A method and system for cooperative tracking control of a heterogeneous UAV-UGV system under DoS attack

Abstract

The application discloses a kind of DoS attack under heterogeneous UAV-UGV system's cooperative tracking control method and system, it is related to tracking control technical field;Method includes: constructing the dynamic model of unmanned aerial vehicle and unmanned vehicle and intermittent DoS attack model;Unknown dynamic of leader is approximated using fuzzy logic system, and distributed fuzzy adaptive estimator is designed to reconstruct leader state and switch estimation structure according to attack state switching signal;Unknown dynamic of each follower is approximated using fuzzy logic, and switching type fuzzy state observer is designed to estimate unmeasured state and switch observation gain according to attack signal;Based on the output of both, introduce preset time filter to establish coordinate transformation relationship, according to which adaptive update law is designed to update fuzzy weight, and virtual control law is constructed, and then the final preset time cooperative controller is constructed, to drive each follower output to track leader output within preset time.The application can realize that follower UGV tracks the preset time cooperative control of leader UAV trajectory.

Description

Technical Field

[0001] This invention relates to the field of tracking and control technology, and in particular to a collaborative tracking and control method and system for heterogeneous UAV-UGV systems under DoS attacks. Background Technology

[0002] In recent years, multi-UAV systems have shown broad application prospects in areas such as cooperative reconnaissance, area coverage, and formation flight. Essentially, they are a type of distributed multi-agent system characterized by strong nonlinearity, state coupling, and networked communication. In such systems, each UAV shares state information and coordinates its actions through a wireless communication network to complete complex group tasks such as cooperative target tracking, dynamic formation formation, and maintenance. However, the position and attitude subsystems of UAVs typically involve unknown aerodynamic nonlinearities, external environmental disturbances, and actuator dynamics, exhibiting typical multi-input multi-output characteristics. When the system communication link suffers intermittent denial-of-service attacks, sensor measurement data cannot be transmitted normally to the controller or neighboring UAVs during the active attack period, leading to periodic loss of critical state information. Traditional backstepping control methods based on continuous state feedback heavily rely on real-time, complete state information. Under the unpredictable state conditions caused by DoS attacks, not only does tracking accuracy drop sharply, but system instability may even occur due to error accumulation.

[0003] From a technical perspective, existing methods face multiple challenges. First, traditional state observers are typically based on the assumption of continuous availability of measurement data. When a DoS attack causes intermittent interruptions in the output signal, the observer experiences estimation error divergence during the active attack period due to a lack of effective correction. Even if communication is restored during the attack interval, it is difficult to converge quickly to compensate for the accumulated error. Second, multi-UAV systems generally face constraints in communication bandwidth and onboard energy. Traditional periodic control update mechanisms are clearly insufficient in a DoS attack environment: maintaining a high update frequency leads to the loss of a large number of data packets during the attack, resulting in wasted communication resources and the risk of control failure; reducing the update frequency introduces significant latency, deteriorating the system's dynamic response performance. Furthermore, existing research lacks a quantitative description of the frequency and duration of intermittent DoS attacks, often simplifying them to a simple on / off model, failing to fully consider the "active-dormant" alternation characteristics exhibited by the attacker's limited energy, thus making it difficult to establish a quantitative correlation between attack modes and controller parameters.

[0004] The current technological system faces three core bottlenecks: First, there is a lack of dynamic models that can accurately describe the temporal characteristics of intermittent DoS attacks, and existing methods struggle to effectively integrate attack duration and frequency information in controller design. Second, under conditions where output measurement signals are discontinuous, robust state estimation methods suitable for nonlinear multi-UAV systems have not yet been developed. Existing switching observers are mostly limited to single-input single-output or linear systems, making it difficult to handle the estimation complexity brought about by high-dimensional nonlinear coupling. Third, event-triggered control mechanisms and security control requirements under DoS attacks have not been effectively coordinated. Traditional event-triggered conditions are prone to false triggering under measurement jumps caused by attacks, and the design of distributed triggering strategies for multi-UAV collaborative scenarios lacks robustness analysis and comprehensive optimization for the intermittency of attacks. These technological bottlenecks collectively reflect the triple coupling challenge of "state estimation-communication scheduling-cooperative control" faced by nonlinear multi-UAV systems when subjected to cyberattacks. Currently, there is an urgent need to develop an integrated security and cooperative control framework that can uniformly handle intermittent data loss, system nonlinear uncertainty, and communication resource constraints. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides a collaborative tracking control method and system for heterogeneous UAV-UGV systems under DoS attacks. This method enables follower UGVs to perform pre-set time-based collaborative tracking control of the leader UAV's trajectory even when the communication link suffers intermittent DoS attacks that result in the loss of state information.

[0006] To achieve the above objectives, this invention provides a cooperative tracking control method for a heterogeneous UAV-UGV system under a DoS attack. The heterogeneous UAV-UGV system includes a leader UAV and multiple follower UGVs. The method includes: Construct dynamic models for the leader UAV and the follower UGV respectively, and establish an intermittent DoS attack model characterized by attack time intervals and attack state switching signals; A distributed fuzzy adaptive estimator is designed by approximating the unknown nonlinear function in the dynamic model of the leader UAV using a fuzzy logic system. The distributed fuzzy adaptive estimator is used to reconstruct the state information of the leader UAV and adjust the estimation structure according to the attack state switching signal. The unknown nonlinear functions in the dynamic models of each follower UGV are approximated using a fuzzy logic system to design their respective switching fuzzy state observers; the switching fuzzy state observers are used to estimate the unmeasurable state of the corresponding follower UGV and adjust the observation gain according to the attack state switching signal. Based on the outputs of the distributed fuzzy adaptive estimator and the switching fuzzy state observer, a preset time nonlinear filter is introduced to perform dynamic surface design, and a coordinate transformation relationship including cooperative tracking error, filtering error and state estimation error is established. Based on the coordinate transformation relationship, an adaptive update law for fuzzy parameters is designed to update the fuzzy weights in the distributed fuzzy adaptive estimator and the switching fuzzy state observer. Based on the coordinate transformation relationship and the preset time convergence condition, a preset time virtual control law is constructed; Based on the preset time virtual control law and the updated fuzzy weights, a final preset time collaborative controller is constructed; the final preset time collaborative controller is used to generate control inputs for each follower UGV so that the output of each follower UGV tracks the output of the leader UAV within a preset time.

