Relay protection security reinforcement method and system for coping with active and passive fixed value attacks

Abstract

The invention discloses a relay protection security reinforcement method and system for coping with active and passive value attacks, and relates to the technical field of relay protection. According to the method, modeling is carried out on active constant value attacks and passive constant value attacks of relay protection of the power distribution system respectively, relay protection vulnerabilities are analyzed, security prevention measures are put forward preliminarily, the attack phenomenon is marginalized strategically, and on the basis of the security prevention measures, the minimum electric quantity expectation value is taken as a target function, so that the security of the relay protection system is improved. According to the method, the prevention budget distribution is combined with a defense strategy, interaction among a defender, an attacker and the power distribution system is simulated, the power distribution system is protected, the power distribution system is protected, and the optimal prevention budget distribution reinforcement model is established, solved and output, and the prevention budget distribution of each relay of the power distribution system and the prevention budget of the whole power distribution system are output. A new thought is provided for relay protection security reinforcement for coping with active and passive fixed value attacks, and the network security of relay protection is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of relay protection, and more specifically, to a relay protection security reinforcement method and system for coping with active and passive value attacks. Background Art

[0002] As the last link of the power system, the distribution system directly faces the end users and plays an important role in the power system. Its perfection is directly related to the reliability and quality of power consumption of the majority of users. The distribution system is equipped with coordinated inverse time overcurrent relay protection to prevent short circuit faults. In this process, the overcurrent relay device completes this task by controlling the circuit breakers and reclosers at the beginning and middle of the distribution feeder.

[0003] Distribution automation covers a combination of technologies such as intelligent electronic devices, sensors, detectors, switch technologies and communication networks to control and monitor the operation of distribution systems. Distribution automation can automate important distribution operation functions such as fault detection, location, isolation and restoration by switching distribution topologies.

[0004] The implementation of distribution automation technology exposes the distribution system to various types of malicious cyber attacks, and various cyber attack incidents against the distribution system are emerging in an endless stream. In terms of cyber attacks on relay protection, there are active and passive fixed-value attacks. When the attacker launches a fixed-value attack, he may adopt different strategies to affect the distribution system. Among them, the attacker's goal in active fixed-value attacks (ARSA) is to access and modify one or more settings so that the relay immediately sends a trip command to the corresponding circuit breaker or recloser, causing the relay to trip under normal circumstances. The attacker's goal in passive fixed-value attacks (PRSA) is to change a set of settings so that the relay works normally under normal circumstances, but not when common faults occur in the power system. In passive fixed-value attacks, the attacker changes the settings and waits for a fault to occur downstream. By manipulating the settings of the relay, the nearest backup relay trips earlier than the manipulated relay, that is, the passive fixed-value attack is to disrupt coordination and make the backup relay trip faster than the main relay, thereby increasing the interruption load. The main difference between active set point attack (ARSA) and passive set point attack (PRSA) lies in the distribution area affected by the attack. If the upstream of the relay contains more loads than the downstream, PRSA causes more interruptions than ARSA.

[0005] Active set-value attacks directly affect the power system, while passive set-value attacks destroy relay coordination and increase the risk of future failures. Both active set-value attacks and passive set-value attacks seriously threaten the operational safety of the power system. Therefore, in order to deal with active and passive set-value attacks, it is necessary to strengthen the relay protection. Summary of the invention

[0006] In order to solve the problem of how to reinforce the security of relay protection when dealing with active and passive fixed value attacks, the present invention proposes a method and system for reinforcing the security of relay protection when dealing with active and passive fixed value attacks, thereby improving the network security of relay protection to resist active and passive fixed value attacks.

[0007] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0008] In the first aspect, the present application proposes a relay protection security reinforcement method for responding to active and passive value attacks, including:

[0009] S1: Model the active set value attack and passive set value attack of distribution system relay protection respectively;

[0010] S2: Based on modeling, analyze the relay protection vulnerabilities and preliminarily propose safety precautions;

[0011] S3: Determine attack prevention identification budget allocation based on security precautions;

[0012] S4: extract attack scenarios, calculate the load loss caused by attack scenarios based on distribution system parameters and attack prevention identification budget allocation;

[0013] S5: Use load loss to construct the expected value of power shortage in the attack scenario;

[0014] S6: Taking the minimum expected value of power shortage as the objective function, and taking the relationship between budget allocation, reinforcement budget allocation and successful attack probability, and system recovery time and budget allocation as constraints, establish and solve the optimal prevention budget allocation reinforcement model, and output the prevention budget allocation of each relay in the distribution system and the prevention budget of the entire distribution system;

[0015] S7: Safety reinforcement of relay protection is performed based on the prevention budget allocation of each relay in the distribution system and the prevention budget of the entire distribution system.

