Multi-six-rotor unmanned aerial vehicle control method and system based on finite time command filter

Abstract

The application provides a multi-six-rotor unmanned aerial vehicle control method and system based on finite time command filtering, and relates to the field of unmanned aerial vehicle control. The application comprises the following steps: establishing a six-rotor unmanned aerial vehicle attitude system dynamics model; establishing state equations of a multi-six-rotor unmanned aerial vehicle leader and follower; modeling an asymmetric nonlinear input dead zone in the follower state equation, and compensating for the error of the nonlinear input dead zone by using the boundedness of the dead zone slope; designing an error compensation signal based on finite time command filtering, compensating for the synchronization error by using the error compensation signal, and approximating unknown nonlinear terms in the multi-six-rotor unmanned aerial vehicle attitude system by using a radial basis function neural network; and designing a virtual controller, an actual controller and an adaptive law of the multi-six-rotor unmanned aerial vehicle under the nonlinear input dead zone. The application realizes finite time consistency control of the multi-six-rotor unmanned aerial vehicle attitude system, compensates for the nonlinear input dead zone of the system by using the boundedness of the dead zone slope, and achieves better control effect.

Description

A Control Method and System for Multi-Sixrotor UAVs Based on Finite-Time Command Filtering Technical Field

[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) control technology, and particularly relates to a control method and system for multi-hexarotor UAVs based on finite-time command filtering. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] In recent years, UAV technology has developed rapidly. Hexacopter UAVs have advantages such as ease of operation, vertical takeoff and landing, simple structure, and hovering capability. Furthermore, with six drive motors, they possess good payload capacity and stability. Therefore, hexacopter UAVs are widely used in military, industrial, and agricultural fields, such as reconnaissance, aerial photography, pesticide spraying, and emergency rescue. However, as mission complexity increases, a single hexacopter UAV has certain limitations and cannot meet complex mission requirements. Compared to a single hexacopter UAV, multi-hexacopter UAVs offer better scalability, flexibility, and hierarchy, effectively solving large-scale complexity problems. In the attitude consistency control of multi-hexacopter UAVs, the leader and followers, and followers among themselves, communicate and coordinate with each other through a network topology. However, the attitude system of a multi-hexacopter UAV is also a typical underactuated system, exhibiting strong coupling and strong nonlinearity, which makes the design of attitude consistency control for multi-hexacopter UAVs extremely difficult.

[0004] For decades, researchers have largely focused on how to achieve faster convergence speeds for systems. The fastest convergence speed is exponential, but this does not yield good convergence performance; this type of control falls under the category of infinite-time control. In the field of multi-hexrotor UAV control, convergence performance is also a very important metric, and finite-time control possesses optimal convergence and stronger fault tolerance than infinite-time control.

[0005] With the development of control technology, some advanced control methods, including neural network control, backpropagation control, and sliding mode control, have achieved certain results in the control of multi-six-rotor UAVs. However, most of these methods neglect the issues of human-machine interaction and input dead zone in practical applications. In actual engineering, multi-six-rotor UAVs require high precision. When performing tasks, human operators are still needed to correct some command signals to complete complex tasks, such as obstacle avoidance and complex route movements. Human operators play a crucial role in the task execution process of multi-six-rotor UAVs. Input dead zone refers to the range of input signals where the corresponding output is zero. In practical applications, the unknown input dead zone of the multi-six-rotor UAV attitude system has a significant impact on control accuracy. Ignoring the input dead zone in the control method design may even lead to instability in the control system. Therefore, effectively solving the problems of human-machine interaction and input dead zone is a key issue in the design of attitude consistency control methods for multi-six-rotor UAVs.

[0006] In the design of attitude consistency control methods for multi-hexagonal UAVs, there exists a class of unknown smooth nonlinear functions. Radial basis function neural networks are used to approximate these unknown smooth nonlinear functions, and then backstepping is employed to design an effective control method. However, to address the "complexity explosion" problem in backstepping control, a new and effective solution is needed. Summary of the Invention

[0007] To overcome the shortcomings of the prior art, this invention provides a control method and system for a multi-six-rotor UAV based on finite-time command filtering. It comprehensively considers the human-machine interaction, nonlinear input dead zone, and finite-time command filtering control of the multi-six-rotor UAV, achieving finite-time consistency control of the attitude system. The boundedness of the dead zone slope compensates for the system's nonlinear input dead zone. A radial basis function neural network is used to approximate the unknown smooth nonlinear function in the system, and accurate approximation is achieved by automatically adjusting the weights. A finite-time command filtering is designed to solve the "complexity explosion" problem, resulting in better control performance.

