Quadrotor unmanned aerial vehicle affine formation control method and system, electronic device and medium

Abstract

The application discloses a quad-rotor unmanned aerial vehicle affine formation control method and system, electronic equipment and medium, and relates to the technical field of unmanned aerial vehicle flight control. The method comprises the following steps: determining a leader, a follower and a communication topology structure diagram in a target unmanned aerial vehicle formation according to affine transformation theory; determining a symbolic Laplace matrix and an entry neighbor node set of each unmanned aerial vehicle according to the diagram; if the leader is a constant acceleration, obtaining a control force according to the matrix, the follower and the physical parameters transmitted by each unmanned aerial vehicle in the set and a first position controller; if the leader is a time-varying acceleration, obtaining the control force according to the matrix, the estimated parameter of each unmanned aerial vehicle in the set, a state observer, the physical parameters and a second position controller; obtaining a control torque according to the control force and an adaptive sliding film attitude controller; and controlling the follower by using the control force and the control torque. The application enables the unmanned aerial vehicle cluster to have stronger anti-interference ability in formation and realize more flexible formation maneuvering.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) flight control technology, and in particular to a method, system, electronic device, and medium for controlling the affine formation of a quadcopter UAV. Background Technology

[0002] The continuous development of multiple disciplines such as control science, computer technology, artificial intelligence, and communication technology has created favorable research conditions and a solid foundation for communication and collaboration among multiple unmanned aerial vehicle (UAV) systems. Currently, the formation control of multiple UAV systems has a wide range of applications, enabling intelligent logistics management in the civilian sector and reconnaissance, positioning, and ground strike capabilities in the military sector. Therefore, UAV formation control research has become a popular research direction. Quadcopter UAVs are currently the most common type of rotary-wing UAV. Due to their ease of operation, simple structure, and superior performance, quadcopter multi-UAV systems are often chosen as research platforms for UAV formation control.

[0003] Current distributed formation control methods based on quadcopter UAVs mostly suffer from insufficient flexibility, thus limiting their application to more extreme environments. Furthermore, input saturation, parameter uncertainty, and environmental disturbances can all affect UAV flight stability. To enhance the anti-interference capabilities and enable more flexible formation maneuvers in UAV formation flight, a new UAV formation control method is needed. Summary of the Invention

[0004] The purpose of this invention is to provide a method, system, electronic device and medium for controlling the affine formation of quadcopter UAVs, which enables UAV swarms to have stronger anti-interference capabilities and achieve more flexible formation maneuvers.

[0005] To achieve the above objectives, the present invention provides the following solution:

[0006] A method for controlling the affine formation of a quadcopter unmanned aerial vehicle (UAV) includes:

[0007] Acquire the target drone formation;

[0008] The leader and followers in the target UAV formation and the communication topology of the target UAV formation are determined according to affine transformation theory; the total number of leaders is greater than or equal to a set number and the formation formed by the leaders is affinely spanned in three-dimensional space; the communication topology of the target UAV formation satisfies the requirements of affine transformation; the communication topology of the target UAV formation is a directed graph.

[0009] Based on the communication topology diagram of the target UAV formation, determine the symbolic Laplace matrix of the communication topology diagram and the set of incoming neighbor nodes for each UAV; the set of incoming neighbor nodes includes all the predecessor neighbors of the UAV.

[0010] If all navigators have constant acceleration, for any follower, the control force of the follower can be obtained based on the symbolic Laplace matrix of the communication topology diagram, the physical parameters of the follower, the physical parameters of each UAV in the follower's ingress neighbor node set, and the follower's first position controller; the physical parameters include velocity and position.

[0011] If all navigators have time-varying acceleration, for any follower, the control force of the follower is obtained based on the symbolic Laplace matrix of the communication topology graph, the estimated parameters of each UAV in the follower's incoming neighbor node set, the follower's state observer, the follower's physical parameters, and the follower's second position controller; the estimated parameters include: an estimate of the desired position, an estimate of the desired velocity, and an estimate of the desired acceleration;

[0012] The control torque of the follower is obtained based on the control force of the follower and the adaptive sliding membrane attitude controller;

[0013] The control force and control torque of the follower are input into the dynamic model of the follower to achieve control of the follower, so that the follower reaches the desired position and desired speed.

[0014] Optionally, the control force of the follower is obtained based on the symbolic Laplace matrix of the communication topology diagram, the estimated parameters of each UAV in the follower's ingress neighbor node set, the follower's state observer, the follower's physical parameters, and the follower's second position controller. Specifically, this includes:

[0015] The estimated parameters of the follower are obtained based on the symbolic Laplace matrix of the communication topology diagram, the estimated parameters of each UAV in the set of incoming neighbor nodes of the follower, and the state observer of the follower.

[0016] The control force of the follower is obtained based on the estimated parameters of the follower, the physical parameters of the follower, and the second position controller of the follower.

[0017] Optionally, the control torque of the follower is obtained based on the control force of the follower and the adaptive sliding membrane attitude controller, specifically including:

[0018] The desired parameters of the follower are obtained based on the follower's control force; the desired parameters include desired posture and desired angular velocity.

[0019] The error parameters of the follower are obtained based on the expected parameters of the follower; the error parameters include attitude error and angular velocity error.

[0020] The control torque of the follower is obtained based on the error parameters of the follower and the adaptive sliding membrane attitude controller.