[0007] Optionally, the expression for the dynamic model of the follower UGV is: In the formula, Number the followers. , The total number of followers; , This refers to the system status; , The state order; For the first The final actual physical control input of the ground-following vehicle; For system output; To be consistent with system status Related unknown smooth nonlinear functions; Represents state variables The first derivative with respect to time, Represents state variables The first derivative with respect to time.

[0008] Optionally, the dynamic model of the leader UAV is expressed as follows: In the formula, , The leader's state order; For the leader One state variable; For the leader One state equation; For leaders to control input; To be with the leader State equations Related unknown smooth nonlinear functions; Output for leaders; Indicates the leader's State variables The first derivative with respect to time; Indicates the leader's State variables The first derivative with respect to time; Indicates the relationship with the leader State equations Related unknown smooth nonlinear functions; This is represented as the first state variable of the leader.

[0009] Optionally, the expression for the distributed fuzzy adaptive estimator is: In the formula, , The leader's state order; For the first Taiwan's intelligent agents and leaders Order-order state estimate; The estimated value of the fuzzy ideal weight vector; , This is the first design parameter; This is the cooperative error term; The coupling coefficient is related to the attack state; Indicates the use of approximating the leader's first A fuzzy basis function vector for nonlinear dynamics with unknown order state; Indicates the use of approximating the leader's first A fuzzy basis function vector for nonlinear dynamics with unknown order state; For leaders to control input; Indicates the first Taiwan's intelligent agents and leaders The estimated value of the first-order state; Indicates the first Taiwan's intelligent agents and leaders The estimated value of the first-order state; Indicates the first The ground-based following agent is the leader's first Order state estimate The time derivative; Indicates the first Taiwan's intelligent agents and leaders Order state estimate The time derivative; This represents the estimated output value for the leader.

[0010] Optionally, the expression for the switching fuzzy state observer is: In the formula, , The state order; For the first Taiwan Followers Internal state estimate of the first order; For the first Taiwan Followers The estimated value of the fuzzy weight of the first-order state; For the first Taiwan Followers The estimated value of the fuzzy weight of the first-order state; For the first Taiwan Followers Observation gain for switching between order states; For the first Taiwan Followers Observation gain for switching between order states; For the first The first-order state output estimation error of the follower; This is a signal to switch attack states. Indicates the use of approximating the first The first follower The fuzzy basis function vector of the unknown nonlinear function of the internal state of the order; Indicates the first Taiwan Followers The fuzzy input vector corresponding to the first-order state estimate; Indicates the first Taiwan Followers The fuzzy input vector corresponding to the first-order state estimate; Indicates the first Taiwan Followers Internal state estimate of the first order The first derivative with respect to time; Indicates the first Taiwan Followers The first derivative of the internal state estimate with respect to time; For the first The final actual physical control input of the ground-following vehicle; Indicates the first Output estimates of the followers; Indicates the first Taiwan follower switching signals The first-order state output value.

[0011] Optionally, the expression for the preset time nonlinear filter is: In the formula, , The state order; Preset convergence time; , The first preset time index, This is the second preset time index; This is the second design parameter. This is the third design parameter; For the first The first follower Design parameters corresponding to the first state; For the first Design parameters corresponding to the first-order state of each follower; Indicates the first Each follower corresponds to the output of a preset time nonlinear filter. The first derivative with respect to time; Indicates the first Auxiliary saturation constraint variables for each follower's first-order filter channel; Indicates the first The first follower Auxiliary saturation constraint variables for the first-order filter channel; Indicates the first The first follower Virtual control signal The first derivative of the output value with respect to time after being filtered by a preset time; Indicates the first The first follower Order state variables; This indicates that the preset time index is Time The first follower Order state variables; This indicates that the preset time index is Time The first follower Order of state variables.

[0012] Optionally, the expression for the fuzzy parameter adaptive update law is: In the formula, , The state order; The coupling coefficient is related to the attack state; Indicates the first The collaborative estimation error vector of the leader's state by each follower; Indicates the use of approximating the leader's first A fuzzy basis function vector for nonlinear dynamics with unknown order state; , This is the first adaptive gain parameter. This is the second adaptive gain parameter; The input matrix; This represents the positive definite symmetric Lyapunov matrix obtained by solving linear matrix inequalities; Indicates the first Fuzzy ideal weight vector estimate of a leader The first derivative with respect to time; Indicates the first The first follower fuzzy weight estimate of the first order state The first derivative with respect to time.

[0013] Optionally, the expression for the preset time virtual control law is: In the formula, Preset convergence time; This is the second design parameter. This is the third design parameter; Indicates the first Auxiliary saturation constraint variables for each follower's first-order filter channel; Indicates the first The first-order preset time virtual control signal for the platform follower; Indicates the first Taiwan Follower First-Order Consistency Tracking Error of Power of; Indicates the first Taiwan Follower First-Order Consistency Tracking Error of Power of; Indicates the use of approximating the first The fuzzy basis function vector of the unknown nonlinear function of the first-order internal state of a follower; Indicates the first The estimated fuzzy weights of the first-order state of each follower; For the first The first-order state output estimation error of the follower; For the first The observation gain of the first-order state switching of the platform follower; This indicates the output corresponding to the preset time nonlinear filter. The first derivative with respect to time.

[0014] Optionally, the expression for the final preset time coordination controller is: In the formula, Preset convergence time; This is the second design parameter. This is the third design parameter; Indicates the first Auxiliary saturation constraint variables for each follower's first-order filter channel; , The state order; Indicates the first The final actual physical control input of the ground-following vehicle; Indicates the first Taiwan Followers Consistency tracking error of Power of; Indicates the first Taiwan Followers Consistency tracking error of Power of; Indicates the use of approximating the first The first follower The fuzzy basis function vector of the unknown nonlinear function of the internal state of the order; Indicates the first The first follower The estimated value of the fuzzy weight of the first-order state; For the first The first-order state output estimation error of the follower; For the first Taiwan Followers Observation gain for switching between order states; Indicates the first Taiwan Followers First-order filter virtual control quantity The first derivative with respect to time.

