Attack detection and optimal control method for mixed traffic system based on vehicle-infrastructure cooperation

Abstract

The application provides a kind of hybrid traffic system attack detection and optimization control method based on vehicle-road cooperation. The method comprises the following steps: a roadside unit obtains the motion information of a vehicle driving on a target road through network connection technology; according to the historical trusted motion information of the driving vehicle and the trusted motion information of the surrounding vehicles of the driving vehicle, the roadside unit generates the motion prediction information of the driving vehicle through the self-updating distributed estimator; according to the motion prediction information of the driving vehicle and the preset vehicle motion model, the composite attack detector of the roadside unit is constructed, and the motion information of the driving vehicle is divided into abnormal data and non-abnormal data through the composite attack detector; according to the abnormal data and the non-abnormal data, the roadside unit generates an attack detection decision and broadcasts the trusted vehicle information and the attack flag information of the driving vehicle; according to the trusted vehicle information and the attack flag information of the driving vehicle, the intelligent network connected automobile dynamically optimizes the vehicle control strategy of itself.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of automatic driving technology and intelligent transportation system, and particularly relates to a mixed traffic system attack detection and optimization control method based on vehicle-road cooperation, an electronic device and a storage medium. BACKGROUND

[0002] Automobiles and transportation systems are developing towards intelligentization and networking, and the application of networking technology provides strong support for vehicle-vehicle cooperation and vehicle-road cooperation. Existing research has shown that appropriate cooperation strategies can effectively suppress the fluctuations of mixed traffic flow, reduce vehicle conflicts, improve road capacity and reduce energy consumption. However, with the gradual landing of intelligent and connected vehicles, there will inevitably be a long transition period of mixed traffic systems containing both intelligent and connected vehicles and manually driven vehicles. At the same time, as a cyber-physical system, networking technology brings connectivity to intelligent and connected vehicles and mixed traffic systems, but also the possibility of cyber attacks.

[0003] Currently, there have been related researches on deception attack detection and defense strategies in transportation systems. Detection methods can be divided into observer-based detection, filter-based detection and partial differential equation-based detection. At the same time, distributed estimators are used for vehicle motion state estimation in multi-sensor networks. However, the above research objects are all pure intelligent and connected vehicle platoons, and few studies mention the network security problem of mixed traffic scenarios. However, the above results cannot be directly used for mixed traffic systems, because: first, the behavior of manually driven vehicles is different from that of intelligent and connected vehicles, and the motion of manually driven vehicles is not affected by networking information, and manually driven vehicles exhibit motion heterogeneity due to different driving styles of drivers; second, the communication range of the vehicle-mounted communication device is limited, and the connectivity of the mixed traffic system is poor due to the presence of manually driven vehicles; third, manually driven vehicles do not have the ability to perceive the surrounding environment, and the information they send only contains their own state information, that is, there is less information about the front and rear vehicles of manually driven vehicles on the network. These specialities make the methods for pure intelligent and connected vehicle platoons not directly applicable to mixed traffic scenarios. SUMMARY

[0004] In view of the above problems, the present application provides a mixed traffic system attack detection and optimization control method based on vehicle-road cooperation, an electronic device and a storage medium, which can at least solve one of the above problems.

[0005] According to a first aspect of the present application, a mixed traffic system attack detection and optimization control method based on vehicle-road cooperation is provided, characterized in that it comprises:

[0006] The roadside unit obtains motion information of a target vehicle on a target road in a mixed traffic system through a network connection technology, wherein the target vehicle includes an artificial driving vehicle and an intelligent connected vehicle, and the motion information includes speed information and position information of the target vehicle;

[0007] According to the historical reliable motion information of the target vehicle and the reliable motion information of surrounding vehicles of the target vehicle, motion prediction information of the target vehicle is generated by a self-updating distributed estimator of the roadside unit;

[0008] According to the motion prediction information of the target vehicle and a preset vehicle motion model, a composite attack detector of the roadside unit is constructed, and the motion information of the target vehicle is divided into abnormal data and non-abnormal data by the composite attack detector;

[0009] According to the abnormal data and the non-abnormal data, an attack detection decision is generated by the roadside unit, and reliable vehicle information and attack flag information of the target vehicle are broadcasted in the mixed traffic system;

[0010] According to the reliable vehicle information and the attack flag information of the target vehicle, a vehicle control strategy of the intelligent connected vehicle is dynamically optimized.

[0011] According to the embodiment of the present application, the roadside unit obtains the motion information of the target vehicle on the target road in the mixed traffic system through the network connection technology, which includes:

[0012] The roadside unit receives the speed information and the position information of the artificial driving vehicle broadcasted by the artificial driving vehicle through a vehicle-mounted network connection device in the mixed traffic system through the network connection technology;

[0013] The roadside unit receives the speed information and the position information of the intelligent connected vehicle and the speed information and the position information of surrounding vehicles broadcasted by the intelligent connected vehicle through a vehicle-mounted network connection device in the mixed traffic system through the network connection technology, wherein the speed information and the position information of the surrounding vehicles are obtained by a vehicle-mounted sensor of the intelligent connected vehicle.

