A method for determining the threat level and defending against electric vehicle charging systems under network attacks
By analyzing the mobile characteristics and charging device factors of electric vehicles, establishing a probability model to calculate the probability and threat level of network attacks, and formulating mitigation strategies, it solves the problem of insufficient identification and defense of electric vehicle network attacks in the smart grid and improves charging System security and reliability.
Patent Information
- Application Number
- CN202111657660.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-12-30
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2041-12-30
AI Technical Summary
The prior art has failed to effectively identify and defend against cyber attacks against electric vehicles in smart grids, especially considering the high mobile characteristics of electric vehicles and the information security threats of distributed charging piles. Traditional isolation protection solutions exist in power services. Issues of insufficient availability constraints and physical backups.
A method for determining and defending the threat level of electric vehicle charging system under cyberattacks is proposed. By analyzing the mobile characteristics, charging equipment number and occupancy of electric vehicles, establishing a probability model to calculate the probability of transmission and threat level of network attacks, and formulating mitigation The strategy of attack propagation includes disconnecting high-risk vehicles when security thresholds are met.
The mobile characteristics of electric vehicles are effectively considered, which improves the rationality of the cyber attack propagation process and the accuracy of threat level calculations, slows down the spread of cyber attacks, and improves the reliability and security of the charging system.
Smart Images

Figure CN114297584B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of power system network attacks, and specifically relates to a method for determining the threat level of an electric vehicle charging system under network attacks and a defensive response method. Background Art
[0002] As electric vehicles become more intelligent and connected, the risk of electric vehicle infrastructure being compromised is increasing. When vulnerabilities in electric vehicle infrastructure are not properly addressed, the security of the smart grid will also be at great risk.
[0003] Take interruption attacks as an example. Interruption attacks damage information transmission, causing users to be unable to stably obtain system information resources. At the same time, they can also damage the system hard disk and communication lines. In charging equipment communication, attackers mainly damage the charging pile body, such as the 4G module and the card reader interface. Since distributed electric vehicle charging piles are mainly based on 4G public network wireless communication during operation, damaging the base stations around the charging piles is also one of the targets of interruption attacks.
[0004] Based on these characteristics, cyberattacks against EVs could spread to the grid infrastructure, starting with the EV supply equipment (EVSE) and all the way to the utility system. Traditional isolation-based protection schemes do not work in smart grids because power services are subject to availability constraints and few components have physical backups.
[0005] At present, the identification and defense of cyber attacks at home and abroad are mainly studied from a wide range of targets on the information side and the power grid side, without fully considering the research on cyber attacks against a specific object, and the analysis, control and planning of the two lack coordination. Adding electric vehicles to the interconnected environment of the equipment requires a strong cyber attack response and preparation strategy. At the same time, the high mobility of electric vehicles may further spread potential cyber attacks, which will also make the information security of widely distributed electric vehicle charging piles more threatened. Various threat factors continue to change over time. Summary of the invention
[0006] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a method for determining and defending the threat level of an electric vehicle charging system under a network attack. First, it is assumed that there are two propagation modes of a network attack, one is the propagation when the electric vehicle and the charging device communicate interactively; the other is the propagation of the network attack between the communication networks of the electric vehicle charging devices. During interactive propagation, the mobility characteristics of the electric vehicle and the load conditions of each charging system are mainly considered; when propagating between communication networks, the probability of the network attack successfully bypassing the firewall and the number of communication relays between the charging systems are mainly considered. This mitigation strategy fully demonstrates the control of the changes in hazards (threats) of the charging system under a network attack based on electric vehicles over time. This method can take into account the influence of the uncertainty factors of the mobility characteristics of electric vehicles, so that the calculation results of the threat level of the charging system under a network attack based on electric vehicles over time are more convincing. Modeling with a probability model method can make the analysis more reasonable and the solution process clearer.