[0015] This invention also provides a cooperative tracking control system for a heterogeneous UAV-UGV system under a DoS attack, wherein the heterogeneous UAV-UGV system includes a leader UAV and multiple follower UGVs, and the control system includes: The model building unit is used to build dynamic models of the leader UAV and the follower UGV respectively, and to establish an intermittent DoS attack model characterized by attack time intervals and attack state switching signals. A state estimation unit is used to approximate the unknown nonlinear function in the dynamic model of the leader UAV using a fuzzy logic system in order to design a distributed fuzzy adaptive estimator; the distributed fuzzy adaptive estimator is used to reconstruct the state information of the leader UAV and adjust the estimation structure according to the attack state switching signal; A state observation unit is used to approximate the unknown nonlinear function in the dynamic model of each follower UGV using a fuzzy logic system, so as to design their respective switching fuzzy state observers; the switching fuzzy state observers are used to estimate the unmeasurable state of the corresponding follower UGV and adjust the observation gain according to the attack state switching signal. The coordinate transformation unit is used to design a dynamic surface by introducing a preset time nonlinear filter based on the output of the distributed fuzzy adaptive estimator and the switching fuzzy state observer, and to establish a coordinate transformation relationship that includes cooperative tracking error, filtering error and state estimation error. An adaptive update unit is used to design an adaptive update law for fuzzy parameters based on the coordinate transformation relationship, so as to update the fuzzy weights in the distributed fuzzy adaptive estimator and the switching fuzzy state observer. Preset time coordination control unit, used for: Based on the coordinate transformation relationship and the preset time convergence condition, a preset time virtual control law is constructed; Based on the preset time virtual control law and the updated fuzzy weights, a final preset time collaborative controller is constructed; the final preset time collaborative controller is used to generate control inputs for each follower UGV so that the output of each follower UGV tracks the output of the leader UAV within a preset time.

[0016] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects: The control method for heterogeneous UAV-UGV systems under DoS attacks provided by this invention effectively overcomes the technical bottleneck of existing methods that cannot stably control heterogeneous UAV-UGV systems under intermittent denial-of-service attack environments due to communication congestion and lack of measurement data. This method enables secure collaborative tracking of heterogeneous UAV-UGV systems within a preset time period. This invention accurately characterizes the attacker's "active-dormant" alternation by constructing an intermittent DoS attack model represented by attack time intervals and attack state switching signals. Based on this, a distributed fuzzy adaptive estimator and a switching-type fuzzy state observer for followers are designed using a fuzzy logic system to approximate unknown nonlinear dynamics. Because the estimator and observer can dynamically adjust the estimation structure and observation gain according to the attack state switching signals, the heterogeneous UAV-UGV system can maintain the effectiveness of state estimation during the active attack period by relying on built-in dynamics, and can quickly correct accumulated errors using intermittent measurements during the dormant attack period. This solves the problem of estimation error divergence and slow recovery caused by the unavailability of measurement data in traditional observers.

[0017] To address the common communication bandwidth and onboard energy constraints in heterogeneous UAV-UGV systems, this method abandons the traditional control approach relying on high-frequency periodic state feedback. Instead, it introduces a nonlinear filter with preset time convergence parameters to implement dynamic surface technology, and establishes a unified coordinate transformation relationship encompassing cooperative tracking error, filtering error, and state estimation error. Based on this, and combined with preset time convergence conditions, a preset-time virtual control law and a final preset-time cooperative controller are constructed sequentially. This allows each follower UGV to quickly track the leader UAV's state within a preset, finite time without establishing complex high-frequency communication links. This saves communication resources and solves the problems of system response lag and dynamic performance degradation caused by reducing the update frequency of traditional control.

[0018] Furthermore, addressing the challenge of accurately modeling unknown nonlinear dynamics in heterogeneous UAV-UGV systems, this invention, after establishing coordinate transformation relationships, further designs a fuzzy parameter adaptive update law. By updating the fuzzy weights in the distributed fuzzy adaptive estimator and the switching fuzzy state observer online, the control structure can adapt in real time to changes in system parameters and uncertainties caused by external disturbances, avoiding the impact of model mismatch on control accuracy. Finally, the control input generated by the aforementioned preset-time cooperative controller ensures the consistent eventual boundedness of all signals in the heterogeneous UAV-UGV system under complex environments with intermittent DoS attacks, and ensures that the cooperative tracking error converges to a small neighborhood near the origin within a preset time. This invention effectively addresses the core challenge posed by the triple coupling of "state estimation-communication scheduling-cooperative control," achieving reliable cooperative control of heterogeneous UAV-UGV systems under complex DoS attack interference. Attached Figure Description

[0019] The above and other objects, features and advantages of the present invention will become more apparent from the more detailed description of exemplary embodiments of the invention in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same parts.

[0020] Figure 1 This is a schematic flowchart illustrating the collaborative tracking control method for heterogeneous UAV-UGV systems under a DoS attack, as shown in an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of a quadcopter drone as shown in an embodiment of the present invention; Figure 3 This is a schematic diagram of the unmanned vehicle (UGV) structure shown in an embodiment of the present invention; Figure 4 This is a communication topology diagram of a heterogeneous UAV-UGV system shown in an embodiment of the present invention; Figure 5 This is a three-dimensional spatial position cooperative trajectory diagram of the system as shown in an embodiment of the present invention; Figure 6 This is a UGV control input curve with a DoS attack shown in an embodiment of the present invention; Figure 7 This is a cooperative estimation error curve under a DoS attack, as shown in an embodiment of the present invention. Figure 8 The observer estimation error curve under a DoS attack is shown in the embodiment of the present invention. Figure 9 This is a position tracking error curve under a DoS attack, as shown in an embodiment of the present invention. Figure 10 The preset time filter error curve is shown in the embodiment of the present invention; Figure 11 This is a schematic diagram of the module structure of the collaborative tracking control system of a heterogeneous UAV-UGV system under a DoS attack, as shown in an embodiment of the present invention. Detailed Implementation

[0021] 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.

[0022] Please see Figure 1 , Figure 1This is a flowchart illustrating the collaborative tracking control method for heterogeneous UAV-UGV systems under DoS attacks.

[0023] This invention combines adaptive backstepping recursion, dynamic surface control, and preset time control theory to propose a fuzzy adaptive preset time-based secure cooperative control strategy for heterogeneous UAV-UGV systems. This strategy achieves consistent eventual boundedness for all system signals under intermittent DoS attack scenarios and effectively compensates for communication blockage and unpredictable state issues caused by attacks.