[0014] According to the embodiment of the present application, the roadside unit generates the motion prediction information of the target vehicle according to the historical motion information of the target vehicle and the reliable motion information of the surrounding vehicles of the target vehicle through the self-updating distributed estimator of the roadside unit, which includes:

[0015] The historical reliable motion information of the target vehicle is processed by using a longitudinal motion model of the target vehicle to obtain acceleration prediction information of the target vehicle;

[0016] Based on the acceleration prediction information of the target vehicle, the motion information of the target vehicle and the reliable motion information of the surrounding vehicles, a vehicle platoon system model with disturbance is constructed, wherein the disturbance conforms to a Gaussian distribution;

[0017] The self-updating distributed estimator of the roadside unit is constructed based on a vehicle platoon system model with perturbation, and the historical reliable motion information of the traveling vehicle is processed by the self-updating distributed estimator to obtain speed prediction information and position prediction information of the traveling vehicle.

[0018] According to the embodiment of the present application, the composite attack detector of the roadside unit is constructed according to the motion prediction information of the traveling vehicle and the preset vehicle motion model, and the motion information of the traveling vehicle is divided into abnormal data and non-abnormal data by the composite attack detector, which includes:

[0019] The composite attack detector of the roadside unit is constructed according to the motion prediction information of the traveling vehicle and the preset vehicle motion model.

[0020] The non-abnormal motion information of the traveling vehicle under the non-attack scene is obtained by the Monte Carlo simulation experiment, and the non-abnormal motion information of the traveling vehicle is processed by the composite attack detector to obtain an attack alarm threshold.

[0021] The motion information of the traveling vehicle is processed by the composite attack detector based on the Kalman filter principle and the chi-square detector principle to obtain a processing result.

[0022] In the case that the processing result is greater than the attack alarm threshold, the motion information of the traveling vehicle is determined as abnormal data.

[0023] In the case that the processing result is less than or equal to the attack alarm threshold, the motion information of each vehicle is determined as non-abnormal data.

[0024] According to the embodiment of the present application, the preset vehicle motion model includes an intelligent driver model of a human-driven vehicle and an improved cooperative intelligent driver model of a smart connected vehicle.

[0025] According to the embodiment of the present application, the attack detection decision is generated by the roadside unit according to the abnormal data and the non-abnormal data, and the reliable vehicle information and the attack flag information of the traveling vehicle are broadcasted in the mixed traffic system, which includes:

[0026] The attack detection decision is generated by the roadside unit by screening the abnormal data and the non-abnormal data, wherein the attack detection decision includes the reliable vehicle information and the attack flag information of the traveling vehicle.

[0027] The reliable vehicle information and the attack flag information of the traveling vehicle are broadcasted by the roadside unit to the traveling vehicle on the target road in the mixed traffic system.

[0028] According to the embodiment of the present application, the intelligent connected vehicle dynamically optimizes its vehicle control strategy according to the reliable vehicle information and the attack flag information of the traveling vehicle, which includes:

[0029] In the case that the attack sign information of the vehicle driving in front of the intelligent connected vehicle is a preset value, the intelligent connected vehicle optimizes the parameters of the vehicle control model of the intelligent connected vehicle according to the trusted vehicle information and outputs a vehicle dynamic control strategy;

[0030] In the case that the intelligent connected vehicle is attacked by other vehicles driving on the target road, the intelligent connected vehicle resets the communication network according to the attack sign information of the attacking vehicle.

[0031] According to the embodiment of the present application, during the process of resetting the communication network by the intelligent connected vehicle, the intelligent connected vehicle is in an attack-free state for a preset time length.

[0032] According to a second aspect of the present application, an electronic device is provided, comprising:

[0033] one or more processors;

[0034] a storage device for storing one or more programs,

[0035] wherein when the one or more programs are executed by the one or more processors, the one or more processors execute the attack detection and optimization control method for the car-road cooperative hybrid traffic system.

[0036] According to a second aspect of the present application, a computer readable storage medium is provided, which stores executable instructions, the instructions being executed by a processor to execute the attack detection and optimization control method for the car-road cooperative hybrid traffic system.

[0037] The attack detection and optimization control method for the car-road cooperative hybrid traffic system provided by the present application can improve the control of the intelligent connected vehicle in the complex hybrid traffic system, improve the response ability of the intelligent connected vehicle to various risks in the complex hybrid traffic system, and improve the safety and traffic efficiency of the hybrid traffic system. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 is a structure diagram of a hybrid traffic system based on car-road cooperation according to an embodiment of the present application;

[0039] Figure 2 is a flowchart of the attack detection and optimization control method for the car-road cooperative hybrid traffic system according to an embodiment of the present application;

[0040] Figure 3 is a flowchart of obtaining the motion information of the vehicle driving on the target road in the hybrid traffic system according to an embodiment of the present application;

[0041] Figure 4is a flowchart of generating motion prediction information of a traveling vehicle according to an embodiment of the present application;

[0042] Figure 5 is a flowchart of dividing motion information of a traveling vehicle into abnormal data and non-abnormal data according to an embodiment of the present application;

[0043] Figure 6 is a flowchart of generating an attack detection decision by a road side unit according to an embodiment of the present application;

