A Method and System for Dual-Timescale Secure Cooperative Control of Unmanned Systems Under FDI Attacks

By employing multi-scale control technology and distributed observers, the ill-conditioned numerical problem in dual-timescale control of traditional unmanned systems under FDI attacks is solved, achieving consistency between the leader and follower outputs, eliminating the impact of attacks in real time, and ensuring system stability.

CN120178671BActive Publication Date: 2026-11-17CHINA UNIV OF MINING & TECH
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202510296032.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-03-13
Publication Date
2026-11-17
Estimated Expiration
2045-03-13

AI Technical Summary

Technical Problem

In traditional FDI attack-based dual-timescale security collaborative control methods and systems, PID controllers are suitable for single-timescale unmanned systems, leading to ill-conditioned numerical problems in dual-timescale systems. This makes it impossible to effectively achieve output consistency between the leader and follower, and data-driven detection models cannot eliminate the impact of FDI attacks in real time.

Method used

By employing multi-scale control technology, a topology graph and dynamic model are established, and a distributed elastic observer and adaptive compensator filter are constructed. The convergence condition of the observation error is analyzed through Lyapunov theory, and fast and slow decomposition is performed to construct a novel composite controller and a dynamic output error system model, thereby eliminating the impact of FDI attacks in real time.

Benefits of technology

It effectively solves the ill-conditioned numerical problem caused by the dual time scale characteristics, realizes the consistency of the output of the leader and the follower in the unmanned system, eliminates the impact of FDI attacks in real time, and ensures that the system can remain stable when an attack occurs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120178671B_ABST
    Figure CN120178671B_ABST
Patent Text Reader

Abstract

This invention discloses a dual-timescale secure cooperative control method and system for unmanned systems under FDI attacks, relating to the field of unmanned system secure cooperative technology. First, the invention establishes a communication topology diagram and a dynamic model of the unmanned system with dual-timescale characteristics. Then, it establishes a bounded FDI attack model and a dynamic observation error model, and constructs a distributed elastic observer. Next, it analyzes the sufficient conditions for observation error convergence. Then, it decomposes the followers into fast and slow groups, and constructs a novel composite controller and a dynamic output error system model. Finally, it analyzes the sufficient conditions for the leader and followers to adjust their outputs in unmanned systems under FDI attacks. This avoids the ill-conditioned numerical problems of the adjustment equations caused by the dual-timescale characteristics, eliminates FDI attacks in real time, and is unaffected by the timing and probability of the attack, effectively achieving output consistency between the leader and followers in the unmanned system.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of unmanned system security collaboration technology, specifically to a method and system for dual-timescale security collaboration control of unmanned systems under FDI attacks. Background Technology

[0002] With the rapid development of artificial intelligence, computer, communication and microelectronics technologies, the potential applications and research of heterogeneous multi-agent systems in sensor information fusion, UAV formation, and power system cooperative control have attracted widespread attention. Among them, the problem of consistent output between the leader and followers of unmanned systems is one of the fundamental problems. It requires the unmanned system to use various consensus protocols or synchronization control algorithms to enable all followers to synchronize with the leader's output. In fact, many unmanned systems have multi-timescale characteristics. Therefore, the problem of consistent output control of unmanned systems with dual-timescale characteristics urgently needs to be solved.

[0003] Traditional methods for dual-timescale safety cooperative control of unmanned systems under FDI attacks use PID controllers to achieve output consistency between the leader and followers, and limit the attack range of FDI by constructing a data-driven detection model. Obviously, this method for dual-timescale safety cooperative control of unmanned systems under FDI attacks based on multi-scale control technology has at least the following shortcomings: 1. The PID controller in the traditional dual-timescale safety cooperative control method for unmanned systems under FDI attacks is suitable for single-scale unmanned systems. When used in dual-timescale unmanned systems, it will lead to ill-conditioned numerical problems when solving the corresponding output regulation equations, and cannot effectively achieve output consistency between the leader and followers in the unmanned system.

[0004] 2. Traditional dual-timescale security collaborative control methods for unmanned systems under FDI attacks and the data-driven detection models in the system are used to limit the attack range of FDI, but they cannot eliminate the impact of FDI attacks in real time. Summary of the Invention

[0005] To address the aforementioned technical shortcomings, the present invention aims to provide a method for safe collaborative control of unmanned systems based on multi-scale control technology under FDI attacks.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: In the first aspect, the present invention provides a dual-time-scale secure collaborative control method for unmanned systems under FDI attacks, including the following steps: Step 1: Establishing a topology graph and dynamic model: Establishing a communication topology graph based on communication nodes, and constructing a dynamic model of the unmanned system with dual-time-scale characteristics.

[0007] Step 2: Constructing the observer and model: Establish a bounded FDI attack model between follower communication channels, construct a distributed resilient observer, design filters with adaptive compensators on each distributed resilient observer, and establish a corresponding dynamic observation error model.

[0008] Step 3: Analyze the sufficient conditions for convergence of the observer estimation error: Using Lyapunov theory, analyze the unmanned system with dual time scale characteristics under FDI attack, and analyze the sufficient conditions for convergence of the observation error.

[0009] Step 4: Perform fast and slow decomposition: For followers with dual time scale characteristics, use multi-scale control techniques to perform fast and slow decomposition to obtain the corresponding fast subsystem and slow subsystem.

[0010] Step 5: Construct the controller and dynamic output error system model: Obtain the state information of the fast subsystem and the slow subsystem observed by the observer, construct a new composite controller, and establish the corresponding dynamic output error system model.

[0011] Step Six: Analyze the sufficient conditions for consistent adjustment outputs: Using Lyapunov stability theory and multi-scale control technology, analyze the unmanned system with dual time-scale characteristics under FDI attack, and analyze the sufficient conditions for consistent adjustment outputs between the leader and follower under FDI attack.

