A collaborative security control method for unmanned aerial vehicle (UAV) swarms based on virtual pipelines

Abstract

This invention discloses a cooperative safety control method for UAV swarms based on virtual pipelines. It constructs a nonlinear UAV model considering unknown external interference and estimates this interference using a nonlinear interference observer. The method obtains the trajectory of the leader UAV in the swarm and the expected trajectories of each follower UAV. A virtual pipeline is constructed with the expected trajectories of the follower UAVs as the axis. An artificial potential field region is constructed within the virtual pipeline to confine each follower UAV to its corresponding virtual pipeline. A formation controller is designed based on the nonlinear interference observer and the virtual pipeline. The constructed formation controller controls the UAV swarm. The virtual pipeline is designed to confine UAVs within their corresponding pipelines, preventing collisions within the UAV swarm. Combining the interference observer and the virtual pipeline, a distributed cooperative safety controller is constructed to solve the cooperative safety control problem of UAV swarms with unknown external interference and internal collision prevention issues.

Description

Technical Field

[0001] This invention relates to the collaborative safety control of unmanned aerial vehicle (UAV) swarms, specifically to a collaborative safety control method for UAV swarms based on virtual pipelines. Background Technology

[0002] Unmanned aerial vehicles (UAVs) are characterized by "zero crew, recoverability, low cost, and high maneuverability," enabling them to operate for extended periods in high-risk or complex airspace. Their applications began in military scenarios, such as intelligence surveillance, emergency rescue, communications relay, and border patrol; and have extended to the civilian market, covering diverse tasks such as traffic monitoring, forest fire monitoring, and plant protection mapping. Despite continuous upgrades in individual UAV capabilities, limitations remain due to factors such as field of view, payload, range, and equipment reliability: obtaining a complete view of the fire scene is difficult during fire observation; stereoscopic multi-angle imaging is impossible during reconnaissance operations; fuel and attachment point constraints during border patrol flights lead to insufficient endurance and limited response capabilities; and a malfunction means mission interruption. Therefore, the industry is generally shifting towards multi-UAV collaborative solutions, significantly expanding mission adaptability and resilience.

[0003] During actual flight, unmanned aerial vehicles (UAVs) are constantly threatened by unknown external disturbances. Severe external disturbances can affect system stability and ultimately lead to mission failure. To improve the robustness of UAVs, numerous disturbance rejection control algorithms have been proposed, such as robust control, sliding mode control, and disturbance observer-based control methods. Among these, disturbance observer-based control methods can accurately estimate unknown external disturbances by collecting UAV state information and feed it forward to the controller to compensate for their adverse effects on the UAV.

[0004] In cooperative control research, safety remains a core concern and a key technical challenge, especially collision avoidance among drone swarms. Due to the inherent physical structure of drones and their susceptibility to unknown external interference during flight, maintaining a sufficient safe distance between drones is crucial to prevent potential collisions during cooperative missions. Commonly used methods for preventing internal collisions include artificial potential field methods, model predictive control, online optimization strategies, and pre-defined performance control methods. While these methods have proven effective in practical applications, they still face common challenges, such as heavy computational burden, high real-time communication requirements, and conservative control strategies. Summary of the Invention

[0005] Purpose of the invention: To address the above-mentioned shortcomings, this invention provides a highly secure collaborative security control method for unmanned aerial vehicle (UAV) swarms based on virtual pipelines.

[0006] Technical Solution: To solve the above problems, this invention adopts a collaborative security control method for unmanned aerial vehicle (UAV) swarms based on virtual pipelines, comprising the following steps:

[0007] (1) Construct a nonlinear model of the UAV that considers unknown external interference, and construct a nonlinear interference observer to estimate the unknown external interference;

[0008] (2) Obtain the trajectory of the leader drone in the drone swarm. Based on the relative distance between each follower drone and the leader drone, obtain the expected trajectory of each follower drone. Construct a virtual pipeline with the expected trajectory of the follower drone as the axis. By constructing an artificial potential field region inside the virtual pipeline, each follower drone is restricted to the corresponding virtual pipeline.

[0009] (3) Design of a formation controller based on a nonlinear disturbance observer and a virtual pipeline;

[0010] (4) Control the drone swarm through the constructed formation controller.