[0008] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions:

[0009] The first aspect of this invention provides a control method for a multi-hexarotor unmanned aerial vehicle based on finite-time command filtering.

[0010] The control method for a multi-hexarotor UAV based on finite-time command filtering includes the following steps:

[0011] Establish a dynamic model of the attitude system of a six-rotor UAV;

[0012] Based on the dynamic model of the attitude system of a hexacopter UAV, state equations for the leader and follower of a multi-hexacopter UAV are established.

[0013] The asymmetric nonlinear input dead zone in the follower state equation is modeled, and the boundedness of the dead zone slope is used to compensate for the error of the nonlinear input dead zone.

[0014] The synchronization error is defined based on the node information of adjacent hexacopter UAVs, the error compensation signal is designed based on finite-time command filtering, the synchronization error is compensated by the error compensation signal, and the unknown nonlinear terms in the attitude system of the multi-hexacopter UAV are approximated by radial basis function neural network.

[0015] Based on the backstepping control framework of finite-time command filtering, a virtual controller, an actual controller, and an adaptive law are designed for a multi-hexrotor UAV with nonlinear input dead zone.

[0016] The second aspect of the present invention provides a control system for a multi-hexarotor unmanned aerial vehicle based on finite-time command filtering.

[0017] A multi-hexarotor UAV control system based on finite-time command filtering includes:

[0018] The dynamics model building module is configured to: build a dynamics model of the attitude system of a six-rotor UAV;

[0019] The state equation establishment module is configured to: establish the state equations of the leader and follower of the multi-hexacopter UAV based on the dynamic model of the attitude system of the hexacopter UAV;

[0020] The nonlinear input dead zone error compensation module is configured to: model the asymmetric nonlinear input dead zone in the follower state equation and compensate for the error of the nonlinear input dead zone by utilizing the boundedness of the dead zone slope;

[0021] The synchronization error compensation and unknown nonlinear term approximation module is configured to: define the synchronization error based on the node information of adjacent hexacopter UAVs, design the error compensation signal based on finite-time command filtering, use the error compensation signal to compensate for the synchronization error, and use a radial basis function neural network to approximate the unknown nonlinear term in the attitude system of the multi-hexacopter UAV.

[0022] The virtual controller, actual controller, and adaptive law design module are configured to design the virtual controller, actual controller, and adaptive law for a multi-hexagonal UAV under nonlinear input dead zone based on a backstepping control framework using finite-time command filtering.

[0023] A third aspect of the present invention provides a computer-readable storage medium having a program stored thereon that, when executed by a processor, implements the steps of the multi-hexarotor unmanned aerial vehicle control method based on finite-time command filtering as described in the first aspect of the present invention.

[0024] The fourth aspect of the present invention provides an electronic device including a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in the multi-hexarotor unmanned aerial vehicle control method based on finite-time command filtering as described in the first aspect of the present invention.

[0025] The above one or more technical solutions have the following beneficial effects:

[0026] 1. This invention proposes a backstepping control framework based on finite-time command filtering, and designs a virtual controller, an actual controller, and an adaptive law for a multi-hexacopter UAV, thereby realizing finite-time consistency control of the attitude system of the multi-hexacopter UAV.

[0027] 2. For a type of attitude system for multi-hexrotor UAVs with nonlinear input dead zones in practical applications, the boundedness of the dead zone slope is used to compensate for the nonlinear input dead zone. A radial basis function neural network is used to approximate the unknown smooth nonlinear function in the system, and accurate approximation is achieved by automatically adjusting the weights, resulting in better control performance.

[0028] 3. Finite-time command filtering technology is introduced into the backstepping method to avoid the "complexity explosion" problem. Finally, simulation examples demonstrate the effectiveness of the proposed control method, verifying that a multi-hexarotor UAV achieves leader-follower consistency control in a finite time. Compared with existing technologies, this invention has advantages such as higher accuracy, better convergence performance, and stronger fault-tolerant control.

[0029] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0030] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0031] Figure 1 is a flowchart of the method according to the first embodiment of the present invention.

