Control method and device for multi-uav suspended load transportation system

Abstract

This disclosure provides a control method and apparatus for a transport system with multiple unmanned aerial vehicles (UAVs) carrying loads, applicable to the field of automation technology. The method includes: processing acquired load desired pose information to obtain virtual desired control force and virtual desired control torque, and calculating virtual control quantities; processing the virtual control quantities to obtain a cable virtual control vector; converting the virtual control quantities into a UAV desired attitude rotation matrix, and processing the virtual control vector and the UAV desired attitude rotation matrix to obtain UAV control torque and UAV thrust values; processing the UAV control torque and UAV thrust values ​​to obtain pose information; processing the pose information of the load, cable, and UAV to obtain the transport system tracking error; and adjusting the desired attitude information of the cable subsystem and UAV subsystem based on feedback information and the acquired load subsystem's desired pose information until the transport system tracking error converges.

Description

Technical Field

[0001] This disclosure relates to the field of automation technology, and in particular to a control method and apparatus for a transport system in which multiple unmanned aerial vehicles (UAVs) carry loads. Background Technology

[0002] Quadrotor drones, with their vertical takeoff and landing, stable hovering, and high maneuverability, are widely used in emergency response, power line inspection, agricultural protection, and cargo transportation. In different application scenarios, the coordinated operation of multiple quadcopter drones is particularly important. Considering the limited payload capacity of a single quadcopter drone during aerial cargo transport, collaborative transportation using multiple quadcopter drones is an excellent solution for various application scenarios.

[0003] However, the control methods for multi-quadrotor UAV collaborative transportation do not fully consider the impact of aerodynamic effects during UAV flight. Although the impact of aerodynamic effects on UAVs is relatively small in hovering mode, the impact of aerodynamic effects cannot be ignored in high-speed transportation scenarios, resulting in low tracking and control performance of multi-quadrotor UAV collaborative transportation. Summary of the Invention

[0004] In view of the above problems, this disclosure provides a control method and apparatus for a transport system in which multiple UAVs carry loads.

[0005] According to the first aspect of this disclosure, a control method for a transport system using multiple unmanned aerial vehicles (UAVs) carrying loads is provided. The transport system includes a load subsystem, a rope subsystem, and a UAV subsystem. The method includes: processing the acquired load desired pose information using the load subsystem to obtain a virtual desired control force and a virtual desired control torque for the load; and solving the virtual desired control force and torque to obtain a virtual control quantity for the load. Processing the acquired virtual desired control force and torque using the rope subsystem to obtain a virtual control vector for the rope, where the virtual control vector is the sum of the virtual control quantity and the virtual control vector for the rope. Solving the acquired virtual control quantity into a desired UAV attitude rotation matrix using the UAV subsystem, and processing the virtual control vector and the desired UAV attitude rotation matrix to obtain a UAV control torque and a UAV thrust value. Processing the acquired UAV control torque and thrust value using the UAV subsystem to obtain the pose information for the load, the rope, and the UAV. The acquired load pose information, rope pose information, and drone pose information are processed by the load subsystem, rope pose information, and drone pose information, respectively, to obtain the transportation system tracking error. This tracking error characterizes the error between the pose information of each load, rope pose information, and drone pose information and their respective desired pose information. Based on the transportation system tracking error and the acquired desired pose information of the load subsystem, the desired pose information of the rope pose information and drone pose information are adjusted until the transportation system tracking error converges.

[0006] The second aspect of this disclosure provides a control device for a drone-borne load transport system. The transport system includes a load subsystem, a rope subsystem, and a drone subsystem. The device includes: a first processing module, used to process the load's desired pose information obtained by the load subsystem to obtain a virtual desired control force and a virtual desired control torque, and to solve the virtual desired control force and torque to obtain a virtual control quantity for the load. A second processing module, used to process the load's virtual desired control force and torque obtained by the rope subsystem to obtain a rope virtual control vector, where the virtual control vector is the sum of the load virtual control quantity and the rope virtual control vector. A solving module, used to solve the obtained virtual control quantity into a drone desired attitude rotation matrix using the drone subsystem, and to process the virtual control vector and the drone desired attitude rotation matrix to obtain the drone control torque and drone thrust value. A third processing module, used to process the drone control torque and drone thrust value obtained by the drone subsystem to obtain the pose information of the load, the rope, and the drone. The fourth processing module processes the acquired load pose information, rope pose information, and drone pose information using the load subsystem, rope pose information, and drone pose information, respectively, to obtain the transportation system tracking error. This tracking error characterizes the error between the pose information of the load, rope pose information, and drone pose information and their respective desired pose information. The adjustment module adjusts the desired pose information of the rope pose information and drone pose information based on the transportation system tracking error and the acquired desired pose information of the load subsystem, until the transportation system tracking error converges.