[0016] Preferably, the power distribution system comprises a group of nodes N = {1, ..., N}, a group of lines L = {ji|i∈N, j∈N} and a group of relay protection devices R = {1, ..., R}, wherein N represents the number of nodes, N(i) represents the number of buses connected to the bus i, and R is the total number of relay protection devices; when a permanent short circuit fault occurs on the line ji∈L, the inverse time characteristic of the overcurrent relay satisfies:

[0017]

[0018] Among them, I r,f represents the fault current seen by relay r; Represents the adjusted pickup current; TD r Indicates the time interval setting value, A r ,B r ,p r All represent constants of the curve characteristics selected by relay r;

[0019] Modeling active constant-value attacks:

[0020] in, is the pickup current of relay r after being attacked by active setting, x a is the active fixed-value attack scalar; γ>0, γ is the attack design parameter considering measurement noise; the attacker launches a successful active fixed-value attack by adjusting the attack design parameter γ;

[0021] Modeling a passive constant-value attack:

[0022] Among them, t r′,f represents the operation time delay of relay r when fault f occurs downstream of relay r′, δt is the coordination time interval, which represents the minimum time difference between two relays; r′=1, r′=R′ are the nearest backup relay and the farthest backup relay of relay r respectively;

[0023]

[0024] in, is the operation delay of relay r after passive fixed value attack; δt p This is a passive fixed-value attack time delay; the attacker changes the adjustable setting of the relay TD r ,A r ,B r ,p r , launching a successful passive fixed-value attack.

[0025] Preferably, the relay protection vulnerability includes: communication network, intermediate relay firmware and local access network; wherein the attacker penetrates the communication network through the connection point, bypasses the firewall to connect to the relay, modifies the adjustable settings of the relay, and implements the attack; the attacker launches a Trojan attack on the firmware installed on the intermediate relay; the attacker connects to the relay through the local access network, modifies the adjustable settings of the relay, and implements the attack;

[0026] The security precautions include: precautions implemented on a single relay and precautions implemented on the entire distribution system; among them, the precautions implemented on a single relay include: installing physical barriers, upgrading firmware security, multi-factor authentication and installing firewalls; the precautions implemented on the entire distribution system include: communication data encryption, firmware update verification and installation of an attack identification system.

[0027] According to the above technical means, basic safety precautions are proposed on the basis of relay protection vulnerability analysis. Although cyber attacks cannot be completely eliminated, these abnormal phenomena can be strategically marginalized through several simultaneous measures. Based on safety precautions, it is convenient to combine prevention budget allocation with defense strategies in the future.

[0028] Preferably, the attack prevention identification budget allocation vector is defined as:

[0029] b=[b1,…,b r ,…,b R ,b s ,b t ],

[0030] Among them, b r Allocate a budget for preventive measures for a single relay r; b s A budget allocation for preventive measures for the entire distribution system; b. t Indicates the allocated budget for the installed attack identification system;

[0031] The defender action space is defined as:

[0032] B represents the total available strengthening budget;

[0033] Relating the probability of successful attack of each relay to the attack prevention identification budget invested, we get:

[0034]

[0035] Among them, π r (b r ,b s ) represents the probability of a successful attack on relay r, which is a function of the reinforcement budget allocated to relay r and the entire distribution system; π r (b r ,b s ) is decomposed into two probabilities: Depends on the precautions of each relay, π s (b s ) depends on the precautions of the entire distribution system; α∈[0,1] indicates the importance of each probability relative to another probability; λ r,μ is the design parameter, and the budget is associated with the reinforcement measures of relay r and the entire distribution system respectively; Initial safety level of relay r and distribution system respectively;

[0036] Distribution system recovery time after installing the attack identification system:

[0037] τ R (b t )=τ i -τ(b t )

[0038] Among them, τ(b t ) represents the recovery time of the power distribution system without the attack identification system installed, τ i represents the total recovery time, τ(b t ) is calculated as:

[0039]

[0040] Where Δτ represents the recovery time reduced by installing the attack identification system, B t is the budget needed to install the system.

[0041] Preferably, let c∈C={1,…,C} be the attack scenario, C be the set of all attack scenarios, C be the total number of attack scenarios, and subset is the maximum set of attacked intermediate relays in attack scenario c;

[0042] In attack scenario c, the attacker can attack at most all relays r∈R c , if the attacker's budget allows to manipulate at most K intermediate relays, whether it is an active fixed value attack or a passive fixed value attack, the total number of attack scenarios C is:

[0043]

[0044] in, represents the sum of m attacked relays in the relays;

[0045] Assume that the load loss caused by attack scenario c is * (R c ) indicates that the calculation expression is:

[0046]

[0047] in, represents the total load loss, represents the load loss caused by non-attack scenarios, and O is the set of operation decision variables.

[0048] Preferably, according to the distribution system parameters and the attack prevention identification budget allocation, the process of calculating the load loss caused by the attack scenario is:

[0049] Construct the objective function, the expression is:

[0050]

[0051] Among them, o∈O={p grid ,q grid ,p,q,y,e,s}, including active power of the distribution system, reactive power active line flow, reactive line flow, service load ratio. Line power-on status and switch status;

[0052] Constraints:

[0053]

[0054]

[0055]

[0056]

[0057]

[0058]

[0059]

[0060]

[0061]

[0062]

[0063]

[0064]

[0065]

[0066]

[0067] Equations (1) to (5) represent the power balance constraints of the distribution system nodes, p ij ,q ij are the active power flow and reactive power flow of line ij; is their maximum value; is the total demand of node i, y i is the load ratio of the service on the node; binary variable eij represents the energy flow from node i to j, which should be zero for isolated lines; Equations (6) to (13) indicate the relay travel required to isolate the system fault according to the state of the operating relay. Equation (14) indicates that if relay r is successfully targeted by an active fixed-value attack, the attacker will immediately open its associated switch s ij ;f ij represents the fault location parameter; ε r Indicates the operating status of relay r; For switch ij The initial state can only trip when there is an operation of the relay protection device; the transformation matrix B r,ij Connect the relay r to the distribution line ij where the relay is located;

[0068] Fault location parameter f ij And relay groups under active customized attacks and passive customized attacks As the input to solve the objective function and constraints, solve for the load loss ζ * (R c ).