[0032] Figure 2 shows the Euler angles of the attitude system of the six-rotor UAV in the first embodiment of the present invention.

[0033] Figure 3 shows the input dead zone of the attitude system of the multi-hexarotor UAV in the first embodiment of the present invention.

[0034] Figure 4 is a directed communication topology diagram of the first embodiment of the multi-hexarotor UAV of the present invention.

[0035] Figure 5 is a schematic diagram of the output trajectory of the multi-hexarotor UAV according to the first embodiment of the present invention.

[0036] Figure 6 shows the output signal of the finite-time command filter of the multi-hexacopter UAV according to the first embodiment of the present invention. and virtual control signals A diagram illustrating the error between them.

[0037] Figure 7 is a schematic diagram of the control input for a follower of a multi-hexagonal UAV according to the first embodiment of the present invention.

[0038] Figure 8 shows the synchronization error of the multi-hexarotor UAV according to the first embodiment of the present invention. Schematic diagram.

[0039] Figure 9 is a system structure diagram of the second embodiment of the present invention. Detailed Implementation

[0040] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0041] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.

[0042] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0043] The overall concept proposed in this invention is as follows:

[0044] This invention addresses the attitude control system of a human-in-the-loop multi-hexrotor UAV with a nonlinear input dead zone, proposing a control method and system based on finite-time command filtering. The overall control flow is as follows: First, the state equations of the multi-hexrotor UAV are designed based on the dynamic model of the attitude system, and the asymmetric nonlinear input dead zone of the attitude system is modeled. Then, a finite-time command filter is designed, and an error compensation signal is designed based on the finite-time command filter. The error compensation signal is used to compensate for synchronization errors, and a radial basis function neural network is used to approximate the unknown smooth nonlinear function in the system. Finally, a virtual controller, an actual controller, and an adaptive law are designed based on the backstepping control framework of the finite-time command filter. The Lyapunov function proves that all signals in the system are semi-globally uniform and bounded, and the system achieves leader-follower consistency control within a finite time.

[0045] Example 1

[0046] This embodiment discloses a control method for a multi-hexarotor unmanned aerial vehicle based on finite-time command filtering.

[0047] The control method for a multi-hexarotor UAV based on finite-time command filtering includes the following steps:

[0048] Establish a dynamic model of the attitude system of a six-rotor UAV;

[0049] Based on the dynamic model of the attitude system of a hexacopter UAV, state equations for the leader and follower of a multi-hexacopter UAV are established.

[0050] The asymmetric nonlinear input dead zone in the follower state equation is modeled, and the boundedness of the dead zone slope is used to compensate for the error of the nonlinear input dead zone.

[0051] The synchronization error is defined based on the node information of adjacent hexacopter UAVs, the error compensation signal is designed based on finite-time command filtering, the synchronization error is compensated by the error compensation signal, and the unknown nonlinear terms in the attitude system of the multi-hexacopter UAV are approximated by radial basis function neural network.

[0052] Based on the backstepping control framework of finite-time command filtering, a virtual controller, an actual controller, and an adaptive law are designed for a multi-hexrotor UAV with nonlinear input dead zone.

[0053] As shown in Figure 1, a control method for a multi-hexcopter UAV based on finite-time command filtering is proposed. This method considers a hexcopter leader and M hexcopter followers. The follower system is affected by a nonlinear input dead zone, while the leader system is controlled by a human. The hexcopter leader is denoted as 0, and the hexcopter followers are denoted as i = 1, ..., M. The method includes the following steps:

[0054] S1: Establishing a dynamic model of the attitude system of a six-rotor UAV:

[0055] S1.1: Define the body coordinate system ε of the hexacopter UAV i,b ={O i,b ,x i,b ,y i,b ,z i,b} and Earth coordinate system ε i,e ={O i,e ,x i,e ,y i,e ,z i,e The dynamic model of the attitude system of the six-rotor UAV is established as follows:

[0056]

[0057] In the above formula, This indicates that the hexacopter drone is in ε i,e Euler angles (roll, pitch, yaw), δ i εi,e and ε i,b The transformation matrix between them, ω i =[ω i,x ,ω i,y ,ω i,z ] T ε i,b The angular velocity of the center of gravity in the middle.

