An air-ground heterogeneous unmanned system anti-interference composite trigger formation contains a control method, device and storage medium

Abstract

This invention discloses an anti-interference composite triggering formation control method, device, and storage medium for an air-ground heterogeneous unmanned system. The method includes: constructing an external disturbance compensator containing auxiliary state variables and outputting a compensation signal; estimating the virtual leader's state through a distributed observer; constructing a composite triggering mechanism based on adaptive coupling weights and internal dynamic variables for state synchronization, and dynamically updating both based on neighbor errors; calculating the actual input based on the observed state and compensation signal to control the unmanned vehicle swarm formation; designing a dual distributed compensator for followers to obtain estimates of the leader's state and formation information; designing an independent triggering mechanism to synchronize the unmanned vehicle's state and dynamically updating weights and variables based on errors; calculating the control input based on the above estimates and compensation signals, and controlling the unmanned vehicle swarm to converge within the dynamic convex hull determined by the unmanned vehicle swarm. This invention achieves high-precision containment control and improves the accuracy of cooperative task execution.

Description

Technical Field

[0001] This invention belongs to the field of multi-agent cooperative control technology, and more specifically, relates to an anti-interference composite trigger formation of an air-ground heterogeneous unmanned system, including a control method, device and storage medium. Background Technology

[0002] With the rapid development of unmanned systems technology, air-to-ground heterogeneous unmanned systems, composed of unmanned aerial vehicles (UAVs) and unmanned ground vehicles (UGVs), have become a research hotspot in the field of control due to their enormous potential in reconnaissance and strike, disaster relief, and intelligent inspection. In this system, multiple UAVs with wide fields of view typically act as leaders, maintaining a specific formation in the air for wide-area searches or communication relays; while multiple UGVs with strong payload capacity and endurance act as followers, performing refined operations on the ground. To ensure the safety and efficiency of air-to-ground collaborative missions, the system needs to achieve formation control objectives: the aerial UAV leaders must maintain a preset time-varying formation, while the ground-based UGV followers must converge and remain within the convex hull region (safe zone) formed by the UAVs' position projections to avoid exposure or leaving the communication range.

[0003] However, in practical air-ground collaborative engineering applications, existing formation control methods face three major challenges: First, there is the contradiction between heterogeneous dynamics and communication resources: UAVs and unmanned vehicles are typical linear heterogeneous multi-agent systems. The former follows higher-order flight dynamics, while the latter is constrained by ground motion. Their state space dimensions and parameter matrices are completely different. Although existing technologies have theoretically solved the heterogeneous synchronization problem, their controller designs mostly rely on continuous-time state feedback. In air-to-ground collaborative scenarios with limited communication bandwidth in the field, continuous data transmission leads to huge data throughput, which can easily cause channel congestion and packet loss. Moreover, continuous signal radiation significantly increases the risk of the system being detected and located by the enemy, making it difficult to meet the tactical requirements of covert penetration.

[0004] Secondly, there is the challenge of active anti-interference in complex environments: The environments of battlefields or disaster areas are complex and ever-changing. UAVs are often disturbed by gusts of wind, and unmanned vehicles are frequently affected by changes in surface friction caused by rugged terrain. These disturbances are typically unknown, time-varying, and continuous (and can be modeled as signals generated by linear external systems). While existing technologies have proposed dynamic event triggering mechanisms, they primarily focus on ideal environmental control of isomorphic systems, lacking independent observation and feedforward compensation mechanisms for unknown external disturbances. Existing passive robust control strategies are insufficient to effectively suppress large-scale external system interference. Without the introduction of observer-based active compensation, formation distortion is highly likely, and followers may even deviate from the preset safe convex hull area due to accumulated interference.

[0005] Third, the adaptability and flexibility of event triggering mechanisms are insufficient: existing event triggering controls mostly use fixed thresholds or simple decaying dynamic thresholds, which are difficult to adapt to the complex dynamic characteristics of air-to-ground collaborative tasks. For example, simple dynamic thresholds often ignore the historical cumulative effect of measurement errors and the dynamic evolution within the system, leading to false triggering (wasting resources) in noisy environments, or trigger lag (reduced accuracy) when the system performs large maneuvers. How to design a composite adaptive triggering mechanism that combines instantaneous errors, historical data, and internal dynamic variables under non-uniform sampling conditions, maximizes the saving of communication resources while ensuring control accuracy, and theoretically strictly eliminates Zeno behavior, is a key problem that urgently needs to be solved.

[0006] Therefore, in order to address the aforementioned limitations, it is of great theoretical significance and practical engineering value to develop a control method that can simultaneously solve the problems of dynamic heterogeneity, strong external interference, and limited communication resources, especially a control scheme for air-ground heterogeneous unmanned systems based on an auxiliary variable interference observer and a composite adaptive event triggering mechanism. Summary of the Invention

[0007] To address the shortcomings of related technologies, the present invention aims to provide an anti-interference composite trigger formation for air-to-ground heterogeneous unmanned systems, including a control method, device, and storage medium, which aims to solve the problems of dynamic heterogeneity, strong external interference, and limited communication resources in the existing collaborative control methods for UAVs and unmanned vehicles.