[0007] A method for determining the threat level of an electric vehicle charging system under a network attack and a defense response method, comprising the following steps:
[0008] Step 1: Considering the mobility characteristics of electric vehicles, based on the battery charge, the number of charging devices in the charging system, and the occupancy rate of charging devices, the possibility of switching charging devices after the first charging failure of the electric vehicle is obtained;
[0009] Step 2: According to the analysis in step 1 and in combination with the distances between different charging systems, the attraction coefficients between different charging systems are determined, and a probability model of the electric vehicle switching charging system is obtained;
[0010] Step 3: Based on the switching system probability model, determine the probability of different charging systems being harmed under network attacks;
[0011] Step 4: Based on the victimization probability, the probability of network attack propagation accompanying electric vehicles is proposed, and the calculation formula of the threat level of the charging system under network attack over time is obtained;
[0012] Step 5: Propose corresponding strategies to mitigate the spread of attacks, analyze whether they conform to the characteristics of network attack propagation, and verify the effectiveness of the strategies.
[0013] Furthermore, in step 1, based on the mobility characteristics of electric vehicles, the main influencing factors for switching charging equipment include the battery charge, the number of charging equipment in the charging system, and the charging equipment occupancy rate. It represents the possibility of the kth electric vehicle to move and switch charging equipment in the i-th charging system (Electric Vehicle Charging System, EVCS). The specific formula is:
[0014]
[0015] N i is the number of charging devices in the i-th EVCS, is the charge of the kth attacked vehicle in the i-th EVCS, O i is the occupancy rate of charging equipment in the i-th EVCS.
[0016] Furthermore, in step 2, the attraction coefficient μ between different charging systems ij It is determined by the number of charging devices in the charging system, the occupancy rate of charging devices and the distance between different charging systems. The specific expression is:
[0017]
[0018] Where: N j is the number of charging devices in the jth EVCS, O j is the occupancy rate of charging equipment in the jth EVCS, d ij is the distance between the i-th charging system and the j-th charging system.
[0019] Furthermore, in step 2, the probability M of the electric vehicle switching the charging system ij The probability of switching charging equipment by electric vehicles is related to the attraction coefficient between different charging systems. The specific formula is:
[0020]
[0021] Furthermore, in step 3, the probability of different charging systems being harmed under network attacks is j (0) is:
[0022]
[0023] Where E j (0) represents the initial damage of the jth EVCS, E j (0) = 1 means victimization, E j (0) = 0 means no harm; P(E j (0) = 1) represents the initial victim probability of the jth EVCS; α represents the probability that a network attack successfully bypasses the firewall, and Θ is a set of EVCSs that detect danger.
[0024] Furthermore, in step 4, the probability of the charging system being harmed at Δt under the network attack is obtained along with the probability of the electric vehicle being attacked. j (Δt) is:
[0025]
[0026] Where β ijrepresents the probability that the attack is transferred from the i-th EVCS to the j-th undamaged EVCS at the communication layer.
[0027] Furthermore, in step 4, with n as the step length and K as the total step length, the victim probability at the current moment changes based on the victim probability at the previous moment, so the victim probability (threat level) of the charging system under network attack changes over time ξ j (nΔt) calculation formula:
[0028]
[0029] Furthermore, in step 5, the strategy to slow down the spread of cyber attacks among electric vehicle charging systems is to keep the vehicles connected when the set safety threshold is met; if the safety threshold is exceeded, the high-risk vehicles are disconnected. ij =1 means connection, B ij =0 means disconnection, Ψ is the set total safety threshold, y j is the algebraic expression in Eq. The simplified expression of Φ is the simplified expression y j Safety threshold:
[0030]
[0031]
[0032]
[0033] B ij ·max{ln(ξ(nΔt)·β ij )}=0 (10)
[0034] The beneficial effects achieved by the present invention are: 1) the mobility factor of electric vehicles is taken into account, so that the network attack propagation process based on electric vehicles is more reasonable; 2) the calculation formula result of the threat level of the charging system under network attacks that changes over time is more reasonable. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] Figure 1 Schematic diagram of the electric vehicle system architecture under network attack in an embodiment of the present invention.