[0058] S1.2: The relationship equations between the system's moment of inertia, rotational moment, and gyroscopic moment are as follows:

[0059]

[0060] In the above formula, I i =diag{I i,x ,I i,y ,I i,z} is the moment of inertia. It is the rotational moment. It is the gyroscope moment.

[0061] S1.3: The dynamic model of the attitude system of the six-rotor UAV is rewritten as follows:

[0062]

[0063] S2: Based on the physical characteristics of hexacopter drones and combined with the dynamic model of the hexacopter drone attitude system, establish the state equations for the leader and followers of the multi-hexacopter drone:

[0064] S2.1: The state equation for the leader of the multi-hexacopter UAV is established as follows:

[0065]

[0066] In the above formula, j = 1, 2, 3, These represent the Euler angles and the angular velocity of the center of gravity of the leader of the six-rotor drones, respectively. Let represent the reciprocals of the moments of inertia in the x, y, and z directions of a known hexacopter drone leader, respectively. 0,j It is the output signal of the leader of the attitude system. It is an unknown nonlinear term in the system, u 0,j It is a non-zero command signal input by a human operator in the attitude system of a multi-rotor UAV.

[0067] S2.2: The state equation for the i-th hexacopter UAV follower with a nonlinear input dead zone is:

[0068]

[0069] In the above formula, j = 1, 2, 3, This indicates the state of a hexacopter drone follower (the follower's Euler angles and angular velocity at the center of gravity). Let represent the reciprocals of the moments of inertia in the x, y, and z directions for a known hexacopter UAV follower, respectively. i,j It is the output signal of the i-th hexacopter UAV attitude system follower. and It is an unknown nonlinear term in the system, u i,j This represents the asymmetric nonlinear input dead zone of the system.

[0070] S3: As shown in Figure 3, the asymmetric nonlinear input dead zone of the attitude system of the multi-hexarotor UAV is modeled, and the boundedness of the dead zone slope is used to compensate for the error of the nonlinear input dead zone.

[0071] S3.1: The asymmetric nonlinear input dead zone model of the attitude system of the multi-hexarotor UAV is established as follows:

[0072]

[0073] Where L i,j (t) is:

[0074]

[0075] λ i,j (t) is:

[0076]

[0077] In the above formula, L represents the control signal indicating the nonlinear input dead zone of the system. i,jr L i,jl ,d i,jr -d i,jl These are the right slope, left slope, right breakpoint, and left breakpoint of the dead-zone nonlinear function. It is a positive scalar that satisfies λ i,j (t) satisfies

[0078] S3.2: Compensating for errors in the dead zone of nonlinear inputs by utilizing the boundedness of the input slope in the dead zone:

[0079] Definition: H i,j =max{L i,jr ,L i,jl}, ci,j =min{L i,jr ,L i,jl}, Among them B i,j (t)≥0,

[0080] The asymmetric nonlinear input dead-zone model of the system is written as follows:

[0081]

[0082] S4: The leader-follower communication topology of the multi-hexacopter UAV is shown in Figure 4. Based on the node information of adjacent hexacopter UAVs, the synchronization error and coordinate transformation are defined. An error compensation signal is designed based on finite-time command filtering. The synchronization error is compensated using the error compensation signal. The unknown smooth nonlinear function in the system is approximated using a radial basis function neural network.

[0083] S4.1: In a multi-hexacopter UAV attitude system, communication between the leader and followers is represented by a directed communication topology. express, represents a point set, It is a node in the topology graph, representing the i-th multi-hexagonal drone follower. Denotes the edge set, Ξ=[a i,l ]∈R F×F Let represent the adjacency matrix. If follower i can receive information from follower l, then follower l is called a neighbor node of follower i. i,l >0, otherwise a i,l =0, where l = 1, ..., M, and l ≠ i.

[0084] The leader adjacency matrix is ​​defined as Ξ0 = diag[b1, b2, ..., bb2]. M When the i-th follower is able to receive the leader's information, b i >0, otherwise b i =0.

[0085] S4.2: Based on the information of adjacent hexacopter UAV nodes, within the framework of finite-time command filtering and backstepping control design, the synchronization error is defined as follows:

[0086]

[0087] In the above formula, y l,j It is the output signal of the l-th hexacopter UAV attitude system follower, and l≠i. It is the output signal of a finite-time command filter.