[0024] First, a heterogeneous multi-agent system consisting of a drone leader and multiple unmanned vehicles as followers is selected as the control object, and a dynamic model incorporating unknown nonlinear dynamics and external disturbances is established. A distributed fuzzy adaptive estimator is introduced to reconstruct the leader's state information, which is unavailable under attack. A fuzzy logic system is used to approximate the unknown nonlinear functions of the leader and followers, thereby constructing a piecewise Lyapunov function. A pre-set time adaptive filtering mechanism is then constructed using dynamic surface techniques. Within a backstepping recursive framework, the system error is decomposed into cooperative tracking error, filtering error, and state estimation error through coordinate transformation. A pre-set time convergence constraint ensures that the system completes tracking within a specified time. The final designed fuzzy adaptive pre-set time controller can dynamically switch observation and control gains according to the attack state, enabling the closed-loop heterogeneous system to maintain stable cooperative operation under intermittent DoS attacks. Furthermore, rigorous theoretical analysis verifies the system's pre-set time stability condition.

[0025] Specifically, the cooperative tracking control method for heterogeneous UAV-UGV systems under DoS attacks includes: S101: Construct dynamic models for both the leader UAV and the follower UGV, and establish an intermittent DoS attack model characterized by attack time intervals and attack state switching signals.

[0026] The control method of this invention targets a heterogeneous multi-agent system consisting of one leader unmanned aerial vehicle (UAV) and N follower unmanned vehicles (UGVs), where each agent (UAV / UGV) exhibits unknown nonlinear dynamics, unmeasurable states, and intermittent communication blocking characteristics. First, a nonlinear rigorous feedback dynamic model is established for each agent. The dynamic model of each follower UAV is described as follows: In the formula, Number the followers. , The total number of followers; , This refers to the system status; , The state order; For the first The final actual physical control input of the ground-following vehicle; For system output; To be consistent with system status Related unknown smooth nonlinear functions; Represents state variables The first derivative with respect to time, Represents state variables The first derivative with respect to time.

[0027] The nonlinear rigorous feedback dynamics model of the leader drone is described as follows: In the formula, , The leader's state order; For the leader One state variable; For the leader One state equation; For leaders to control input; To be with the leader State equations Related unknown smooth nonlinear functions; Output for leaders; Indicates the leader's State variables The first derivative with respect to time; Indicates the leader's State variables The first derivative with respect to time; Indicates the relationship with the leader State equations Related unknown smooth nonlinear functions; This is represented as the first state variable of the leader.

[0028] The establishment of the above dynamic model provides a model foundation for subsequent fuzzy approximation, state observation and control law design, enabling the control method to effectively cope with the unknown nonlinear characteristics in the system.

[0029] Simultaneously, a mathematical model for intermittent denial-of-service (DoS) attacks was established. This DoS attack model consists of attack time intervals. It means that the first Communication is completely interrupted during the attack; the attack exhibits an alternating "active-dormant" characteristic: during the active attack period, sensor measurement data cannot be transmitted normally to the controller or neighboring agents, and critical state information is periodically lost; during the dormant attack period, the communication link returns to normal, and measurement data can be transmitted normally. To quantify the temporal characteristics of the attack, let an arbitrary interval be defined. Within, the total duration of the attack activity was The total hibernation time is And satisfy: in, This represents the attack period. The inequality provides an upper bound on the attack's active duration within any time interval, reflecting the attacker's limited energy and providing a quantitative basis for the switching logic in subsequent controller design. The actual sensor measurement output affected by the attack can be expressed as: in, This is a signal to switch attack states. When =1 indicates active attack and communication interruption. A value of 0 indicates that the attack is in hibernation and communication is normal. This switching signal will serve as a trigger mechanism for subsequent switching between the estimator and observer structures, enabling each part of the control system to dynamically adjust its operating mode according to changes in the communication status.

[0030] S102: A distributed fuzzy adaptive estimator is designed by approximating the unknown nonlinear function in the dynamic model of the leader UAV using a fuzzy logic system. The distributed fuzzy adaptive estimator is used to reconstruct the state information of the leader UAV and adjust the estimation structure according to the attack state switching signal.

[0031] To accurately obtain the state information of the leader drone under communication congestion conditions, a distributed fuzzy adaptive estimator needs to be designed. The dynamic model of the leader drone contains unknown nonlinear functions, which can be approximated using a fuzzy logic system. Based on the approximation characteristics of the fuzzy logic system, the unknown smooth nonlinear function in the dynamic model of the leader drone can be expressed as follows: in, Indicates the use of approximating the leader's first An ideal weight vector for an unknown nonlinear dynamic; Indicates the leader's The fuzzy approximation error of unknown nonlinear dynamics.

[0032] Based on the aforementioned fuzzy approximation, a distributed fuzzy adaptive estimator is constructed to reconstruct the state information of the leader drone under communication congestion conditions. This estimator utilizes information interaction between local neighboring agents, combined with its own estimate of the leader's state, thus enabling accurate reconstruction of the leader's state even when attacks prevent direct access to the leader's state information. The specific structure of the estimator is as follows: In the formula, , The leader's state order; For the first Taiwan's intelligent agents and leaders Order-order state estimate; The estimated value of the fuzzy ideal weight vector; , This is the first design parameter; This is the cooperative error term; The coupling coefficient is related to the attack state; Indicates the use of approximating the leader's first A fuzzy basis function vector for nonlinear dynamics with unknown order state; Indicates the use of approximating the leader's first A fuzzy basis function vector for nonlinear dynamics with unknown order state; For leaders to control input; Indicates the first Taiwan's intelligent agents and leaders The estimated value of the first-order state; Indicates the first Taiwan's intelligent agents and leaders The estimated value of the first-order state; Indicates the first The ground-based following agent is the leader's first Order state estimate The time derivative; Indicates the first Taiwan's intelligent agents and leaders Order state estimate The time derivative; This represents the estimated output value for the leader.

[0033] When the attack is in a dormant phase and the communication link is normal, the estimator uses neighbor information to correct the estimated value in real time to ensure estimation accuracy. When the attack is in an active phase and the communication link is interrupted, the estimator switches to open-loop estimation mode, relying on its own dynamic equation and the learned fuzzy weights to continue outputting the estimated value of the leader state, avoiding estimation divergence caused by the loss of measurement data. Thus, the distributed fuzzy adaptive estimator realizes the function of dynamically adjusting the estimation structure according to the attack state switching signal, thereby maintaining reliable reconstruction of the leader state throughout the entire process of intermittent DoS attacks.

[0034] S103: Use a fuzzy logic system to approximate the unknown nonlinear functions in the dynamic model of each follower UGV in order to design their respective switching fuzzy state observers; the switching fuzzy state observers are used to estimate the unmeasurable state of the corresponding follower UGV and adjust the observation gain according to the attack state switching signal.