[0036] Figure 2 It is a flowchart of the steps of a method for mitigating the threat level of an electric vehicle charging system under a network attack in an embodiment of the present invention.
[0037] Figure 3 This is the threat infection status of P2, C2, and C3 when no measures are taken in the embodiment of the present invention.
[0038] Figure 4: The probability of attack propagation under different parameter settings taking P2 as an example in the embodiment of the present invention.
[0039] Figure 5 This is the threat infection situation after the cutting measures are taken in the embodiment of the present invention. DETAILED DESCRIPTION
[0040] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings.
[0041] like Figure 2 The embodiment of the present invention provides a threat level determination and defense response strategy for an electric vehicle charging system under a network attack, including the following steps:
[0042] (1) First, analyze the mobility characteristics of electric vehicles. Analyze the indicators that affect the mobility characteristics of electric vehicles, and perform principal component analysis on the main elements such as the number of charging piles, the occupancy rate of charging piles, the current state of charge of the vehicle, the interval time of charging piles, the charging power, and the interval time of charging piles to obtain the component matrix:
[0043] Ingredient Matrix:
[0044]
[0045]
[0046] In the first column, the maximum values are 0.882, 0.802, and 0.705, which correspond to the distance between charging piles, the occupancy rate of charging piles, and the number of charging piles, respectively. Therefore, the mobility factor of electric vehicles can be attributed to the load of charging piles. In the second column, the maximum value of 0.798 corresponds to the current state of charge, and the mobility factor can be attributed to the power of electric vehicles.
[0047] use It represents the possibility of the kth electric vehicle moving and switching charging equipment in the i-th charging system (Electric Vehicle Charging System, EVCS):
[0048]
[0049] Where: N i is the number of charging devices in the i-th EVCS, is the charge of the kth attacked vehicle in the i-th EVCS, O i is the occupancy rate of charging equipment in the i-th EVCS.
[0050] (2) The probability of an electric vehicle switching from the i-th EVCS to the j-th EVCS is:
[0051]
[0052] Where μ ij is the attraction coefficient of the jth EVCS to the i-th EVCS:
[0053]
[0054] Where: N j is the number of charging devices in the jth EVCS, O j is the occupancy rate of charging equipment in the jth EVCS, d ij is the distance between the i-th charging system and the j-th charging system.
[0055] (3) Assume that at time t = 0, x electric vehicle charging equipment (EVCE) in the i-th EVCS are detected to be under network attack, while the remaining EVCEs may be attacked but not detected. The initial victim probability of the i-th EVCS that has been attacked is ξ i (0) = 1, then the probability of the jth charging system being harmed is proportional to the probability of the vehicle transferring from the ith to the jth, and Θ represents a set of detected dangerous EVCS. Based on formula (2), the probability of the jth charging system being harmed is:
[0056]
[0057] Where α represents the probability that a network attack successfully bypasses the firewall.
[0058] (4) Let the probability of attacking the jth EVCS at time Δt be ξ j (Δt)=P(E j (Δt) = 1), because the probability of the attack succeeding at the jth EVCS is composed of the probability of infection at Δt when the initial infection is present and the probability of infection at Δt when the initial infection is not present, so let the probability of infection at Δt when the initial infection is not present be U j =P{E j (Δt)=1|E j (0)=0}, and the initial infection time is Δt and the infection time is P{E j (Δt)=1|E j (0)=1}=1, so ξ j (Δt)=U j ×P(E j (0) = 0) + 1 × P (E j (0) = 1), where:
[0059]
[0060] Where β ijrepresents the probability that the attack is transferred from the i-th EVCS to the j-th undamaged EVCS at the communication layer:
[0061]
[0062] Where: γ represents the probability that a transmission attack successfully compromises the EVCE at the destination; η represents the probability that the attack propagates through the communication relay in the communication network; D ij Represents the hop distance, which is the minimum number of relays connecting the i-th EVCS and the j-th EVCS. A relay can be a small cell base station or a large cell base station in a wireless access network, or a router in a backhaul. In the example, γ is set to 0.05 and η is set to 0.05.