[0088] S4.3: The finite-time command filter design is as follows:

[0089]

[0090]

[0091] In the above formula, It is a virtual controller. These are design parameters. It is in S4.2 The derivative of .

[0092] S4.4: The design compensates for the tracking error as follows: and in and This represents the error compensation signal, which satisfies the following relationship:

[0093]

[0094] In the above formula, γ i satisfy It is a positive constant. and These are positive design parameters.

[0095] S4.5: Using radial basis function neural networks to approximate unknown smooth nonlinear functions in the system.

[0096] The radial basis function neural network is designed as follows:

[0097]

[0098] In the above formula, k = 1, 2, This is the input to the radial basis function neural network. This represents the weight vector of a radial basis function neural network. Let Gausky function be denoted as , where The center vector of the hidden layer neurons. denoted as the width of the Gaussian function.

[0099] S4.6: Based on steps S2.2, S4.1, S4.2, S4.3, and S4.4, the synchronization error can be written as:

[0100]

[0101] Define an unknown smooth nonlinear function Approximating an unknown smooth nonlinear function using a radial basis function neural network.

[0102]

[0103] Therefore, we can conclude that:

[0104]

[0105] In the above formula, This is the input to the radial basis function neural network. Let r represent the weight vector of the radial basis function neural network, where r represents the number of radial basis function neural networks. These are the basis functions of a radial basis function neural network. Indicates the approximation error. It is a constant greater than 0.

[0106] Similarly, we can obtain

[0107]

[0108] In the above formula, This is the input to the radial basis function neural network. Let r represent the weight vector of the radial basis function neural network, and r represent the number of radial basis function neural networks. These are the basis functions of a radial basis function neural network. Represents the approximation error, satisfying It is a constant greater than 0.

[0109] S5: Based on a backstepping control framework using finite-time command filtering, the virtual controller, actual controller, and adaptive law for a multi-hexagonal UAV under nonlinear input dead zone are designed as follows:

[0110]

[0111]

[0112]

[0113] In the above formula, k = 1, 2, Indicates a virtual controller. Indicates the actual controller, This represents the adaptive law. and All design parameters are positive. It is the weight vector of the radial basis function neural network. The estimated value is 0 < μ < 1.

[0114] Virtual Controller It is a control input in the control design process of the follower subsystem. In traditional control methods, the synchronization error is defined as... in Representing a virtual controller, and then utilizing synchronization errors. and error compensation signal The tracking error was obtained. Then, the Lyapunov function is designed, and the stability of the system is proven through the Lyapunov function. This method will cause the "complexity explosion" problem.

[0115] To avoid the "complexity explosion" problem in traditional control methods, this invention designs a finite-time command filter. To approximate the virtual controller Redefining synchronization error in The output signal of the finite-time command filter, virtual controller It was only introduced in S4.6, which led to the conclusion that The expression is used to design a virtual controller. Using the function expression and the Lyapunov function designed below, we prove the stability of the system. This method can effectively solve the "complexity explosion" problem in traditional control methods.

[0116] Actual controller Essentially, the system follows the control signal with a nonlinear input dead zone, and the design of the actual controller is required. The function expression will be used to design the actual controller. Prove the stability of the system using the Lyapunov function designed below.

[0117] Adaptive law These are the weight vectors of the radial basis function neural network in S4.5 and S4.6. The derivative of, where k = 1, 2, is obtained through the adaptive law designed in S5. The function expression can continuously adjust the weight vector of the radial basis function neural network. The value is updated to achieve a more accurate approximation of unknown smooth nonlinear functions in the system by the radial basis function neural network.

[0118] Prove tracking error and adaptive error Design the first Lyapunov function that is bounded in finite time:

[0119]

[0120] By differentiating the Lyapunov function in the above equation, we obtain:

[0121]

[0122] Combined with the virtual controllers designed in S4.4, S4.6, and S5 Actual controller Adaptive rate Calculation yields:

[0123]

[0124] In the above formula, Where 0 < δ < 1.