[0035] Under conditions where communication congestion causes intermittent interruptions in the output signal, the unmeasurable state of the follower autonomous vehicle cannot be obtained through direct measurement. Therefore, a switching fuzzy state observer needs to be designed for each follower autonomous vehicle. The dynamic model of each follower autonomous vehicle contains an unknown nonlinear function, which is approximated using a fuzzy logic system. Based on the approximation characteristics of the fuzzy logic system, the unknown smooth nonlinear function in the dynamic model of each follower can be expressed as follows: in, For the first Taiwan Intelligent Agent An ideal weight vector for a fuzzy logic system This is used to approximate the error.

[0036] Based on the aforementioned fuzzy approximation, a switching fuzzy state observer is constructed for each follower autonomous vehicle to estimate its own unmeasurable state under the condition of output signal interruption caused by an attack. The specific structure of the observer is as follows: In the formula, , The state order; For the first Taiwan Followers Internal state estimate of the first order; For the first Taiwan Followers The estimated value of the fuzzy weight of the first-order state; For the first Taiwan Followers The estimated value of the fuzzy weight of the first-order state; For the first Taiwan Followers Observation gain for switching between order states; For the first Taiwan Followers Observation gain for switching between order states; For the first The first-order state output estimation error of the follower; This is a signal to switch attack states. Indicates the use of approximating the first The first follower The fuzzy basis function vector of the unknown nonlinear function of the internal state of the order; Indicates the first Taiwan Followers The fuzzy input vector corresponding to the first-order state estimate; Indicates the first Taiwan Followers The fuzzy input vector corresponding to the first-order state estimate; Indicates the first Taiwan Followers Internal state estimate of the first order The first derivative with respect to time; Indicates the first Taiwan Followers The first derivative of the internal state estimate with respect to time; For the first The final actual physical control input of the ground-following vehicle; Indicates the first Output estimates of the followers; Indicates the first Taiwan follower switching signals The first-order state output value.

[0037] The observation gain of the aforementioned observer can dynamically switch according to the attack state switching signal. When the attack is in the dormant phase and the communication link is normal, the actual measurement output is available, and the observer uses the output estimation error to correct the state estimation in real time, ensuring that the observation error converges quickly to high accuracy. When the attack is in the active phase and the communication link is interrupted, the actual measurement output is set to zero. At this time, the observer disconnects the measurement correction channel and relies on its own dynamic equation and the learned fuzzy weights to perform open-loop state estimation, avoiding estimation divergence caused by the injection of erroneous measurement values. Thus, the switching fuzzy state observer realizes the function of dynamically adjusting the observation gain according to the attack state switching signal, thereby maintaining a reliable estimation of the unmeasurable state of the follower throughout the entire process of an intermittent DoS attack.

[0038] After completing the design of the distributed fuzzy adaptive estimator and the switching fuzzy state observer, the dynamic equations for the estimation error and the observation error can be uniformly given. Let the estimation error be... , The observation error is ;in, For the first The first intelligent agent The estimation error of each parameter, For the first The first intelligent agent The ideal values ​​for each parameter For the first The first intelligent agent An ideal weight vector for a fuzzy logic system For the first The first intelligent agent The weight estimation error of a fuzzy logic system; then the dynamics of the estimation and observation errors can be expressed as: In the formula, To estimate the error dynamics, and further, to derive the stability of the estimation error system, the Lyapunov function is constructed as follows: in, To estimate the error vector; This is the communication topology matrix; It is a positive definite matrix related to the attack state; For Kronecker product.

[0039] To derive the stability of the observation error system, the Lyapunov function is constructed as follows: in, This is for attack-related positive definite matrices.

[0040] Constructing the global Lyapunov function: In the formula, For the first Lyapunov function of an agent's observation error system For the first The Lyapunov functions corresponding to each agent in the no-attack region. And the attack-active regions respectively. With dormancy zone Segmented Lyapunov stability analysis was performed.

[0041] S104: Based on the output of the distributed fuzzy adaptive estimator and the switching fuzzy state observer, a preset time nonlinear filter is introduced to design the dynamic surface and establish the coordinate transformation relationship including the cooperative tracking error, the filtering error and the state estimation error.

[0042] To transform the state information output by the estimator and observer into error feedback variables that can be used for controller design, a unified coordinate transformation relationship needs to be established. Specifically, by defining error variables such as cooperative tracking error, filtering error, and state estimation error, the tracking control problem of each order of dynamics is transformed into a feedback control problem with errors as state variables. This allows for the subsequent construction of virtual control laws and the final control law layer by layer based on these error variables during the adaptive backstepping recursion process. The coordinate transformation relationship is designed by introducing a first-order nonlinear filter as follows: in, For tracking error; This is the filtering error; The filter output for the leader. The filter output for the virtual control law; For virtual control laws; Output an estimate for the leader.

[0043] Within the dynamic surface design framework, to address the differential explosion problem arising from repeated differentiation of the virtual control law in the backstep recursion, a pre-defined time-nonlinear filter needs to be constructed. The pre-defined time-nonlinear filter is designed as follows: In the formula, , The state order; Preset convergence time; , The first preset time index, This is the second preset time index; This is the second design parameter. This is the third design parameter; For the first The first follower Design parameters corresponding to the first state; For the first Design parameters corresponding to the first-order state of each follower; Indicates the first Each follower corresponds to the output of a preset time nonlinear filter. The first derivative with respect to time; Indicates the first Auxiliary saturation constraint variables for each follower's first-order filter channel; Indicates the first The first follower Auxiliary saturation constraint variables for the first-order filter channel; Indicates the first The first follower Virtual control signal The first derivative of the output value with respect to time after being filtered by a preset time; Indicates the first The first follower Order state variables; This indicates that the preset time index is Time The first follower Order state variables; This indicates that the preset time index is Time The first follower Order of state variables.

[0044] S105: Design an adaptive update law for fuzzy parameters based on coordinate transformation relationships to update the fuzzy weights in the distributed fuzzy adaptive estimator and the switching fuzzy state observer.

[0045] To enable online adjustment of unknown fuzzy weights in the distributed fuzzy adaptive estimator and the switching fuzzy state observer, an adaptive update law for fuzzy parameters needs to be designed. Since the unknown nonlinear functions in the leader and follower dynamics models are approximated using fuzzy logic systems, and the fuzzy ideal weight vector is an unknown constant, an adaptive law is needed to update the weight estimates online, gradually bringing them closer to the ideal weights, thereby ensuring the accuracy of the fuzzy approximation.