[0007] According to the control method and apparatus of the UAV-borne load transportation system provided in this disclosure, the load virtual expected control force and load virtual expected control torque are obtained by processing the load expected pose information obtained by the load subsystem, and the load virtual expected control force and load virtual expected control torque are calculated to obtain the load virtual control quantity, taking into account the aerodynamic effect of the load in the transportation scenario.

[0008] The virtual load control force and load control torque obtained by processing the rope subsystem are then used to obtain the rope virtual control vector, taking into account the aerodynamic effects of the rope in the transportation scenario.

[0009] The acquired virtual control variables are then used by the UAV subsystem to calculate the desired UAV attitude rotation matrix. The virtual control vector and the desired UAV attitude rotation matrix are then processed to obtain the UAV control torque and thrust values, making these values ​​more accurate. By processing the acquired UAV control torque and thrust values ​​using the UAV subsystem, the pose information of the load, cable, and UAV is obtained, further improving the accuracy of this individual pose information.

[0010] Finally, the acquired load pose information, rope pose information, and drone pose information are processed by the load subsystem, rope pose information, and drone pose information, respectively, to obtain the transportation system tracking error. This tracking error characterizes the error between the pose information of the load, rope pose information, and drone pose information and their respective desired pose information. Based on the transportation system tracking error and the acquired desired pose information of the load subsystem, the desired pose information of the rope subsystem and drone pose information are adjusted until the transportation system tracking error converges, thus improving the tracking and control performance of the drone-borne load transportation system. Attached Figure Description

[0011] The foregoing contents, as well as other objects, features, and advantages of this disclosure, will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0012] Figure 1 A flowchart illustrating a control method for a multi-UAV sling-load transport system according to an embodiment of the present disclosure is shown schematically.

[0013] Figure 2 A schematic diagram of a multi-UAV sling transport system structure according to an embodiment of the present disclosure is shown.

[0014] Figure 3 A schematic block diagram of the control structure of a multi-UAV sling transport system according to an embodiment of the present disclosure is shown.

[0015] Figure 4 This diagram illustrates a system simulation verification schematic of a multi-UAV sling transport system according to an embodiment of the present disclosure.

[0016] Figure 5 This schematically illustrates a virtual narrow slit diagram used in the simulation verification of a multi-UAV sling transport system according to an embodiment of the present disclosure;

[0017] Figure 6 The diagram illustrates the desired trajectory curve of the suspended load during simulation verification according to an embodiment of the present disclosure.

[0018] Figure 7 The diagram illustrates the position tracking error curve of the suspended load during simulation verification according to an embodiment of the present disclosure.

[0019] Figure 8 The diagram schematically illustrates the attitude tracking error curve of the suspended load during simulation verification according to an embodiment of the present disclosure.

[0020] Figure 9 The diagram schematically illustrates the desired trajectory curve of the suspended load's attitude (Euler angles) during simulation verification according to an embodiment of the present disclosure.

[0021] Figure 10 The diagram illustrates the tracking error curve of the suspension rope during simulation verification according to an embodiment of the present disclosure.

[0022] Figure 11 The diagram schematically illustrates the attitude tracking error curve of a quadcopter drone during simulation verification according to an embodiment of the present disclosure.

[0023] Figure 12 The diagram illustrates the position change curve of a quadcopter drone during simulation verification according to an embodiment of the present disclosure.

[0024] Figure 13 The diagram illustrates the thrust variation curve of a quadcopter drone during simulation verification according to an embodiment of the present disclosure.

[0025] Figure 14 The diagram illustrates the torque variation curve of a quadcopter drone during simulation verification according to an embodiment of the present disclosure. Detailed Implementation

[0026] The embodiments of the present disclosure will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the disclosure. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of the present disclosure for ease of explanation. However, it will be apparent that one or more embodiments may be practiced without these specific details. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concepts of the present disclosure.

[0027] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0028] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0029] When using expressions such as "at least one of A, B, and C", they should generally be interpreted in accordance with the meaning that is commonly understood by a person skilled in the art (e.g., "a system having at least one of A, B, and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B, and C, etc.).

[0030] In the technical solution of this invention, the user information (including but not limited to user personal information, user image information, user device information, such as location information) and data (including but not limited to data used for analysis, stored data, and displayed data) involved are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of related data all comply with the relevant laws, regulations, and standards of the relevant countries and regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entry points for users to choose to authorize or refuse.

