A dynamic output feedback vehicle platoon control method for coping with network interference

Through the dynamic output feedback control method, a model of sensor saturation and spoofing attacks was established, and the controller parameters were derived using matrix analysis technology, which solved the robustness of the fleet system in complex network environments and achieved rapid recovery and stability of the system .

CN117215231BActive Publication Date: 2024-11-22NANTONG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202311142773.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-09-06
Publication Date
2024-11-22
Estimated Expiration
2043-09-06

AI Technical Summary

Technical Problem

The lack of robust feedback queue controller designs in the prior art to effectively respond to sensor saturation and spoofing attacks in complex network environments leads to degradation of system performance.

Method used

A dynamic output feedback vehicle queue control method is designed. By establishing a state space model of a three-vehicle collaborative queue, characterizing sensor saturation and spoofing attacks, designing a dynamic feedback controller, and using matrix analysis technology to solve linear matrix inequality and derive controller parameters. Ensure system stability and H infinite performance.

Benefits of technology

In the presence of sensor saturation and spoofing attacks, the fleet system is quickly restored to balanced position and speed, ensuring the stability of the system and the satisfaction of performance indicators.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117215231B_ABST
    Figure CN117215231B_ABST
Patent Text Reader

Abstract

The present invention provides a dynamic output feedback automobile queue control method for coping with network interference, which relates to the field of vehicle networking and automatic driving technology. The control method comprises the following steps: raising a problem based on a collaborative adaptive cruise control system and making necessary assumptions about the problem; establishing a state space model of a three-vehicle collaborative queue; establishing a model that can simultaneously characterize random sensor saturation and random deception attacks; designing a dynamic feedback controller; providing conditions to ensure the stability of the augmented system and sufficient conditions for H-infinity performance requirements; using matrix analysis technology to solve a set of linear matrix inequalities to derive controller parameters; and using a three-vehicle collaborative queue model to verify the effectiveness of a robust dynamic output feedback control algorithm. In this application, based on a robust dynamic output feedback control algorithm, the control problem of a new collaborative adaptive cruise control system under sensor saturation, parameter uncertainty and deception attacks is solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of vehicle networking and autonomous driving technology, and in particular to a dynamic output feedback automobile queue control method for coping with network interference. Background Art

[0002] Intelligent connected vehicles have long played a key role in maintaining traffic capacity and stability using advanced sensing and communication technologies. Among them, the collaborative adaptive cruise control system achieves automatic following and adaptive driving control between vehicles in a convoy through collaborative cooperation between vehicles, which is of great help in improving the driving efficiency of the convoy, reducing the distance between vehicles, reducing the risk of traffic accidents, and improving driving comfort.

[0003] In reality, due to specific physical limitations, on-board sensors will randomly experience sensor saturation. This is mainly caused by factors such as component aging, sudden changes in the environment (sudden acceleration or deceleration), and limited components (vehicle speed exceeds the sensor scale). As a result, all or part of the transmission is forced to remain within a bounded set. If these characteristics are not handled properly, system performance will decline or even degrade. With the rapid development of global network systems, network security has become an increasingly important and realistic issue, especially for V2V networks in cooperative adaptive cruise control systems. There are three main types of attacks in the field of network security, namely denial of service attacks, replay attacks, and spoofing attacks. Among them, random spoofing attacks are the most dangerous attack behaviors, which are characterized by injecting more general malicious data into real information to interfere with the transmission between vehicle queues, thereby seriously reducing system performance. Summary of the invention

[0004] The purpose of the present invention is to solve the problem in the prior art that there is a lack of a design of a robust feedback queue controller that can effectively cope with complex network environments.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A dynamic output feedback vehicle queue control method for coping with network interference comprises the following steps:

[0007] S1: Based on the cooperative adaptive cruise control system, a new control problem of a cooperative adaptive cruise control system with parameter uncertainty, randomly occurring sensor saturation and randomly occurring deception attacks is proposed, and necessary assumptions are made for the problem.

[0008] S2: Establish a state space model of a three-vehicle cooperative platoon;

[0009] S3: Build a model that can simultaneously characterize randomly occurring sensor saturation and randomly occurring spoofing attacks;

[0010] S4: Designed a dynamic feedback controller;

[0011] S5: Provide conditions to ensure the stability of the augmented system and the sufficient conditions for the H-infinity performance requirements;

[0012] S6: Using matrix analysis techniques, a set of linear matrix inequalities are solved to derive controller parameters;

[0013] S7: The effectiveness of the robust dynamic output feedback control algorithm is verified using a three-vehicle cooperative platoon model.