[0011] Furthermore, the nonlinear model of the UAV is as follows:

[0012] ;

[0013] in, For the first The location of the drone Let be the speed of the i-th drone. For the first The nonlinear term of the drone model, For the first The system control input for the drone, For the first Unknown external interference with the drone For the first The adjustment matrix of the drone.

[0014] Furthermore, the nonlinear disturbance observer is:

[0015] ;

[0016] in, For the first Observations of unknown external interference from drones. as auxiliary variables The estimated value, It is a parameter matrix.

[0017] Furthermore, the auxiliary variables The expression is:

[0018] ;

[0019] Auxiliary variables The update law is:

[0020] .

[0021] Furthermore, the artificial potential field region includes a repulsive field, the expression of which is:

[0022] ;

[0023] ;

[0024] ;

[0025] in, In the first drone A repulsive field along the axial direction. , , , These are the three coordinate axes of the spatial coordinate system. For the first The maximum distance between the center and the boundary of the virtual pipeline constructed by the drone. Let be the radius of the virtual pipe. This represents the outermost distance of the virtual pipeline's upper boundary. This represents the outermost distance of the lower boundary of the virtual pipe. Indicates the first The radius of the safe zone within the virtual pipeline constructed by the drone. For the first The desired trajectory of the drone. For the first Tracking error of the drone For the first A drone in The position of the axis.

[0026] Furthermore, the boundary of the virtual pipeline is adjusted in real time using a positive signal generated by a self-adjusting auxiliary system, and the boundary expression of the virtual pipeline is:

[0027] ;

[0028] ;

[0029] ;

[0030] in, This represents the upper boundary of the virtual pipe. This represents the lower boundary of the virtual pipe. and For the state variables of the self-regulating auxiliary system, This represents the width of the artificial potential field region within the virtual pipe;

[0031] The expression for the positive signal generated by the self-regulating auxiliary system is:

[0032] ;

[0033] in, and Indicates in virtual pipeline State variables of a self-adjusting auxiliary system under boundary conditions Indicates in virtual pipeline The system input of the self-adjusting auxiliary system under boundary conditions. and Indicates in virtual pipeline System parameters of the self-adjusting auxiliary system under boundary conditions. , This indicates the upper boundary condition of the virtual pipeline. This indicates the lower boundary condition of the virtual pipe.

[0034] Furthermore, the formation controller is:

[0035] ;

[0036] ;

[0037] ;

[0038] in, The adjustment matrix for all follower drones, , and For the parameter matrix, For formation tracking error, To address formation tracking errors Find the first derivative. For the nonlinear term of the model for all follower drones, Estimates of unknown external interference for all follower drones. For the trajectory of the leader's drone Find the second derivative. For the system control input of all follower drones, The repulsive force experienced by all follower drones. and As an auxiliary variable, For the speed error of all follower drones, These are design parameters.

[0039] Furthermore, the repulsive force experienced by all follower drones is:

[0040] ;

[0041] ;

[0042] in, The repulsive force experienced by the first drone To represent the transpose of a matrix, The repulsive force experienced by the second drone For the first The repulsive force experienced by the drone For the first A drone in The repulsive force in the axial direction.

[0043] The present invention also employs a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above method.

[0044] The present invention also employs a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above method.

[0045] Beneficial Effects: Compared with existing technologies, the significant advantage of this invention is the design of virtual pipelines to confine drones within corresponding pipelines, preventing collisions within the drone swarm. By combining an interference observer with the virtual pipelines, a distributed cooperative safety controller is constructed to solve the cooperative safety control problem of drone swarms with unknown external interference and internal collision avoidance issues. The interference observer is simple in design and can quickly and accurately estimate unknown external interference, facilitating engineering implementation. Verification shows that the cooperative safety control algorithm designed in this invention can ensure that drones track the desired trajectory and maintain the desired formation. This invention effectively solves the cooperative safety control problem of drone swarms during mission execution. Attached Figure Description

[0046] Figure 1 This is a flowchart illustrating the control method of the present invention.