[0021] A quadcopter unmanned aerial vehicle (UAV) affine formation control system, comprising:

[0022] The acquisition module is used to acquire the target drone formation;

[0023] The follower-leader communication topology diagram determination module is used to determine the leader and followers in the target UAV formation and the communication topology diagram of the target UAV formation based on affine transformation theory; the total number of leaders is greater than or equal to a set number and the formation formed by the leaders is affinely spanned in three-dimensional space; the communication topology diagram of the target UAV formation satisfies the affine transformation requirements; the communication topology diagram of the target UAV formation is a directed graph.

[0024] The matrix set determination module is used to determine the symbolic Laplace matrix of the communication topology diagram and the set of incoming neighbor nodes for each UAV based on the communication topology diagram of the target UAV formation; the set of incoming neighbor nodes includes all the predecessor neighbors of the UAV.

[0025] The first control force result determination module is used to determine the control force of any follower if all navigators have constant acceleration, based on the symbolic Laplace matrix of the communication topology diagram, the physical parameters of the follower, the physical parameters of each UAV in the follower's ingress neighbor node set, and the follower's first position controller; the physical parameters include velocity and position.

[0026] The second control force result determination module is used to determine the control force of any follower if all navigators have time-varying acceleration, based on the symbolic Laplace matrix of the communication topology graph, the estimated parameters of each UAV in the follower's incoming neighbor node set, the follower's state observer, the follower's physical parameters, and the follower's second position controller; the estimated parameters include: an estimate of the desired position, an estimate of the desired velocity, and an estimate of the desired acceleration;

[0027] A control torque determination module is used to obtain the control torque of the follower based on the control force of the follower and the adaptive sliding membrane attitude controller.

[0028] The control module is used to input the control force and control torque of the follower into the dynamic model of the follower to control the follower, so that the follower reaches the desired position and desired speed.

[0029] Optionally, the second control force result determination module specifically includes:

[0030] The estimator parameter determination unit is used to obtain the estimator parameters of the follower based on the symbolic Laplace matrix of the communication topology diagram, the estimator parameters of each UAV in the set of incoming neighbor nodes of the follower, and the state observer of the follower.

[0031] The control force determination unit is used to obtain the control force of the follower based on the estimated parameters of the follower, the physical parameters of the follower, and the second position controller of the follower.

[0032] Optionally, the control torque determination module specifically includes:

[0033] A desired parameter determination unit is used to obtain the desired parameters of the follower based on the control force of the follower; the desired parameters include desired attitude and desired angular velocity;

[0034] An error parameter determination unit is used to obtain the error parameters of the follower based on the expected parameters of the follower; the error parameters include attitude error and angular velocity error.

[0035] A control torque determination unit is used to obtain the control torque of the follower based on the error parameters of the follower and the adaptive sliding membrane attitude controller.

[0036] An electronic device, comprising:

[0037] A memory and a processor, the memory for storing a computer program, the processor for running the computer program to cause the electronic device to perform the affine formation control method for a quadcopter UAV as described above.

[0038] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned quadcopter UAV affine formation control method.

[0039] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects: The present invention determines the leader, follower, and communication topology diagram in the target UAV formation based on affine transformation theory, and determines the movement trajectory of the follower based on the communication topology diagram. Compared with traditional formation methods, it can achieve flexible maneuverability of the formation while realizing target formation convergence, and has stronger anti-interference capability. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a schematic diagram showing the structure of a quadcopter UAV and the correspondence between its body coordinate system and inertial coordinate system.

[0042] Figure 2 This is a schematic diagram of affine formation transformation;

[0043] Figure 3 This is a schematic diagram of the quadcopter UAV formation control system of the present invention;

[0044] Figure 4 This is a schematic diagram of the desired formation and formation communication topology;

[0045] Figure 5 The error curves of five follower quadcopter drones in an embodiment of the present invention when the navigator has a constant acceleration are shown.

[0046] Figure 6 The error curves of five follower quadcopter drones in an embodiment of the present invention when the navigator has time-varying acceleration;

[0047] Figure 7 The above are the attitude response curves of five follower quadcopter UAVs in an embodiment of the present invention when the navigator has time-varying acceleration.

[0048] Figure 8 The image shows the three-dimensional trajectory curves of nine quadcopter drones in an embodiment of the present invention when the navigator has time-varying acceleration.

[0049] Figure 9 The three-dimensional trajectory curves of nine quadcopter UAVs performing formation maneuvers in an embodiment of the present invention;

[0050] Figure 10 A flowchart of a quadcopter UAV affine formation control method provided in an embodiment of the present invention. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0053] like Figure 10 As shown, this embodiment of the invention provides a method for affine formation control of quadrotor UAVs, specifically a distributed real-time control method for affine formation of quadrotor UAVs under directed communication topology constraints, which includes:

[0054] Step 101: Obtain the target drone formation. Each drone in the target drone formation has its own number.

[0055] Step 102: Determine the leader and followers in the target UAV formation and the communication topology of the target UAV formation based on affine transformation theory. The total number of leaders is greater than or equal to a set number (set to 4), and the formation formed by the leaders is affinely spanned in three-dimensional space; the communication topology of the target UAV formation satisfies the affine transformation requirements; the communication topology of the target UAV formation is a directed graph with UAVs as nodes. The leader flies according to the planned desired path, and the followers obtain information through the communication topology to plan their flight paths.