[0012] Preferably, the process of establishing a communication topology map based on the communication nodes is as follows: Establishing the topology map: S11, obtaining each communication node from the database.

[0013] S12. Establish a communication topology diagram for an unmanned system with dual time scale characteristics.

[0014]

[0015] In the formula, G represents the communication topology of an unmanned system with dual time scales, and V represents the set of communication nodes, then V = {v1, v2, v3, ..., v N}, v i Let i represent the i-th follower, where i represents the follower's ID, i = 1, 2, 3, ..., N, and N represents the total number of followers. Let the set of edges in the communication topology graph representing an unmanned system with dual time-scale characteristics be given.

[0016] Establish a dynamic model: S21. Use sensors to obtain the leader's state vector, the leader's output vector, the follower's output vector, the follower's slow state vector, and the fast state vector, and obtain the constant matrices from the database.

[0017] S22. Construct a dynamic model of an unmanned system with dual time scale characteristics:

[0018]

[0019] In the formula, x0(t) represents the leader's state vector at time t, y0(t) represents the leader's output vector at time t, and A0 and B0 are both constant matrices. Let represent the derivative of the leader's state vector at time t with respect to t. The derivative of the slow state vector of the i-th follower at time t with respect to t, x i (t) represents the slow state vector of the i-th follower at time t. z represents the derivative of the fast state vector of the i-th follower at time t with respect to t. i (t) represents the fast state vector of the i-th follower at time t, u i (t) represents the control input of the i-th follower at time t, y i (t) represents the output vector of the i-th follower at time t, and εi represents the degree of separation of the dynamics of the fast and slow state vectors of the i-th follower on the fast and slow time scales. and Each represents a constant matrix of the i-th follower, t represents the number of each running time, t = 1, 2, 3, ..., d, d represents the total number of running times, i represents the number of each follower, and i ∈ V, t, d and i are all positive integers.

[0020] Preferably, the process of establishing a bounded FDI attack model is as follows: S31, construct a weighted adjacency matrix, denoted as A, then A = [a ij ], where i and j represent the follower numbers, (i,j) represents the communication channel between the i-th follower and the j-th follower, i∈V,j∈V, and i≠j, when At that time, a ij =1, when At that time, a ij =0.

[0021] S32. Use the Lapses matrix L to describe G, denoted as L = [l ij ]∈R N×N When i≠j, l ij =-a ij ,otherwise

[0022] S33. Construct the coupling weights, denoted as B, then B = diag{b1,b2,b3,...,b...} i}, where b i b represents the return value indicating the status received by the i-th follower. b is the value when the i-th follower can receive the leader's status information. i =1, when the i-th follower cannot receive the leader's status information, b i =0.

[0023] S34. Construct the neighbor set of each follower, denoted as D. Where k represents the ID of each neighboring follower following the i-th follower, and v k ∈V, and k≠i.

[0024] S35. Based on the idea of ​​adaptive filtering, obtain the estimates of each state vector and the spurious data vector.

[0025] S36. Establish a bounded FDI attack model:

[0026] η ij (t)=η j (t)+δ ij (t),

[0027]

[0028] In the formula η ij (t) represents the state vector of the leader estimated by the i-th follower after the j-th follower suffers an FDI attack on channel (i,j), where η is the final state vector of the leader. j (t) represents the state vector estimated by the j-th follower for the leader at time t, δ ij (t) represents the spoofed data vector generated by the (i,j) channel under FDI attack at runtime t. δ i (t) represents the vector of the sum of spurious data received by the i-th follower from each channel at time t, δ i0 (t) represents the specific spurious data vector of the i-th follower.

[0029] Preferably, the simultaneous construction of the distributed resilient observer involves the following process: using sensors to obtain the estimated state vectors of each follower after being subjected to an FDI attack, then:

[0030]

[0031]

[0032] In the formula η i(t) represents the state vector of the leader observed by the observer of the i-th follower at time t. Represents η i The derivative of (t) with respect to t, Let represent the state of the filter with adaptive compensator on the i-th follower at time t. represent The derivative with respect to t, η i0 (t) represents the leader's state information received by the i-th follower after it is attacked by FDI at runtime t. This represents the compensation sum of the spurious data injected into each channel by the i-th follower at runtime t. M represents the compensation term for spurious data by the i-th follower at time t on channel (i,j). i The diagonal block matrix represents the i-th follower. This represents the estimated upper bound of the attack signal on channel (i,j) at runtime t. Let represent any continuously positive definite bounded function of channel (i,j) at time t. represent The derivative with respect to t, where c represents a fixed constant and T represents the transpose operator.

[0033] Preferably, the process of establishing the corresponding dynamic observation error model is as follows: S41, Establish the dynamic observation error vector:

[0034]

[0035] In the formula Represents the dynamic observation error vector at the t-th running time;

[0036] S42. Establish a dynamic observation error model: Obtain the true values ​​of the upper bound of the attack signal on each channel from the database, and combine them with the observer and filter information to obtain:

[0037]

[0038] H = L + B,

[0039]

[0040] In the formula Represents the dynamic observation error at time t. represent The derivative with respect to t, This represents the filter vector with adaptive filter at time t. represent The derivative with respect to t, where H represents the sum of the Laplace matrix and the coupling weight matrix, I p This represents a p-order identity matrix, where p is a positive integer. The compensator vector represents the vector at the t-th running time. This represents the upper bound error of the attack signal on channel (i,j) at runtime t. represent The derivative with respect to t, This represents the true value of the upper bound of the attack signal on the (i,j) channel at the t-th running time.

[0041] Preferably, the sufficient condition for the convergence of the analysis observer estimation error is as follows: S51, Construct the Lyapunov equation:

[0042]

[0043] M = diag{M1, M2, ..., M} i},

[0044] In the formula, M represents a diagonal matrix, Z represents a symmetric positive definite matrix, and M i This represents the diagonal matrix of the i-th follower.