[0008] To achieve the above objectives, in a first aspect, the present invention provides a method for anti-interference composite triggering formation control of air-to-ground heterogeneous unmanned systems, comprising: S100. Based on the external environmental interference experienced by the air-ground heterogeneous unmanned system, an external disturbance compensator containing auxiliary state variables is constructed, and a compensation signal is output to the heterogeneous unmanned system; the air-ground heterogeneous unmanned system includes a drone swarm as the leadership layer and an unmanned vehicle swarm as the follower layer, and the unmanned vehicles only receive instructions from the drones in one direction. S200. Estimate the virtual leader state using a distributed observer; construct a composite adaptive triggering mechanism based on adaptive coupling weights and internal dynamic variables. When the sampling deviation of the weighted UAV local neighbor error exceeds the lower bound of the threshold jointly formed by the local neighbor error and the internal dynamic variables, trigger communication between UAV local neighbors to perform state synchronization updates; update the adaptive coupling weights and internal dynamic variables in real time based on the dynamic changes of the UAV local neighbor error; calculate the actual control input of the UAV based on the virtual leader state and the compensation signal to control the UAV swarm formation. S300: Design a dual distributed compensator for each follower, and estimate the leader state and leader formation information through the dual distributed compensator; the dual distributed compensator has an independent triggering mechanism to synchronize the state updates among the local neighbors of the unmanned vehicle; according to the dynamic changes of the local neighbor error of the unmanned vehicle, update the coupling weights and internal dynamic variables in the two independent triggering mechanisms in real time; calculate the actual control input of the unmanned vehicle based on the leader state estimation result, the leader formation information estimation result and the compensation signal, so as to control the unmanned vehicle group to converge to the dynamic convex hull range determined by the unmanned vehicle group.

[0009] Optionally, step S100 specifically includes: S110. To address the persistent external environmental interference of heterogeneous air-ground unmanned systems during air-ground collaborative operations, this interference is modeled as a signal generated by a linear external system; the interference signal is defined. Produced by the linear external system:

[0010] in, The state vector of the external system that is interfering; and For the interference model matrix, the matrix is ​​selected as follows: and To simulate gust interference with specific frequency characteristics; for unmanned vehicles, a matrix is ​​selected. and To simulate constant or slowly varying surface friction disturbances; S120, Introducing intermediate auxiliary variables Based on real-time status measurements of drones and unmanned vehicles and actual control input Construct the observer dynamic equations:

[0011] in, For the first Intermediate auxiliary variables for each agent and For the first The system matrix and input matrix of each agent. The observer gain matrix is... For the first Real-time state measurements of each agent For the first The actual control input of an intelligent agent; When the value is 1 to M, it indicates a drone. When the value of is from M+1 to N, it represents an unmanned vehicle, where M and N are both positive integers; S130, According to the system matrix and interference model matrix Design the observer gain matrix The gain matrix can be solved using the pole placement method or by solving linear matrix inequalities. This ensures that the observation error of the external disturbance compensator dynamically satisfies the Hurwitz stability condition and the fast convergence constraint. S140, Based on the auxiliary variable and the Real-time state measurement of each agent Reconstruct the estimated value of external disturbances It is then output as a compensation signal to the formation controller of the UAV swarm and the inclusion controller of the unmanned vehicle swarm in the heterogeneous unmanned system.

[0012] Optionally, step S200 specifically includes: S210. Set up a distributed observer on each UAV and calculate the local neighbor error based on the communication topology. ;in, For the first The estimated value of the virtual leader's state by a drone. It is a virtual leader state. For communication topology weights; S220. Based on the local neighbor error of the UAV, the update law of the distributed observer is: ;in, The observer gain matrix is... The latest event trigger time, For adaptive coupling weights; S230, Define Measurement Error Introducing internal dynamic variables Construct a composite adaptive triggering mechanism, with the triggering function as follows:

[0013] in, These are preset positive parameters. It is a positive definite weighted matrix; when When the trigger time is reached, communication between the drone's local neighbors is initiated, the state of the drone's local neighbors is synchronized and updated, and the trigger time is updated. ; S240. Based on the dynamic changes of the UAV's local neighbor error, coupling weights and internal dynamic variables The adaptive law is updated online according to the following formulas:

[0014]

[0015] in, , , , All are preset positive design parameters; , The local neighbor error at the latest trigger time; The positive definite weighting matrix is ​​defined as follows: the coupling weights automatically increase as the local neighbor error of the UAV increases, thereby tightening the trigger threshold; conversely, the weights decrease as the error decreases, in order to reduce communication. S250, Observation state obtained based on the distributed observer Estimated values ​​of external disturbances Calculate the first Actual control input of the drone :

[0016] in, This is a preset time-varying formation vector; For feedback control gain; and It is a solution to the regulator equation; According to the actual control input Control the corresponding drones to control the drone swarm formation.

[0017] Optionally, step S300 specifically includes: S310, each follower is designed with a state fusion compensator. and formation information compensator The system obtains the leadership state estimate and leadership formation information estimate accordingly; and calculates the local neighbor error of state fusion based on the communication topology. Local neighbor error of formation information :

[0018] The dynamic update law for constructing a dual distributed compensator:

[0019] in, For the virtual leader system matrix, Generate a matrix for the formation. Here is the feedback gain matrix of the state fusion compensator. The feedback gain matrix for the formation information compensator; S320, respectively, is the state fusion compensator. and formation information compensator Two independent triggering mechanisms are designed; each triggering mechanism includes adaptive coupling weights and internal dynamic variables; the triggering function is:

[0020] Among them, subscript These correspond to the state fusion and formation information channels, respectively; when the trigger function... When the time is right, the corresponding channel's communication event is triggered and the time is updated. ; S330. Based on the dynamic changes in the local neighbor error of the unmanned vehicle, the internal dynamic variable... and adaptive coupling weights The update law is:

[0021]

[0022] in, This represents the measurement error for each channel; S340, Estimated values ​​based on leadership state estimation, leadership formation information estimation, and external disturbances. Calculate the first The actual control input of an unmanned vehicle :

[0023] in, For calming effect; and For feedforward gain, , ; According to the actual control input The unmanned vehicle is controlled to counteract external environmental interference and drive it to converge within the dynamic convex hull range determined by the drone swarm.