[0046] Based on the coordinate transformation relationship established above, an adaptive update law for fuzzy parameters is designed based on the error variables after coordinate transformation. The design of the adaptive update law for fuzzy parameters is as follows: In the formula, , The state order; The coupling coefficient is related to the attack state; Indicates the first The collaborative estimation error vector of the leader's state by each follower; Indicates the use of approximating the leader. A fuzzy basis function vector for nonlinear dynamics with unknown order state; , This is the first adaptive gain parameter. This is the second adaptive gain parameter; The input matrix; This represents the positive definite symmetric Lyapunov matrix obtained by solving linear matrix inequalities; Indicates the first Fuzzy ideal weight vector estimate of a leader The first derivative with respect to time; Indicates the first The first follower fuzzy weight estimate of the first order state The first derivative with respect to time.

[0047] By adjusting the fuzzy weight estimates online, the fuzzy logic system can continuously and accurately approximate the unknown nonlinear dynamics of the system. The updated fuzzy weights are fed back to the distributed fuzzy adaptive estimator and the switching fuzzy state observer, forming a closed-loop adaptive mechanism. This ensures that the estimator and observer can maintain accurate estimates of the leader's state and the unpredictable state of the followers even when communication is interrupted by intermittent DoS attacks.

[0048] S106: Based on the coordinate transformation relationship and the preset time convergence condition, construct the preset time virtual control law.

[0049] After completing the coordinate transformation and adaptive update law design, it is necessary to construct the preset time virtual control law layer by layer based on the preset time convergence condition. The virtual control law is constructed using preset time control as follows: In the formula, Preset convergence time; This is the second design parameter. This is the third design parameter; Indicates the first Auxiliary saturation constraint variables for each follower's first-order filter channel; Indicates the first The first-order preset time virtual control signal for the platform follower; Indicates the first Taiwan Follower First Order Consistency Tracking Error of Power of; Indicates the first Taiwan Follower First-Order Consistency Tracking Error of Power of; Indicates the use of approximating the first The fuzzy basis function vector of the unknown nonlinear function of the first-order internal state of a follower; Indicates the first The estimated fuzzy weights of the first-order state of each follower; For the first The first-order state output estimation error of the follower; For the first The observation gain of the first-order state switching of the platform follower; This indicates the output corresponding to the preset time nonlinear filter. The first derivative with respect to time.

[0050] S107: Construct the final preset time cooperative controller based on the preset time virtual control law and the updated fuzzy weights; the final preset time cooperative controller is used to generate the control input of each follower UGV so that the output of each follower UGV tracks the output of the leader UAV within a preset time.

[0051] After constructing the virtual control law for the preset time, it is necessary to further design the final preset time coordination controller that actually operates on the system. This controller is the final output stage of the entire control design scheme, and its output directly serves as the control input for each follower autonomous vehicle. The specific expression of the final preset time coordination controller is as follows: In the formula, Preset convergence time; This is the second design parameter. This is the third design parameter; Indicates the first Auxiliary saturation constraint variables for each follower's first-order filter channel; , The state order; Indicates the first The final actual physical control input of the ground-following vehicle; Indicates the first Taiwan Followers Consistency tracking error of Power; Indicates the first Taiwan Followers Consistency tracking error of Power; Indicates the use of approximating the first The first follower The fuzzy basis function vector of the unknown nonlinear function of the internal state of the order; Indicates the first The first follower The estimated value of the fuzzy weight of the first-order state; For the first The first-order state output estimation error of the follower; For the first Taiwan Followers Observation gain for switching between order states; Indicates the first Taiwan Followers First-order filter virtual control quantity The first derivative with respect to time.

[0052] Under the control of the final preset time-coordinated controller, the control inputs of each follower drone are generated in real time and applied to the controlled system, driving the state of each follower drone to gradually track the state of the leader drone. The entire control link, from the state estimation and observation at the front end, to the coordinate transformation and virtual control law construction in the middle, and finally to the output of the controller, forms a complete closed-loop control circuit.

[0053] Through Lyapunov's pre-set time stability analysis, it is proved that under the combined action of the distributed fuzzy adaptive estimator, the switching fuzzy observer, and the pre-set time virtual control law provided above, all signals in the closed-loop system are uniformly and eventually bounded, and the cooperative tracking error and the state estimation error converge to a small neighborhood near the origin within a pre-set time. The convergence time satisfies: in The total attack duration is shown. The system remained stable under intermittent DoS attacks without any abnormal triggering behavior. Simulations verified the effectiveness and robustness of the method in the heterogeneous UAV-UGV collaborative system.

[0054] This invention relates to a heterogeneous UAV-UGV air-to-ground cooperative control architecture. This system achieves safe and stable operation under DoS attacks through distributed state estimation, switching state observation, and preset-time cooperative control. The heterogeneous system mainly comprises four parts: the controlled objects, namely the dynamic models of the UAV leader and UGV followers; sensors and communication modules, affected by intermittent DoS attacks; a distributed fuzzy adaptive estimator and a switching fuzzy state observer; and an adaptive backstepping cooperative controller based on a preset time.

[0055] Taking a heterogeneous multi-agent system consisting of a drone leader and three unmanned vehicles as the control object as an example, the simulation results verify the effectiveness and robustness of the control scheme provided by this invention. Figure 2 This is a schematic diagram of a quadcopter drone. Figure 3 This is a schematic diagram of the unmanned vehicle structure, clearly showing the motion and sensing components of UAV and UGV. Figure 4 The communication topology of the heterogeneous UAV-UGV system is shown, which adopts a directed connectivity topology, in which some followers can directly receive information from the leader. Figure 5 The system demonstrated coordinated trajectories in three-dimensional space, with all UGV followers accurately tracking the UAV leader's trajectory, maintaining good coordinated formation even under intermittent DoS attacks. Figure 6The UGV control input curve for the DoS attack is smooth and bounded, without severe chattering or divergence. Figure 7 The collaborative estimation error curve under DoS attack is presented. The leader state estimation error converges quickly and remains within a small bounded interval. Figure 8 To estimate the error curve for the observer in an attack environment, the switching fuzzy observer can maintain estimation accuracy during communication interruptions. Figure 9 The system's position tracking error curve is shown, and each follower converges quickly and stably tracks the leader's output within a preset time. Figure 10 The preset time filter has a small filtering error curve, which converges quickly and effectively suppresses the backstep differential explosion problem.