[0045] S52. Analyze the convergence condition of the observation error of the i-th agent under FDI attack:

[0046]

[0047] When the above matrix inequality holds, it means that the estimation error of the i-th follower of the leader's state can converge to a preset range; when the above matrix inequality does not hold, it means that the estimation error of the i-th follower of the leader's state cannot converge to a preset range.

[0048] Preferably, the process of using multi-scale control technology to perform fast-slow decomposition to obtain the corresponding fast and slow subsystems is as follows: Using multi-scale control technology to obtain the state variables of the slow subsystem and the fast subsystem, then:

[0049]

[0050]

[0051] In the formula Let represent the slow state quantity of the slow subsystem of the i-th follower at time t. Let represent the control quantity of the slow subsystem of the i-th follower at time t. represent The derivative with respect to t, where s represents the sign of the variables related to the slow subsystem. Let represent the state of the fast subsystem of the i-th follower at time t. Let represent the control quantity of the fast subsystem of the i-th follower at time t. represent The derivative with respect to t, f represents the sign of the variables related to the tachyon system.

[0052] Preferably, the specific process of constructing the controller and dynamic output error system model is as follows: S61, Constructing a novel composite controller: Using sensors to acquire the state variables of each follower, and obtaining the constant matrix in the output equation of each follower and the constant matrix in the output equation of the leader from the database, then:

[0053]

[0054] In the formula u i (t) represents the control input of the i-th follower at time t, x i (t) represents the system's fast / slow state variables of the i-th follower at time t, z i (t) represents the state of the i-th follower at time t. and Both represent gain matrices, and and Both are by Hurwiz.

[0055] Solve according to the matrix adjustment equation. and :

[0056]

[0057]

[0058] In the formula C i Ci represents the constant matrix in the output equation of the i-th follower, and C0 represents the constant matrix in the output equation of the leader.

[0059] S62. Establish the corresponding dynamic output error system model:

[0060]

[0061] In the formula Let X represent the error vector of the i-th follower slow subsystem at time t. s i The constant matrix representing the i-th follower is obtained from S61.

[0062] Preferably, the sufficient condition for consistent analysis and adjustment outputs is as follows: S71, Construct the corresponding Lyapunov equation:

[0063]

[0064] In the formula W i (t) represents the Lyapunov function of the i-th follower at time t, R i The symmetric positive definite matrix represents the i-th follower.

[0065] S72. A sufficient condition for the leader and followers to adjust their outputs to be consistent under an FDI attack is:

[0066]

[0067] In the formula, He is an operator that represents taking the Hermite part of the matrix, and μ represents a given scalar. When the above inequality is true, it means that the i-th follower and the leader adjust their output in the same way under the FDI attack. When the above inequality is false, it means that the i-th follower and the leader adjust their output in different ways under the FDI attack.

[0068] Secondly, the present invention provides a dual-timescale security collaborative control system for unmanned systems under FDI attacks, including the following modules: a topology graph and dynamic model building module for building a communication topology graph based on communication nodes and constructing a dynamic model of the unmanned system with dual-timescale characteristics.

[0069] The observer and model building module is used to establish a bounded FDI attack model between follower communication channels, while building a distributed resilient observer and designing filters with adaptive compensators on each distributed resilient observer to establish a corresponding dynamic observation error model.

[0070] The module for analyzing sufficient conditions for convergence of observer estimation error is used to analyze unmanned systems with dual timescale characteristics under FDI attacks using Lyapunov theory, and to analyze sufficient conditions for convergence of observer estimation error.

[0071] The fast-slow decomposition module is used to decompose followers with dual time-scale characteristics into fast and slow subsystems using multi-scale control techniques.

[0072] The controller and dynamic output error system model building module is used to obtain the state information of the fast subsystem and the slow subsystem observed by the observer, build a new composite controller, and establish the corresponding dynamic output error system model.

[0073] The sufficient condition analysis module for regulating output consistency is used to analyze unmanned systems with dual time scale characteristics under FDI attacks using Lyapunov stability theory and multi-scale control technology, and to analyze the sufficient conditions for the leader and followers to regulate output consistency under FDI attacks.

[0074] The database is used to store the constant matrices, the control inputs of each follower, the gain matrices, the control quantities of each follower's slow subsystem, the control quantities of each follower's fast subsystem, the symmetric positive definite matrices, the diagonal matrices of each follower, the estimated values ​​of the upper bound of the attack signal, the continuously positive definite bounded functions, and the fixed constants.

[0075] The beneficial effects of this invention are as follows: 1. This invention provides a dual-timescale secure cooperative control method and system for unmanned systems under FDI attacks. First, a communication topology diagram and a dynamic model of the unmanned system with dual-timescale characteristics are established. Then, a boundary-based FDI attack model and a dynamic observation error model are established, and a distributed elastic observer is constructed. Next, the sufficient conditions for convergence of observation errors are analyzed. Then, the fast and slow decomposition of followers is performed, and a novel composite controller and a dynamic output error system model are constructed. Finally, the sufficient conditions for the leader and followers to adjust their outputs in a consistent manner under FDI attacks are analyzed. This avoids the ill-conditioned numerical problem of the adjustment equation caused by the dual-timescale characteristics, eliminates FDI attacks in real time, and is not affected by the time and probability of the attack. It effectively realizes the output consistency of the leader and followers in the unmanned system.

[0076] 2. This invention uses multi-scale control technology to decompose each follower with dual time-scale characteristics into fast and slow subsystems, thereby avoiding the ill-conditioned numerical problem of the regulation equation caused by dual time-scale characteristics.

[0077] 3. This invention combines the information of the slow subsystems of each follower with the information of the fast subsystems of each follower to design a novel composite controller. This novel composite controller can effectively achieve output consistency between the leader and the followers in an unmanned system. Attached Figure Description

[0078] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0079] Figure 1 This is a schematic diagram of the implementation steps of the method of the present invention.

[0080] Figure 2This is a schematic diagram of the system structure connection of the present invention.