[0014] The necessary assumptions in S1 include the following:

[0015] (1) The controlled vehicle can monitor the position and speed of other vehicles without delay;

[0016] (2) The controlled vehicles are fully automated and cooperative: information is shared via V2V and V2I communications, so human drivers and non-cooperative behavior are not considered;

[0017] (3) Control decisions are updated at constant time intervals;

[0018] (4) Control the vehicle and execute control decisions simultaneously.

[0019] The target system state vector x in S2 t The definition is as follows:

[0020]

[0021] In the above formula, s i,t is the distance between the i-th car and the (i-1)-th car, represents the equilibrium spacing, τ represents the expected time interval, d0 is the minimum stagnation spacing, v i,t Represents the desired speed (leading vehicle speed) and the current speed v i (t), based on which the state space model is obtained:

[0022]

[0023] in:

[0024]

[0025] Control Signal where f j,t It represents the external force of the jth vehicle to adjust the distance and speed deviation, m i Represents the mass of the i-th car.

[0026] The sensor saturation model in S3 is as follows:

[0027]

[0028] where a Bernoulli random variable β k The random occurrence of sensor saturation is simulated. k = 1, the sensor is saturated and receives the measured value When β k = 0, the true measurement value is y k Sensor saturation does not occur, and the deception attack model is as follows:

[0029]

[0030] in is a non-zero value sent by the adversary The injection signal introduces the Bernoulli random variable α k To describe the success rate of the attack. Assume that the signal ξ from the opponent k is energy-limited and satisfies the constraint

[0031] The S4 dynamic feedback controller:

[0032]

[0033]

[0034] in is the estimated value of the state in the system, u k is the control input, and are the parameters of the controller.

[0035] In S5, the disturbance attenuation level γ>0 and the controller parameter A are set c , B c and C c , if there exists a positive definite matrix P>0 and a scalar ∈1,∈2>0 that satisfies the following matrix inequality, then the H-infinity performance requirement defined in the following formula is achieved:

[0036]

[0037] Then there is

[0038]

[0039] Where γ>0 is a specific scalar.

[0040] The S6 matrix analysis technique is as follows:

[0041] For a given constant γ>0, if there exist positive definite matrices Y, Ω, V and W, and scalars ∈1>0, ∈2>0 and ∈3>0, for a given matrix X>0, the following inequality holds:

[0042]

[0043] Then there is

[0044]

[0045] The method for controlling a dynamic output feedback automobile queue in response to network interference according to claim 1 is characterized in that: S7 comprises the following steps:

[0046] A1: Build a simulation experiment platform;

[0047] A2: Take the three-car system as the test model and assume that the speed of the lead car is constant;

[0048] A3: Set the initial value x0 = [1.2 0.8 1.0 0.9 0.6 0.5] T , the expected time interval τ1=τ2=τ3=1s, the initial accelerations of the three following vehicles are all 0, that is, u0=[0 0 0] T The parameters in A and Υ are positively correlated with the response time. The analysis shows that the robust dynamic output feedback control algorithm can respond quickly and effectively, allowing the vehicle-to-vehicle system to quickly return to the equilibrium position and equilibrium speed.

[0049] Compared with the prior art, this application has the following beneficial effects:

[0050] 1) Based on the robust H infinite dynamic output feedback control algorithm, the control problem of a new cooperative adaptive cruise control system considering sensor saturation, parameter uncertainty and deception attack is solved;

[0051] 2) A new sufficient condition is introduced to ensure the stability of the augmented system and the corresponding performance indicators exist simultaneously;

[0052] 3) Using the decomposition method, the controller parameters are derived using the solutions of linear matrix inequalities. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A flowchart of a dynamic output feedback vehicle queue control method for coping with network interference in one embodiment of the present invention;

[0054] Figure 2 A schematic diagram of a framework of an intelligent connected vehicle equipped with a cooperative adaptive cruise control system in one embodiment of the present invention;

[0055] Figure 3 A schematic diagram of a closed-loop control system in one embodiment of the present invention;

[0056] Figure 4 is a control signal of each vehicle in one embodiment of the present invention;

[0057] Figure 5 is the deviation between the actual position and the equilibrium position of each vehicle in one embodiment of the present invention;

[0058] Figure 6 is the deviation between each vehicle and the leading vehicle in one embodiment of the present invention. DETAILED DESCRIPTION

[0059] The present invention is further described in detail below in conjunction with specific embodiments.