[0024] Optionally, after step S300, the method further includes: S400. Determine whether the performance indicators of the air-ground heterogeneous unmanned system meet the preset requirements; the performance indicators include: external interference estimation error, UAV swarm formation tracking error, and the inclusion error of the unmanned vehicle swarm relative to the dynamic convex hull; If all indicators meet the preset requirements, the current control parameters are maintained, and the system enters steady-state operation or waits for the next trigger. If any indicator fails to meet the preset requirements, the design parameters of the corresponding observer or controller will be optimized according to the type of performance indicator that fails to meet the requirements.

[0025] Optionally, the optimization of design parameters includes: When the external disturbance estimation error is not up to standard, optimize the observer gain matrix. Adjust the pole placement; When the drone swarm formation tracking error fails to meet the standard, optimize the leadership feedback control gain. and adaptive trigger parameters; When the inclusion error of the autonomous vehicle swarm relative to the dynamic convex hull is not up to standard, optimize the feedback gain matrix of the dual distributed compensator. and .

[0026] In a second aspect, the present invention also provides an anti-interference composite triggering formation control device for air-to-ground heterogeneous unmanned systems, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the anti-interference composite triggering formation control method for air-to-ground heterogeneous unmanned systems as described in the first aspect.

[0027] Thirdly, the present invention also provides a computer-storable medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the anti-interference composite triggering formation control method for air-to-ground heterogeneous unmanned systems as described in the first aspect.

[0028] Compared with the prior art, the above-described technical solutions conceived in this invention can achieve the following beneficial effects: 1. This invention provides an anti-interference composite triggering formation inclusion control method for air-ground heterogeneous unmanned systems. For air-ground heterogeneous unmanned systems composed of UAVs and unmanned vehicles, a dual distributed compensator architecture is designed. This effectively overcomes the dynamic differences between air and ground units, allowing the ground unmanned vehicle to accurately reconstruct the macroscopic state and geometric configuration of the aerial formation using only local interaction information. This breaks through the bottleneck of heterogeneous system collaborative control, achieving high-precision inclusion control and ensuring that it always converges and remains within the dynamic safety convex hull formed by the UAV position projection, significantly improving the execution accuracy of cross-domain collaborative tasks. By introducing internal dynamic variables and adaptive coupling weights to design a composite adaptive event triggering mechanism, instantaneous measurement errors, historical trigger data, and internal dynamic evolution of the system are organically combined. Compared with traditional fixed thresholds or simple dynamic thresholds, this mechanism can not only encrypt communication to maintain stability when the error is large and significantly reduce communication to save bandwidth in steady state, making the communication resource scheduling of air-ground heterogeneous unmanned systems more intelligent, but also theoretically guarantee that there is a strict positive lower bound between any two triggering times (excluding Zeno behavior), thereby ensuring the feasibility of the control scheme in physical implementation and effectively avoiding channel congestion in low-bandwidth environments in the field.

[0029] 2. This invention provides an anti-interference composite triggering formation control method for air-to-ground heterogeneous unmanned systems. Addressing persistent environmental interference generated by external systems, such as gusts and rugged terrain, this invention designs an external disturbance compensator based on auxiliary variables. Compared to traditional extended-dimensional observers or differentiators, this method avoids directly differentiating noisy state signals, thus achieving accurate reconstruction of unknown heterogeneous interference without amplifying measurement noise. It possesses active anti-interference capability based on auxiliary variables and exhibits low noise sensitivity. The compensation signal output from the external disturbance compensator is used as a feedforward compensation term and output to the formation controller of the UAV swarm and the inclusion controller of the unmanned vehicle swarm, actively offsetting the influence of the external environment and ensuring the input state stability (ISS) and robustness of the air-to-ground formation system under strong interference environments.

[0030] 3. This invention provides an anti-interference composite triggering formation control method for air-to-ground heterogeneous unmanned systems. The control protocol of this invention's air-to-ground heterogeneous unmanned system relies only on local information interaction between neighbors and is fully applicable to directed communication topologies. This hierarchical distributed architecture reduces dependence on fully connected networks, enhances the system's survivability in situations of asymmetric or partial failure of communication links, and makes it easier to deploy in complex real-world field network environments. Attached Figure Description

[0031] Figure 1 This is a general flowchart of the control method for the anti-interference composite triggering formation of the air-to-ground heterogeneous unmanned system provided in an embodiment of the present invention; Figure 2 A design flowchart of an external disturbance compensator based on auxiliary variables provided for an embodiment of the present invention; Figure 3 A design flowchart of a drone leadership formation controller provided for an embodiment of the present invention; Figure 4 A design flowchart of the controller included in the unmanned vehicle following layer provided in an embodiment of the present invention; Figure 5 This is a topology diagram of an air-to-ground unmanned system provided in an embodiment of the present invention; Figure 6 The triggering time diagram is provided for the composite adaptive triggering mechanism in the embodiments of the present invention; wherein, (a) is the triggering time of the leadership event triggering controller, (b) is the triggering time of the follower event triggering controller 1, and (c) is the triggering time of the follower event triggering controller 2. Figure 7 The diagram illustrates the trajectory evolution of the air-ground heterogeneous unmanned system at different times, as provided in this embodiment of the invention. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0033] The following description, in conjunction with a preferred embodiment, illustrates the content involved in the above embodiments.