[0069] Preferably, using load loss, the expected value of power shortage in the attack scenario is constructed as follows:

[0070] in,

[0071] u c (b) represents the expected value of insufficient power in the attack scenario; ψ c (b) represents the expected load loss, which is a function of the probability vector of the relay attack probability π(b), π c (b) means all The successful attack probability vector is a combined probability function.

[0072] Preferably, the optimal prevention budget allocation reinforcement model is:

[0073] Objective function:

[0074] Constraints:

[0075]

[0076]

[0077] τ R (b t )=τ i -τ(b t ), (17)

[0078]

[0079]

[0080] Among them, formula (15) represents the budget constraint allocation constraint, formula (16) represents the constraint between the enhanced budget allocation and the probability of successful attack, formula (17) to formula (18) represent the relationship constraint between the system recovery time and the budget allocation; formula (19) represents the attacker's expected value of power shortage in C for different attack scenarios.

[0081] Preferably, an auxiliary variable E is introduced to transform the optimal prevention budget allocation reinforcement model into:

[0082] Objective function:

[0083] Constraints:

[0084]

[0085]

[0086] τ R (b t )=τ i -τ(b t ), (17)

[0087]

[0088]

[0089]

[0090] The two-layer optimization consisting of the process of calculating the load loss caused by the attack scenario and the process of calculating the optimal prevention budget allocation reinforcement model is converted into a single-layer optimization.

[0091] In a second aspect, the present application further proposes a relay protection security reinforcement system for responding to active and passive value attacks, the system being used to implement the relay protection security reinforcement method for responding to active and passive value attacks, comprising:

[0092] A modeling unit, used for modeling active set value attacks and passive set value attacks of relay protection of distribution system respectively;

[0093] Vulnerability analysis unit, used to analyze relay protection vulnerabilities and preliminarily propose safety precautions;

[0094] An attack prevention identification budget allocation unit determines attack prevention identification budget allocation based on security precautions;

[0095] A load loss calculation unit, used to extract attack scenarios and calculate the load loss caused by the attack scenarios according to the distribution system parameters and attack prevention identification budget allocation;

[0096] The power shortage expected value building unit uses load loss to build the power shortage expected value in the attack scenario;

[0097] The optimal prevention budget allocation unit takes the minimum expected value of power shortage as the objective function, and uses the relationship constraints between budget allocation, reinforcement budget allocation and successful attack probability, and system recovery time and budget allocation to establish and solve the optimal prevention budget allocation reinforcement model, and output the prevention budget allocation of each relay in the distribution system and the prevention budget of the entire distribution system;

[0098] The safety reinforcement unit performs safety reinforcement on relay protection based on the prevention budget allocation of each relay in the distribution system and the prevention budget of the entire distribution system.

[0099] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0100] The present invention proposes a method and system for strengthening the security of relay protection for responding to active and passive value attacks, respectively models the active fixed value attacks and passive fixed value attacks of the relay protection of the distribution system, analyzes the vulnerabilities of the relay protection, and preliminarily proposes safety precautions to strategically marginalize the attack phenomenon. Based on the safety precautions, the minimum expected value of insufficient power is used as the objective function, and the relationship constraints between budget allocation, strengthening budget allocation and successful attack probability, and system recovery time and budget allocation are used to establish and solve the optimal prevention budget allocation reinforcement model, and output the prevention budget allocation of each relay in the distribution system and the prevention budget of the entire distribution system. The present invention combines the prevention budget allocation with the defense strategy, simulates the interaction between the defender, the attacker and the distribution system, and provides a new idea for the safety reinforcement of relay protection for responding to active and passive fixed value attacks, thereby improving the network security of relay protection. BRIEF DESCRIPTION OF THE DRAWINGS

[0101] Figure 1 A schematic diagram of a flow chart of a relay protection security reinforcement method for responding to active and passive value attacks proposed in an embodiment of the present invention;

[0102] Figure 2 A schematic diagram showing the impact areas of active fixed value attack and passive fixed value attack proposed in an embodiment of the present invention;

[0103] Figure 3 A schematic diagram showing a relay protection security reinforcement system for coping with active and passive value attacks proposed in an embodiment of the present invention. DETAILED DESCRIPTION

[0104] The drawings are for illustrative purposes only and should not be construed as limiting the present patent;

[0105] In order to better illustrate the present embodiment, some parts in the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product;

[0106] It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0107] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0108] Embodiment 1 This embodiment proposes a relay protection security reinforcement method for dealing with active and passive value attacks. The flowchart of the method is shown in Figure 1 ,like Figure 1 As shown, the steps process includes:

[0109] S1: Model the active set value attack and passive set value attack of distribution system relay protection respectively;

[0110] S2: Based on modeling, analyze the relay protection vulnerabilities and preliminarily propose safety precautions;

[0111] S3: Determine attack prevention identification budget allocation based on security precautions;

[0112] S4: extract attack scenarios, calculate the load loss caused by attack scenarios based on distribution system parameters and attack prevention identification budget allocation;

[0113] S5: Use load loss to construct the expected value of power shortage in the attack scenario;

[0114] S6: Taking the minimum expected value of power shortage as the objective function, and taking the relationship between budget allocation, reinforcement budget allocation and successful attack probability, and system recovery time and budget allocation as constraints, establish and solve the optimal prevention budget allocation reinforcement model, and output the prevention budget allocation of each relay in the distribution system and the prevention budget of the entire distribution system;

[0115] S7: Safety reinforcement of relay protection is performed based on the prevention budget allocation of each relay in the distribution system and the prevention budget of the entire distribution system.