[0047] Figure 2 This is a schematic diagram of the virtual pipeline in this invention. Detailed Implementation

[0048] In this embodiment, the drone swarm is composed of... A formation system consisting of drones, with internal communication transmission between drones via a diagram. It means that, among them, and These are the sets of nodes and the sets of edges, respectively. Indicates the first The set of neighboring nodes of a given node. Furthermore, the Laplace matrix is ​​defined as... , Let be an adjacency matrix, and each element therein be a non-negative number. If the ... frame and the first If a drone can perform data interaction, then For formation strategies involving leader drones, a graph is introduced. To represent data transmission between leaders and followers, a diagonal matrix is ​​used to indicate whether data transmission exists between the leader and follower drones. It means that if the first If a follower can receive data from the leader, then ;otherwise, .

[0049] like Figure 1 As shown in this embodiment, a collaborative security control method for UAV swarms based on virtual pipelines specifically includes the following steps:

[0050] Step 1: Construct a nonlinear model of the UAV that considers unknown external disturbances.

[0051] Assuming the drone is a rigid body, the nonlinear model of its position subsystem is constructed as follows:

[0052] (1)

[0053] in, , and The first A drone in , and Displacement of the axis, , and The first The roll angle, pitch angle, and yaw angle of the drone. For the first The quality of the drone For the first The total thrust of the drone , and For the first The air damping coefficient of the drone, , and For the first Unknown external interference experienced by the drone It represents the acceleration due to gravity.

[0054] Among them, the Total thrust of the drone for:

[0055] (2)

[0056] in, , and The first A drone in , and The thrust of the shaft.

[0057] From equations (1) and (2), the nonlinear model of the UAV can be obtained as follows:

[0058] (3)

[0059] in, , , , and The first The position, velocity, model nonlinearities, system control inputs, and unknown external disturbances of the UAV. , For the first The mass of the drone. This model is a second-order nonlinear model.

[0060] The purpose of this invention is to design a cooperative safety controller that enables a swarm of drones with unknown external interference and internal collision avoidance to track a desired trajectory and maintain a specific formation. The following assumptions are given before designing the safety control scheme.

[0061] Assumption 1. Communication topology diagram between drones It is undirected and connected, therefore the matrix It is reversible. In addition, there exists a directed spanning tree with the leader drone as the root node, and all follower drones can directly or indirectly receive information from the leader drone.

[0062] Assumption 2. The desired trajectory of the leader drone and its first derivative Second derivative With third derivative The norms of all are bounded.

[0063] Assumption 3. Unknown external disturbances and the norm of their derivatives are bounded, i.e., there exist positive constants. and make and Established.

[0064] Step 2: Construct a nonlinear disturbance observer to estimate unknown external disturbances.

[0065] In order to effectively suppress unknown external interference To assess the adverse effects of drones, the following auxiliary variables are introduced:

[0066] (4)

[0067] in, and Let be the parameter matrix to be designed. Taking the derivative of equation (4), we get:

[0068] (5)

[0069] in, The update law is designed as follows:

[0070] (6)

[0071] Therefore, the interference observer can be obtained as follows:

[0072] (7)

[0073] in, Indicates the impact of unknown external interference. The estimate.

[0074] Step 3: Obtain the trajectory of the leader drone in the drone swarm. Based on the relative distance between each follower drone and the leader drone, obtain the expected trajectory of each follower drone. Construct a virtual pipeline with the expected trajectory of the follower drone as the axis. By constructing an artificial potential field region inside the virtual pipeline, each follower drone is confined to the corresponding virtual pipeline.

[0075] Specifically, the first Desired trajectory of a drone As the first The center of the virtual pipeline has a radius designed as follows: , No. The upper and lower boundaries of the virtual pipes are respectively and . , Indicates the first The virtual pipe and the first The minimum distance between virtual pipes Indicates the minimum safe distance between drones. This represents the width of the artificial potential field region within the virtual pipe. Indicates the first The radius of the safe zone within a virtual pipeline.

[0076] An artificial potential field region is constructed inside the virtual pipeline, confining each drone within its corresponding pipeline. The repulsive force field within this artificial potential field region... The specific expression is:

[0077] (8)

[0078] in, , , and These represent the outer distances of the upper and lower boundaries, respectively. , For the first Tracking error of the drone.