[0044] Figure 7 is a schematic diagram of an attack detection framework based on vehicle infrastructure cooperation according to an embodiment of the present application;

[0045] Fig. 8(a) is a schematic diagram of an attacker injecting positive false data to vehicle 3 without using attack detection and optimization control strategy according to an embodiment of the present application;

[0046] Fig. 8(b) is a schematic diagram of motion changes of vehicles 2-5 when an attacker injects positive false data to vehicle 3 without using attack detection and optimization control strategy according to an embodiment of the present application;

[0047] Fig. 8(c) is a schematic diagram of an attacker injecting negative false data to vehicle 3 without using attack detection and optimization control strategy according to an embodiment of the present application;

[0048] Fig. 8(d) is a schematic diagram of motion changes of vehicles 2-5 when an attacker injects negative false data to vehicle 3 without using attack detection and optimization control strategy according to an embodiment of the present application;

[0049] Fig. 9(a) is a schematic diagram of detection probabilities of a self-updating distributed estimator and a composite attack detector when tampering with current vehicle state data output by each vehicle according to an embodiment of the present application;

[0050] Fig. 9(b) is a schematic diagram of detection probabilities of a self-updating distributed estimator and a composite attack detector when tampering with front / rear vehicle state data output by a smart connected vehicle according to an embodiment of the present application;

[0051] Fig. 10(a) is a schematic diagram of an attacker injecting positive false data to vehicle 3 using attack detection and optimization control strategy according to an embodiment of the present application;

[0052] Fig. 10(b) is a schematic diagram of motion changes of vehicles 2-5 when an attacker injects positive false data to vehicle 3 using attack detection and optimization control strategy according to an embodiment of the present application;

[0053] Fig. 10(c) is a schematic diagram of an attacker injecting negative false data to vehicle 3 using attack detection and optimization control strategy according to an embodiment of the present application;

[0054] Figure 10(d) is a schematic diagram of the motion changes of vehicles 2 to 5 when an attacker injects negative false data into vehicle 3 using an attack detection and optimization control strategy according to an embodiment of the present invention.

[0055] Figure 11 The diagram illustrates an electronic device suitable for implementing a vehicle-road cooperative hybrid transportation system attack detection and optimization control method according to an embodiment of the present invention. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0057] In the technical solution disclosed in this invention, the acquisition of vehicle driving data and road data involved has been authorized by the relevant parties, and the above data is processed, applied and stored with the permission of the relevant parties. The relevant process complies with the provisions of laws and regulations, and necessary and reliable confidentiality measures have been taken, which is in line with the requirements of public order and good morals.

[0058] For spoofed data injection attacks, estimator-based detection methods are a widely accepted solution. However, in mixed traffic systems, manually driven vehicles lack perception and decision-making capabilities, resulting in poor connectivity and posing a challenge to attack detection. Furthermore, the behavior of manually driven vehicles exhibits variability and randomness, making state estimation a crucial research area. Therefore, this invention provides an attack detection and optimization control method for mixed traffic systems based on vehicle-road cooperation, addressing the existing problems in this technical field. This method also has significant engineering application value.

[0059] Figure 1 This is a schematic diagram of a hybrid transportation system based on vehicle-road cooperation according to an embodiment of the present invention.

[0060] like Figure 1 As shown, both manually driven cars and intelligent connected cars have connectivity capabilities, enabling them to communicate with surrounding vehicles and roadside units. The information sent by a manually driven car is defined as... The information sent by intelligent connected vehicles is defined as follows Where v n (k) and p n (k) represents the vehicle's speed and position information, v n-1,n (k) and p n-1,n (k) represents the speed and position measurements of the vehicle ahead by the intelligent connected vehicle, v n+1,n (k) and p n+1,n(k) is the measurement value of the intelligent connected vehicle to the rear vehicle speed and position. The measurement information source of vehicle n is defined as The communication range between vehicles is L C The vehicle set in this distance range is denoted as N n = {n-N c,f ,…,n-1,n+1,…,n+N c,r}, N c,f is the number of vehicles in the front communication range, and N c,r is the number of vehicles in the rear communication range. The road side unit with a communication unit and a computing unit connects the vehicle with the edge cloud. Generally, the road side unit has higher attack resistance. In the present application, the possibility of the road side unit being attacked by network is not considered, and only the scenario of the connected device on the vehicle side being attacked by false data injection is considered.

[0061] Figure 2 is a flowchart of a cooperative vehicle infrastructure system attack detection and optimization control method based on the embodiment of the present application.

[0062] As shown in Figure 2 , the cooperative vehicle infrastructure system attack detection and optimization control method based on the above includes operation S210 to operation S250.

[0063] In operation S210, the road side unit obtains the motion information of the vehicles driving on the target road in the hybrid traffic system through the connected technology, wherein the driving vehicles include artificial driving cars and intelligent connected cars, and the motion information includes the speed information and position information of the driving vehicles.

[0064] The road side unit, i.e. Road Side Unit, is installed on both sides of the road, uses C-V2X (Cellular-Vehicle To Everything) technology, communicates with the on-board unit (OBU, On Board Unit), and realizes vehicle identity recognition, electronic deduction, and data interaction. As shown in Figure 1 , the road side unit can also interact with the edge cloud for data processing, analysis and decision-making, and edge computing.