[0081] Figure 3 This is a control structure block diagram of a safe collaborative control method for unmanned systems based on multi-scale control technology under FDI attacks, provided by the present invention.

[0082] Figure 4 This is a schematic diagram of FDI attack modeling in an embodiment of the present invention.

[0083] Figure 5 This is a schematic diagram of the communication topology between four multi-agents in an embodiment of the present invention.

[0084] Figure 6 This is a state curve diagram of the observer without a compensator in an embodiment of the present invention.

[0085] Figure 7 This is a state curve diagram of the observer when a compensator is present in an embodiment of the present invention.

[0086] Figure 8 This is a graph showing the observation error without a compensator in an embodiment of the present invention.

[0087] Figure 9 This is a graph showing the observation error when a compensator is present in an embodiment of the present invention.

[0088] Figure 10 This is a diagram showing the consistent output effect without a compensator in an embodiment of the present invention.

[0089] Figure 11 This is a diagram illustrating the consistent output effect of adding a compensator in an embodiment of the present invention. Detailed Implementation

[0090] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0091] Please see Figure 1 As shown, in the first aspect, the present invention provides a dual-timescale security cooperative control method for unmanned systems under FDI attacks, including the following steps: Step 1: Establishing a topology graph and dynamic model: Establishing a communication topology graph based on communication nodes, and constructing a dynamic model of the unmanned system with dual-timescale characteristics.

[0092] In a specific embodiment, the process of establishing a communication topology graph based on communication nodes is as follows: Establishing the topology graph: S11, obtaining each communication node from the database.

[0093] S12. Establish a communication topology diagram for an unmanned system with dual time scale characteristics:

[0094]

[0095] In the formula, G represents the communication topology of an unmanned system with dual time scales, and V represents the set of communication nodes, then V = {v1, v2, v3, ..., v N}, v i Let i represent the i-th follower, where i represents the follower's ID, i = 1, 2, 3, ..., N, and N represents the total number of followers. Let the set of edges in the communication topology graph representing an unmanned system with dual time-scale characteristics be given.

[0096] Establish a dynamic model: S21. Use sensors to obtain the leader's state vector, the leader's output vector, the follower's output vector, the follower's slow state vector, and the fast state vector, and obtain the constant matrices from the database.

[0097] S22. Construct a dynamic model of an unmanned system with dual time scale characteristics:

[0098]

[0099] In the formula, x0(t) represents the leader's state vector at time t, y0(t) represents the leader's output vector at time t, and A0 and B0 are both constant matrices. Let represent the derivative of the leader's state vector at time t with respect to t. The derivative of the slow state vector of the i-th follower at time t with respect to t, x i (t) represents the slow state vector of the i-th follower at time t. z represents the derivative of the fast state vector of the i-th follower at time t with respect to t. i (t) represents the fast state vector of the i-th follower at time t, u i (t) represents the control input of the i-th follower at time t, y i (t) represents the output vector of the i-th follower at time t, and εi represents the degree of separation of the dynamics of the fast and slow state vectors of the i-th follower on the fast and slow time scales. and Each represents a constant matrix of the i-th follower, t represents the number of each running time, t = 1, 2, 3, ..., d, d represents the total number of running times, i represents the number of each follower, and i ∈ V, t, d and i are all positive integers.

[0100] It should be noted that each communication node represents a follower.

[0101] It should be noted that the constant matrices mentioned above were obtained by the designers based on experiments.

[0102] It should also be noted that the derivatives mentioned above were calculated by the designers.

[0103] It should also be noted that the separation of the dynamics of the fast and slow state vectors of the follower on the fast and slow time scales was set by the designers with reference to existing data and conclusions.

[0104] It should also be noted that, The first constant matrix representing the slow state vectors of each follower. The first constant matrix representing the control input vectors of each follower. The second constant matrix represents the slow state vector of each follower. The second constant matrix represents the control input vector of each follower. The first constant matrix representing the fast state vectors of each follower. The second constant matrix represents the fast state vector of each follower.

[0105] Step 2: Constructing the observer and model: Establish a bounded FDI attack model between follower communication channels, construct a distributed resilient observer, design filters with adaptive compensators on each distributed resilient observer, and establish a corresponding dynamic observation error model.

[0106] In a specific embodiment, the process of establishing a bounded FDI attack model is as follows: S31, construct a weighted adjacency matrix, denoted as A, then A = [a ij ], where i and j represent the follower numbers, (i,j) represents the communication channel between the i-th follower and the j-th follower, i∈V,j∈V, and i≠j, when At that time, a ij =1, when At that time, a ij =0.

[0107] It should be noted that, This indicates that there is a communication channel between the i-th follower and the j-th follower. This indicates that there is no communication channel between the i-th follower and the j-th follower.

[0108] S32. Use the Lapsian matrix L to describe G, denoted as When i ≠ j, l ij =-a ij ,otherwise

[0109] S33. Construct the coupling weights, denoted as B, then B = diag{b1,b2,b3,...,b...} i}, where b i b represents the return value indicating the status received by the i-th follower. b is the value when the i-th follower can receive the leader's status information. i =1, when the i-th follower cannot receive the leader's status information, b i =0.

[0110] It should be noted that diag represents extracting the diagonal elements of the matrix.

[0111] S34. Construct the neighbor set of each follower, denoted as D. Where k represents the ID of each neighboring follower following the i-th follower, and v k ∈V, and k≠i.

[0112] S35. Based on the idea of ​​adaptive filtering, obtain the estimates of each state vector and the spurious data vector.

[0113] S36. Establish a bounded FDI attack model:

[0114] η ij (t)=η j (t)+δ ij (t),

[0115]

[0116] In the formula η ij (t) represents the state vector of the leader estimated by the i-th follower after the j-th follower suffers an FDI attack on channel (i,j), where η is the final state vector of the leader. j (t) represents the state vector estimated by the j-th follower for the leader at time t, δ ij (t) represents the spoofed data vector generated by the (i,j) channel under FDI attack at runtime t. δ i (t) represents the vector of the sum of spurious data received by the i-th follower from each channel at time t, δ i0 (t) represents the specific spurious data vector of the i-th follower.