[0034] Example 1 This invention provides a method for anti-interference composite triggering formation control of air-to-ground heterogeneous unmanned systems, including: S100. Based on the external environmental interference experienced by the air-ground heterogeneous unmanned system, an external disturbance compensator containing auxiliary state variables is constructed, and a compensation signal is output to the heterogeneous unmanned system; the air-ground heterogeneous unmanned system includes a drone swarm as the leadership layer and an unmanned vehicle swarm as the follower layer, and the unmanned vehicles only receive instructions from the drones in one direction. S200. Estimate the virtual leader state using a distributed observer; construct a composite adaptive triggering mechanism based on adaptive coupling weights and internal dynamic variables. When the sampling deviation of the weighted UAV local neighbor error exceeds the lower bound of the threshold jointly formed by the local neighbor error and the internal dynamic variables, trigger communication between UAV local neighbors to perform state synchronization updates; update the adaptive coupling weights and internal dynamic variables in real time based on the dynamic changes of the UAV local neighbor error; calculate the actual control input of the UAV based on the virtual leader state and the compensation signal to control the UAV swarm formation. S300: Design a dual distributed compensator for each follower, and estimate the leader state and leader formation information through the dual distributed compensator; the dual distributed compensator has an independent triggering mechanism to synchronize the state updates among the local neighbors of the unmanned vehicle; according to the dynamic changes of the local neighbor error of the unmanned vehicle, update the coupling weights and internal dynamic variables in the two independent triggering mechanisms in real time; calculate the actual control input of the unmanned vehicle based on the leader state estimation result, the leader formation information estimation result and the compensation signal, so as to control the unmanned vehicle group to converge to the dynamic convex hull range determined by the unmanned vehicle group.

[0035] This invention primarily addresses the formation inclusion control challenge faced by heterogeneous air-to-ground unmanned systems (UAVs) performing collaborative tasks under complex communication and environmental conditions in the field. Existing research on formation inclusion control largely assumes a homogeneous system model and often relies on continuous-time state feedback, requiring constant monitoring and transmission of neighbor states. However, in practical air-to-ground collaborative engineering applications, UAVs and unmanned vehicles constitute typical linear heterogeneous multi-agent systems with fundamentally different dynamic characteristics. Furthermore, the bandwidth of tactical networks in the field is severely limited, and frequent data interactions easily lead to channel congestion and increase exposure risks. More critically, existing methods generally lack proactive estimation and compensation mechanisms for persistent environmental disturbances generated by external systems. Relying solely on passive robustness is insufficient to maintain high-precision formations, and there is even a risk that followers may deviate from the safe convex hull region.

[0036] To address the characteristics of linear heterogeneous multi-agent systems, this invention employs a fully distributed dual-state compensation control architecture. Compared to traditional continuous-time control and single-threshold event-triggered control, it introduces a composite adaptive event-triggered mechanism based on an auxiliary variable-based disturbance observer and internal dynamic variables. This method not only utilizes auxiliary variables to avoid differentiating noise states, thereby achieving accurate reconstruction and feedforward compensation for unknown heterogeneous disturbances, but also dynamically optimizes communication frequencies under non-uniform sampling conditions by combining system instantaneous errors, historical data, and internal dynamic behavior. While significantly reducing communication resource consumption and theoretically eliminating Zeno behavior, it effectively overcomes the synchronization difficulties caused by heterogeneous dynamics, ensuring the stability of the input state and the consistent eventual boundedness of the control within the air-to-ground formation system under strong interference environments.

[0037] The research scheme of the anti-interference composite triggering formation including control method of the air-ground heterogeneous unmanned system proposed in this invention is as follows: Figure 1 As shown, it includes three parts: using Algorithm 1 to construct an extended-dimensional state observer, estimate and compensate for external environmental disturbances in real time, and ensure the robustness of the system; using Algorithm 2 for the UAV leader, adjust the control signal through an adaptive event triggering strategy to achieve time-varying formation tracking of the virtual leader; and using Algorithm 3 for the UAV follower, design a controller and triggering mechanism to make it converge to the dynamic convex hull formed by the leader.

[0038] To address the persistent environmental disturbances encountered by heterogeneous air-to-ground unmanned systems during actual operations (such as gusts of wind encountered by UAVs and terrain friction encountered by unmanned vehicles), these disturbances are modeled as unknown bounded signals generated by a linear external system. To avoid directly differentiating the noisy state signals to obtain disturbance information, this invention constructs a disturbance observer based on auxiliary state variables to supplement external disturbances.

[0039] like Figure 2As shown, optionally, step S100 specifically includes: S110. To address the persistent external environmental interference of heterogeneous air-ground unmanned systems during air-ground collaborative operations, this interference is modeled as a signal generated by a linear external system; the interference signal is defined. Produced by the linear external system:

[0040] in, The state vector of the external system that is interfering; and For the interference model matrix, the matrix is ​​selected as follows: and To simulate gust interference with specific frequency characteristics; for unmanned vehicles, a matrix is ​​selected. and To simulate constant or slowly varying surface friction disturbances; S120, Introducing intermediate auxiliary variables Based on real-time status measurements of drones and unmanned vehicles and actual control input Construct the observer dynamic equations:

[0041] in, For the first Intermediate auxiliary variables for each agent and For the first The system matrix and input matrix of each agent. The observer gain matrix is... For the first Real-time state measurements of each agent For the first The actual control input of an intelligent agent; When the value is 1 to M, it indicates a drone. When the value of is from M+1 to N, it represents an unmanned vehicle, where M and N are both positive integers; Furthermore, the equation is solved in real time in an embedded processor using numerical integration (such as the Euler method or the Runge-Kutta method) to obtain auxiliary variables. The current value.