[0056] The simulation results show that the designed control scheme can ensure that the cooperative tracking error converges to the origin neighborhood within a preset time under intermittent DoS attacks, and that both the state estimation error and adaptive parameters remain bounded. The system also exhibits good anti-interference capability and cooperative control accuracy. During the active attack period, the system relies on a switching observer to maintain stability, while during the attack dormancy period, a preset time mechanism is used to improve the convergence speed. This achieves integrated optimization of anti-attack security control, preset time convergence, and heterogeneous system cooperation.

[0057] Corresponding to the aforementioned application function implementation method embodiments, the present invention also provides a control system for a heterogeneous UAV-UGV system under DoS attack and corresponding embodiments.

[0058] Please see Figure 11 , Figure 11 This is a schematic diagram of the module structure of the control system of a heterogeneous UAV-UGV system under a DoS attack.

[0059] A heterogeneous UAV-UGV system consists of a leader UAV and multiple follower UGVs. The control system of a heterogeneous UAV-UGV system under a DoS attack includes: Model building unit 11 is used to build dynamic models of the leader UAV and follower UGV respectively, and to establish an intermittent DoS attack model characterized by attack time intervals and attack state switching signals. The state estimation unit 12 is used to approximate the unknown nonlinear function in the dynamic model of the leader UAV using a fuzzy logic system in order to design a distributed fuzzy adaptive estimator; the distributed fuzzy adaptive estimator is used to reconstruct the state information of the leader UAV and adjust the estimation structure according to the attack state switching signal. The state observation unit 13 is used to approximate the unknown nonlinear function in the dynamic model of each follower UGV using a fuzzy logic system in order to design their respective switching fuzzy state observers; the switching fuzzy state observers are used to estimate the unmeasurable state of the corresponding follower UGV and adjust the observation gain according to the attack state switching signal. The coordinate transformation unit 14 is used to design a dynamic surface by introducing a preset time nonlinear filter based on the output of the distributed fuzzy adaptive estimator and the switching fuzzy state observer, and to establish a coordinate transformation relationship that includes cooperative tracking error, filtering error and state estimation error. The adaptive update unit 15 is used to design an adaptive update law for fuzzy parameters based on the coordinate transformation relationship, so as to update the fuzzy weights in the distributed fuzzy adaptive estimator and the switching fuzzy state observer. The preset time coordination control unit 16 is used for: Based on coordinate transformation relationships and preset time convergence conditions, a preset time virtual control law is constructed. Based on the preset time virtual control law and the updated fuzzy weights, a final preset time cooperative controller is constructed. The final preset time cooperative controller is used to generate the control input of each follower UGV so that the output of each follower UGV tracks the output of the leader UAV within a preset time.

[0060] Regarding the system in the above embodiments, the specific manner in which each unit module performs operations has been described in detail in the embodiments related to the method, and will not be elaborated further here.

[0061] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A cooperative tracking control method for heterogeneous UAV-UGV systems under DoS attack, characterized in that, The heterogeneous UAV-UGV system includes a leader UAV and multiple follower UGVs, and the method includes: Construct dynamic models for the leader UAV and the follower UGV respectively, and establish an intermittent DoS attack model characterized by attack time intervals and attack state switching signals; A distributed fuzzy adaptive estimator is designed by approximating the unknown nonlinear function in the dynamic model of the leader UAV using a fuzzy logic system. The distributed fuzzy adaptive estimator is used to reconstruct the state information of the leader UAV and adjust the estimation structure according to the attack state switching signal. The unknown nonlinear functions in the dynamic models of each follower UGV are approximated using a fuzzy logic system to design their respective switching fuzzy state observers; the switching fuzzy state observers are used to estimate the unmeasurable state of the corresponding follower UGV and adjust the observation gain according to the attack state switching signal. Based on the outputs of the distributed fuzzy adaptive estimator and the switching fuzzy state observer, a preset time nonlinear filter is introduced to perform dynamic surface design, and a coordinate transformation relationship including cooperative tracking error, filtering error and state estimation error is established. Based on the coordinate transformation relationship, an adaptive update law for fuzzy parameters is designed to update the fuzzy weights in the distributed fuzzy adaptive estimator and the switching fuzzy state observer. Based on the coordinate transformation relationship and the preset time convergence condition, a preset time virtual control law is constructed; Based on the preset time virtual control law and the updated fuzzy weights, a final preset time collaborative controller is constructed; the final preset time collaborative controller is used to generate control inputs for each follower UGV so that the output of each follower UGV tracks the output of the leader UAV within a preset time.

2. The cooperative tracking control method for heterogeneous UAV-UGV systems under DoS attack according to claim 1, characterized in that, The expression for the dynamic model of the follower UGV is: In the formula, Number the followers. , The total number of followers; , This refers to the system status; , The state order; For the first The final actual physical control input of the ground-following vehicle; For system output; To be consistent with system status Related unknown smooth nonlinear functions; Represents state variables The first derivative with respect to time, Represents state variables The first derivative with respect to time.

3. The cooperative tracking control method for heterogeneous UAV-UGV systems under DoS attack according to claim 1, characterized in that, The expression for the dynamic model of the leader UAV is: In the formula, , The leader's state order; For the leader One state variable; For the leader One state equation; For leaders to control input; To be with the leader State equations Related unknown smooth nonlinear functions; Output for leaders; Indicates the leader's State variables The first derivative with respect to time; Indicates the leader's State variables The first derivative with respect to time; Indicates the relationship with the leader State equations Related unknown smooth nonlinear functions; This is represented as the first state variable of the leader.

4. The cooperative tracking control method for heterogeneous UAV-UGV systems under DoS attack according to claim 1, characterized in that, The expression for the distributed fuzzy adaptive estimator is: In the formula, , The leader's state order; For the first Taiwan's intelligent agents and leaders Order-order state estimate; The estimated value of the fuzzy ideal weight vector; , This is the first design parameter; This is the cooperative error term; The coupling coefficient is related to the attack state; Indicates the use of approximating the leader's first A fuzzy basis function vector for nonlinear dynamics with unknown order state; Indicates the use of approximating the leader's first A fuzzy basis function vector for nonlinear dynamics with unknown order state; For leaders to control input; Indicates the first Taiwan's intelligent agents and leaders The estimated value of the first-order state; Indicates the first Taiwan's intelligent agents and leaders The estimated value of the first-order state; Indicates the first The ground-based following agent is the leader's first Order state estimate The time derivative; Indicates the first Taiwan's intelligent agents and leaders Order state estimate The time derivative; This represents the estimated output value for the leader.