[0117] It should be noted that the vector of spoofed data generated by each follower under FDI attack on each channel is summed to obtain the vector of the sum of spoofed data received by each follower on each channel.

[0118] It should also be noted that the specific spoof data vectors for each follower are set by the designers.

[0119] In another specific embodiment, the simultaneous construction of the distributed resilient observer is carried out as follows: Using sensors to acquire the state vectors of each follower after being subjected to an FDI attack, then:

[0120]

[0121]

[0122] In the formula η i (t) represents the state vector of the leader observed by the observer of the i-th follower at time t. Represents η i The derivative of (t) with respect to t, Let represent the state of the filter with adaptive compensator on the i-th follower at time t. represent The derivative with respect to t, η i0 (t) represents the leader's state information received by the i-th follower after it is attacked by FDI at runtime t. This represents the compensation sum of the spurious data injected into each channel by the i-th follower at runtime t. M represents the compensation term for spurious data by the i-th follower at time t on channel (i,j). i The diagonal block matrix represents the i-th follower. This represents the estimated upper bound of the attack signal on channel (i,j) at runtime t. Let represent any continuously positive definite bounded function of channel (i,j) at time t. represent The derivative with respect to t, where c represents a fixed constant and T represents the transpose operator.

[0123] It should be noted that the derivatives mentioned above were obtained by the designers through calculation.

[0124] It should also be noted that the compensation for the attack signal, the compensation term for the spurious data, the diagonal block matrix, the estimated value of the upper bound of the attack signal, the continuous positive definite bounded function, and the fixed constant are all set by the designers.

[0125] It should also be noted that the compensations for the attack signal from the filters with adaptive compensators on each follower are summed to obtain the compensation sum of the spurious data injected by each follower in each channel.

[0126] In another specific embodiment, the establishment of the corresponding dynamic observation error model is carried out as follows: Establishing the dynamic observation error model: Obtain the true values ​​of the upper bound of the attack signal on each channel from the database, and combine this with the observer and filter information to obtain:

[0127]

[0128] H = L + B,

[0129]

[0130] In the formula Represents the dynamic observation error at time t. represent The derivative with respect to t, This represents the filter vector with adaptive filter at time t. represent The derivative with respect to t, where H represents the sum of the Laplace matrix and the coupling weight matrix, I p This represents a p-order identity matrix, where p is a positive integer. The compensator vector represents the vector at the t-th running time. This represents the upper bound error of the attack signal on channel (i,j) at runtime t. represent The derivative with respect to t, This represents the true value of the upper bound of the attack signal on the (i,j) channel at the t-th running time.

[0131] It should be noted that the derivatives mentioned above were calculated by the designers.

[0132] It should also be noted that when using machine learning algorithms to obtain the true value of the upper bound of the attack signal on each channel, a feature selection algorithm is first used to select the most representative and discriminative feature subset from the extracted features, then redundant and irrelevant features are removed, and finally a regression model is selected to predict the upper bound of the attack signal strength.

[0133] Step 3: Analyze the sufficient condition for convergence of the observer estimation error: Using Lyapunov theory, analyze the unmanned system with dual time scale characteristics under FDI attack, and give the sufficient condition for convergence of the observer estimation error.

[0134] In a specific embodiment, the sufficient condition for the convergence of the analysis observer estimation error is as follows: S51, S51, Construct the Lyapunov equation:

[0135]

[0136] M = diag{M1, M2, ..., M} i},

[0137] In the formula, M represents a diagonal matrix, Z represents a symmetric positive definite matrix, and M i This represents the diagonal matrix of the i-th follower.

[0138] It should be noted that the symmetric positive definite matrix and the diagonal matrix of each follower are both set by the designers.

[0139] S52. Analyze the convergence condition of the observation error of the i-th agent under FDI attack:

[0140]

[0141] When the above matrix inequality holds, it means that the estimation error of the i-th follower of the leader's state can converge to a preset range; when the above matrix inequality does not hold, it means that the estimation error of the i-th follower of the leader's state cannot converge to a preset range.

[0142] Step 4: Perform fast and slow decomposition: For followers with dual time scale characteristics, use multi-scale control techniques to perform fast and slow decomposition to obtain the corresponding fast subsystem and slow subsystem.

[0143] In a specific embodiment, the process of using multi-scale control technology to perform fast-slow decomposition to obtain the corresponding fast and slow subsystems is as follows: Using multi-scale control technology to obtain the state variables of the slow subsystem and the fast subsystem, then:

[0144]

[0145] In the formula Let represent the slow state quantity of the slow subsystem of the i-th follower at time t. Let represent the control quantity of the slow subsystem of the i-th follower at time t. represent The derivative with respect to t, where s represents the sign of the variables related to the slow subsystem. Let represent the state of the fast subsystem of the i-th follower at time t. Let represent the control quantity of the fast subsystem of the i-th follower at time t. represent The derivative with respect to t, f represents the sign of the variables related to the tachyon system.

[0146] It should be noted that the control quantities of the slow subsystems and the fast subsystems of each follower are both designed and controlled by the designers.

[0147] Step 5: Construct the controller and dynamic output error system model: Obtain the state information of the fast subsystem and the slow subsystem observed by the observer, construct a new composite controller, and establish the corresponding dynamic output error system model.

[0148] In a specific embodiment, the process of constructing the controller and dynamic output error system model is as follows: S61, S61, Constructing a novel composite controller: Using sensors to acquire the state variables of each follower, and obtaining the constant matrix from the output equation of each follower and the constant matrix from the output equation of the leader from the database:

[0149]

[0150] In the formula u i (t) represents the control input of the i-th follower at time t, x i (t) represents the system's fast / slow state variables of the i-th follower at time t, z i (t) represents the state of the i-th follower at time t. and Both represent gain matrices, and and Both are by Hurwiz.