[0042] S130, According to the system matrix and interference model matrix Design the observer gain matrix The gain matrix can be solved using the pole placement method or by solving linear matrix inequalities. This ensures that the observation error of the external disturbance compensator dynamically satisfies the Hurwitz stability condition and the fast convergence constraint. The gain matrix must satisfy the matrix... The condition for the Hurwitz matrix is ​​to obtain the matrix. The eigenvalues ​​are all located in the left half of the complex plane, and the absolute value of the real part of its poles should be greater than the absolute value of the real part of the dominant pole of the system, so as to ensure the stability of the observation error dynamics of the external disturbance compensator in the Lyapunov sense.

[0043] S140, Based on the auxiliary variable and the Real-time state measurement of each agent Reconstruct the estimated value of external disturbances It is then output as a compensation signal to the formation controller of the UAV swarm and the inclusion controller of the unmanned vehicle swarm in the heterogeneous unmanned system.

[0044] The obtained interference estimate Introduced as a feedforward compensation term into the subsequent controller design, it generates a reverse-action term in the control law. To counteract the effects of heterogeneous environmental interference.

[0045] To address the leader set in a drone swarm and reduce air-to-air communication frequency while adapting to complex dynamic environments, this invention proposes a composite adaptive event triggering mechanism based on internal dynamic variables. This mechanism estimates the virtual leader state using a distributed observer and dynamically adjusts the triggering threshold using an adaptive law.

[0046] like Figure 3 As shown, optionally, step S200 specifically includes: S210. Set up a distributed observer on each UAV and calculate the local neighbor error based on the communication topology. ;in, For the first The estimated value of the virtual leader's state by a drone. It is a virtual leader state. For communication topology weights; S220. Based on the local neighbor error of the UAV, the update law of the distributed observer is: ;in, The observer gain matrix is... The latest event trigger time, For adaptive coupling weights.

[0047] Furthermore, in order to achieve on-demand communication and reduce the frequency of air-to-air communication while ensuring control accuracy, a composite adaptive triggering mechanism is designed, which includes adaptive coupling weights and internal dynamic variables.

[0048] S230, Define measurement error Introducing internal dynamic variables Construct a composite adaptive triggering mechanism, with the triggering function as follows:

[0049] in, These are preset positive parameters. It is a positive definite weighted matrix; when When the trigger time is reached, communication between the drone's local neighbors is initiated, the state of the drone's local neighbors is synchronized and updated, and the trigger time is updated. .

[0050] Internal dynamic variables were introduced. Used to further filter transient noise and unnecessary triggers.

[0051] S240. Based on the dynamic changes of the UAV's local neighbor error, coupling weights and internal dynamic variables The adaptive law is updated online according to the following formulas:

[0052]

[0053] in, , , , All are preset positive design parameters; , The local neighbor error at the latest trigger time; The positive definite weighting matrix is ​​defined as follows: the coupling weights automatically increase as the local neighbor error of the UAV increases, thereby tightening the trigger threshold; conversely, the weights decrease as the error decreases, in order to reduce communication. S250, Observation state obtained based on the distributed observer Estimated values ​​of external disturbances Calculate the first Actual control input of the drone :

[0054] in, This is a preset time-varying formation vector; For feedback control gain; and It is a solution to the regulator equation; According to the actual control input Control the corresponding drones to control the drone swarm formation.

[0055] To address the issue of unmanned vehicle swarms as a set of followers, and to ensure that the swarm converges and remains within the dynamic convex hull formed by the leader's position, this invention employs a dual distributed compensator structure, coupled with a composite adaptive event triggering mechanism, to independently optimize the communication resources of each channel.

[0056] like Figure 4 As shown, optionally, step S300 specifically includes: S310, each follower is designed with a state fusion compensator. and formation information compensator The system obtains the leadership state estimate and leadership formation information estimate accordingly; and calculates the local neighbor error of state fusion based on the communication topology. Local neighbor error of formation information :

[0057] The dynamic update law for constructing a dual distributed compensator:

[0058] in, For the virtual leader system matrix, Generate a matrix for the formation. Here is the feedback gain matrix of the state fusion compensator. The feedback gain matrix for the formation information compensator; Design of the state fusion compensator and formation information compensator A convex combination used to obtain the leadership state and a convex combination of formation geometry information.

[0059] S320, respectively, is the state fusion compensator. and formation information compensator Two independent triggering mechanisms are designed; each triggering mechanism includes adaptive coupling weights and internal dynamic variables; the triggering function is:

[0060] Among them, subscript These correspond to the state fusion and formation information channels, respectively; when the trigger function... When the time is right, the corresponding channel's communication event is triggered and the time is updated. ; For state fusion compensator and formation information compensator Two independent triggering mechanisms are designed to independently adjust the communication frequency of the two compensators. This decoupling design can allocate communication resources on demand according to the different dynamic characteristics of the macroscopic state and formation geometry, thereby avoiding unnecessary global synchronization caused by the error exceeding the limit of a single channel, and further significantly saving communication bandwidth.

[0061] S330. Based on the dynamic changes in the local neighbor error of the unmanned vehicle, the internal dynamic variable... and adaptive coupling weights The update law is:

[0062]

[0063] in, This represents the measurement error for each channel; Coupling weights Dynamic updates enable adaptive tightening of trigger thresholds in complex terrain areas to ensure control accuracy; while internal dynamic variables... The dynamic update provides a non-negative relaxation buffer for the triggering conditions, effectively filtering out frequent false triggers caused by transient error fluctuations, and further saving communication resources.

[0064] S340, Estimated values ​​based on leadership state estimation, leadership formation information estimation, and external disturbances. Calculate the first The actual control input of an unmanned vehicle :

[0065] in, For calming effect; and For feedforward gain, , ; According to the actual control input The unmanned vehicle is controlled to counteract external environmental interference and drive it to converge within the dynamic convex hull range determined by the drone swarm.