[0116] In this embodiment, active fixed-value attacks and passive fixed-value attacks on relay protection of distribution system are modeled respectively, the loopholes of relay protection are analyzed, and preliminary safety precautions are proposed to strategically marginalize the attack phenomenon. Based on safety precautions, the minimum expected value of power shortage is used as the objective function, and the relationship constraints between budget allocation, reinforcement budget allocation and successful attack probability, and system recovery time and budget allocation are used to establish and solve the optimal prevention budget allocation reinforcement model, and output the prevention budget allocation of each relay in the distribution system and the prevention budget of the entire distribution system. The present invention combines the prevention budget allocation with the defense strategy, simulates the interaction between the defender, the attacker and the distribution system, and provides a new idea for the safety reinforcement of relay protection to deal with active and passive fixed-value attacks, thereby improving the network security of relay protection.

[0117] Example 2

[0118] This embodiment specifically illustrates the method proposed in Embodiment 1.

[0119] When modeling, the target distribution system includes a set of nodes N = {1, ..., N}, a set of lines L = {ji|i∈N, j∈N} and a set of relay protection devices R = {1, ..., R}, where N represents the number of nodes, N(i) represents the number of buses connected to bus i, and R is the total number of relay protection devices; when a permanent short circuit fault occurs on line ji∈L, the inverse time characteristic of the overcurrent relay satisfies:

[0120]

[0121] Among them, I r,f represents the fault current seen by relay r; Represents the adjusted pickup current; TD r Indicates the time interval setting value, A r ,B r ,p r All represent constants of the curve characteristics selected by relay r;

[0122] In actual implementation, in order to solve possible faults in relay operation, they will provide protection support in a hierarchical coordination manner.

[0123] When launching a fixed-value attack, attackers may adopt different strategies to affect the distribution system.

[0124] (1) Modeling active fixed-value attacks:

[0125] In an active protection setting attack (ARSA), the attacker's goal is to access and modify one or more settings to cause the relay to immediately send a trip command to the corresponding circuit breaker or recloser. For example, reduce the pickup current With a current less than normal, the relay trips under normal circumstances. The model is as follows:

[0126]

[0127] in, is the pickup current of relay r after being attacked by active setting, x a is the active fixed-value attack scalar; γ>0, γ is the attack design parameter considering measurement noise; the attacker launches a successful active fixed-value attack by adjusting the attack design parameter γ; usually, a successful ARSA on the relay will cause the interruption of the downstream load of the relay, such as Figure 2 As shown, the red marked area is affected by ARSA on relay γ>0.

[0128] (2) Modeling of passive fixed-value attacks:

[0129] In a passive protection setting attack (PRSA), the attacker's goal is to change a set of settings so that the relay works properly under normal conditions, but not when common faults occur in the distribution system. In this strategy, the attacker changes the settings and waits for a fault to occur downstream.

[0130] In the overcurrent relay coordination problem, the goal is to determine the time-to-dial current setting value of each relay so that the total operating time of the primary relay is minimized. As a set of backup relays for relay r, the time delay constraint between the main relay and the backup relay is defined as:

[0131]

[0132] Among them, t r′,f represents the operation time delay of relay r when fault f occurs downstream of relay r′, δt is the coordination time interval, which represents the minimum time difference between two relays; r′=1, r′=R′ are the nearest backup relay and the farthest backup relay of relay r respectively;

[0133] Therefore, the attacker attempts to manipulate t r,f To violate the time delay constraint between the main relay and the backup relay:

[0134]

[0135] in, is the operation delay of relay r after passive fixed value attack; δt p This is a passive fixed-value attack time delay; the attacker changes the adjustable setting of the relay TD r ,A r ,B r,p r , launching a successful passive fixed-value attack.

[0136] The main purpose of PRSA is to disrupt coordination and cause the backup relay to trip faster than the primary relay, thereby increasing the interruption load. Figure 2 As shown, a PRSA is launched on an automation device. This attack does not cause an immediate load interruption, but if a fault occurs downstream of the automation device, the relay cannot trip on time, causing the protection device to trip and everything downstream of the protection device to be interrupted. The loads that are interrupted due to a PRSA on the automation device are all loads in the system, excluding loads downstream of the automation device, which are interrupted due to system faults even if the automation device is not compromised. The main difference between ARSA and PRSA is the distribution area affected by the attack. If the upstream of the relay contains more loads than the downstream, PRSA causes more interruptions than ARSA.