[0079] To provide more buffer space for drones while ensuring no collisions occur within the drone swarm, flexible virtual pipes are constructed, such as... Figure 2 As shown, the first The boundary design of each virtual pipeline is as follows:

[0080] (9)

[0081] in, , This represents the upper boundary of the virtual pipe. This represents the lower boundary of the virtual pipe.

[0082] In addition, in order to change and Design a self-adjusting auxiliary system:

[0083] (10)

[0084] in, , , This indicates the upper boundary condition of the virtual pipeline. This indicates the lower boundary condition of the virtual pipe. , and These represent the two states of the system and the system input, respectively. and These are the parameters to be designed.

[0085] After applying the amplitude limiting process, the final result is... and .

[0086] Step 4: Design a formation controller based on a nonlinear disturbance observer and a virtual pipeline.

[0087] Define the formation tracking error as:

[0088] (11)

[0089] in, Indicates the first frame and the first The expected distance vector of the drone. Indicates the first The location of the drone Indicates the leader's expected trajectory. For leaders and the The expected distance vector of the drone.

[0090] Based on the flight patterns of the leader and followers, expanding the formation tracking error yields the following:

[0091] (12)

[0092] in, , , and

[0093] Based on the formation tracking error, the auxiliary variable is constructed as follows:

[0094] (13)

[0095] in, , The parameter matrix to be designed, For the speed error of all follower drones, The parameter matrix to be designed, , , , , , It is a repulsive force, and:

[0096] (14)

[0097] in, .

[0098] The following is a distributed formation controller designed based on an interference observer and a virtual pipeline:

[0099] (15)

[0100] in, The adjustment matrix for all follower drones, and The parameter matrix to be designed, , , For the nonlinear term of the model for all follower drones, Estimates of unknown external interference for all follower drones. , , , , These are the parameters to be designed.

[0101] Step 5: Control the drone swarm using the constructed formation controller.

[0102] Based on the above analysis and discussion, the following conclusions are drawn:

[0103] Conclusion 1: Consider a... A formation system consisting of drones, of which the first The model of the UAV is shown in Equation (1). Under the condition of unknown external interference, if assumptions 1 and 2 hold, and the designed interference observer is used as shown in Equation (7), the artificial potential field is used as shown in Equation (8), the flexible virtual pipeline boundary is used as shown in Equation (9), and the distributed formation controller is used as shown in Equation (15), then when the parameters are appropriately selected so that the following inequality holds: , , , , and If so, then all signals within the closed-loop system are bounded.

[0104] Proof: Choose the following Lyapunov function:

[0105] (16)

[0106] in, , , , , , , and .

[0107] From equation (16), we can obtain:

[0108] (17)

[0109] From the artificial potential field function, we can obtain:

[0110] (18)

[0111] Then for:

[0112] (19)

[0113] From equations (16)-(19), we can obtain:

[0114] (20)

[0115] in, , , , , , , , , , , , , , and .

[0116] From Young's inequality, we can obtain:

[0117] (twenty one)

[0118] (twenty two)

[0119] (twenty three)

[0120] (twenty four)

[0121] (25)

[0122] (26)

[0123] (27)

[0124] (28)

[0125] (29)

[0126] in, , , , , , , and For positive integers, , , , , , and We can conclude that:

[0127] (30)

[0128] in, .

[0129] The proof process will now be divided into three different scenarios:

[0130] Scenario 1: satisfy and ,and, and satisfy We can obtain:

[0131] (31)

[0132] Choose the appropriate , , , , and Let the following inequalities apply:

[0133]

[0134]

[0135]

[0136]

[0137]

[0138]

[0139] If true, then we can obtain .

[0140] Scenario 2: satisfy and ,and, and satisfy We can obtain:

[0141]

[0142] Choose the appropriate , , , , and Let the following inequalities apply:

[0143]

[0144]

[0145]

[0146]

[0147]

[0148]

[0149] If true, then we can obtain .

[0150] Case 3: If cases 1 and 2 are not satisfied, then:

[0151]

[0152] Choose the appropriate , , , , , , and Let the following inequalities apply:

[0153]

[0154]

[0155]

[0156]

[0157]

[0158]

[0159]

[0160]

[0161] If true, then we can obtain .

[0162] Combining cases 1-3 with Lyapunov's extended theory, we can obtain , , , , , , and Bounded. Furthermore... Boundedness means that all drones are confined within their corresponding conduits, ensuring that collisions do not occur within the formation system. This confirms the above conclusion.