5. The cooperative tracking control method for heterogeneous UAV-UGV systems under DoS attack according to claim 1, characterized in that, The expression for the switching fuzzy state observer is: In the formula, , The state order; For the first Taiwan Followers Internal state estimate of the first order; For the first Taiwan Followers The estimated value of the fuzzy weight of the first-order state; For the first Taiwan Followers The estimated value of the fuzzy weight of the first-order state; For the first Taiwan Followers Observation gain for switching between order states; For the first Taiwan Followers Observation gain for switching between order states; For the first The first-order state output estimation error of the follower; This is a signal to switch attack states. Indicates the use of approximating the first The first follower The fuzzy basis function vector of the unknown nonlinear function of the internal state of the order; Indicates the first Taiwan Followers The fuzzy input vector corresponding to the first-order state estimate; Indicates the first Taiwan Followers The fuzzy input vector corresponding to the first-order state estimate; Indicates the first Taiwan Followers Internal state estimate of the first order The first derivative with respect to time; Indicates the first Taiwan Followers The first derivative of the internal state estimate with respect to time; For the first The final actual physical control input of the ground-following vehicle; Indicates the first Output estimates of the followers; Indicates the first Taiwan follower switching signals The first-order state output value.

6. The cooperative tracking control method for heterogeneous UAV-UGV systems under DoS attack according to claim 1, characterized in that, The expression for the preset time nonlinear filter is: In the formula, , The state order; Preset convergence time; , The first preset time index, This is the second preset time index; This is the second design parameter. This is the third design parameter; For the first The first follower Design parameters corresponding to the first state; For the first Design parameters corresponding to the first-order state of each follower; Indicates the first Each follower corresponds to the output of a preset time nonlinear filter. The first derivative with respect to time; Indicates the first Auxiliary saturation constraint variables for each follower's first-order filter channel; Indicates the first The first follower Auxiliary saturation constraint variables for the first-order filter channel; Indicates the first The first follower Virtual control signal The first derivative of the output value with respect to time after being filtered by a preset time; Indicates the first The first follower Order state variables; This indicates that the preset time index is Time The first follower Order state variables; This indicates that the preset time index is Time The first follower Order of state variables.

7. The cooperative tracking control method for heterogeneous UAV-UGV systems under DoS attack according to claim 1, characterized in that, The expression for the adaptive update law of the fuzzy parameters is: In the formula, , The state order; The coupling coefficient is related to the attack state; Indicates the first The collaborative estimation error vector of the leader's state by each follower; Indicates the use of approximating the leader's first A fuzzy basis function vector for nonlinear dynamics with unknown order state; , This is the first adaptive gain parameter. This is the second adaptive gain parameter; The input matrix; This represents the positive definite symmetric Lyapunov matrix obtained by solving linear matrix inequalities; Indicates the first Fuzzy ideal weight vector estimate of a leader The first derivative with respect to time; Indicates the first The first follower fuzzy weight estimate of the first order state The first derivative with respect to time.

8. The cooperative tracking control method for heterogeneous UAV-UGV systems under DoS attack according to claim 1, characterized in that, The expression for the preset time virtual control law is: In the formula, Preset convergence time; This is the second design parameter. This is the third design parameter; Indicates the first Auxiliary saturation constraint variables for each follower's first-order filter channel; Indicates the first The first-order preset time virtual control signal for the platform follower; Indicates the first Taiwan Follower First-Order Consistency Tracking Error of Power of; Indicates the first Taiwan Follower First-Order Consistency Tracking Error of Power of; Indicates the use of approximating the first The fuzzy basis function vector of the unknown nonlinear function of the first-order internal state of a follower; Indicates the first The estimated fuzzy weights of the first-order state of each follower; For the first The first-order state output estimation error of the follower; For the first The observation gain of the first-order state switching of the platform follower; This indicates the output corresponding to the preset time nonlinear filter. The first derivative with respect to time.

9. The control method for a heterogeneous UAV-UGV system under a DoS attack according to claim 1, characterized in that, The expression for the final preset time coordination controller is: In the formula, Preset convergence time; This is the second design parameter. This is the third design parameter; Indicates the first Auxiliary saturation constraint variables for each follower's first-order filter channel; , The state order; Indicates the first The final actual physical control input of the ground-following vehicle; Indicates the first Taiwan Followers Consistency tracking error of Power of; Indicates the first Taiwan Followers Consistency tracking error of Power of; Indicates the use of approximating the first The first follower The fuzzy basis function vector of the unknown nonlinear function of the internal state of the order; Indicates the first The first follower The estimated value of the fuzzy weight of the first-order state; For the first The first-order state output estimation error of the follower; For the first Taiwan Followers Observation gain for switching between order states; Indicates the first Taiwan Followers First-order filter virtual control quantity The first derivative with respect to time.

10. A cooperative tracking control system for heterogeneous UAV-UGV systems under DoS attack, characterized in that, The heterogeneous UAV-UGV system includes a leader UAV and multiple follower UGVs, and the control system includes: The model building unit is used to build dynamic models of the leader UAV and the follower UGV respectively, and to establish an intermittent DoS attack model characterized by attack time intervals and attack state switching signals. A state estimation unit is used to approximate the unknown nonlinear function in the dynamic model of the leader UAV using a fuzzy logic system in order to design a distributed fuzzy adaptive estimator; the distributed fuzzy adaptive estimator is used to reconstruct the state information of the leader UAV and adjust the estimation structure according to the attack state switching signal. A state observation unit is used to approximate the unknown nonlinear function in the dynamic model of each follower UGV using a fuzzy logic system, so as to design their respective switching fuzzy state observers; the switching fuzzy state observers are used to estimate the unmeasurable state of the corresponding follower UGV and adjust the observation gain according to the attack state switching signal. The coordinate transformation unit is used to design a dynamic surface by introducing a preset time nonlinear filter based on the output of the distributed fuzzy adaptive estimator and the switching fuzzy state observer, and to establish a coordinate transformation relationship that includes cooperative tracking error, filtering error and state estimation error. An adaptive update unit is used to design an adaptive update law for fuzzy parameters based on the coordinate transformation relationship, so as to update the fuzzy weights in the distributed fuzzy adaptive estimator and the switching fuzzy state observer. Preset time coordination control unit, used for: Based on the coordinate transformation relationship and the preset time convergence condition, a preset time virtual control law is constructed; Based on the preset time virtual control law and the updated fuzzy weights, a final preset time collaborative controller is constructed; the final preset time collaborative controller is used to generate control inputs for each follower UGV so that the output of each follower UGV tracks the output of the leader UAV within a preset time.

Images