[0137] Based on the manipulation of the setting parameters of the relay device by the attacker considered in the modeling process, there are several ways for the attacker to access the relay protection and cause failures. The threats can be mainly divided into three categories, namely, the relay protection vulnerabilities include: communication network, intermediate relay firmware and local access network; among them, the attacker penetrates the communication network through the connection point, bypasses the firewall to connect to the relay, and modifies the adjustable settings of the relay to achieve the attack; the attacker launches a Trojan attack on the firmware installed on the intermediate relay; the attacker connects to the relay through the local access network and modifies the adjustable settings of the relay to achieve the attack;

[0138] The security precautions proposed in this embodiment include: precautions implemented on a single relay and precautions implemented on the entire distribution system; wherein, the precautions implemented on a single relay include: installing physical barriers, upgrading firmware security, multi-factor authentication and installing firewalls; the precautions implemented on the entire distribution system include: communication data encryption, firmware update verification and installation of an attack identification system.

[0139] According to the above technical means, basic safety precautions are proposed on the basis of relay protection vulnerability analysis. Although cyber attacks cannot be completely eliminated, these abnormal phenomena can be strategically marginalized through several simultaneous measures. Based on safety precautions, it is convenient to combine prevention budget allocation with defense strategies in the future.

[0140] To strengthen cybersecurity measures, after evaluating relay protection vulnerabilities, it is necessary to explore strengthening measures to optimize the allocation of prevention budgets. Although cyber attacks cannot be completely eliminated, these anomalies can be strategically marginalized through several simultaneous measures. This patent focuses on strengthening measures based on relay protection vulnerabilities: multi-factor authentication, communication data encryption, and firewall installation are strengthening measures that can be used to reduce the possibility of cyber attacks on relays through communication networks. In addition, verification of updated firmware and firmware security upgrades can also be deployed to mitigate Trojan attacks on relay firmware. In addition, creating physical barriers can reduce the risk of cyber attacks on relays through local access. Finally, if there is still a successful attack, implementing a cyber attack identification system will reduce recovery time.

[0141] Harden the security of relays in the distribution system to minimize the expected brownouts caused by malicious attacks, given available budget and resources. Defenders can reduce the expected brownouts by 1) preventing successful attacks on relays and 2) identifying successful attacks after they occur, reducing the time it takes to restore the distribution system. Attack prevention is achieved by using hardening measures to reduce the probability of successful attacks on intermediate relays. Prevention can be implemented on individual relays or across the entire distribution system.

[0142] The attack prevention identification budget allocation vector is defined as:

[0143] b=[b1,…,b r ,…,b R ,b s ,b t ],

[0144] Among them, b r Allocate a budget for preventive measures for a single relay r; b s A budget allocation for preventive measures for the entire distribution system; b. t Indicates the allocated budget for the installed attack identification system;

[0145] The defender action space is defined as:

[0146] B represents the total available strengthening budget;

[0147] Security precautions are implemented either on a single relay or on the entire distribution system. Taking each security precaution changes the probability of a successful attack on the target relay and the expected value of underpowering, which is directly related to this probability. This embodiment proposes a mapping function that assumes that the probability of a successful attack decays exponentially with respect to the investment in strengthening the relays and the entire distribution system. Security precautions that affect the entire distribution system affect all relays simultaneously, while local measures affect only a single relay. Therefore, the following function is proposed to relate the probability of a successful attack on each relay to the invested reinforcement budget:

[0148]

[0149] Among them, π r (b r ,b s ) represents the probability of a successful attack on relay r, which is a function of the reinforcement budget allocated to relay r and the entire distribution system; π r (b r ,b s ) is decomposed into two probabilities: Depends on the precautions of each relay, π s (b s ) depends on the precautions of the entire distribution system; α∈[0,1] indicates the importance of each probability relative to another probability; λ r ,μ is the design parameter, and the budget is associated with the reinforcement measures of relay r and the entire distribution system respectively; μ i Initial safety level of relay r and distribution system respectively;

[0150] If an attack is launched into the distribution system, there is no way to identify it as an attack and it may be mistaken for a natural fault, prolonging the restoration time and load interruption time. A simple and practical method can be used to identify the attack and reduce the impact of the attack on the distribution system.

[0151] To identify attacks faster, a database of intermediate relay settings can be created in the outage management system (OMS) and compared with the intermediate relay settings after each event. When a relay trips, the settings of that relay and other relays in the distribution system must be compared with the correct settings in the outage management system to identify the attack.

[0152] Distribution system recovery time after installing the attack identification system:

[0153] τ R (b t )=τ i -τ(b t )

[0154] Among them, τ(b t ) represents the recovery time of the power distribution system without the attack identification system installed, τ i represents the total recovery time, τ(b t ) is calculated as:

[0155]

[0156] Where Δτ represents the recovery time reduced by installing the attack identification system, B t is the budget needed to install the system.

[0157] The attacker can select one or more intermediate relays and initiate ARSA or PRSA targeting them. Let c∈C={1,…,C} be the attack scenario, C be the set of all attack scenarios, C be the total number of attack scenarios, and subset is the maximum set of attacked intermediate relays in attack scenario c;

[0158] In attack scenario c, the attacker can attack at most all relays r∈R c , if the attacker's budget allows to manipulate at most K intermediate relays, whether it is an active fixed value attack or a passive fixed value attack, the total number of attack scenarios C is:

[0159]

[0160] in, represents the sum of m attacked relays in the relays;

[0161] Assume that the load loss caused by attack scenario c is * (R c ) indicates that the calculation expression is:

[0162]

[0163] in, represents the total load loss, represents the load loss caused by non-attack scenarios, and O is the set of operation decision variables.