[0065] Through V2I (Vehicle to Infrastructure) communication, the road side unit can obtain the state information sent by each vehicle on the target road in the hybrid traffic system and the measurement information of the intelligent connected car to the front and rear vehicles.

[0066] Figure 3 is a flowchart of obtaining the motion information of the vehicles driving on the target road in the hybrid traffic system according to the embodiment of the present application.

[0067] As Figure 3 indicated above, the roadside unit obtains the motion information of the target vehicle driving on the target road in the mixed traffic system through the network connection technology, including operation S310 to operation S320.

[0068] In operation S310, the roadside unit receives the speed information and position information of the manually driven vehicle broadcasted by the manually driven vehicle through the vehicle-mounted network connection device in the mixed traffic system through the network connection technology.

[0069] In operation S320, the roadside unit receives the speed information and position information of the intelligent network connection vehicle and the speed information and position information of the surrounding vehicles broadcasted by the intelligent network connection vehicle through the vehicle-mounted network connection device in the mixed traffic system through the network connection technology.

[0070] The surrounding vehicles are vehicles located within a certain range of the target driving vehicle and may affect the motion of the target driving vehicle.

[0071] The speed information and position information of the surrounding vehicles are obtained through the vehicle-mounted sensors of the intelligent network connection vehicle.

[0072] In operation S220, the motion prediction information of the driving vehicle is generated by the self-updating distributed estimator of the roadside unit according to the historical trusted motion information of the driving vehicle and the trusted motion information of the surrounding vehicles of the driving vehicle.

[0073] The trusted motion information means that the motion information is true value.

[0074] Figure 4 is a flowchart for generating the motion prediction information of the driving vehicle according to an embodiment of the present application.

[0075] As Figure 4 indicated above, the motion prediction information of the driving vehicle is generated by the self-updating distributed estimator of the roadside unit according to the historical motion information of the driving vehicle and the trusted motion information of the surrounding vehicles of the driving vehicle, including operation S410 to operation S430.

[0076] In operation S410, the historical trusted motion information of the driving vehicle is processed by using the longitudinal motion model of the driving vehicle to obtain the acceleration prediction information of the driving vehicle.

[0077] In operation S420, based on the acceleration prediction information of the driving vehicle, the motion information of the driving vehicle and the trusted motion information of the surrounding vehicles, a vehicle queue system model with disturbance is constructed, wherein the disturbance conforms to Gaussian distribution.

[0078] In operation S430, a self-updating distributed estimator of the roadside unit is constructed based on the vehicle platoon system model with disturbance, and historical trusted motion information of the traveling vehicle is processed by the self-updating distributed estimator to obtain speed prediction information and position prediction information of the traveling vehicle.

[0079] The operations S410-S430 are further described in detail below in combination with specific embodiments.

[0080] Each vehicle follows kinematic characteristics, and in the longitudinal direction, satisfies equations shown in formulas (1) and (2):

[0081]

[0082]

[0083] wherein, is the acceleration of the vehicle, and ΔT represents a sampling period. v,n (k) and μ p,n (k) represent disturbances that can occur in the process of vehicle motion.

[0084] Therefore, the vehicle platoon system with disturbance is represented by equations shown in formulas (3) and (4):

[0085] x n (k+1) = Ax n (k) + Bu n (k) + μ n (k) (3),

[0086] y n (k) = Cx n (k) + ψ n (k) (4),

[0087] wherein, x n (k) = [v n (k), p n (k)] T , y n (k) represents state output of the system, which can be obtained through V2I. and represent disturbances and observation noise that can exist in the process of vehicle motion, which are subject to Gaussian distribution, i.e., μ n ~ N(0, Q), and ψ n ~ N(0, R).

[0088] At the same time, the system matrix is shown in formula (5):

[0089]

[0090] The state output of vehicle n obtained from the intelligent connected vehicle is shown in formula (6):

[0091] y n,j (k)=Cx n (k)+ξ n,j (k),j∈M n (6),

[0092] in, And ξ n ~N(0,E n ).

[0093] In this invention, an intelligent driver model and a collaborative intelligent driver model are used to describe the following behavior of manually driven cars and intelligent connected cars, respectively. These two different behaviors can be expressed in a general form, as shown in formula (7):

[0094]

[0095] in, and Let N be the distance from the front of vehicle n. f (N f ≤N c,f The predicted distance and relative speed obtained from the distance and speed of the vehicles are shown in formulas (8) and (9):

[0096]

[0097]

[0098] when This is referred to as intelligent connected vehicles. This indicates a manually driven car. n (t)=p n-1 (t)-p n (t) and Δv n (t)=v n (t)-v n-1 (t) represents the distance and relative speed between vehicle n and the vehicle in front. In the following behavior of intelligent connected vehicles, Δv f,nj =v n -v n-j This differs from common cooperative intelligent driver models. Through vehicle motion evolution analysis, this study found that common cooperative intelligent driver models exhibit slow convergence speeds and overshoot when approaching equilibrium. This invention uses Δv f,nj This can mitigate the delay caused by the backward propagation of queue speed fluctuations, improve the convergence speed of intelligent connected vehicle motion, and help it reach steady state more quickly. α nj With β njis the weight coefficient, satisfying is the desired inter-vehicle distance, whose mathematical expression is shown in equation (10):

[0099]

[0100] the maximum acceleration a n , the comfortable deceleration b n , the desired speed v 0,n , the minimum distance s 0,n , and the inter-vehicle time distance T n are parameters of vehicle n, whose values reflect the driving style of the vehicle. In addition, τ n is the reaction delay of the vehicle.