[0066] Furthermore, after controlling the air-to-ground heterogeneous unmanned system using the above method, it is also necessary to determine whether it has been adjusted to the target range. Therefore, after step S300, the following steps are also included: S400. Determine whether the performance indicators of the air-ground heterogeneous unmanned system meet the preset requirements; the performance indicators include: external interference estimation error, UAV swarm formation tracking error, and the inclusion error of the unmanned vehicle swarm relative to the dynamic convex hull; If all indicators meet the preset requirements, the current control parameters are maintained, and the system enters steady-state operation or waits for the next trigger. If any indicator fails to meet the preset requirements, the design parameters of the corresponding observer or controller will be optimized according to the type of performance indicator that fails to meet the requirements.

[0067] Optionally, the optimization of design parameters includes: When the external disturbance estimation error is not up to standard, optimize the observer gain matrix. Adjust the pole placement; When the drone swarm formation tracking error fails to meet the standard, optimize the leadership feedback control gain. and adaptive trigger parameters; When the inclusion error of the autonomous vehicle swarm relative to the dynamic convex hull is not up to standard, optimize the feedback gain matrix of the dual distributed compensator. and .

[0068] Through the above judgments and parameter optimizations, precise collaborative control of the air-ground heterogeneous unmanned system can be guaranteed. It not only meets the control accuracy of aerial formation and achieves high-precision containment control, ensuring that the unmanned vehicle always converges and remains within the dynamic safety convex hull formed by the UAV's position projection, but also maximizes the saving of communication resources while ensuring control accuracy, significantly improving the reliability and adaptability of this control method in actual engineering deployment.

[0069] This invention provides an anti-interference composite triggering formation inclusion control method for air-ground heterogeneous unmanned systems. Targeting air-ground heterogeneous unmanned systems composed of UAVs and unmanned vehicles, a dual distributed compensator architecture is designed. This effectively overcomes the dynamic differences between air and ground units, allowing the ground unmanned vehicle to accurately reconstruct the macroscopic state and geometric configuration of the aerial formation using only local interaction information. This breaks through the bottleneck of heterogeneous system collaborative control, achieving high-precision inclusion control and ensuring that it always converges and remains within the dynamic safety convex hull formed by the UAV position projection, significantly improving the execution accuracy of cross-domain collaborative tasks.

[0070] To verify the effectiveness of the proposed adaptive event-based anti-interference composite triggering formation control method for air-to-ground heterogeneous unmanned systems, a multi-agent simulation platform was built; such as Figure 5 As shown, a heterogeneous air-ground unmanned system with 10 agents was constructed in the simulation scenario, in which agents numbered 1 to 4 were set as the drone leaders ( (i takes values ​​from 1 to 4), agents numbered 5 to 10 are designated as followers of the autonomous vehicle. (where i takes values ​​from 5 to 10). The communication topology is designed as a directed graph structure (e.g., ...). Figure 5(As indicated by the middle arrow) to simulate the asymmetry of actual air-to-ground communication links. For example, the unmanned vehicle (node ​​5) can only receive instructions from the drone (node ​​1) in one direction, reflecting the hierarchical command relationship from the leadership to the followers.

[0071] To address the heterogeneity between drones and unmanned vehicles (UAVs), a linear system model conforming to the motion characteristics of aircraft is selected for the leadership layer (drones). A linear system model conforming to the motion characteristics of ground vehicles is selected for the following layer (UAVs). This heterogeneity is reflected in the simulation parameter initialization phase, where the system matrices of the leader and followers are initialized. The system is configured with matrices of different dimensions or parameters to rigorously simulate the heterogeneous dynamic characteristics. Simultaneously, all selected system parameters satisfy controllability and observability conditions.

[0072] External interference environment simulation: To test the system's anti-interference performance, interference signals simulating the external environment were introduced into the simulation. The interference signal was set to a time-varying signal generated by an external system (such as a superposition of sine waves or bounded random noise) to simulate the force of high-altitude gusts on the UAV and the friction and disturbance caused by the rugged road surface on the UAV.

[0073] Based on the above, an anti-interference composite triggering formation control method for air-to-ground heterogeneous unmanned systems based on adaptive events is constructed. Numerical simulations are performed under the above simulation environment and parameter settings, and the following experimental results are analyzed: Figure 6 The simulation shows the distribution of trigger times for the drone leader and the unmanned vehicle follower under the composite adaptive event triggering mechanism within the simulation time (0-10s). Figure 6 (a) shows the triggering time of the leadership event triggering controller, reflecting the time distribution of communication between UAVs in order to collaboratively estimate the virtual leader state; Figure 6 (b) shows the triggering time of the follower layer event trigger controller 1 (DTC1), which corresponds to the state fusion compensator of the autonomous vehicle follower and demonstrates the synchronization communication frequency between the local neighbors of the autonomous vehicle for the "macro motion state of the leadership layer". Figure 6 (c) shows the triggering time of the follower layer event trigger controller 2 (DTC2), which corresponds to the formation information compensator of the autonomous vehicle follower and shows the synchronization communication frequency between local neighbors of the autonomous vehicle for the "leader formation geometry".

[0074] from Figure 6As can be seen from (a)-(c), the triggering times of each channel controller are not continuously distributed, but exhibit obvious sparsity. This indicates that the agents only communicate and update control when necessary (i.e., when the weighted local neighbor error sampling deviation exceeds the dynamic threshold), rather than periodically transmitting data continuously. In the early stages of system operation (e.g., t=0s to t=2s) or when state changes drastically, triggering is relatively dense to ensure rapid system convergence; however, after the system stabilizes, the triggering interval increases significantly. Compared with traditional time-triggered control, the independent dual-channel triggering mechanism designed in this invention avoids unnecessary global synchronization, significantly reduces data transmission frequency, and maximizes the saving of air-to-ground communication bandwidth resources. Furthermore, the time interval between any two adjacent triggering times in the figure is strictly greater than zero, verifying that the system does not exhibit Zeno behavior and is physically safe and feasible.