[0151] It should be noted that Hurwiz is a matrix that represents the asymptotic stability of the corresponding system and is a condition for ensuring the stability of the novel composite controller design.

[0152] It should also be noted that, The gain matrix representing the slow state vector of each follower. The gain matrix representing the observations of each follower on the leader's state vector is given by the observer. Represents the first gain matrix. Represents the second gain matrix. This represents the third gain matrix.

[0153] Solve according to the matrix adjustment equation. and :

[0154]

[0155] In the formula C iCi represents the constant matrix in the output equation of the i-th follower, and C0 represents the constant matrix in the output equation of the leader.

[0156] It should be noted that the constant matrices in the output equations of each follower and the constant matrices in the output equation of the leader are both set by the designers.

[0157] S62. Establish the corresponding dynamic output error system model:

[0158]

[0159]

[0160] In the formula Let represent the error vector of the i-th follower slow subsystem at time t. The constant matrix representing the i-th follower is obtained from S61.

[0161] It should also be noted that the difference between the estimated state vector of each follower slow subsystem and the state vector of each follower slow subsystem is used as the error vector of each follower slow subsystem.

[0162] It should also be noted that the constant matrix of each follower is set by the designers.

[0163] Step Six: Analyze the sufficient conditions for consistent adjustment outputs: Using Lyapunov stability theory and multi-scale control technology, analyze the unmanned system with dual time-scale characteristics under FDI attack, and analyze the sufficient conditions for consistent adjustment outputs between the leader and follower under FDI attack.

[0164] In a specific embodiment, the sufficient condition for consistent analysis and regulation outputs is as follows: S71, Construct the corresponding Lyapunov equation:

[0165]

[0166] In the formula W i (t) represents the Lyapunov function of the i-th follower at time t, R i The symmetric positive definite matrix represents the i-th follower.

[0167] It should be noted that the symmetric positive definite matrices of each follower are set by the designers.

[0168] S72. A sufficient condition for the leader and followers to adjust their outputs to be consistent under an FDI attack is:

[0169]

[0170] In the formula, He is an operator that represents taking the Hermite part of the matrix, and μ represents a given scalar. When the above inequality is true, it means that the i-th follower and the leader adjust their output in the same way under the FDI attack. When the above inequality is false, it means that the i-th follower and the leader adjust their output in different ways under the FDI attack.

[0171] It should be noted that the given scalar is set by the designer.

[0172] The second aspect of the present invention provides a dual-timescale security collaborative control system for unmanned systems under FDI attacks, comprising the following modules: a topology graph and dynamic model building module for building a communication topology graph based on communication nodes and constructing a dynamic model of the unmanned system with dual-timescale characteristics.

[0173] The observer and model building module is used to establish a bounded FDI attack model between follower communication channels, while building a distributed resilient observer and designing filters with adaptive compensators on each distributed resilient observer to establish a corresponding dynamic observation error model.

[0174] The module for analyzing sufficient conditions for convergence of observer estimation error is used to analyze unmanned systems with dual timescale characteristics under FDI attacks using Lyapunov theory, and to analyze sufficient conditions for convergence of observer estimation error.

[0175] The fast-slow decomposition module is used to decompose followers with dual time-scale characteristics into fast and slow subsystems using multi-scale control techniques.

[0176] The controller and dynamic output error system model building module is used to obtain the state information of the fast subsystem and the slow subsystem observed by the observer, build a new composite controller, and establish the corresponding dynamic output error system model.

[0177] The sufficient condition analysis module for regulating output consistency is used to analyze unmanned systems with dual time scale characteristics under FDI attacks using Lyapunov stability theory and multi-scale control technology, and to analyze the sufficient conditions for the leader and followers to regulate output consistency under FDI attacks.

[0178] The database is used to store the constant matrices, the control inputs of each follower, the gain matrices, the control quantities of each follower's slow subsystem, the control quantities of each follower's fast subsystem, the symmetric positive definite matrices, the diagonal matrices of each follower, the estimated values ​​of the upper bound of the attack signal, the continuously positive definite bounded functions, and the fixed constants.

[0179] This invention first establishes a communication topology diagram and a dynamic model of an unmanned system with dual time-scale characteristics. Then, it establishes a bounded FDI attack model and a dynamic observation error model, and constructs a distributed elastic observer. Next, it analyzes the sufficient conditions for the convergence of observation errors. Then, it decomposes the followers into fast and slow groups, and constructs a novel composite controller and a dynamic output error system model. Finally, it analyzes the sufficient conditions for the leader and followers to adjust their outputs in a consistent manner under FDI attacks. This avoids the ill-conditioned numerical problem of the adjustment equation caused by the dual time-scale characteristics, eliminates FDI attacks in real time, and is not affected by the time and probability of the attack. It effectively achieves the output consistency between the leader and followers in the unmanned system.

[0180] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the protection scope of the present invention.