[0056] Step 103: Determine the symbolic Laplace matrix of the communication topology diagram and the set of incoming neighbor nodes for each UAV based on the communication topology diagram of the target UAV formation. The set of incoming neighbor nodes includes all predecessor neighbors of the UAV. A directed graph G = (V, E) is used to represent the communication network among n UAVs in the formation, where the node set V = {1, 2, ..., n} and the edge set... When there exists an edge (j,i)∈E, drone i can obtain information from drone j. In this case, drone j is called the in-neighbor of drone i, that is, drone j is the predecessor neighbor of drone i. Define N i ={j|(j,i)∈E} is the set of incoming neighbor nodes of node i.

[0057] In position control, the navigator drone flight trajectory can be pre-designed according to mission requirements or the navigator can be controlled in real time. According to the different acceleration information of the navigator, it can be divided into two cases: constant acceleration or time-varying acceleration. Based on this, two types of position controllers are designed: the first position controller and the second position controller.

[0058] Step 104: If all navigators have constant acceleration, for any follower, the control force of the follower is obtained based on the symbolic Laplace matrix of the communication topology diagram, the physical parameters of the follower, the physical parameters of each UAV in the follower's ingress neighbor node set, and the follower's first position controller. The physical parameters include velocity and position.

[0059] Step 105: If all leaders have time-varying acceleration, for any follower, the control force of the follower is obtained based on the symbolic Laplace matrix of the communication topology graph, the estimated parameters of each UAV in the follower's incoming neighbor node set, the follower's state observer, the follower's physical parameters, and the follower's second position controller. The estimated parameters include: an estimate of the desired position, an estimate of the desired velocity, and an estimate of the desired acceleration.

[0060] Step 106: Obtain the control torque of the follower based on the control force of the follower and the adaptive sliding membrane attitude controller.

[0061] Step 107: Input the control force and control torque of the follower into the dynamic model of the follower to control the follower, so that the follower reaches the desired position and desired speed. The control method described in steps 104 to 107 is used for real-time control at each moment.

[0062] In practical applications, a drone formation consisting of n quadcopter drones is used. Based on the Euler-Newton formula, a dynamic model of the i-th follower drone is established:

[0063]

[0064] in, It is a unit quaternion, where, and q i0 and q iv For q i The components. In polar coordinates, the attitude of the UAV can be expressed as the orientation after rotating around a unit rotation axis ρ = [x, y, z] by an angle μ. The relationship between the unit quaternion and the rotation axis ρ and the rotation angle μ is as follows:

[0065] for x × =[0,-x3,x2; x3,0,-x1;-x2,x1,0] is a skew-symmetric matrix of column vectors x. From the body coordinate system F bi to inertial coordinate system F I The rotation matrix, It is q iv A skew-symmetric matrix, It is an identity matrix.

[0066] and The i-th follower drone is in the inertial coordinate system F I Given the position and velocity below, e3 = [0,0,1] T This is the vector used in the calculation, where g is the acceleration due to gravity. T i m i For thrust magnitude and drone mass, It is control. It is the desired attitude of the i-th follower drone. From the desired body coordinate system to the inertial coordinate system F I The rotation matrix.

[0067] J i ∈R 3×3 The matrix of rotational inertia is positive definite. For the UAV in the body coordinate system F bi The angular velocity, control torque, and disturbance torque of the three axes are used. The first two formulas in the dynamic model of the i-th follower UAV are for position motion, and the last three formulas are for attitude motion. The values ​​of u are obtained from the first or second position controller. i Input the position and motion model of the i-th follower to obtain the derivative of the position of the i-th follower. That is, the velocity, and the derivative of the velocity of the i-th follower. That is, acceleration, which will be determined by the control torque τ of the i-th follower obtained from the adaptive sliding membrane attitude controller. i The derivative of the angular velocity of the i-th follower is obtained by inputting the attitude motion model. That is, angular acceleration.

[0068] In practical applications, when the symbolic Laplace matrix L satisfies the condition The communication topology diagram of the formation satisfies the requirements of affine transformation. Therefore, when determining the communication topology diagram of the target UAV formation based on affine transformation theory, L can be made to satisfy the above formula, where, It is the identity matrix. Indicates the Kronecker product. These represent the positions of all navigators and all followers, respectively. This is a matrix representing the positions of each drone in a drone swarm.

[0069] In practical applications, the symbolic Laplace matrix of the communication topology diagram of the target UAV formation is determined as follows:

[0070] L = [l ij The symbolic Laplace matrix is ​​defined as follows:

[0071]

[0072] Among them, l ij The weights in L are real numbers that can be positive or negative. The elements in L correspond to the communication topology of the formation. When UAV i cannot obtain information from UAV j, l... ij =0, otherwise l ij ≠0, Let i represent the set of incoming neighbor nodes of drone i.

[0073] After confirming the formation's communication network, the symbolic Laplace matrix L can now be divided into: Within the formation, the navigator does not receive information from other drones, therefore This indicates that the Navigator drone does not receive information from other Navigator drones. This indicates that the leader drone does not receive information from follower drones via a communication network. This refers to the communication network through which follower drones receive information from leader drones. n represents the communication network through which a follower drone receives information from other follower drones. l n f It represents the total number of navigator drones and follower drones.

[0074] To ensure the stability of the proposed formation algorithm, a stable diagonal matrix D needs to be introduced such that L s All non-zero eigenvalues ​​have positive real parts.