[0075] Figure 7 This paper demonstrates the evolution of the two-dimensional planar motion trajectory of a heterogeneous air-ground unmanned system at different time points (t=0s, 1s, ..., 5s). The quadrilateral formed by the connecting lines in the figure represents a virtual formation of four UAVs (diamond-shaped dots in the figure). Despite the introduction of simulated high-altitude gusts as external interference in the simulation, the UAV swarm can quickly adjust from the initial chaotic state (t=0s) and stably maintain the preset rectangular formation after t=3s, proving the effectiveness of the disturbance compensator and formation controller in this method. The circular dots in the figure represent three unmanned vehicles. It can be seen that as time progresses, all unmanned vehicles successfully converge and always remain within the convex hull region (safe zone) formed by the UAV positions. Even under simulated ground rough interference, the motion trajectory of the unmanned vehicles remains smooth, without exceeding the convex hull boundary, proving that the follower layer state inclusion controller in this method can effectively counteract ground environment interference and achieve high-precision air-ground cooperative inclusion control.

[0076] Simulation results fully verify the superiority of the proposed solution in achieving heterogeneous air-ground collaboration, reducing communication load, and resisting external interference.

[0077] Taking joint search and rescue in disaster areas as an example: an air-ground heterogeneous collaborative rescue system is composed of multiple drones (as the leadership layer) and multiple unmanned vehicles (as the following layer). During the mission, multiple drones maintain a pre-set specific formation (such as a rectangular formation) in the air, using onboard vision equipment to conduct wide-area searches of the disaster area and act as communication relay nodes; at this time, the drones in the high altitude will be affected by strong and unknown gusts of wind. Meanwhile, the ground unmanned vehicles, fully loaded with relief supplies or fine detection equipment, travel on rugged ruins or wild roads, facing disturbances such as constantly changing surface friction.

[0078] In this typical application scenario, the method proposed in this invention is applied as follows: First, through an external disturbance compensator, the UAV and unmanned vehicle can actively estimate and offset the physical interference caused by high-altitude gusts and rugged ground, avoiding formation distortion caused by passive anti-interference. Second, the ground unmanned vehicle swarm, through a dual distributed compensator, only needs to rely on local communication interaction to accurately converge and always travel within the "safe convex hull area" formed by the position projection of the aerial UAV swarm, ensuring that the unmanned vehicle swarm is always within the communication coverage and field of vision protection range of the UAV. Finally, considering the damage to communication base stations in disaster areas and the extreme limitation of wireless network bandwidth, the composite adaptive triggering mechanism adopted by the system ensures that the air-ground nodes only send status data when the sampling deviation of local neighbor errors exceeds a dynamic threshold (e.g., when encountering sudden strong winds that cause a sharp increase in error), remaining silent during smooth driving, thereby greatly saving valuable air-ground communication bandwidth resources and effectively avoiding channel congestion.

[0079] Example 2 The present invention also provides an anti-interference composite triggering formation control device for air-to-ground heterogeneous unmanned systems, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the anti-interference composite triggering formation control method for air-to-ground heterogeneous unmanned systems as described in any one of the embodiments.

[0080] The present invention provides an anti-interference composite triggering formation control device for air-to-ground heterogeneous unmanned systems, used to execute the anti-interference composite triggering formation control method for air-to-ground heterogeneous unmanned systems as described in any one of Embodiment 1, and has the same beneficial effects.

[0081] Example 3 The present invention also provides a computer-storable medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the anti-interference composite triggering formation control method for air-to-ground heterogeneous unmanned systems as described in any one of Embodiment 1.

[0082] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A control method for anti-interference composite triggering formation of an air-to-ground heterogeneous unmanned system, characterized in that, include: S100. Based on the external environmental interference experienced by the air-ground heterogeneous unmanned system, an external disturbance compensator containing auxiliary state variables is constructed, and a compensation signal is output to the heterogeneous unmanned system; the air-ground heterogeneous unmanned system includes a drone swarm as the leadership layer and an unmanned vehicle swarm as the follower layer, and the unmanned vehicles only receive instructions from the drones in one direction. S200. Estimate the virtual leader state using a distributed observer; construct a composite adaptive triggering mechanism based on adaptive coupling weights and internal dynamic variables. When the sampling deviation of the weighted UAV local neighbor error exceeds the lower bound of the threshold jointly formed by the local neighbor error and the internal dynamic variables, trigger communication between UAV local neighbors to perform state synchronization updates; update the adaptive coupling weights and internal dynamic variables in real time based on the dynamic changes of the UAV local neighbor error; calculate the actual control input of the UAV based on the virtual leader state and the compensation signal to control the UAV swarm formation. S300: Design a dual distributed compensator for each follower, and estimate the leader state and leader formation information through the dual distributed compensator; the dual distributed compensator has an independent triggering mechanism to synchronize the state updates among the local neighbors of the unmanned vehicle; according to the dynamic changes of the local neighbor error of the unmanned vehicle, update the coupling weights and internal dynamic variables in the two independent triggering mechanisms in real time; calculate the actual control input of the unmanned vehicle based on the leader state estimation result, the leader formation information estimation result and the compensation signal, so as to control the unmanned vehicle group to converge to the dynamic convex hull range determined by the unmanned vehicle group.