[0101] According to the above model, the acceleration of vehicle n at k is shown in equation (11):

[0102]

[0103] The definition of the self-updating distributed estimator is shown in equation (12):

[0104]

[0105] wherein, and are the estimated values updated according to the historical detection results, and are the model inputs obtained using the trusted historical state. The Kalman gain of the estimator is shown in equation (13):

[0106]

[0107] The residual of the estimator is shown in equations (14)-(16):

[0108]

[0109]

[0110]

[0111] wherein, κ n,j (k+1) indicates whether the measurement value of the intelligent connected vehicle to the front or rear vehicle is abnormal, as shown in equation (17):

[0112]

[0113] The attack-free state output is shown in equation (18):

[0114]

[0115] In operation S230, a composite attack detector of the roadside unit is constructed according to the motion prediction information of the running vehicle and a preset vehicle motion model, and the motion information of the running vehicle is divided into abnormal data and non-abnormal data by the composite attack detector.

[0116] According to the embodiment of the present application, the preset vehicle motion model includes an intelligent driver model of a human-driven vehicle and an improved cooperative intelligent driver model of a smart connected vehicle.

[0117] The determination condition of the motion information is shown in formula (19) and formula (20):

[0118]

[0119]

[0120] wherein ρ1 and ρ2 are prior alarm thresholds. When G n (k) < ρ1, it indicates that the self-state output by the vehicle n is normal, otherwise it indicates that there is an abnormality. When G n,j (k) < ρ2, it indicates that the measurement state of the smart connected vehicle on the vehicle n is normal, otherwise it indicates that there is an abnormality.

[0121] Figure 5 is a flowchart for dividing the motion information of the running vehicle into abnormal data and non-abnormal data according to the embodiment of the present application.

[0122] As shown in Figure 5 , the composite attack detector of the roadside unit is constructed according to the motion prediction information of the running vehicle and the preset vehicle motion model, and the motion information of the running vehicle is divided into abnormal data and non-abnormal data by the composite attack detector, which includes operation S510 to operation S550.

[0123] In operation S510, the composite attack detector of the roadside unit is constructed according to the motion prediction information of the running vehicle and the preset vehicle motion model.

[0124] In operation S520, non-abnormal motion information of the running vehicle under the attack-free scene is obtained by Monte Carlo simulation experiment, and the non-abnormal motion information of the running vehicle is processed by the composite attack detector to obtain an attack alarm threshold.

[0125] In operation S530, the motion information of the running vehicle is processed by the composite attack detector based on the Kalman filtering principle and the chi-square detector principle to obtain a processing result.

[0126] In operation S540, in the case that the processing result is greater than the attack alarm threshold, the motion information of the running vehicle is determined as abnormal data.

[0127] In operation S550, in the case that the processing result is less than or equal to the attack alarm threshold, the motion information of each vehicle is determined as non-abnormal data.

[0128] In operation S240, according to the abnormal data and the non-abnormal data, the attack detection decision is generated by the road side unit, and the trusted vehicle information and the attack flag information of the running vehicle are broadcast in the mixed traffic system.

[0129] Figure 6 is a flowchart of generating an attack detection decision by a road side unit according to an embodiment of the application.

[0130] As shown in Figure 6 , the above-mentioned generating an attack detection decision by a road side unit according to the abnormal data and the non-abnormal data, and broadcasting the trusted vehicle information and the attack flag information of the running vehicle in the mixed traffic system include operation S610 to operation S620.

[0131] In operation S610, the abnormal data and the non-abnormal data are screened by the road side unit to generate an attack detection decision, wherein the attack detection decision includes the trusted vehicle information and the attack flag information of the running vehicle.

[0132] In operation S620, the trusted vehicle information and the attack flag information of the running vehicle are broadcast by the road side unit to the running vehicle on the target road in the mixed traffic system.

[0133] The road side unit generates an attack detection decision, and sends the trusted vehicle information and the attack state of each vehicle. The trusted vehicle motion state data stream is represented as wherein At the same time, the vehicle attack flag bit (or flag information) is generated if formula (21) is shown:

[0134]

[0135] In operation S250, according to the trusted vehicle information and the attack flag information of the running vehicle, the intelligent connected vehicle dynamically optimizes its own vehicle control strategy.