[0164] In the following embodiment, according to the power distribution system parameters and the attack prevention identification budget allocation, the load loss caused by the attack scenario is calculated, and the process is as follows:

[0165] Construct the objective function, the expression is:

[0166]

[0167] Among them, o∈O={p grid ,q grid,p,q,y,e,s}, including active power of the distribution system, reactive power active line flow, reactive line flow, service load ratio. Line power-on status and switch status;

[0168] Constraints:

[0169]

[0170]

[0171]

[0172]

[0173]

[0174]

[0175]

[0176]

[0177]

[0178]

[0179]

[0180]

[0181]

[0182]

[0183] Equations (1) to (5) represent the power balance constraints of the distribution system nodes, p ij ,q ij are the active power flow and reactive power flow of line ij; is their maximum value; is the total demand of node i, y i is the load ratio of the service on the node; binary variable e ij represents the energy flow from node i to j, which should be zero for isolated lines. Note that e ij and e ji do not share the same value, each of them shows a specific flow direction in the ij row. ij and p ji Equations (6) to (13) indicate that the relay travel required to isolate the system fault and the isolation requirements of the fault line are determined according to the state of the operating relay, where fij = 1, as in (6), while (7) ensures that any line connected to the faulted line by a closed switch is de-energized. Constraint (8) ensures that line ij is energized if it is directly connected to another energized upstream line. Essentially, if s ij and ji are all closed and the connected line ki is also energized, e ij is necessarily 1. The power-on status of the substation node is checked by equation (9). Equation (10) represents the radial constraint. In equation (11), ε r Indicates the operating state of relay r. When ε r = 1, the relay r is available and operational, but when the relay is attacked, ε r = 0. Formula (12) is used to check the possibility of tripping, where For switch ij The initial state of the circuit breaker is tripped only when the protective device is operated; the transformation matrix B r,ij Relay r is connected to the distribution line ij where the relay is located. It should be noted that in the formula proposed in this patent, s ij is a continuous variable, but due to the constraints of equations (7) and (9), it only takes binary values. Equation (14) indicates that if relay r is successfully targeted by an active fixed-value attack, the attacker will immediately open its associated switch s ij ;f ij represents the fault location parameter; ε r Indicates the operating status of relay r; For switch ij The initial state can only trip when there is an operation of the relay protection device; the transformation matrix B r,ij Connect the relay r to the distribution line ij where the relay is located;

[0184] Fault location parameter f ij And relay groups under active customized attacks and passive customized attacks As the input to solve the objective function and constraints, solve for the load loss ζ * (R c ) and have the same priority in calculating load loss.

[0185] Using load loss, the expected value of power shortage in the attack scenario is constructed as follows:

[0186] in,

[0187] u c (b) represents the expected value of insufficient power in the attack scenario; ψ c(b) represents the expected load loss, which is a function of the probability vector of the relay attack probability π(b), π c (b) means all The success probability vector of is a combined probability function, whose operation will be explained by the following example. c = {r}), function ψ c (·) is defined as:

[0188] ψ c (b) = π r (b r ,b s ) * ({r})

[0189] However, for simultaneous attacks on multiple relays, the formula will be more complicated. In this case, combinations of successful attacks on all subsets of the target relays should be considered. For example, when two relays are attacked simultaneously, r,′r″(R c ={r,′r″})), its expected load loss function is given by:

[0190] ψ c (b) = π r′ (b r′ ,b s )π r″ (b r″ ,b s ) * ({r,′r″})

[0191] +π r′ (b r′ ,b s )(1-π r″ (b r″ ,b s ))ζ * (r′)

[0192] +(1-π r′ (b r′ ,b s ))π r″ (b r″ ,b s ) * (r″).

[0193] The same method is also applicable to calculate the expected load loss when there are more simultaneous attacks. It is worth mentioning that in formula (7), we choose π r (b r ,b s ) can ensure that in all attack scenarios, ψ c(b) are all convex; b represents a preventive action.

[0194] The optimal prevention budget allocation reinforcement model is:

[0195] Objective function:

[0196] Constraints:

[0197]

[0198]

[0199] τ R (b t )=τ i -τ(b t ), (17)

[0200]

[0201]

[0202] Among them, formula (15) represents the budget constraint allocation constraint, formula (16) represents the constraint between the enhanced budget allocation and the probability of successful attack, formula (17) to formula (18) represent the relationship constraint between the system recovery time and the budget allocation; formula (19) represents the attacker's expected value of power shortage in C for different attack scenarios.

[0203] By introducing auxiliary variable E, the optimal prevention budget allocation reinforcement model is transformed into:

[0204] Objective function:

[0205] Constraints:

[0206]

[0207]

[0208] τ R (b t )=τ i -τ(b t ), (17)

[0209]

[0210]

[0211]

[0212] The double-layer optimization of the optimal prevention budget allocation reinforcement model is converted into a single-layer optimization. When solving the problem, a relatively mature existing optimization solver can be used.

