power cyber-physical system is defined as the integrity of the power cyber-physical system and its ability to maintain system functions after a cyber attack; the coupling network node loss rate of the power cyber-physical system after a cyber attack is used to evaluate the topological structure vulnerability of the power cyber-physical system after the attack, and the power system load loss rate after the attack is used to evaluate the power supply capacity vulnerability of the power cyber-physical system after the attack, respectively. n and p d Indicates that the topological vulnerability p after the kth attack nk and power supply vulnerability p dk As follows: Where N0 represents the number of valid nodes in the power cyber-physical system before the attack, N k represents the number of valid nodes remaining in the power cyber-physical system after the kth attack, D0 represents the total load demand of the power system during normal operation, s is the set of subsystems of the power system after the kth attack, z represents the subsystem number, and ΔL z Denotes the load shedding of subsystem z, D z represents the load demand of subsystem z, C z represents the remaining generator capacity of subsystem z, iso represents the set of load nodes that become isolated and fail after the attack, L b represents the load of the island load node numbered b; S44: After a fault occurs, all nodes in the power grid and their connected edges are traversed, and an optimal load reduction model based on DC power flow calculation is used to assess the power system losses caused by the cascading fault. The function of this optimal load reduction model is to adjust the generator output and load demand of the power system after removing the power-over-limit route, and minimize the load reduction when the power system power balance is achieved based on the DC power flow equation. This optimal load reduction model is expressed as follows: s.t.F=A F-P P |F1|≤F lmax -P Gi ≤ΔP Gi ≤P Gimax 0≤ΔP Fi ≤P Fi Where M is the number of power nodes, P Gi and P Fi are the generator output and load controlled by power node i; ΔP Gi is the generator output adjustment value of power node i; ΔP Fi is the load reduction amount of power node i; F is the branch power flow vector, A F-P is the node admittance matrix, P is the node injection power vector; F1 is the branch l power flow, F lmax is the power flow limit of branch l, P Gimax is the power output limit of the generator node i; when an information node i fails, its corresponding power node is uncontrollable, ΔP Gi , ΔP Fi Both are 0.
Citation Information
Patent Citations
Fault hazard index calculation and vulnerability evaluation method for smart power grid
CN113344743A
Electric power information physical system cascading failure modeling method and device
CN116366349A
Validating power network models for monitoring and correcting operation of electric power networks
US20170373535A1