[0063] Substituting equation (5) into ξ j (Δt), we get:
[0064]
[0065] Until the network patches the vulnerability, the threat level will increase over time. Therefore, the general threat level formula is obtained in a similar manner to formula (7), as shown in formula (8):
[0066]
[0067] When the set safety threshold is met, the connection is maintained; if the safety threshold is exceeded, the high-risk vehicle is disconnected to slow down the spread of network attacks and ensure the reliability and safety of the charging system. ij Indicates the connection between the car from the i-th charging system and the charging pile of the j-th charging system, B ij =1 means connection, B ij = 0 means disconnection. The objective function is that the threat level of network attack is within the safety threshold Ψ, and the constraint is that as many cars as possible remain connected:
[0068]
[0069] in Yes j (nΔt) is positively correlated, so Taking the logarithm, we have:
[0070]
[0071] So let At the same time, in order to keep as many cars connected as possible, the constraints are:
[0072]
[0073]
[0074] B ij ·max{ln(ξ(nΔt)·β ij )}=0 (13)
[0075] A specific example is given below.
[0076] Take 5 EVCE systems as an example, the initial parameters are: η = 0.05, γ = 0.05, Δt = 1s, α = 0.1, the danger threshold is 0.2, and the solution is obtained by MATLAB. The results are as follows:
[0077] Table 1 Proportion of electric vehicle movement between charging systems
[0078]
[0079]
[0080] Table 2 Hop distance in EVCS network
[0081] EVCE P1 P2 C1 C2 C3 P1 - 1 2 1 3 P2 1 - 2 3 1 C1 2 2 - 1 1 C2 1 3 1 - 1 C3 3 1 1 1 -
[0082] Table 3 Probability of an attack being transmitted from the i-th damaged EVCS to the j-th undamaged EVCS within Δt = 1s
[0083] EVCS P1 P2 C1 C2 C3 Other P1 0.0204 0.0275 0.0275 0.0371 0.0319 0.0430 P2 0.0275 0.0371 0.0204 0.0371 0.0371 0.0275 C1 0.0275 0.0204 0.0275 0.0371 0.0371 0.0371 C2 0.0371 0.0371 0.0371 0.0371 0.0204 0.0204 C3 0.0319 0.0371 0.0371 0.0204 0.0275 0.0319 Other 0.0430 0.0275 0.0371 0.0204 0.0319 0.0275
[0084] Table 4 The probability β of the attack being transmitted from the i-th damaged EVCS to the j-th undamaged EVCS at the initial time (t=0) ij
[0085]
[0086]
[0087] Table 5 Initial threat ratio of charging piles
[0088] <![CDATA[β ij ]]> 0.1 0.2 0.3 0.4 0.5 0.05 P1 1 1 1 1 1 1 P2 0.050594 0.05107 0.051545 0.05202 0.052495 0.050357 C1 1 1 1 1 1 1 C2 0.010217 0.015438 0.020656 0.025871 0.031085 0.007606 C3 0.00799 0.013468 0.018942 0.02441 0.029873 0.005249
[0089] By the attached Figure 3 It can be seen that if no corresponding measures are taken when the network attack spreads, the network attack will spread between vehicles and between vehicles and charging systems, eventually causing all devices to be harmed. Figure 4 It can be seen that the initial sensitivity to network attacks is different, and the propagation speed of network attacks is also different. The lower the initial setting value, the slower the network attack propagation speed. However, if no measures are taken, all devices will eventually be harmed. Figure 5This is the infection situation after the cut-off measures are taken. It can be seen from the figure that cutting off high-risk vehicles after the danger value reaches the set safety threshold can effectively slow down the spread of threats and reduce the threat level. This mitigation strategy fully demonstrates the control of the changes in the hazards (threats) of cyber attacks based on electric vehicles over time, which shows the superiority of this method.
[0090] Moreover, the use of probabilistic models for modeling is more adaptable and more convincing in both the calculation process and the results when the calculation contains uncertain factors such as the mobility characteristics of electric vehicles.