[0213] Example 3

[0214] This embodiment proposes a relay protection security reinforcement system for dealing with active and passive value attacks. This system is used to implement the relay protection security reinforcement method for dealing with active and passive value attacks proposed in embodiments 1 and 2. The structure is shown in Figure 3 ,include:

[0215] A modeling unit, used for modeling active set value attacks and passive set value attacks of relay protection of distribution system respectively;

[0216] Vulnerability analysis unit, used to analyze relay protection vulnerabilities and preliminarily propose safety precautions;

[0217] An attack prevention identification budget allocation unit determines attack prevention identification budget allocation based on security precautions;

[0218] A load loss calculation unit, used to extract attack scenarios and calculate the load loss caused by the attack scenarios according to the distribution system parameters and attack prevention identification budget allocation;

[0219] The power shortage expected value building unit uses load loss to build the power shortage expected value in the attack scenario;

[0220] The optimal prevention budget allocation unit takes the minimum expected value of power shortage as the objective function, and uses the relationship constraints between budget allocation, reinforcement budget allocation and successful attack probability, and system recovery time and budget allocation to establish and solve the optimal prevention budget allocation reinforcement model, and output the prevention budget allocation of each relay in the distribution system and the prevention budget of the entire distribution system;

[0221] The safety reinforcement unit performs safety reinforcement on relay protection based on the prevention budget allocation of each relay in the distribution system and the prevention budget of the entire distribution system.

[0222] The terms used to describe the positional relationship in the accompanying drawings are only used for illustrative purposes and should not be construed as limiting this patent; it is clear that the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not limitations on the implementation methods of the present invention. For ordinary technicians in the relevant field, other different forms of changes or modifications can be made on the basis of the above description. It is not necessary and impossible to list all implementation methods here. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the claims of the present invention.

Claims

1. A method for reinforcing relay protection security against active and passive value attacks, characterized in that: include: S1: Model the active set value attack and passive set value attack of distribution system relay protection respectively; S2: Based on modeling, analyze the relay protection vulnerabilities and preliminarily propose safety precautions; S3: Determine attack prevention identification budget allocation based on security precautions; S4: extract attack scenarios, calculate the load loss caused by attack scenarios based on distribution system parameters and attack prevention identification budget allocation; S5: Use load loss to construct the expected value of power shortage in the attack scenario; S6: Taking the minimum expected value of power shortage as the objective function, and taking the relationship between budget allocation, reinforcement budget allocation and successful attack probability, and system recovery time and budget allocation as constraints, establish and solve the optimal prevention budget allocation reinforcement model, and output the prevention budget allocation of each relay in the distribution system and the prevention budget of the entire distribution system; S7: Safety reinforcement of relay protection is performed based on the prevention budget allocation of each relay in the distribution system and the prevention budget of the entire distribution system.

2. The method for reinforcing relay protection security against active and passive value attacks according to claim 1 is characterized in that: The distribution system includes a set of nodes N = {1, ..., N}, a set of lines L = {ji|i∈N, j∈N} and a set of relay protection devices R = {1, ..., R}, where N represents the number of nodes, N(i) represents the number of buses connected to bus i, and R is the total number of relay protection devices; when a permanent short circuit fault occurs on line ji∈L, the inverse time characteristic of the overcurrent relay satisfies: Among them, I r,f represents the fault current seen by relay r; Represents the adjusted pickup current; TD r Indicates the time interval setting value, A r ,B r ,p r All represent constants of the curve characteristics selected by relay r; Modeling active constant-value attacks: in, is the pickup current of relay r after being attacked by active setting, x a is the active fixed-value attack scalar; γ>0, γ is the attack design parameter considering measurement noise; the attacker launches a successful active fixed-value attack by adjusting the attack design parameter γ; Modeling passive fixed-value attacks: t r′,f -t r,f ≥δt, Among them, t r′,f represents the operation time delay of relay r when fault f occurs downstream of relay r′, δt is the coordination time interval, which represents the minimum time difference between two relays; r′=1, r′=R′ are the nearest backup relay and the farthest backup relay of relay r respectively; δt p >δt+γ, in, is the operation delay of relay r after passive fixed value attack; δt p This is a passive fixed-value attack time delay; the attacker changes the adjustable setting of the relay TD r ,A r ,B r ,p r , launching a successful passive fixed-value attack.

3. The method for reinforcing relay protection security against active and passive value attacks according to claim 2 is characterized in that: The relay protection vulnerabilities include: communication network, intermediate relay firmware and local access network; wherein, the attacker penetrates the communication network through the connection point, bypasses the firewall to connect to the relay, and modifies the adjustable settings of the relay to achieve the attack; the attacker launches a Trojan attack on the firmware installed on the intermediate relay; the attacker connects to the relay through the local access network and modifies the adjustable settings of the relay to achieve the attack; The security precautions include: precautions implemented on a single relay and precautions implemented on the entire distribution system; among them, the precautions implemented on a single relay include: installing physical barriers, upgrading firmware security, multi-factor authentication and installing firewalls; the precautions implemented on the entire distribution system include: communication data encryption, firmware update verification and installation of an attack identification system.

4. The method for reinforcing relay protection security against active and passive value attacks according to claim 3 is characterized in that: The attack prevention identification budget allocation vector is defined as: b=[b1,…,b r ,…,b R ,b s ,b t ], Among them, b r Allocate a budget for preventive measures for a single relay r; b s A budget allocation for preventive measures for the entire distribution system; b. t Indicates the allocated budget for the installed attack identification system; The defender action space is defined as: B represents the total available strengthening budget; Relating the probability of successful attack of each relay to the attack prevention identification budget invested, we get: Among them, π r (b r ,b s ) represents the probability of a successful attack on relay r, which is a function of the reinforcement budget allocated to relay r and the entire distribution system; π r (b r ,b s ) is decomposed into two probabilities: Depends on the precautions of each relay, π s (b s ) depends on the precautions of the entire distribution system; α∈[0,1] indicates the importance of each probability relative to another probability; λ r ,μ is the design parameter, and the budget is associated with the reinforcement measures of relay r and the entire distribution system respectively; μ i Initial safety level of relay r and distribution system respectively; Distribution system recovery time after installing the attack identification system: t R (b) t )=τ i -t(b t ) Among them, τ(b t ) represents the recovery time of the power distribution system without the attack identification system installed, τ i represents the total recovery time, τ(b t ) is calculated as: Where Δτ represents the recovery time reduced by installing the attack identification system, B t is the budget needed to install the system.