[0060] Please refer to Table 1. The following table is the interpretation of some English symbols in the application.

[0061] Table 1 List of symbols

[0062]

[0063]

[0064] See also Figure 1 , a dynamic output feedback vehicle queue control method for coping with network interference, comprising the following steps:

[0065] S1: Based on the cooperative adaptive cruise control system, a new control problem of a cooperative adaptive cruise control system with parameter uncertainty, randomly occurring sensor saturation and randomly occurring deception attacks is proposed, and necessary assumptions are made for the problem.

[0066] In one embodiment, the assumptions include the following:

[0067] (1) The controlled vehicle can monitor the position and speed of other vehicles without delay;

[0068] (2) The controlled vehicles are fully automated and cooperative: information is shared via V2V and V2I communications, so human drivers and non-cooperative behavior are not considered;

[0069] (3) Control decisions are updated at constant time intervals;

[0070] (4) Control the vehicle and execute control decisions simultaneously.

[0071] S2: Establish a state space model of a three-vehicle cooperative platoon.

[0072] In one embodiment, the target system state vector x tThe definition is as follows:

[0073]

[0074] In the above formula, s i,t is the distance between the i-th car and the (i-1)-th car, represents the equilibrium spacing, τ represents the expected time interval, d0 is the minimum stagnation spacing, v i,t Represents the desired speed (leading vehicle speed) and the current speed v i (t). Then we get the state space model:

[0075]

[0076] in:

[0077]

[0078] Control Signal where f j,t It is the external force for adjusting the distance and speed deviation of the jth vehicle, m i Represents the mass of the i-th car.

[0079] The collaborative adaptive cruise control system consists of an automation system and a communication system. In order to ensure the stable and safe operation of the vehicle platoon, key parameters such as position and speed are essential.

[0080] Deviation of vehicle balance distance Δs i,t and the speed deviation Δυ i,t Usually measurable and defined in The final measurement results are as follows:

[0081] y t =C′x t (3)

[0082] The dynamics of vehicle motion is continuous, while in practice the cooperative adaptive cruise control problem is usually described by a discrete system.

[0083] Using the discretionary scheme and considering the uncertainty of disturbance and structure, the following formula is obtained:

[0084]

[0085] Where AA = e A′T , CC′, where T>0 is the sampling period, and the disturbance w k ∈l2[0,∞), is a matrix with compatible dimensions. Assume that the uncertain parameter ΔA satisfies the following admissible conditions:

[0086]

[0087] Δs i,k and Δv i,k Essential for ensuring traffic capacity and safety. Figure 2 , considering the safety of vehicles, the distance between vehicles should not be too small, and at the same time, the distance between two adjacent vehicles should not be too large, because it will reduce the traffic capacity. In fact, the distance between two adjacent vehicles should tend to the equilibrium distance, that is, Δs i,k → 0. In addition, to avoid collision, each vehicle should enter a speed close to v, i.e., Δv i,k →0.

[0088] Under ideal conditions, the dynamics formula of cooperative adaptive cruise control is expressed as However, in actual situations, the environment is complex and the interference from the external environment cannot be simply ignored. Unwanted disturbances (such as wet road surface, wind action, tire friction, etc.) will seriously affect the dynamics of cooperative adaptive cruise control and there are some obstacles to obtaining an accurate model. Therefore, the modeling of the uncertain random system in formula (4) more reasonably reflects the actual dynamics of cooperative adaptive cruise control.

[0089] S3: Build a model that can simultaneously characterize randomly occurring sensor saturation and randomly occurring spoofing attacks:

[0090] Through the dedicated short-range technology of the V2V network, information can be effectively transmitted between vehicles, so that the operating status can be adjusted in real time to ensure that the collaborative vehicle platoon system is in a reliable and stable operating state.

[0091] Communication in cooperative adaptive cruise control systems plays a key role in localizing or adjusting speed and position (dedicated short-range communication protocol). Considering the ranging sensor (receiving or transmitting measurement) operating under unfavorable conditions experiencing sensor saturation, the formula is as follows:

[0092]

[0093] where D2 is a matrix of compact dimension.