[0125] There exists a constant satisfy Make It can be written in the following two forms:

[0126] Format 1:

[0127]

[0128] Form 2:

[0129]

[0130] According to form 1, if but By reducing V, the tracking error can be reduced. and adaptive error It converges to the region within a finite time T1. The finite time T1 satisfies:

[0131]

[0132] According to form 2, if but By reducing V, the tracking error can be reduced. and adaptive error It converges to the region within a finite time T2. The finite time T2 satisfies:

[0133]

[0134] In summary, the tracking error Will be within a limited time Convergence to

[0135] As shown in Figure 6, there exists a positive parameter o. i,j It can be achieved within a limited time. Prove the error compensation signal Given that the time interval is bounded, we can design the second Lyapunov function as follows:

[0136]

[0137] Calculations show that:

[0138]

[0139] In the above formula, Through design Making Λ-N>0, the error compensation mechanism reaches stability in a finite time T3, which satisfies:

[0140]

[0141] In summary, the synchronization error It converges to the maximum {T1,T3} in a finite time interval t ≥ max{T1,T3}.

[0142] Based on Lyapunov's stability theory, by designing parameters It can make all signals of the system semi-globally consistent and bounded, and realize finite-time consistency control of the attitude system of a multi-hexarotor UAV.

[0143] Simulation experiment:

[0144] The control objective of this invention is to enable a follower in a human-in-the-loop multi-hexcopter UAV attitude system with a nonlinear input dead zone to track the leader's attitude trajectory within a finite time, achieving leader-follower attitude consistency control. The leader's input signal u in the multi-hexcopter UAV attitude system... 0,j Provided by a human operator, with the following conditions:

[0145] When \(0\leq t\leq20s\), u 0,θ \( = 0.35\sin(0.5t)\), \(u\) 0,ψ \( = 0.1\cos(0.85t)\);

[0146] When \(20s\lt t\leq40s\), u 0,θ \( = 0.35\sin(10)+0.75\sin(0.1\times(t - 2))\cos(0.1\times(t - 2))\), \(u\) 0,ψ \( = 0.1\cos(17)+0.5\sin(0.5\times(t - 2))\);

[0147] When \(40s\lt t\), u 0,θ \( = 0.35\sin(10)+0.75\sin(2)\cos(2)+0.5\sin(t - 40)\), \(u\) 0,ψ \( = 0.1\cos(10)+0.5\sin(10)+0.25\sin(t - 40)\cos(t - 40)\).

[0148] The inertial parameters are:

[0149] I 1,x \( = I\) 2,x \( = I\) 3,x \( = I\) 0,x \( = I\) 1,y \( = I\) 2,y \( = I\) 3,y \( = I\) 0,y \( = 1.2416\),

[0150] I 1,z \( = I\) 2,z \( = I\) 3,z \( = I\) 0,z \( = 2.4832\);

[0151] The parameter design of the non - linear input dead zone is as follows:

[0152] L i,1r \( = L\) i,2r \( = L\) i,3r \( = 2.5\), \(L\) i,1l \( = L\) i,2l \( = L\) i,3l \( = 2\);

[0153] d i,1r \( = d\) i,2r \( = d\) i,3r \( = d\) i,1l \( = d\) i,2l \( = d\) i,3l=5;

[0154]

[0155] Other parameters are designed as follows:

[0156]

[0157]

[0158]

[0159]

[0160]

[0161]

[0162] Simulation results are shown in Figures 5-8. Figure 5 shows the output trajectory diagram, indicating that the multi-hexagonal UAV follower can track the given leader trajectory within a finite time. Figure 6 shows the virtual control signal... The output signal can be filtered by command within a finite time. As estimated in Figure 7, the follower's control input is bounded within a finite time interval, as shown in Figure 8 regarding the synchronization error. As shown in the schematic diagram, the synchronization error is between -0.1 and 0.05 for the first two seconds, and then the synchronization error approaches 0, and gets closer to 0 the longer the time. Therefore, the simulation example proves the effectiveness of the control method proposed in this invention.

[0163] In summary, the present invention has the following advantages:

[0164] (1) Establish a nonlinear input dead zone model and use the boundedness of the dead zone slope to compensate for the error of the nonlinear input dead zone. For the unknown smooth nonlinear function in the system, use a radial basis function neural network to approximate it. Accurate approximation is achieved by continuously updating the weights. This method has the advantages of high accuracy and wide applicability.

[0165] (2) By introducing the finite-time command filtering technique into the backstepping method and using the output of the command filter to estimate the virtual control signal, the "complexity explosion" problem can be effectively solved. Compared with infinite-time control, the finite-time control designed in this invention has the advantages of better convergence performance and stronger fault-tolerant control.