5. The method for reinforcing relay protection security against active and passive value attacks according to claim 4 is characterized in that: Let c∈C={1,…,C} be the attack scenario, C be the set of all attack scenarios, C be the total number of attack scenarios, and subset is the maximum set of attacked intermediate relays in attack scenario c; In attack scenario c, the attacker can attack at most all relays r∈R c , if the attacker's budget allows to manipulate at most K intermediate relays, whether it is an active fixed value attack or a passive fixed value attack, the total number of attack scenarios C is: in, represents the sum of m attacked relays in the relays; Assume that the load loss caused by attack scenario c is * (R c ) indicates that the calculation expression is: in, represents the total load loss, represents the load loss caused by non-attack scenarios, and O is the set of operation decision variables.

6. The method for reinforcing relay protection security against active and passive value attacks according to claim 5 is characterized in that: According to the distribution system parameters and attack prevention identification budget allocation, the process of calculating the load loss caused by the attack scenario is: Construct the objective function, the expression is: Among them, o∈O={p grid ,q grid ,p,q,y,e,s}, including active power of the distribution system, reactive power active line flow, reactive line flow, service load ratio. Line power-on status and switch status; Constraints: Equations (1) to (5) represent the power balance constraints of the distribution system nodes, p ij ,q ij are the active power flow and reactive power flow of line ij; is their maximum value; is the total demand of node i, y i is the load ratio of the service on the node; binary variable e ij represents the energy flow from node i to j, which should be zero for isolated lines; Equations (6) to (13) indicate the relay travel required to isolate the system fault according to the state of the operating relay. Equation (14) indicates that if relay r is successfully targeted by an active fixed-value attack, the attacker will immediately open its associated switch s ij ;f ij represents the fault location parameter; ε r Indicates the operating status of relay r; For switch ij The initial state can only trip when there is an operation of the relay protection device; the transformation matrix B r,ij Connect the relay r to the distribution line ij where the relay is located; Fault location parameter f ij And relay groups under active customized attacks and passive customized attacks As the input to solve the objective function and constraints, solve for the load loss ζ * (R c ).

7. The method for reinforcing relay protection security against active and passive value attacks according to claim 6 is characterized in that: Using load loss, the expected value of power shortage in the attack scenario is constructed as follows: you c (b)=ψ c (b)·t R (b) t ), Among them, u c (b) represents the expected value of insufficient power in the attack scenario; ψ c (b) represents the expected load loss, which is a function of the probability vector of the relay attack probability π(b), π c (b) represents all r∈R c ; The successful attack probability vector is a combined probability function.

8. The method for reinforcing relay protection security against active and passive value attacks according to claim 7 is characterized in that: The optimal prevention budget allocation reinforcement model is: Objective function: Constraints: t R (b) t )=τ i -t(b t ), (17) Among them, formula (15) represents the budget constraint allocation constraint, formula (16) represents the constraint between the enhanced budget allocation and the probability of successful attack, formula (17) to formula (18) represent the relationship constraint between the system recovery time and the budget allocation; formula (19) represents the attacker's expected value of power shortage in C for different attack scenarios.

9. The method for reinforcing relay protection security against active and passive value attacks according to claim 8 is characterized in that: The auxiliary variable E transforms the optimal prevention budget allocation reinforcement model into: Objective function: Constraints: t R (b) t )=τ i -t(b t ), (17) The two-level optimization consisting of the optimal prevention budget allocation reinforcement model is converted into a single-level optimization.

10. A relay protection safety reinforcement system for responding to active and passive value attacks, characterized in that: The system is used to implement the relay protection security reinforcement method for responding to active and passive value attacks as described in claim 1, comprising: A modeling unit, used for modeling active set value attacks and passive set value attacks of relay protection of distribution system respectively; Vulnerability analysis unit, used to analyze relay protection vulnerabilities and preliminarily propose safety precautions; An attack prevention identification budget allocation unit determines attack prevention identification budget allocation based on security precautions; A load loss calculation unit, used to extract attack scenarios and calculate the load loss caused by the attack scenarios according to the distribution system parameters and attack prevention identification budget allocation; The power shortage expected value building unit uses load loss to build the power shortage expected value in the attack scenario; The optimal prevention budget allocation unit takes the minimum expected value of power shortage as the objective function, and uses the relationship constraints between budget allocation, reinforcement budget allocation and successful attack probability, and system recovery time and budget allocation to establish and solve the optimal prevention budget allocation reinforcement model, and output the prevention budget allocation of each relay in the distribution system and the prevention budget of the entire distribution system; The safety reinforcement unit performs safety reinforcement on relay protection based on the prevention budget allocation of each relay in the distribution system and the prevention budget of the entire distribution system.

Images