[0094] In one embodiment, the sensor saturation is defined as follows:

[0095]

[0096] in r iis the i-th element of the vector r, sign(·) represents the sign function, v i Represents the saturation level. According to the saturation function , it is not difficult to verify whether there is a diagonal matrix Λ that satisfies 0≤A≤I and the following formula:

[0097]

[0098] See also Figure 2 , cyber attacks usually occur by forging or blocking data packets transmitted in the V2V network, or by installing a device on the vehicle that can attack the measurement value. Substitute for real measurement value characteristics of malicious hardware or software, The definition is as follows:

[0099]

[0100] in is a non-zero value sent by the adversary The injected signal satisfies the following conditions:

[0101] (ξ k -Yx k ) T (ξ k -Yx k )≤δ 2 x k T x k (9)

[0102] For a particular positive scalar δ and a particular matrix Υ of suitable dimensions, the binary 0 and 1 Bernoulli random variables α k and β k The definition is as follows:

[0103]

[0104]

[0105] in and are all specific positive scalars.

[0106] In cooperative adaptive cruise control systems, sensor saturation often occurs due to factors such as component aging, sudden changes in the environment (sudden acceleration or deceleration), and limited components (vehicle speed exceeds the sensor size). In addition, sensor saturation itself may suffer from random mutations, for example, random sensor failures lead to intermittent saturation. Considering the above situation, we use a Bernoulli random variable β in (5) kThe random occurrence of sensor saturation is simulated. k = 1, the sensor is saturated and receives the measured value When β k = 0, the true measurement value is y k Sensor saturation did not occur.

[0107] At the same time, the communication system between vehicles and infrastructure (especially in V2V networks) presents remote attack access to malicious adversaries, who exploit system vulnerabilities for remote attack access. These attacks have a significant impact on system performance and cause some security issues, including three types of attacks: denial of service attacks, replay attacks, and spoofing attacks. Among them, spoofing attacks are the most dangerous, mainly because of their special structure, including real data and injected data. In formula (8), the Bernoulli random variable α is introduced k To describe the success rate of the attack. Assume that the signal ξ from the opponent k is energy-limited and satisfies the constraint It should be pointed out that the results are conservative to a certain extent because the system information is not taken into account.

[0108] S4: A dynamic feedback controller is designed:

[0109] See also Figure 3 , in order to keep the cooperative adaptive cruise control running safely and efficiently, we built the following output feedback controller:

[0110]

[0111]

[0112] in is the estimated value of the state in the system, u k is the control input, and are the parameters of the controller.

[0113] The controlled output value of the closed-loop system is z k =Mx k Also set Based on this, the following augmented system is derived in this application:

[0114]

[0115]

[0116] in:

[0117]

[0118]

[0119]

[0120] For w k = 0, the initial condition is For an augmented system of , if there exist constants λ>0 and ζ∈(0,1), then the differential solution is exponentially stable in the mean square sense, as follows:

[0121]

[0122] S5: Provides sufficient conditions to ensure the stability of the augmented system and the H-infinity performance requirements:

[0123] The robust H-infinity output feedback controller needs to meet the following conditions:

[0124] 1) Augmented system in w k =0 means exponential stability in the mean square sense;

[0125] 2) Under zero initial conditions, when w k is non-zero, control output z k Satisfies the following formula:

[0126]

[0127] Where γ>0 is a specific scalar.

[0128] Assume that the disturbance attenuation level γ>0 and the controller parameter A c , B c and C c The H-infinity performance requirement defined in (13) is achieved if there exists a positive definite matrix P>0 and a scalar ∈1,∈2>0 that satisfies the following matrix inequality:

[0129]

[0130] in:

[0131]

[0132]

[0133]

[0134]

[0135]

[0136] The augmented system is exponentially stable in the mean square sense, and the control output z k Satisfies H infinite index.

[0137] S6: Using matrix analysis techniques, a set of linear matrix inequalities are solved to derive controller parameters:

[0138] Consider the augmented system in (12). For a given constant γ>0, if there exist positive definite matrices Y, Ω, V and W, and scalars ∈1>0, ∈2>0 and ε3>0, for a given matrix X>0, the following inequality holds:

[0139]

[0140] Then there is

[0141]

[0142] S7: The effectiveness of the robust H infinite dynamic output feedback control algorithm is verified using a three-vehicle cooperative platoon model.

[0143] Specifically, in one embodiment, the verification scheme includes the following steps:

[0144] A1: Build an experimental platform:

[0145] Specifically, in one embodiment, by constructing a simulation experiment to verify the advantages and feasibility of the theoretical results, the designed robust H infinite dynamic output feedback control algorithm is demonstrated in MATLAB (R2016a) and Intel Core CPU i5-8265.