Claims

1. A method for safe cooperative control of unmanned systems under FDI attack, characterized in that, Includes the following steps: Step 1: Establish topology and dynamic model: Establish a communication topology based on the communication nodes, and construct a dynamic model of the unmanned system with dual time scale characteristics; Step 2: Constructing the Observer and Model: Establish a bounded FDI attack model between follower communication channels, and simultaneously construct distributed resilient observers. Design filters with adaptive compensators on each distributed resilient observer, and establish corresponding dynamic observation error models. The process of constructing the distributed resilient observers is as follows: Use sensors to obtain the estimated state vectors of each follower towards the leader after being subjected to an FDI attack. Then: , , , , , In the formula Representing the The observer of the follower in the first The leader's state vector is observed at each runtime. represent about The derivative, Representing the The filter with adaptive compensator on the follower is in the first... The filter's own state at each running time. represent about The derivative, Representing the The first follower in the The leader's status information received after a runtime event is subjected to an FDI attack. Representing the The first follower in the Compensation for spurious data injected into each channel at each runtime moment, Representative at the The running time of the first running time One follower Compensation items for false data on the channel, Representing the A diagonal block matrix of followers Representative at the Each running time An estimate of the upper bound of the attack signal on the channel. represent The channel is in Any continuously positive definite bounded function at each running time, represent about The derivative, Represents a fixed constant. Represents the transpose operator; Step 3: Analyze the sufficient conditions for convergence of the observer estimation error: Using Lyapunov theory, analyze the unmanned system with dual time scale characteristics under FDI attack, and analyze the sufficient conditions for convergence of the observation error. Step 4: Perform fast and slow decomposition: For followers with dual time scale characteristics, use multi-scale control techniques to perform fast and slow decomposition to obtain the corresponding fast subsystem and slow subsystem; Step 5: Construct the controller and dynamic output error system model: Obtain the state information of the fast subsystem and the slow subsystem observed by the observer, construct a novel composite controller, and establish the corresponding dynamic output error system model. The process of constructing the novel composite controller is as follows: Use sensors to obtain the state variables of each follower, and obtain the constant matrix from the output equation of each follower and the constant matrix from the output equation of the leader from the database. , , , In the formula Representing the The first follower in the Control inputs at each running moment Representing the The first follower in the The system's fast and slow states at each runtime moment. Representing the The first follower in the The state variables at each runtime moment , , , and Both represent gain matrices, and and Both are from Hurwiz; Step Six: Analyze the sufficient conditions for consistent adjustment outputs: Using Lyapunov stability theory and multi-scale control technology, analyze the unmanned system with dual time-scale characteristics under FDI attack, and analyze the sufficient conditions for consistent adjustment outputs between the leader and follower under FDI attack.

2. The dual-timescale secure cooperative control method for unmanned systems under FDI attacks according to claim 1, characterized in that, The specific process of establishing a communication topology map based on communication nodes is as follows: Create a topology graph: S11. Retrieve each communication node from the database; S12. Establish a communication topology diagram for an unmanned system with dual time scale characteristics: , In the formula A diagram representing the communication topology of an unmanned system with dual time-scale characteristics. The set representing all communication nodes, then , Representing the One follower The numbers representing the followers, =1,2,3,..., , Represents the total number of followers. Let the set of edges in the communication topology graph representing an unmanned system with dual time-scale characteristics be given. ; Establish a dynamic model: S21. Use sensors to obtain the leader's state vector, the leader's output vector, the follower's output vector, the follower's slow state vector, and the fast state vector, and obtain each constant matrix from the database. S22. Construct a dynamic model of an unmanned system with dual time scale characteristics: , , In the formula Representative leader at The state vector at each running time point Representative leader at The output vector at each runtime time step and Both represent constant matrices. Representative leader at The state vector at each running time point is related to The derivative, No. The first follower in the The slow state vector at each running time step is about The derivative, Representing the The first follower in the The slow state vector at each running time step Representing the The first follower in the The fast state vector at each running time step is about The derivative, Representing the The first follower in the A fast state vector at each runtime. Representing the The first follower in the Control inputs at each running moment Representing the The first follower The output vector at each runtime time step Representing the The degree of separation of the dynamics of the fast and slow state vectors of a follower on the fast and slow time scales. , , , , and All represent the first A constant matrix of followers The number representing each running moment, =1,2,3,..., , Represents the total number of running times. Represents the number of each follower, and , , and All are positive integers.

3. The dual-timescale secure cooperative control method for unmanned systems under FDI attacks according to claim 2, characterized in that, The process of establishing a bounded FDI attack model is as follows: S31. Construct a weighted adjacency matrix, denoted as... ,but ,in and Each of these represents the number of the follower. Representing the The first follower and the first Communication channels between followers ,and ,when hour, ,when hour, ; S32, Using the Lapses matrix To describe , recorded as ,when hour, ,otherwise ; S33. Construct the coupling weights, denoted as... ,but ,in Representing the The follower receives the return value of the state, when the first follower receives the return value of the state. When a follower can receive status information from the leader. When the first When a follower cannot receive status information from the leader. ; S34. Construct the neighbor set of each follower, denoted as . ,but ,in Representative followed the first The numbers of each follower's neighboring followers. ,and ; S35. Based on the idea of ​​adaptive filtering, obtain the estimates of each state vector and the spurious data vector; S36. Establish a bounded FDI attack model: , , In the formula Representing the The follower eventually received the first... One follower The estimated state vector of the leader after the channel is subjected to an FDI attack. Representing the The first follower in the The estimated state vector of the leader at each running time step. represent The channel is in A spoofed data vector generated by an FDI attack at any given runtime. , Representing the The first follower in the A vector representing the sum of spurious data received by each channel at each runtime moment. Representing the A specific spoofed data vector of a follower.

4. The dual-timescale secure cooperative control method for unmanned systems under FDI attacks according to claim 1, characterized in that, The specific process for establishing the corresponding dynamic observation error model is as follows: S41. Establish the dynamic observation error vector: , , In the formula Representing the The dynamic observation error vector at each running time; S42. Establish a dynamic observation error model: Obtain the true values ​​of the upper bound of the attack signal on each channel from the database, and combine them with the observer and filter information to obtain: , , , , , , , In the formula Representing the The dynamic observation error at each running moment, represent about The derivative, Represents the number of adaptive filters in the first... The filter vector at each running time step. represent about The derivative, This represents the sum of the Laplacian matrix and the coupling weight matrix. represent An identity matrix of order 1. It is a positive integer. Representing the The compensator vector at each runtime, Representative at the Each running time Upper bound error of attack signal on the channel represent About derivative, Representative at the Each running time The true value of the upper bound of the attack signal on the channel.