[0166] Example 2

[0167] This embodiment discloses a control system for a multi-hexarotor unmanned aerial vehicle based on finite-time command filtering.

[0168] As shown in Figure 9, the control system for a multi-hexarotor UAV based on finite-time command filtering includes:

[0169] The dynamics model building module is configured to: build a dynamics model of the attitude system of a six-rotor UAV;

[0170] The state equation establishment module is configured to: establish the state equations of the leader and follower of the multi-hexacopter UAV based on the dynamic model of the attitude system of the hexacopter UAV;

[0171] The nonlinear input dead zone error compensation module is configured to: model the asymmetric nonlinear input dead zone in the follower state equation and compensate for the error of the nonlinear input dead zone by utilizing the boundedness of the dead zone slope;

[0172] The synchronization error compensation and unknown nonlinear term approximation module is configured to: define the synchronization error based on the node information of adjacent hexacopter UAVs, design the error compensation signal based on finite-time command filtering, use the error compensation signal to compensate for the synchronization error, and use a radial basis function neural network to approximate the unknown nonlinear term in the attitude system of the multi-hexacopter UAV.

[0173] The virtual controller, actual controller, and adaptive law design module are configured to design the virtual controller, actual controller, and adaptive law for a multi-hexagonal UAV under nonlinear input dead zone based on a backstepping control framework using finite-time command filtering.

[0174] Example 3

[0175] The purpose of this embodiment is to provide a computer-readable storage medium.

[0176] A computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the steps in the multi-hexarotor unmanned aerial vehicle control method based on finite-time command filtering as described in Embodiment 1 of this disclosure.

[0177] Example 4

[0178] The purpose of this embodiment is to provide an electronic device.

[0179] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in the multi-hexarotor unmanned aerial vehicle control method based on finite-time command filtering as described in Embodiment 1 of this disclosure.

[0180] The steps and methods involved in the apparatuses of Embodiments 2, 3, and 4 above correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0181] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0182] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A control method for a multi-hexrotor unmanned aerial vehicle based on finite-time command filtering, characterized in that, The process includes the following steps: establishing a dynamic model of the attitude system of a hexacopter UAV; and based on the dynamic model of the attitude system of the hexacopter UAV, establishing state equations for the leader and followers of the multi-hexacopter UAV. The asymmetric nonlinear input dead zone in the follower state equation is modeled, and the boundedness of the dead zone slope is used to compensate for the error of the nonlinear input dead zone. The synchronization error is defined based on the node information of adjacent hexacopter UAVs, the error compensation signal is designed based on finite-time command filtering, the synchronization error is compensated by the error compensation signal, and the unknown nonlinear terms in the attitude system of the multi-hexacopter UAV are approximated by radial basis function neural network. Based on a backstepping control framework using finite-time command filtering, a virtual controller, an actual controller, and an adaptive law are designed for a multi-hexrotor UAV with a nonlinear input dead zone. The asymmetric nonlinear input dead zone model of the system is written as follows: in, The control signal representing the nonlinear input dead zone of the system. ， ， 、 、 、 These are the right slope, left slope, right breakpoint, and left breakpoint of the dead-zone nonlinear function, respectively. satisfy ， In a multi-hexacopter UAV attitude system, communication between the leader and followers is represented by a directed communication topology. express, Display points collection, It is a node in the topological graph, representing the first node. More than 100 six-rotor drone followers Represents an edge set. Represents the adjacency matrix, when the follower Can receive from followers The message sent out will be followed by... called followers The neighboring nodes, ,otherwise ,in, ,and The leader adjacency matrix is ​​defined as follows: When the first When a follower is able to receive information from the leader ,otherwise Based on the information of adjacent six-rotor UAV nodes, and within the framework of finite-time command filtering and backstepping control design, the synchronization error is defined as: in, It is the first The output signal of the attitude system follower of a six-rotor UAV It is the first The output signal of the attitude system follower of a six-rotor UAV It is the output signal of the leader of the attitude system. This indicates the status of a hexacopter drone follower. This is the output signal of the finite-time command filter; the finite-time command filter design is as follows: in, It is a virtual controller. ， These are the design parameters; the radial basis function neural network design is as follows: in, ， This is the input to the radial basis function neural network. This represents the weight vector of a radial basis function neural network. Let Gausky function be denoted as , where The center vector of the hidden layer neurons. denoted as the width of the Gaussian function.