[0091] The above description is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiment. Any equivalent modifications or changes made by ordinary technicians in this field based on the contents disclosed by the present invention should be included in the protection scope recorded in the claims.
Claims
1. A method for determining the threat level of an electric vehicle charging system under a network attack and a defense response method, characterized in that: The method comprises the following steps: Step 1: Considering the mobility characteristics of electric vehicles, based on the battery charge, the number of charging devices in the charging system, and the occupancy rate of charging devices, the possibility of switching charging devices after the first charging failure of the electric vehicle is obtained; In step 1, based on the mobility characteristics of electric vehicles, the main influencing factors for switching charging equipment include the battery charge, the number of charging devices in the charging system, and the charging device occupancy rate, which are expressed as δ i k (t) represents the possibility of the kth electric vehicle moving and switching charging equipment in the i-th charging system EVCS. The specific formula is: N i is the number of charging devices in the i-th EVCS, is the charge of the kth attacked vehicle in the i-th EVCS, O i is the occupancy rate of charging equipment in the i-th EVCS; Step 2: According to the analysis in step 1 and in combination with the distances between different charging systems, the attraction coefficients between different charging systems are determined, and a probability model of the electric vehicle switching charging system is obtained; In step 2, the attraction coefficient μ between different charging systems ij It is determined by the number of charging devices in the charging system, the occupancy rate of charging devices and the distance between different charging systems. The specific expression is: Where: N j is the number of charging devices in the jth EVCS, O j is the occupancy rate of charging equipment in the jth EVCS, d ij is the distance between the i-th charging system and the j-th charging system; In step 2, the probability M of the electric vehicle switching charging system ij The probability of switching charging equipment by electric vehicles is related to the attraction coefficient between different charging systems. The specific formula is: Step 3: Based on the switching system probability model, determine the probability of different charging systems being harmed under network attacks; In step 3, the probability of different charging systems being harmed under network attacks is j (0) is: Where E j (0) represents the initial damage of the jth EVCS, E j (0) = 1 means victimization, E j (0) = 0 means no harm; P(E j (0)=1) represents the initial victimization probability of the jth EVCS; α represents the probability of a network attack successfully bypassing the firewall, and Θ is a set of EVCSs that detect danger; Step 4: Based on the victimization probability, the probability of network attack propagation accompanying electric vehicles is proposed, and the calculation formula of the threat level of the charging system under network attack over time is obtained; In step 4, the probability of the charging system being harmed at Δt under the network attack is obtained along with the probability of the network attack propagation of the electric vehicle. j (Δt) is: Where β ij represents the probability that the attack is transferred from the i-th EVCS to the j-th undamaged EVCS at the communication layer; In step 4, n is the step length and K is the total step length. The victim probability at the current moment changes based on the victim probability at the previous moment. Therefore, the victim probability of the charging system under network attack changes with time ξ j (nΔt) calculation formula: Step 5: Propose corresponding strategies to mitigate the spread of attacks, analyze whether they conform to the characteristics of network attack propagation, and verify the effectiveness of the strategies.
2. According to the method of claim 1, the threat level determination and defense response method of an electric vehicle charging system under a network attack is characterized by: In step 5, the strategy to mitigate the spread of cyber attacks among EV charging systems is to keep the vehicles connected while meeting the set security threshold; If the safety threshold is exceeded, the high-risk vehicle will be disconnected, B ij =1 means connection, B ij =0 means disconnection, Ψ is the set total safety threshold, y j is the algebraic expression in Eq. The simplified expression of Φ is the simplified expression y j Safety threshold: B ij ·max{ln(ξ(nΔt)·β ij )}=0 (10).
Citation Information
Patent Citations
Charging pile attack clustering detection method based on GRU model
CN112398862A
Attack and defense system and method based on electric vehicle power distribution network information physical system
CN112688315A
Technologies for detecting abnormal activities in an electric vehicle charging station
US20200162487A1