5. The dual-timescale secure cooperative control method for unmanned systems under FDI attacks according to claim 4, characterized in that, The sufficient condition for convergence of the analysis observer estimation error is as follows: S51. Construct the Lyapunov equation: , , In the formula Represents a diagonal matrix. Represents a symmetric positive definite matrix. Representing the A diagonal matrix of followers; S52. Analysis of the situation under FDI attack, the first Convergence condition for individual agent observation errors: , When the above matrix inequality holds, it represents the first... The estimation error of the leader's state by each follower can converge to a preset range when the above matrix inequality does not hold, representing the... The estimation error of a follower's state of the leader cannot converge to the preset range.

6. The dual-timescale secure cooperative control method for unmanned systems under FDI attacks according to claim 2, characterized in that, The process of using multi-scale control techniques to perform fast-slow decomposition to obtain the corresponding fast and slow subsystems is as follows: Using multi-scale control techniques to obtain the state variables of the slow subsystem and the fast subsystem, then: , , , , , , In the formula Representing the The slow subsystem of the follower is in the first The system slow state quantity at each runtime moment Representing the The slow subsystem of the follower is in the first Control variables at each runtime moment represent about The derivative, Symbols representing variables relevant to the slow subsystem. Representing the The fast subsystem of the follower in the first... The state variables at each runtime moment Representing the The fast subsystem of the follower in the first... Control variables at each runtime moment represent about The derivative, Symbols representing variables related to the identification of tachy subsystems.

7. The dual-timescale secure cooperative control method for unmanned systems under FDI attacks according to claim 6, characterized in that, The specific process for constructing the controller and dynamic output error system model is as follows: S61. Solve according to the matrix adjustment equation. and : , , In the formula Representing the The constant matrix in the output equation of each follower The constant matrix representing the leader's output equation; S62. Establish the corresponding dynamic output error system model: , , In the formula Representing the The follower slow subsystem in the first... The error vector at each running time step. Representing the The constant matrix of the followers is obtained by S61.

8. The dual-timescale secure cooperative control method for unmanned systems under FDI attacks according to claim 6, characterized in that, The sufficient condition for consistent analysis and adjustment outputs is as follows: S71. Construct the corresponding Lyapunov equation: , In the formula Representing the The first follower in the Lyapunov functions at each runtime, Representing the A symmetric positive definite matrix of followers; S72. A sufficient condition for the leader and followers to adjust their outputs to be consistent under an FDI attack is: , In the formula It is an operator that represents retrieving the Hermite portion of a matrix. Represents a given scalar, and when the above inequality holds, represents the value of the first scalar under an FDI attack. The number of followers and the leader adjust their outputs to be consistent. When the above inequality does not hold, it means that under an FDI attack, the number of followers and the leader adjust their outputs to be consistent. The output of the followers and the leader is inconsistent.

9. A cooperative control system implementing the dual-timescale secure cooperative control method for unmanned systems under FDI attack as described in any one of claims 1-8, characterized in that, include: The topology graph and dynamic model building module is used to build a communication topology graph based on communication nodes and construct a dynamic model of the unmanned system with dual time scale characteristics. The observer and model building modules are used to establish a bounded FDI attack model between follower communication channels, and simultaneously construct distributed resilient observers. Filters with adaptive compensators are designed on each distributed resilient observer to establish a corresponding dynamic observation error model. The process of constructing the distributed resilient observers is as follows: using sensors to obtain the estimated state vectors of each follower to the leader after being subjected to an FDI attack, then: , , , , , In the formula Representing the The observer of the follower in the first The leader's state vector is observed at each runtime. represent about The derivative, Representing the The filter with adaptive compensator on the follower is in the first... The filter's own state at each running time. represent about The derivative, Representing the The first follower in the The leader's status information received after a runtime event is subjected to an FDI attack. Representing the The first follower in the Compensation for spurious data injected into each channel at each runtime moment, Representative at the The running time of the first running time One follower Compensation items for false data on the channel, Representing the A diagonal block matrix of followers Representative at the Each running time An estimate of the upper bound of the attack signal on the channel. represent The channel is in Any continuously positive definite bounded function at each running time, represent about The derivative, Represents a fixed constant. Represents the transpose operator; The module for analyzing sufficient conditions for convergence of observer estimation error is used to analyze unmanned systems with dual time-scale characteristics under FDI attacks using Lyapunov theory, and to analyze sufficient conditions for convergence of observer estimation error. The fast-slow decomposition module is used to decompose followers with dual time-scale characteristics into fast and slow subsystems using multi-scale control techniques. The controller and dynamic output error system model building module is used to acquire the state information of the fast and slow subsystems observed by the observer, construct a novel composite controller, and establish the corresponding dynamic output error system model. The process of constructing the novel composite controller involves: using sensors to acquire the state variables of each follower, and retrieving the constant matrix from the output equations of each follower and the constant matrix from the leader's output equation from the database. , , , In the formula Representing the The first follower in the Control inputs at each running moment Representing the The first follower in the The system's fast and slow states at each runtime moment. Representing the The first follower in the The state variables at each runtime moment , , , and Both represent gain matrices, and and Both are from Hurwiz; The sufficient condition analysis module for regulating output consistency is used to analyze unmanned systems with dual time scale characteristics under FDI attacks by using Lyapunov stability theory and multi-scale control technology, and to analyze the sufficient conditions for the leader and followers to regulate output consistency under FDI attacks. The database is used to store the constant matrices, the control inputs of each follower, the gain matrices, the control quantities of each follower's slow subsystem, the control quantities of each follower's fast subsystem, the symmetric positive definite matrices, the diagonal matrices of each follower, the estimated values ​​of the upper bound of the attack signal, the continuously positive definite bounded functions, and the fixed constants.

Citation Information

Patent Citations

  • Multi-unmanned aerial vehicle system safety control method based on unknown input observer

    CN116540665A

  • State estimation and security immunity control method for cluster unmanned system under spoofing attack

    CN118818973A