[0146] A2: Parameter settings:

[0147] See also Figure 2 In one embodiment, a three-vehicle cooperative platoon system (one leading vehicle and three following vehicles) is selected as a test model, and it is assumed that the speed of the leading vehicle is constant.

[0148] The controller of the cooperative adaptive cruise control makes the vehicle reach a balanced state (including position and relative speed). The system parameters are set to 2n x =6,2n y =6,n u =3. To simplify the calculation, the sampling period is set to T = 0.05s. The sensor saturation parameter is set to A = diag{0.6, 0.6, 0.6, 0.6, 0.6}. The deception attack parameters are set to δ = 1, Υ = diag{2, 2, 2, 2, 2, 2}. and The probability of is 0.5 and 0.6 respectively. The disturbance attenuation level γ = 1, the disturbance amount is w k =0.1sin(k). Other parameter settings are as follows:

[0149] D1=D2=[0.050.050.050.050.050.05] T ,

[0150] M=diag{1, 1, 1, 1, 1, 1}, E=[0.10.10.10.10.10.1],

[0151] H=[0.10.10.10.10.10.1] T , F = 0.01 sin(k),

[0152] X=diag{0.7, 0.65, 0.6, 0.55, 0.45, 0.38}.

[0153] A3: Set the initial value:

[0154] In one embodiment, the initial value of x0 is set to x0 = [1.2 0.8 1.0 0.9 0.6 0.5] T , the expected time interval τ1=τ2=τ3=1s, the initial accelerations of the three following vehicles are all 0, i.e. u01[0 0 0] T .

[0155] The results of the simulation system can be found in Figures 4 to 6 . Figure 4 The control input signals of vehicles 1, 2, and 3 are depicted in Figure 5 The first sub-figure depicts the difference between the position of vehicle 1 and the equilibrium distance, and the second and third sub-figures depict the difference between the positions of vehicles 2 and 3 and the corresponding equilibrium distances, respectively. Figure 6 The first sub-graph in the figure depicts the difference between the speed of vehicle 1 and the equilibrium speed, and the second and third sub-graphs depict the difference between the speed of vehicles 2 and 3 and the equilibrium speed. It is not difficult to find that under the controller we designed, the vehicle quickly returns to the equilibrium position, and the control input converges to 0 in about 5 seconds.

[0156] Table 2 Response time (step length) of control signals with different parameters

[0157]

[0158] Refer to Table 2, which shows the effects of sensor saturation and spoofing attacks on the cooperative adaptive cruise control system. In order to eliminate the influence of different variables on the results, the same initial condition x0 = [0.5 0.5 0.5 0.5 0.5] is set in this application. T , assuming that the fluctuation amplitude of the system is less than 3.0×10 -3At the same time, Table 2 shows the response time of the system to maintain stability under different parameter conditions, and plots the response time of vehicle 1, vehicle 2, and vehicle 3 under different A i and i The control signal under A i They are Λ1=0.3I6,Λ2=0.5I5,Λ3=0.7I6,γ i They are γ1=I6, γ2=2I6, γ3=3I6 respectively. Obviously, the parameters in Table 2Λ and Υ are positively correlated with the response time. This means that the designed robust dynamic output feedback control algorithm can react quickly and effectively, so that the fleet system can quickly return to the equilibrium position and equilibrium speed. That is, the controller provided in this application has a good control effect, so that the system can quickly return to a stable state. It is concluded that:

[0159] 1. The designed control algorithm can perform well even in the presence of parameter instability, randomly occurring sensor saturation and randomly occurring spoofing attacks.

[0160] 2. The control algorithm has a fast response speed, which can restore the fleet system to a stable state in a short time.

[0161] The present application provides a dynamic output feedback vehicle platoon control method for coping with network interference, which is based on a robust H-infinity dynamic output feedback control algorithm, solves the control problem of a new type of cooperative adaptive cruise control system that simultaneously considers sensor saturation, parameter uncertainty, and deception attacks, and proposes a new sufficient condition to ensure that the stability of the augmented system and the corresponding performance indicators exist simultaneously. The decomposition method is used to derive the parameters of the controller using the solution of the linear matrix inequality.