[0075] Method for determining the elements of a stable diagonal matrix D: Select ±1 as elements of the diagonal matrix D, and the selected elements must make D"L ff Right now The eigenvalues ​​of all have positive real parts, and the calculation process is shown below:

[0076]

[0077] In practical applications, the follower's desired position and speed can be uniquely determined by the leader's position and speed: in p represents the desired position and desired velocity of the followers. l v l For the position and speed of the navigator, It is the identity matrix. It represents the Kronecker product.

[0078] In practical applications, the navigator typically has variable acceleration in general mission scenarios. However, when the mission is simpler, the navigator can be set to have constant acceleration. In this case, the controller calculation process is simpler, and the hardware requirements for the drone are lower. Therefore, the first position controller of the i-th follower drone is:

[0079]

[0080] For the For each follower, the control force u of the i-th follower drone can be calculated based on the first position controller. i Where g is the acceleration due to gravity, and e3 = [0, 0, 1] T tanh() is the tangent function, η i ψ i , and η j This is an auxiliary variable for the i-th follower drone in the auxiliary system. Its initial value is set by the user and will be automatically updated later. It is the set of ingress neighbor nodes of the i-th follower drone. and Representing the auxiliary variable η respectively i and η j The first derivative, express The first derivative, p i v i p represents the position and velocity of the i-th follower drone in the inertial coordinate system. j v j The position and velocity of the j-th UAV in the inertial coordinate system are given. This indicates the relationship between the auxiliary variable η and the variable η. i The second derivative, the introduced auxiliary variable, can guarantee the control force u. i Boundedness, α, β, k η k p and k v The gain is a manually adjusted constant, d. i Let l be the i-th diagonal element in the stable matrix D. ij For each element in the symbolic Laplace matrix L, when l ijWhen the value is not 0, UAV i can obtain information from UAV j. The first controller can ensure that the follower UAV can track the desired position and speed when the leader has a fixed acceleration, thereby ensuring the formation convergence and maneuverability of the entire formation.

[0081] In practical applications, the control force of the follower is obtained based on the symbolic Laplace matrix of the communication topology diagram, the estimated parameters of each UAV in the follower's ingress neighbor node set, the follower's state observer, the follower's physical parameters, and the follower's second position controller. Specifically, this includes:

[0082] The estimated parameters of the follower are obtained based on the symbolic Laplace matrix of the communication topology diagram, the estimated parameters of each UAV in the set of incoming neighbor nodes of the follower, and the state observer of the follower.

[0083] The control force of the follower is obtained based on the estimated parameters of the follower, the physical parameters of the follower, and the second position controller of the follower.

[0084] In practical applications, when the leader has time-varying acceleration, the state observer of the i-th follower UAV is:

[0085]

[0086]

[0087]

[0088]

[0089] in, These are the expected positions. Expected speed and expected acceleration The estimated values ​​of the three desired states, with a dot above them, represent the first derivative. For the navigator (j∈V) l For example, using the actual value as the estimate, that is... χ i It is an auxiliary variable, tanh() is the tangent function, l ij d represents the elements that can be positive or negative in the symbolic Laplace matrix L. i Let k be the i-th diagonal element in the stable matrix D. β β k (k = 1, 2, 3, 4) are positive control parameters, and the variable Λ i = in, function κ(t)=c mexp(-v m t), This indicates that off-diagonal elements are 0, and diagonal elements are 0. The matrix, It is χ i The k-th component, c m v m Here, t is a positive constant, exp() is time, and exp() is an exponential function. For a follower drone, this observer can eventually achieve the effect that the state estimate tends to the expected value, i.e.

[0090] The second position controller for the i-th follower drone is:

[0091]

[0092] Where e3 = [0,0,1] T tanh() is the tangent function, η i ψ i It is an auxiliary variable introduced into the auxiliary system, l p l v These are normal control parameters.

[0093] Based on the introduced state observer, the proposed second position controller does not need to use the acceleration information of the follower drone during operation. Therefore, the follower drone does not need to measure its own acceleration information in practical applications, which reduces the hardware requirements of the follower drone. The second position controller can drive the follower drone to track the desired position and velocity when the navigator has time-varying acceleration. Furthermore, based on affine transformation theory, the route of the navigator drone can be designed to achieve the scaling, slicing, rotation and translation transformation effects of the drone formation.

[0094] In practical applications, the UAV control system is a dual closed-loop control system of position control and attitude control. In formation control, it is necessary to extract the target attitude of the UAV based on its position control, and then perform attitude control based on the target attitude to ensure the effectiveness of formation control during flight. Therefore, the control torque of the follower is obtained based on the control force of the follower and the adaptive sliding membrane attitude controller, specifically including:

[0095] The desired parameters of the follower are obtained based on the follower's control force; the desired parameters include desired posture and desired angular velocity.

[0096] The error parameters of the follower are obtained based on the expected parameters of the follower; the error parameters include attitude error and angular velocity error.

[0097] The control torque of the follower is obtained based on the error parameters of the follower and the adaptive sliding membrane attitude controller.

[0098] In practical applications, the expected parameters of the follower are obtained based on the control force of the follower, and the error parameters of the follower are obtained based on the expected parameters of the follower. Specifically, this includes: based on the position control force of the i-th follower UAV. Representing control force u i The components of the x, y, and z coordinate axes in the inertial coordinate system are used to obtain the desired attitude of the i-th follower UAV by introducing the hierarchical control framework proposed in other methods. and expected angular velocity As a tracking target in attitude control and for The weight of the input formula is the control force of the follower.