Claims

1. A collaborative safety control method for unmanned aerial vehicle (UAV) swarms based on virtual pipelines, characterized in that, Includes the following steps: (1) Construct a nonlinear model of the UAV that considers unknown external interference, and construct a nonlinear interference observer to estimate the unknown external interference; (2) Obtain the trajectory of the leader drone in the drone swarm. Based on the relative distance between each follower drone and the leader drone, obtain the desired trajectory of each follower drone. Construct a virtual pipeline with the desired trajectory of the follower drone as the axis. By constructing an artificial potential field region inside the virtual pipeline, each follower drone is confined to the corresponding virtual pipeline. The artificial potential field region includes a repulsive field, the expression of which is: ； ； ； in, In the first drone A repulsive field along the axial direction. , , , These are the three coordinate axes of the spatial coordinate system. For the first The maximum distance between the center and the boundary of the virtual pipeline constructed by the drone. Let be the radius of the virtual pipe. This represents the outermost distance of the virtual pipeline's upper boundary. This represents the outermost distance of the lower boundary of the virtual pipe. Indicates the first The radius of the safe zone within the virtual pipeline constructed by the drone. For the first The desired trajectory of the drone. For the first Tracking error of the drone For the first A drone in The position of the axis; The boundary of the virtual pipeline is adjusted in real time using a positive signal generated by a self-adjusting auxiliary system. The boundary expression of the virtual pipeline is as follows: ； ； ； in, This represents the upper boundary of the virtual pipe. This represents the lower boundary of the virtual pipe. and For the state variables of the self-regulating auxiliary system, This represents the width of the artificial potential field region within the virtual pipe; The expression for the positive signal generated by the self-regulating auxiliary system is: ； in, and Indicates in virtual pipeline State variables of a self-adjusting auxiliary system under boundary conditions Indicates in virtual pipeline The system input of the self-adjusting auxiliary system under boundary conditions. and Indicates in virtual pipeline System parameters of the self-adjusting auxiliary system under boundary conditions. , This indicates the upper boundary condition of the virtual pipeline. Indicates the lower boundary condition of the virtual pipe; (3) Design of a formation controller based on a nonlinear disturbance observer and a virtual pipeline; (4) Control the drone swarm through the constructed formation controller.

2. The method for collaborative security control of unmanned aerial vehicle swarms based on virtual pipelines according to claim 1, characterized in that, The nonlinear model of the UAV is: ； in, For the first The location of the drone For the first The speed of the drone For the first The nonlinear term of the drone model, For the first The system control input for the drone, For the first Unknown external interference with the drone For the first The adjustment matrix of the drone.

3. The method for collaborative safety control of unmanned aerial vehicle swarms based on virtual pipelines according to claim 2, characterized in that, The nonlinear disturbance observer is: ； in, For the first Observations of unknown external interference from unmanned aerial vehicles (UAVs). as auxiliary variables The estimated value, It is a parameter matrix.

4. The method for collaborative safety control of unmanned aerial vehicle swarms based on virtual pipelines according to claim 3, characterized in that, The auxiliary variable The expression is: ； Auxiliary variables The update law is: 。 5. The method for collaborative safety control of unmanned aerial vehicle swarms based on virtual pipelines according to claim 4, characterized in that, The formation controller is: ； ； ； in, The adjustment matrix for all follower drones, , and For the parameter matrix, For formation tracking error, To address formation tracking errors Find the first derivative. For the nonlinear term of the model for all follower drones, Estimates of unknown external interference for all follower drones. For the trajectory of the leader's drone Find the second derivative. The parameter matrix to be designed, The repulsive force experienced by all follower drones. and As an auxiliary variable, For the speed error of all follower drones, These are design parameters.

6. The method for collaborative safety control of unmanned aerial vehicle swarms based on virtual pipelines according to claim 5, characterized in that, The repulsive force experienced by all follower drones is: ； ； in, The repulsive force experienced by the first drone Represents the transpose of a matrix. The repulsive force experienced by the second drone For the first The repulsive force experienced by the drone For the first A drone in The repulsive force in the axial direction.

7. A computer 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 the steps of the method according to any one of claims 1 to 6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images