2. The method as described in claim 1, characterized in that, Step S100 specifically includes: S110. To address the persistent external environmental interference of heterogeneous air-ground unmanned systems during air-ground collaborative operations, this interference is modeled as a signal generated by a linear external system; the interference signal is defined. Produced by the linear external system: in, The state vector of the external system that is interfering; and For the interference model matrix, the matrix is ​​selected as follows: and To simulate gust interference with specific frequency characteristics; for unmanned vehicles, a matrix is ​​selected. and To simulate constant or slowly varying surface friction disturbances; S120, Introducing intermediate auxiliary variables Based on real-time status measurements of drones and unmanned vehicles and actual control input Construct the observer dynamic equations: in, For the first Intermediate auxiliary variables for each agent and For the first The system matrix and input matrix of each agent. The observer gain matrix is... For the first Real-time state measurements of each agent For the first The actual control input of an intelligent agent; When the value is 1 to M, it indicates a drone. When the value of is from M+1 to N, it represents an unmanned vehicle, where M and N are both positive integers; S130, According to the system matrix and interference model matrix Design the observer gain matrix The gain matrix can be solved using the pole placement method or by solving linear matrix inequalities. This ensures that the observation error of the external disturbance compensator dynamically satisfies the Hurwitz stability condition and the fast convergence constraint. S140, Based on the auxiliary variable and the Real-time state measurement of each agent Reconstruct the estimated value of external disturbances It is then output as a compensation signal to the formation controller of the UAV swarm and the inclusion controller of the unmanned vehicle swarm in the heterogeneous unmanned system.

3. The method as described in claim 2, characterized in that, Step S200 specifically includes: S210. Set up a distributed observer on each UAV and calculate the local neighbor error based on the communication topology. ;in, For the first The estimated value of the virtual leader's state by a drone. It is a virtual leader state. For communication topology weights; S220. Based on the local neighbor error of the UAV, the update law of the distributed observer is: ;in, The observer gain matrix is... The latest event trigger time, For adaptive coupling weights; S230, Define measurement error Introducing internal dynamic variables Construct a composite adaptive triggering mechanism, with the triggering function as follows: in, These are preset positive parameters. It is a positive definite weighted matrix; when When the trigger time is reached, communication between the drone's local neighbors is initiated, the state of the drone's local neighbors is synchronized and updated, and the trigger time is updated. ; S240. Based on the dynamic changes of the UAV's local neighbor error, coupling weights and internal dynamic variables The adaptive law is updated online according to the following formulas: in, , , , All are preset positive design parameters; , The local neighbor error at the latest trigger time; The positive definite weighting matrix is ​​defined as follows: the coupling weights automatically increase as the local neighbor error of the UAV increases, thereby tightening the trigger threshold; conversely, the weights decrease as the error decreases, in order to reduce communication. S250, Observation state obtained based on the distributed observer Estimated values ​​of external disturbances Calculate the first Actual control input of the drone : in, This is a preset time-varying formation vector; For feedback control gain; and It is a solution to the regulator equation; According to the actual control input Control the corresponding drones to control the drone swarm formation.

4. The method as described in claim 3, characterized in that, Step S300 specifically includes: S310, each follower is designed with a state fusion compensator. and formation information compensator The system obtains the leadership state estimate and leadership formation information estimate accordingly; and calculates the local neighbor error of state fusion based on the communication topology. Local neighbor error of formation information : The dynamic update law for constructing a dual distributed compensator: in, For the virtual leader system matrix, Generate a matrix for the formation. Here is the feedback gain matrix of the state fusion compensator. The feedback gain matrix for the formation information compensator; S320, respectively, is the state fusion compensator. and formation information compensator Two independent triggering mechanisms are designed; each triggering mechanism includes adaptive coupling weights and internal dynamic variables; the triggering function is: Among them, subscript These correspond to the state fusion and formation information channels, respectively; when the trigger function... When the time is right, the corresponding channel's communication event is triggered and the time is updated. ; S330. Based on the dynamic changes in the local neighbor error of the unmanned vehicle, the internal dynamic variable... and adaptive coupling weights The update law is: in, This represents the measurement error for each channel; S340, Estimated values ​​based on leadership state estimation, leadership formation information estimation, and external disturbances. Calculate the first The actual control input of an unmanned vehicle : in, For calming effect; and For feedforward gain, , ; According to the actual control input The unmanned vehicle is controlled to counteract external environmental interference and drive it to converge within the dynamic convex hull range determined by the drone swarm.

5. The method as described in claim 4, characterized in that, Following step S300, the method further includes: S400. Determine whether the performance indicators of the air-ground heterogeneous unmanned system meet the preset requirements; the performance indicators include: external interference estimation error, UAV swarm formation tracking error, and the inclusion error of the unmanned vehicle swarm relative to the dynamic convex hull; If all indicators meet the preset requirements, the current control parameters are maintained, and the system enters steady-state operation or waits for the next trigger. If any indicator fails to meet the preset requirements, the design parameters of the corresponding observer or controller will be optimized according to the type of performance indicator that fails to meet the requirements.

6. The method as described in claim 5, characterized in that, The optimization of design parameters includes: When the external disturbance estimation error is not up to standard, optimize the observer gain matrix. Adjust the pole placement; When the drone swarm formation tracking error fails to meet the standard, optimize the leadership feedback control gain. and adaptive trigger parameters; When the inclusion error of the autonomous vehicle swarm relative to the dynamic convex hull is not up to standard, optimize the feedback gain matrix of the dual distributed compensator. and .

7. A heterogeneous air-to-ground unmanned system anti-interference composite trigger formation includes a control device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements claim 1. The anti-interference composite trigger formation of the air-to-ground heterogeneous unmanned system described in any one of the 6 claims includes a control method.

8. A computer-storable medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements claim 1. The anti-interference composite trigger formation of the air-to-ground heterogeneous unmanned system described in any one of the 6 claims includes a control method.

Images