[0099]

[0100]

[0101] get and Then according to the formula Calculate the attitude error of the i-th follower drone. and yes The components, according to Calculate the angular velocity error of the i-th follower drone. in, Let T represent the error rotation matrix. i m i Let be the thrust of the i-th follower drone and the mass of the i-th follower drone, and g be the acceleration due to gravity. It is control force u i The derivative, At this point, q i ω i γ represents the current attitude and angular velocity of the i-th UAV. i,1 and γ i,2 ⊙ is an intermediate variable, and ⊙ represents quaternion multiplication.

[0102] In practical applications, to facilitate the design of the attitude controller, a linear operator Υ(x) is introduced to simplify the attitude motion model for any vector.

[0103] At this point, the attitude motion model of the UAV can be simplified using linear operators to the following form:

[0104]

[0105]

[0106] Where, τ i It is the control torque of the i-th follower drone. Indicates to Find the first derivative. Indicates to Find the first derivative. Represents ω i The skew-symmetric matrix, c i It is the external disturbance that is affected, ω i , It is angular velocity and angular velocity error. The matrix represents the identity matrix, and the superscript T indicates finding the transpose. The inertial matrix of the UAV is... Inertial parameter θ i =[J i,11 J i,22 J i,33 J i,23 J i,13 J i,12 ] T It is a column vector composed of elements from the inertia matrix. These are variables introduced for the sake of simplicity in the expression of the equation. Υ(ω i ), Υ(Ξ) i The introduced linear operator Y(x) simplifies the expression of the attitude model but has no physical meaning. i,1 Y i,2 These variables were introduced to simplify the expression of control torque in the drone model and subsequent design.

[0107] Taking into account the uncertainty of the UAV's inertial parameters and the impact of external disturbances, the designed adaptive sliding membrane attitude controller for the i-th follower is as follows:

[0108]

[0109] in, S represents attitude error and angular velocity error. i For the designed sliding surface, sgn() is the sign function. Let θ be the inertial parameter i Estimated value. For the self-determined positive definite gain matrix, k g k d k q The normal control parameters are determined autonomously, and k d It must be greater than the external disturbance c.i The upper bound of the absolute value.

[0110] Introduced into attitude control torque The adaptive module for inertial parameter estimation (the last formula in the adaptive sliding membrane attitude controller) enables the UAV's attitude control to track the desired attitude under conditions of external disturbances and uncertainties in internal parameters, thereby enhancing the robustness of the UAV's attitude control.

[0111] In practical applications, the control force and control torque of the follower are input into the follower's dynamic model to control the follower, enabling it to reach a desired position and speed. Specifically, this includes:

[0112] The control force of the follower is input into the positional motion model in the dynamic model of the follower to obtain the acceleration and velocity of the follower.

[0113] The control torque of the follower is input into the attitude motion model in the dynamic model of the follower to obtain the angular acceleration of the follower.

[0114] The follower is controlled based on its acceleration, velocity, and angular acceleration to achieve the desired position and velocity.

[0115] This invention first establishes a motion model of a quadcopter UAV formation, and then establishes a directed topology communication structure for the formation system based on affine transformation theory. In the design of the position controller, two types of position controllers are designed according to the different accelerations of the navigator. Subsequently, a robust attitude controller is designed under the introduced hierarchical control framework, thereby achieving stable flight of the UAV formation under disturbances. The formation can achieve convergence of the desired formation and flexible formation transformations, and the robust control law ensures the stability of the formation. Finally, the desired cooperative performance of the formation is achieved. The desired trajectory of the formation can be designed according to mission requirements, and the formation maneuver operations during the formation process can be designed based on affine transformation theory, enabling the formation to achieve translation, scaling, oblique cutting, and rotational formation maneuvers. Afterwards, only the navigator needs to be controlled to fly along the desired trajectory. The position controller and attitude controller of this invention can drive the followers in the formation to track the desired turntable in a distributed control manner, thereby ensuring that the formation ultimately achieves the desired formation trajectory and maneuver effects.

[0116] This invention provides a more specific embodiment to describe the above method in detail, as follows:

[0117] A drone formation consists of n quadcopter drones, and the states of the drones must be defined in the inertial coordinate system {O}. E X E Y EZ E} and body coordinate system {O B X B Y B Z B Represented under two standard systems, such as Figure 1 As shown, the two coordinate systems are defined as follows:

[0118] (1) Inertial coordinate system {O E X E Y E Z E Also known as a ground coordinate system, the origin O can be chosen arbitrarily. E X E The axis points east, Y E The axis points north, Z E The orientation of the axis is determined by the right-hand rule.

[0119] (2) Body coordinate system {O B X B Y B Z B}: Fixed to the drone body, origin O B Since the origin point coincides with the center of gravity of the drone, its coordinates also change with the change of the drone's center of mass. B The axis points in the direction the drone is moving, and the right side of the drone's direction of movement is selected as Y. B Axis direction, Z B The orientation of the axis is determined by the right-hand rule.

[0120] exist Figure 2 The diagram illustrates the maneuvering effects achievable through affine transformations in formation control. Figure 2 (a) in the diagram represents the original desired formation. Figure 2 (b) Figure 2 (c) and Figure 2 In the diagram, (d) represents formation transformations performed on the original formation, including rotation, scaling, and skewing. Figure 2 In this paper, (e) represents the result of slicing the vertical component to 0. Based on this affine transformation theory, this invention proposes an affine formation control method for quadrotor UAVs. The control system structure is as follows: Figure 3 As shown, the system is divided into two layers: position control and attitude control.