[0136] According to the embodiment of the present application, the above-mentioned intelligent connected vehicle dynamically optimizes the vehicle control strategy according to the trusted vehicle information and the attack flag information of the running vehicle, which includes: in the case that the attack flag information of the running vehicle in front of the intelligent connected vehicle is a preset value, the intelligent connected vehicle optimizes the parameters of the vehicle control model and outputs the vehicle dynamic control strategy according to the trusted vehicle information; in the case that the intelligent connected vehicle is attacked by other running vehicles on the target road, the intelligent connected vehicle resets the communication network according to the attack flag information of the attacking vehicle.

[0137] According to the embodiment of the present application, in the process of resetting the communication network of the intelligent connected vehicle, the intelligent connected vehicle is in a preset time length of non-attack state.

[0138] When the intelligent connected vehicle is informed that the vehicle is subjected to a false data injection attack and there is no trusted data in the vehicle or adjacent front and rear vehicles, i.e. i (k) = 0, i ∈ {n-1, n, n+1}, which will trigger the network reset process of the vehicle.

[0139] In the network reset phase, the vehicle stops sending information; after the network reset, the vehicle sends the reset flag CR n (k) = 1.

[0140] The intelligent connected vehicle updates the input of the motion model according to the attack detection result sent by the road side unit, as shown in formula (22) and formula (23):

[0141]

[0142]

[0143] wherein, and θ n The binary variable representing the detection result is shown in formula (24) and formula (25):

[0144]

[0145]

[0146] The modified following vehicle model based on the attack detection result is shown in formula (26):

[0147]

[0148] The attack detection and optimization control method based on vehicle-road cooperation provided by the application can improve the control of intelligent networked vehicles in a complex hybrid traffic system, improve the response ability of intelligent networked vehicles to various risks in the complex hybrid traffic system, and improve the safety and traffic efficiency of the hybrid traffic system.

[0149] To better illustrate the advantages and beneficial effects of the above-mentioned method provided by the application, the above-mentioned method will be further described in detail below. Figure 7 to Figure 1 0.

[0150] Figure 7 Fig. 1 is a schematic diagram of an attack detection framework based on vehicle-road cooperation according to an embodiment of the application.

[0151] As shown in Fig. 1, the motion information of each running vehicle in the hybrid traffic system is obtained and input into the roadside unit for processing to obtain an attack detection decision; the attack detection decision is fed back to the intelligent networked vehicle, and the intelligent networked vehicle makes an optimization control strategy, thereby ensuring the safety and efficient traffic of the entire hybrid traffic system. Figure 7

[0152] Fig. 8(a) is a schematic diagram of an attacker injecting positive false data into vehicle 3 without using the attack detection and optimization control strategy according to an embodiment of the application.

[0153] Fig. 8(b) is a schematic diagram of the motion change of vehicles 2-5 when an attacker injects positive false data into vehicle 3 without using the attack detection and optimization control strategy according to an embodiment of the application.

[0154] Fig. 8(c) is a schematic diagram of an attacker injecting negative false data into vehicle 3 without using the attack detection and optimization control strategy according to an embodiment of the application.

[0155] Fig. 8(d) is a schematic diagram of the motion change of vehicles 2-5 when an attacker injects negative false data into vehicle 3 without using the attack detection and optimization control strategy according to an embodiment of the application.

[0156] Fig. 9(a) is a schematic diagram of the detection probability of the self-updating distributed estimator and the composite attack detector when the current vehicle state data output by each vehicle is tampered with according to an embodiment of the application.

[0157] Fig. 9(b) is a schematic diagram of the detection probability of the self-updating distributed estimator and the composite attack detector when the front / rear vehicle state data output by the intelligent networked vehicle is tampered with according to an embodiment of the application.

[0158] Fig. 10(a) is a schematic diagram of an attacker injecting positive false data into vehicle 3 using the attack detection and optimization control strategy according to an embodiment of the application.

[0159] ​Fig. 10(b) is a schematic diagram of the motion changes of vehicles 2-5 when an attacker injects positive false data into vehicle 3 using the attack detection and optimal control strategy according to an embodiment of the present application.

[0160] Fig. 10(c) is a schematic diagram of an attacker injecting negative false data into vehicle 3 using the attack detection and optimal control strategy according to an embodiment of the present application.

[0161] Fig. 10(d) is a schematic diagram of the motion changes of vehicles 2-5 when an attacker injects negative false data into vehicle 3 using the attack detection and optimal control strategy according to an embodiment of the present application.

[0162] For the hybrid traffic system described in the present application, there are two forms of false data injection attacks: tampering with the state data output by each vehicle, and tampering with the state data output by the front / rear vehicle of the intelligent connected vehicle. The mathematical expression forms are shown in equations (27) and (28):

[0163]

[0164]

[0165] The false data injection attack expressed in equation (27) will affect the decision and control of the rear intelligent connected vehicle, and will cause a decrease in safety or traffic efficiency. The false data injection attack expressed in equation (28) will confuse common distributed estimators and attack detectors. In the present application, only the front / rear vehicle measurement information that has been verified as non- abnormal data is input to the distributed estimator, and the estimator is updated in real time using trusted historical trajectory data, so that the detection performance of the distributed estimator and the attack detector can be avoided from being disturbed by the attack described in equation (28).