Claims

1. A dynamic output feedback vehicle platoon control method to cope with network interference, Features: The following steps are involved: S1: Based on the cooperative adaptive cruise control system, a new control problem of cooperative adaptive cruise control system with parameter uncertainty, random sensor saturation and random deception attack is proposed, and necessary assumptions are made for the problem; S2: Establish a state space model of a three-vehicle cooperative platoon; The target system state vector x in S2 t The definition is as follows: In the above formula, x i,t represents the state of the i-th vehicle, T represents the sampling period; S i,t is the distance between the i-th car and the (i-1)-th car, represents the equilibrium interval, τ represents the expected time interval, and d 0 is the minimum stagnation distance, v i,t represents the speed of the i-th vehicle, Δs i,t and Δv i,t They represent the deviation of the i-th vehicle from the equilibrium distance and the deviation from the equilibrium speed, respectively, and the state space model is obtained based on this: in: Control Signal where f j,t It represents the external force of the jth vehicle to adjust the distance and speed deviation, m i represents the mass of the i-th car; S3: Build a model that can simultaneously characterize randomly occurring sensor saturation and randomly occurring spoofing attacks; The sensor saturation model in S3 is as follows: in represents the saturation function, w k represents the disturbance, D 2 Represented as a compact matrix, through a Bernoulli random variable β k The random occurrence of sensor saturation is simulated. k = 1, the sensor is saturated and receives the measured value When β k = 0, the true measurement value is y k Sensor saturation does not occur, and the deception attack model is as follows: in is a non-zero value sent by the adversary The injection signal introduces the Bernoulli random variable α k To describe the success rate of the attack, assuming that the signal ξ from the opponent k is energy-limited and satisfies the constraint S4: Designed a dynamic feedback controller; S5: Provide conditions to ensure the stability of the augmented system and the sufficient conditions for the H-infinity performance requirements; S6: Using matrix analysis techniques, a set of linear matrix inequalities are solved to derive controller parameters; S7: The effectiveness of the robust dynamic output feedback control algorithm is verified using a three-vehicle cooperative platoon model.

2. A dynamic output feedback vehicle queue control method for coping with network interference according to claim 1, Features: The necessary assumptions in S1 include the following: (1) The controlled vehicle can monitor the position and speed of other vehicles without delay; (2) The controlled vehicles are fully automated and cooperative: information is shared via V2V and V2I communications, so human drivers and non-cooperative behavior are not considered; (3) Control decisions are updated at constant time intervals; (4) Control the vehicle and execute control decisions simultaneously.

3. According to claim 1, a dynamic output feedback vehicle queue control method for coping with network interference, Features: The S4 dynamic feedback controller: in is the estimated value of the state in the system, u k is the control input, and are the parameters of the controller; The controlled output value of the closed-loop system is z k =Mx k Also set Based on this, the following augmented system is derived: in: In the above formula, A=e A′T C=C′, is a matrix with compatible dimensions, ΔA is an uncertain parameter, For w k = 0, the initial condition is For an augmented system of , if there exist constants λ>0 and ζ∈(0,1), then the differential solution is exponentially stable in the mean square sense, as follows:

4. The method for controlling a dynamic output feedback vehicle queue in response to network interference according to claim 1, Features: The robust H-infinity output feedback controller in S5 needs to meet the following conditions: 1) Augmented system in w k =0 means exponential stability in the mean square sense; 2) Under zero initial conditions, when w k is non-zero, control output z k Satisfies the following formula: where γ>0 is a specific scalar, Assume that the disturbance attenuation level γ>0 and the controller parameter A c , B c and C c , if there exists a positive definite matrix P>0 and a scalar ∈ 1 ,∈ 2 > 0 satisfies the following matrix inequality, thus achieving the H-infinity performance requirement defined in equation (13):

5. The method for controlling a dynamic output feedback vehicle queue in response to network interference according to claim 1, Features: The S6 matrix analysis technique is as follows: Considering the augmented system in (12), for a given constant γ>0, if there exist positive definite matrices Y, Ω, V and W, and scalar ∈ 1 >0,∈ 2 >0 and∈ 3 >0, for a given matrix X>0, the following inequality holds: Then there is 6. A dynamic output feedback vehicle queue control method for coping with network interference according to claim 1, Features: The S7 comprises the following steps: A1: Build a simulation experiment platform; A2: Take the three-car system as the test model and assume that the speed of the lead car is constant; A3: Set the initial value x 0 =[1.2 0.8 1.0 0.9 0.6 0.5] T , the expected time interval τ 1 =τ 2 =τ 3 =1s, the initial acceleration of the three following vehicles is 0, that is, u 0 =[0 0 0] T , the parameters in Λ and γ are positively correlated with the response time. The analysis shows that the robust dynamic output feedback control algorithm can respond quickly and effectively, so that the vehicle-to-vehicle system can quickly return to the equilibrium position and equilibrium speed.