[0121] The model parameters of the quadcopter drone are set as follows: g = 9.8 (m / s²) 2 According to step 102, drones 1-4 are selected as the navigators of the formation, and the remaining five drones, 5-9, are the followers in the formation. The specific directed communication topology of the formation is as follows: Figure 4 The given figure and the corresponding Laplace matrix L can be obtained according to step 103. The initial settings for the five follower drones are shown in Table 1:

[0122] Table 1 Initial state of the follower drone

[0123] 5 (2,1,3) (0,0,0) 6 (-7,-6,3) (0,0,0) 7 (-6,2,-3) (0,0,0) 8 (8,-3,-4) (0,0,0) 9 (0,3,2) (0,0,0)

[0124] First simulation scenario:

[0125] First, we simulate the situation where the navigator in a drone formation has a constant acceleration. The initial states and constant accelerations of navigators 1-4 are set as shown in Table 2:

[0126] Table 2 Initial State and Constant Acceleration Settings of Navigator UAV

[0127] 1 [-5；5；5] [0.5；0.1；0.1] [1；0；0] 2 [5；-5；5] [0.5；0.1；0.1] [1；0；0] 3 [5；5；-5] [0.5；0.1；0.1] [1；0；0] 4 [-5；-5；-5] [0.5；0.1；0.1] [1；0；0]

[0128] At this point, the control law input of the follower is the same as that of the first position controller, and the control parameter is set to k. η =1, k p =3, k v =5, α=1, β=3, at this point the UAV formation can converge to the target formation, and the simulation results are as follows. Figure 5 As shown, Figure 5 Part (a) shows the position error diagram of the five followers under constant acceleration. Figure 5 Part (b) shows the velocity error diagram of the five followers under constant acceleration. With the leader having constant acceleration, it can be seen that the position controller of the present invention can enable the followers in the formation to follow the desired position and speed, thereby making the position error and velocity error approach 0.

[0129] Second simulation scenario:

[0130] The simulation depicts a drone formation where the leader has time-varying acceleration. The initial states and constant accelerations for leaders 1-4 are shown in Table 3.

[0131] Table 3 Initial State and Constant Acceleration Settings of Navigator UAV

[0132]

[0133]

[0134] At this point, the state observer from the above embodiment is introduced:

[0135]

[0136]

[0137]

[0138]

[0139] in These are the expected positions. Expected speed and expected acceleration The estimates of the three desired states for the navigator (j∈V) l For example, using the actual value as the estimate, that is... X i It is an auxiliary variable, tanh() is the tangent function, l ij d represents the elements that can be positive or negative in the symbolic Laplace matrix L. i Let k be the i-th diagonal element in the stable matrix D. β β k (k = 1, 2, 3, 4) are positive control parameters, variables in κ(t)=c m exp(-v m t), c m v m It is a positive constant, t is time, and exp() is an exponential function.

[0140] Based on this observer, the second position controller proposed in the above embodiments is used.

[0141]

[0142] Where u i This is the control input force for the i-th follower drone in the design, e3 = [0,0,1] T g is the acceleration due to gravity, tanh() is the tangent function, p i v i The position and velocity of the i-th UAV in the inertial coordinate system are given. yes and Estimates of the three desired states, η i θ i k is an auxiliary variable introduced into the auxiliary system. η l p l v These are normal control parameters.

[0143] Then, based on the control force u output by the position controller i Calculate the desired attitude of the UAV Desired angular velocity Attitude error and angular velocity error

[0144] To facilitate subsequent control of torque τ i The design expresses the drone attitude model in the following form:

[0145]

[0146]

[0147] Where the inertial parameter θ i =[J i,11 J i,22 J i,33 J i,23 J i,13 J i,12 ] T It is a column vector composed of elements from the inertia matrix.

[0148] At this point, an adaptive sliding membrane attitude controller is used:

[0149]

[0150] τ i It is the control torque of the design, s i For the designed sliding surface, sgn() is the sign function, J i It is the inertia matrix. Let θ be the inertial parameter i Estimated value. Let k be a positive definite gain matrix. g k d k q The control parameter is a positive constant, and k d It must be greater than the upper bound of the absolute value of the external disturbance.

[0151] At this point, the inertial parameter θ is set. i =[1,1.5,1,0,0.3,0] T The initial attitude of the system is: q(0)=[0.707,0.3,0.4,0.5] T The initial angular velocity is ω(0) = [0,0,0]. T (rad / s), initial estimate of inertial parameters Control parameter k q =3, k g =10, k d =0.25>0.05, adaptive gain matrix Γ=diag(1,1,1), external disturbance torque and inertial parameter perturbation are as follows:

[0152]

[0153] ΔJ=0.5*diag{sin(π*t / 10),1+sin(π*t / 10),1+sin(π*t / 10)}.

[0154] Simulation results are as follows Figure 6 , Figure 7 and Figure 8 As shown, Figure 6 Part (a) shows the position error diagram of the five followers under time-varying acceleration. Figure 6 Part (b) shows the velocity error diagrams of the five followers under time-varying acceleration. Figure 7 Part (a) shows the posture error diagrams of the five followers. Figure 7 Part (b) shows the angular velocity error diagrams of the five followers, from... Figure 6 Follower formation error curve and Figure 7 The follower attitude response curve shows that the position controller and adaptive sliding membrane attitude controller of this invention effectively suppress the impact of disturbances on the formation under the influence of disturbance torque and inertial parameter disturbances, tracks the time-varying desired position and desired attitude with very small tracking error, and meets the control accuracy requirements. Figure 8 The trajectory diagram of the formation shows that the formation eventually achieves stable and expected performance.