[0166] A hybrid traffic queue is set up, and the vehicles are arranged as [0, 1, 1, 0, 1, 0, 1, 0, 0, 1], where 0 represents a manually driven vehicle and 1 represents an intelligent connected vehicle.

[0167] Fig. 8 is a schematic diagram of the motion changes of the vehicles adjacent to the attacked vehicle after the false data injection attack (27) without using the attack detection and optimal control strategy. Figs. 8(a) and (c) respectively represent the false data (i.e., m3(k)) injected by the attacker into vehicle 3, and Figs. 8(b) and (d) respectively represent the motion changes of vehicles 2-5 in the two attack scenarios. According to the motion evolution of each vehicle on the way, it can be seen that the attack can cause the intelligent connected vehicle and the front vehicle to decrease or increase the distance between them through false data injection attack, i.e., to reduce the safety and efficiency of the corresponding vehicle.

[0168] This invention uses Monte Carlo simulation to obtain the alarm thresholds for outliers in the proposed attack detection method. The sample data consists of 10,000 samples, and the alarm thresholds with a 99% distribution are selected as follows: the alarm threshold for the self-state output of each vehicle is 3.3785; the alarm threshold for the measurement state of the preceding vehicle output by the intelligent connected vehicle is 5.6304; and the alarm threshold for the measurement state of the following vehicle output by the intelligent connected vehicle is 5.4738.

[0169] Figure 9 shows the detection probabilities of the self-updating distributed estimator and the composite attack detector when using the above-mentioned outlier alarm threshold. Figures 9(a) and (b) correspond to the detection probability distributions when different false data are used in the attacks described in formulas (27) and (28), respectively. For the attack described in formula (27), 100% detection can be achieved when the absolute value of the false velocity value is greater than 0.5 m / s; for the attack described in formula (28), 100% detection can be achieved when the absolute value of the false velocity value is greater than 0.3 m / s.

[0170] Figure 10 shows the motion changes of vehicles adjacent to the attacked vehicle after being subjected to a false data injection attack (27) when using the attack detection and optimization control strategy proposed in this invention. The definitions of each subgraph are the same as in Figure 8. The results show that the cars in the queue were not affected by the false data, and the security and efficiency of the queue were maintained.

[0171] Figure 11 The diagram illustrates an electronic device suitable for implementing a vehicle-road cooperative hybrid transportation system attack detection and optimization control method according to an embodiment of the present invention.

[0172] like Figure 11 As shown, an electronic device 1100 according to an embodiment of the present invention includes a processor 1101, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1102 or a program loaded from a storage portion 1108 into a random access memory (RAM) 1103. The processor 1101 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 1101 may also include onboard memory for caching purposes. The processor 1101 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present invention.

[0173] In the RAM 1103, various programs and data required for the operation of the electronic device 1100 are stored. The processor 1101, the ROM 1102, and the RAM 1103 are connected to each other via the bus 1104. The processor 1101 performs various operations of the method flow according to the embodiments of the present application by executing the programs in the ROM 1102 and / or the RAM 1103. It should be noted that the programs can also be stored in one or more memories other than the ROM 1102 and the RAM 1103. The processor 1101 can also perform various operations of the method flow according to the embodiments of the present application by executing the programs stored in the one or more memories.

[0174] According to the embodiments of the present application, the electronic device 1100 can further include an input / output (I / O) interface 1105, which is also connected to the bus 1104. The electronic device 1100 can further include one or more of the following components connected to the I / O interface 1105: an input part 1106 including a keyboard, a mouse, etc.; an output part 1107 including a display such as a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage part 1108 including a hard disk, etc.; and a communication part 1109 including a network interface card such as a LAN card, a modem, etc. The communication part 1109 performs communication processing via a network such as the Internet. A drive 1110 is also connected to the I / O interface 1105 as necessary. A removable recording medium 1111 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc. is attached to the drive 1110 as necessary, so that a computer program read out therefrom is installed in the storage part 1108 as necessary.

[0175] The present application also provides a computer readable storage medium, which can be included in the device / apparatus / system described in the above embodiments; or can exist separately without being assembled into the device / apparatus / system. The above computer readable storage medium carries one or more programs, when the one or more programs are executed, the method according to the embodiments of the present application is implemented.

[0176] According to embodiments of the present application, the computer readable storage medium can be a non-transitory computer readable storage medium, such as, for example, without limitation, a portable computer diskette, a hard disk, random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), a portable compact disc read-only memory (CD-ROM), optical storage, magnetic storage, or any suitable combination of the foregoing. In the present application, a computer readable storage medium can be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. For example, according to embodiments of the present application, a computer readable storage medium can include the ROM 1102 and / or the RAM 1103 described above, and / or one or more other memory devices that are not within the ROM 1102 and the RAM 1103.

[0177] The computer readable program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer implemented process, such that the instructions which execute on the computer, other programmable data processing apparatus, or other device implement the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0178] The specific embodiments described above have been disclosed by way of example only and equivalents modifications and substitutes are intended to be included. Accordingly, the application is not limited by the foregoing description, but is only limited by the scope of the appended claims.