2. The control method for a multi-hexacopter UAV based on finite-time command filtering as described in claim 1, characterized in that, Define the body coordinate system, the Earth coordinate system, and the transformation matrix between the body coordinate system and the Earth coordinate system for a hexacopter UAV. Based on the relationship equations of moment of inertia, rotational moment, and gyroscopic moment, establish a dynamic model of the attitude system of the hexacopter UAV.

3. The multi-hexarotor UAV control method based on finite-time command filtering as described in claim 1, characterized in that, Based on the physical characteristics of hexacopter UAVs and combined with the dynamic model of the hexacopter UAV attitude system, state equations for the leader and follower of the multi-hexacopter UAV are established. The state equations of both the leader and the follower contain unknown nonlinear terms, and the state equation of the follower contains an asymmetric nonlinear input dead zone.

4. The control method for a multi-hexarotor UAV based on finite-time command filtering as described in claim 1, characterized in that, The virtual controller, the actual controller, and the adaptive law are respectively: in, ， Indicates a virtual controller. Indicates the actual controller, This represents the adaptive law. 、 and All design parameters are positive. ， It is the weight vector of the radial basis function neural network. The estimated value, ； and This is the synchronization error; and To compensate for tracking errors; This refers to known hexacopter drone followers. ， ， The reciprocal of the moment of inertia in the direction; and These are positive design parameters.

5. A control system for a multi-hexacopter unmanned aerial vehicle based on finite-time command filtering, characterized in that: include: The dynamics model building module is configured to: build a dynamics model of the attitude system of a hexacopter UAV; the state equation building module is configured to: build state equations for the leader and followers of the multi-hexacopter UAV based on the dynamics model of the attitude system of the hexacopter UAV. The nonlinear input dead zone error compensation module is configured to: model the asymmetric nonlinear input dead zone in the follower state equation and compensate for the error of the nonlinear input dead zone by utilizing the boundedness of the dead zone slope; The synchronization error compensation and unknown nonlinear term approximation module is configured to: define the synchronization error based on the node information of adjacent hexacopter UAVs, design an error compensation signal based on finite-time command filtering, compensate the synchronization error using the error compensation signal, and approximate the unknown nonlinear term in the attitude system of the multi-hexacopter UAV using a radial basis function neural network; the virtual controller, actual controller, and adaptive law design module is configured to: design the virtual controller, actual controller, and adaptive law of the multi-hexacopter UAV under nonlinear input dead zone based on the backstepping control framework of finite-time command filtering; and write the asymmetric nonlinear input dead zone model of the system as follows: in, The control signal representing the nonlinear input dead zone of the system. ， ， 、 、 、 These are the right slope, left slope, right breakpoint, and left breakpoint of the dead-zone nonlinear function, respectively. satisfy ， In a multi-hexacopter UAV attitude system, communication between the leader and followers is represented by a directed communication topology. express, Display points collection, It is a node in the topological graph, representing the first node. More than 100 six-rotor drone followers Represents an edge set. Represents the adjacency matrix, when the follower Can receive from followers The message sent out will be followed by... called followers The neighboring nodes, ,otherwise ,in, ,and The leader adjacency matrix is ​​defined as follows: When the first When a follower is able to receive information from the leader ,otherwise Based on the information of adjacent six-rotor UAV nodes, and within the framework of finite-time command filtering and backstepping control design, the synchronization error is defined as: in, It is the first The output signal of the attitude system follower of a six-rotor UAV It is the first The output signal of the attitude system follower of a six-rotor UAV It is the output signal of the leader of the attitude system. This indicates the status of a hexacopter drone follower. This is the output signal of the finite-time command filter; the finite-time command filter design is as follows: in, It is a virtual controller. ， These are the design parameters; the radial basis function neural network design is as follows: in, ， This is the input to the radial basis function neural network. This represents the weight vector of a radial basis function neural network. Let Gausky function be denoted as , where The center vector of the hidden layer neurons. denoted as the width of the Gaussian function.

6. A computer-readable storage medium having a program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps in the multi-hexarotor UAV control method based on finite-time command filtering as described in any one of claims 1-4.

7. An electronic device, comprising a memory, a processor, and a program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the multi-hexarotor unmanned aerial vehicle control method based on finite-time command filtering as described in any one of claims 1-4.

Images