[0155] The third simulation scenario:

[0156] Given a task requiring a formation to sequentially pass through a small hole, a slit, and around a sphere before passing through another slit, it is possible to design the desired trajectory of the followers, thereby enabling the formation to exhibit flexible maneuverability. The achieved maneuverability effect is as follows: Figure 9 As shown, the sequence of actions demonstrates the convergence of the target formation; the formation contracting and then expanding after passing through the small hole; the formation obliquely cutting through the slit in the horizontal direction and then recovering; the formation rotating around the sphere; and the formation obliquely cutting through the second slit in the vertical direction and finally recovering. The entire maneuver fully demonstrates the translational, contraction, rotational, and oblique cutting maneuvers of the affine formation.

[0157] This invention provides a quadcopter UAV affine formation control system corresponding to the above method, comprising:

[0158] The acquisition module is used to acquire the target drone formation.

[0159] The follower-leader communication topology diagram determination module is used to determine the leader and followers in the target UAV formation and the communication topology diagram of the target UAV formation based on affine transformation theory; the total number of leaders is greater than or equal to a set number and the formation formed by the leaders is affinely spanned in three-dimensional space; the communication topology diagram of the target UAV formation satisfies the affine transformation requirements; the communication topology diagram of the target UAV formation is a directed graph.

[0160] The matrix set determination module is used to determine the symbolic Laplace matrix of the communication topology diagram and the set of incoming neighbor nodes for each UAV based on the communication topology diagram of the target UAV formation; the set of incoming neighbor nodes includes all the predecessor neighbors of the UAV.

[0161] The first control force result determination module is used to determine the control force of any follower if all navigators have constant acceleration, based on the symbolic Laplace matrix of the communication topology diagram, the physical parameters of the follower, the physical parameters of each UAV in the follower's ingress neighbor node set, and the follower's first position controller; the physical parameters include velocity and position.

[0162] The second control force result determination module is used to determine the control force of any follower if all navigators have time-varying acceleration, based on the symbolic Laplace matrix of the communication topology diagram, the estimated parameters of each UAV in the follower's incoming neighbor node set, the follower's state observer, the follower's physical parameters, and the follower's second position controller; the estimated parameters include: an estimate of the desired position, an estimate of the desired velocity, and an estimate of the desired acceleration.

[0163] The control torque determination module is used to obtain the control torque of the follower based on the control force of the follower and the adaptive sliding membrane attitude controller.

[0164] The control module is used to input the control force and control torque of the follower into the dynamic model of the follower to control the follower, so that the follower reaches the desired position and desired speed.

[0165] As an optional implementation, the second control force result determination module specifically includes:

[0166] The estimator parameter determination unit is used to obtain the estimator parameters of the follower based on the symbolic Laplace matrix of the communication topology diagram, the estimator parameters of each UAV in the set of incoming neighbor nodes of the follower, and the state observer of the follower.

[0167] The control force determination unit is used to obtain the control force of the follower based on the estimated parameters of the follower, the physical parameters of the follower, and the second position controller of the follower.

[0168] As an optional implementation, the control torque determination module specifically includes:

[0169] The expected parameter determination unit is used to obtain the expected parameters of the follower based on the control force of the follower; the expected parameters include the expected posture and the expected angular velocity.

[0170] An error parameter determination unit is used to obtain the error parameters of the follower based on the expected parameters of the follower; the error parameters include attitude error and angular velocity error.

[0171] A control torque determination unit is used to obtain the control torque of the follower based on the error parameters of the follower and the adaptive sliding membrane attitude controller.

[0172] This invention provides an electronic device, comprising:

[0173] A memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to cause the electronic device to perform the quadcopter UAV affine formation control method according to the above method embodiments.

[0174] This invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the quadcopter UAV affine formation control method described in the above-described method embodiments.

[0175] Compared with existing drone formation methods, the present invention has the following advantages:

[0176] 1. The position control method of the present invention introduces the theory of affine transformation, which, compared with the traditional formation method, can achieve flexible maneuvering of the formation while achieving convergence of the target formation, and has greater application value.

[0177] 2. The position controller of this invention incorporates an observer module, which eliminates the need to measure the acceleration information of the followers during the formation of UAVs, thereby reducing the hardware requirements for the followers UAVs.

[0178] 3. The position controller proposed in this invention introduces an auxiliary system to ensure the boundedness of the input control force signal and satisfy the saturation characteristics of the UAV's output force.

[0179] 4. The attitude controller design of this invention introduces a hierarchical control framework, which is more in line with the physical characteristics of UAV flight. It also introduces an adaptive control module for inertial parameters, which can achieve high-precision UAV flight control under uncertainties in its own parameters and the effects of external disturbances, thereby improving the robustness of UAV formation.