Claims

1. A method for attack detection and optimal control of a vehicle-road cooperation based hybrid traffic system, characterized in that, The method comprises the following steps: A roadside unit obtains motion information of a target vehicle on a target road in a mixed traffic system through network connection technology, wherein the target vehicle comprises an artificial driving vehicle and an intelligent connected vehicle, and the motion information comprises speed information and position information of the target vehicle; According to historical reliable motion information of the target vehicle and reliable motion information of surrounding vehicles of the target vehicle, a self-updating distributed estimator of the roadside unit generates motion prediction information of the target vehicle; According to the motion prediction information of the target vehicle and a preset vehicle motion model, a composite attack detector of the roadside unit is constructed, and the motion information of the target vehicle is divided into abnormal data and non-abnormal data by the composite attack detector; According to the abnormal data and the non-abnormal data, the roadside unit generates an attack detection decision, and broadcasts reliable vehicle information and attack flag information of the target vehicle in the mixed traffic system; According to the reliable vehicle information and the attack flag information of the target vehicle, the intelligent connected vehicle dynamically optimizes its vehicle control strategy; According to the motion prediction information of the target vehicle and a preset vehicle motion model, a composite attack detector of the roadside unit is constructed, and the motion information of the target vehicle is divided into abnormal data and non-abnormal data, which comprises: According to the motion prediction information of the target vehicle and a preset vehicle motion model, a composite attack detector of the roadside unit is constructed; Through Monte Carlo simulation experiment, non-abnormal motion information of the target vehicle under an attack-free scene is obtained, and the non-abnormal motion information of the target vehicle is processed by the composite attack detector to obtain an attack alarm threshold; Based on Kalman filter principle and chi-square detector principle, the motion information of the target vehicle is processed by the composite attack detector to obtain a processing result; In the case that the processing result is greater than the attack alarm threshold, the motion information of the target vehicle is determined as abnormal data; In the case that the processing result is less than or equal to the attack alarm threshold, the motion information of each vehicle is determined as non-abnormal data.

2. The method of claim 1, wherein, The roadside unit obtains motion information of a target vehicle on a target road in a mixed traffic system through network connection technology, which comprises: The roadside unit receives speed information and position information of the artificial driving vehicle sent in a broadcast manner by a vehicle-mounted network connection device of the artificial driving vehicle in the mixed traffic system through network connection technology; The roadside unit receives speed information and position information of the intelligent connected vehicle and speed information and position information of surrounding vehicles sent in a broadcast manner by a vehicle-mounted network connection device of the intelligent connected vehicle in the mixed traffic system through network connection technology, wherein the speed information and position information of the surrounding vehicles are obtained by a vehicle-mounted sensor of the intelligent connected vehicle.

3. The method of claim 1, wherein, According to historical motion information of the target vehicle and reliable motion information of surrounding vehicles of the target vehicle, a self-updating distributed estimator of the roadside unit generates motion prediction information of the target vehicle, which comprises: processing the historical trusted motion information of the traveling vehicle by using a longitudinal motion model of the traveling vehicle to obtain acceleration prediction information of the traveling vehicle; constructing a vehicle platoon system model with disturbance based on the acceleration prediction information of the traveling vehicle, the motion information of the traveling vehicle, and the trusted motion information of the surrounding vehicle, wherein the disturbance conforms to a Gaussian distribution; constructing a self-updating distributed estimator of the road side unit based on the vehicle platoon system model with disturbance, and processing the historical trusted motion information of the traveling vehicle by the self-updating distributed estimator to obtain speed prediction information and position prediction information of the traveling vehicle.

4. The method of claim 1, wherein, The preset vehicle motion model includes an intelligent driver model of the human-driven vehicle and an improved collaborative intelligent driver model of the intelligent connected vehicle.

5. The method of claim 1, wherein, According to the abnormal data and the non-abnormal data, an attack detection decision is generated by the road side unit, and trusted vehicle information and attack flag information of the traveling vehicle are broadcast in the mixed traffic system, including: The abnormal data and the non-abnormal data are screened by the road side unit to generate an attack detection decision, wherein the attack detection decision includes trusted vehicle information and attack flag information of the traveling vehicle; The trusted vehicle information and the attack flag information of the traveling vehicle are broadcast by the road side unit to the traveling vehicles on the target road in the mixed traffic system.

6. The method of claim 1, wherein, According to the trusted vehicle information and the attack flag information of the traveling vehicle, the intelligent connected vehicle dynamically optimizes its vehicle control strategy, including: In the case that the attack flag information of the vehicle in front of the intelligent connected vehicle is a preset value, according to the trusted vehicle information, the intelligent connected vehicle optimizes the parameters of its vehicle control model and outputs a vehicle dynamic control strategy; In the case that the intelligent connected vehicle is attacked by other traveling vehicles on the target road, the intelligent connected vehicle resets its communication network according to the attack flag information of the attacking vehicle.

7. The method of claim 6, wherein, During the process of resetting the communication network of the intelligent connected vehicle, the intelligent connected vehicle is in a preset length of time without attack.

8. An electronic device, comprising: one or more processors; a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors perform the method according to any one of claims 1-7.

9. A computer-readable storage medium having stored thereon executable instructions that, when executed by a processor, cause the processor to perform the method according to any one of claims 1-7.

Images