[0180] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0181] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for affine formation control of a quadcopter unmanned aerial vehicle, characterized in that, include: Acquire the target drone formation; The leader and followers in the target UAV formation and the communication topology of the target UAV formation are determined according to affine transformation theory; the total number of leaders is greater than or equal to a set number and the formation formed by the leaders is affinely spanned in three-dimensional space; the communication topology of the target UAV formation satisfies the requirements of affine transformation; the communication topology of the target UAV formation is a directed graph. Based on the communication topology diagram of the target UAV formation, determine the symbolic Laplace matrix of the communication topology diagram and the set of incoming neighbor nodes for each UAV; the set of incoming neighbor nodes includes all the predecessor neighbors of the UAV. If all navigators have constant acceleration, for any follower, the control force of the follower can be obtained based on the symbolic Laplace matrix of the communication topology diagram, the physical parameters of the follower, the physical parameters of each UAV in the follower's ingress neighbor node set, and the follower's first position controller; the physical parameters include velocity and position. If all navigators have time-varying acceleration, for any follower, the control force of the follower can be obtained based on the symbolic Laplace matrix of the communication topology diagram, the estimated parameters of each UAV in the follower's ingress neighbor node set, the follower's state observer, the follower's physical parameters, and the follower's second position controller. The estimated parameters include: an estimate of the desired position, an estimate of the desired velocity, and an estimate of the desired acceleration; The control torque of the follower is obtained based on the control force of the follower and the adaptive sliding membrane attitude controller; The control force and control torque of the follower are input into the dynamic model of the follower to achieve control of the follower, so that the follower reaches the desired position and desired speed.

2. The affine formation control method for quadcopter UAVs according to claim 1, characterized in that, Based on the symbolic Laplace matrix of the communication topology diagram, the estimated parameters of each UAV in the follower's ingress neighbor node set, the follower's state observer, the follower's physical parameters, and the follower's second position controller, the control force of the follower is obtained, specifically including: The estimated parameters of the follower are obtained based on the symbolic Laplace matrix of the communication topology diagram, the estimated parameters of each UAV in the set of incoming neighbor nodes of the follower, and the state observer of the follower. The control force of the follower is obtained based on the estimated parameters of the follower, the physical parameters of the follower, and the second position controller of the follower.

3. The affine formation control method for quadcopter UAVs according to claim 1, characterized in that, The control torque of the follower is obtained based on the control force of the follower and the adaptive sliding membrane attitude controller, specifically including: The desired parameters of the follower are obtained based on the follower's control force; the desired parameters include desired posture and desired angular velocity. The error parameters of the follower are obtained based on the expected parameters of the follower; the error parameters include attitude error and angular velocity error. The control torque of the follower is obtained based on the error parameters of the follower and the adaptive sliding membrane attitude controller.

4. A quadcopter unmanned aerial vehicle (UAV) affine formation control system, characterized in that, include: The acquisition module is used to acquire the target drone formation; The follower-leader communication topology diagram determination module is used to determine the leader and followers in the target UAV formation and the communication topology diagram of the target UAV formation based on affine transformation theory; the total number of leaders is greater than or equal to a set number and the formation formed by the leaders is affinely spanned in three-dimensional space; the communication topology diagram of the target UAV formation satisfies the affine transformation requirements; the communication topology diagram of the target UAV formation is a directed graph. The matrix set determination module is used to determine the symbolic Laplace matrix of the communication topology diagram and the set of incoming neighbor nodes for each UAV based on the communication topology diagram of the target UAV formation; the set of incoming neighbor nodes includes all the predecessor neighbors of the UAV. The first control force result determination module is used to determine the control force of any follower if all navigators have constant acceleration, based on the symbolic Laplace matrix of the communication topology diagram, the physical parameters of the follower, the physical parameters of each UAV in the follower's ingress neighbor node set, and the follower's first position controller; the physical parameters include velocity and position. The second control force result determination module is used to determine the control force of any follower if all navigators have time-varying acceleration, based on the symbolic Laplace matrix of the communication topology diagram, the estimated parameters of each UAV in the follower's ingress neighbor node set, the follower's state observer, the follower's physical parameters, and the follower's second position controller. The estimated parameters include: an estimate of the desired position, an estimate of the desired velocity, and an estimate of the desired acceleration; A control torque determination module is used to obtain the control torque of the follower based on the control force of the follower and the adaptive sliding membrane attitude controller. The control module is used to input the control force and control torque of the follower into the dynamic model of the follower to control the follower, so that the follower reaches the desired position and desired speed.

5. The quadcopter UAV affine formation control system according to claim 4, characterized in that, The second control force result determination module specifically includes: The estimator parameter determination unit is used to obtain the estimator parameters of the follower based on the symbolic Laplace matrix of the communication topology diagram, the estimator parameters of each UAV in the set of incoming neighbor nodes of the follower, and the state observer of the follower. The control force determination unit is used to obtain the control force of the follower based on the estimated parameters of the follower, the physical parameters of the follower, and the second position controller of the follower.

6. The quadcopter UAV affine formation control system according to claim 4, characterized in that, The control torque determination module specifically includes: A desired parameter determination unit is used to obtain the desired parameters of the follower based on the control force of the follower; the desired parameters include desired attitude and desired angular velocity; An error parameter determination unit is used to obtain the error parameters of the follower based on the expected parameters of the follower; the error parameters include attitude error and angular velocity error. A control torque determination unit is used to obtain the control torque of the follower based on the error parameters of the follower and the adaptive sliding membrane attitude controller.

7. An electronic device, characterized in that, include: A memory and a processor, the memory for storing a computer program, the processor for running the computer program to cause the electronic device to perform the quadcopter unmanned aerial vehicle affine formation control method according to any one of claims 1 to 3.

8. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the quadcopter UAV affine formation control method as described in any one of claims 1 to 3.

Images