[0031] For multi-quadcopter UAV sling transport systems, the position and attitude control of the load depend on the traction force applied by the quadcopter UAV via the sling. Therefore, the tracking speed and accuracy of the UAV attitude controller directly affect the performance of the entire system. Most existing research does not consider the impact of aerodynamic effects during UAV flight. While the impact of aerodynamic effects on UAVs is relatively small in hovering mode, their influence becomes significant in high-speed transport scenarios. Therefore, in practical applications, the system places high demands on the attitude tracking and control performance of the quadcopter UAV. Thus, in multi-quadcopter UAV sling transport systems, it is necessary to design the quadcopter UAV attitude controller based on aerodynamic effects and conduct a closed-loop system stability analysis.

[0032] Figure 1 A flowchart illustrating a control method for a multi-UAV sling-load transport system according to an embodiment of the present disclosure is shown.

[0033] like Figure 1 As shown, the control method of the multi-UAV load-carrying transportation system in this embodiment includes operations S110 to S160.

[0034] In operation S110, the load desired pose information obtained by the load subsystem is processed to obtain the load virtual desired control force and load virtual desired control torque. The load virtual desired control force and load virtual desired control torque are then solved to obtain the load virtual control quantity.

[0035] In operation S120, the virtual load control force and the virtual load control torque obtained by the rope subsystem are processed to obtain the rope virtual control vector.

[0036] During operation S130, the UAV subsystem is used to solve the acquired virtual control quantities into the desired attitude rotation matrix of the UAV, and the virtual control vector and the desired attitude rotation matrix of the UAV are processed to obtain the control torque and thrust values ​​of the UAV.

[0037] During operation of S140, the UAV control torque and UAV thrust values ​​obtained by the UAV subsystem are processed to obtain the attitude information of the load, the cable, and the UAV.

[0038] During operation of S150, the attitude information of the acquired load, rope, and drone is processed by the load subsystem, rope subsystem, and drone subsystem respectively to obtain the tracking error of the transportation system.

[0039] In operation S1 60, based on the tracking error of the transportation system and the desired pose information of the load subsystem, the desired pose information of the rope subsystem and the UAV subsystem are adjusted until the tracking error of the transportation system converges.

[0040] According to embodiments of this disclosure, the transportation system includes a load subsystem, a rope subsystem, and an unmanned aerial vehicle (UAV) subsystem.

[0041] According to embodiments of this disclosure, the drone can be a quadcopter drone, a fixed-wing drone, etc. The type of drone is not limited. The rope can be a single strand, or each drone can be connected to a separate rope; the ropes are used to transport the load.

[0042] According to embodiments of this disclosure, the virtual control vector is the sum of the load virtual control vector and the rope virtual control vector.

[0043] According to embodiments of this disclosure, the tracking error of the transportation system characterizes the error between the pose information of the load, the rope, and the drone and their respective desired pose information.

[0044] According to embodiments of this disclosure, the transportation system adjusts the desired pose information of the load subsystem, rope subsystem, and UAV subsystem based on the transportation system tracking error until the transportation system tracking error converges. This adjustment can be made by the load subsystem, rope subsystem, and UAV subsystem of the transportation system based on the transportation system tracking error, or by other subsystems of the transportation system, such as the processing subsystem.

[0045] According to embodiments of this disclosure, the thrust value of the drone can be the upward lift value of the drone.

[0046] According to embodiments of this disclosure, the pose information of the load, the rope, and the drone can be the position information, rotation angle information, etc. of the load, the rope, and the drone.

[0047] According to embodiments of this disclosure, the rope virtual control vector can be a control component parallel to the rope direction, and the load virtual control vector can be a control component perpendicular to the rope direction.

[0048] According to embodiments of this disclosure, the load virtual expected control force and load virtual expected control torque are obtained by processing the load expected pose information obtained by the load subsystem, and the load virtual expected control force and load virtual expected control torque are solved to obtain the load virtual control quantity, taking into account the aerodynamic effects of the load in the transportation scenario.

[0049] The virtual load control force and load control torque obtained by processing the rope subsystem are then used to obtain the rope virtual control vector, taking into account the aerodynamic effects of the rope in the transportation scenario.

[0050] The acquired virtual control variables are then used by the UAV subsystem to calculate the desired UAV attitude rotation matrix. The virtual control vector and the desired UAV attitude rotation matrix are then processed to obtain the UAV control torque and thrust values, making these values ​​more accurate. By processing the acquired UAV control torque and thrust values ​​using the UAV subsystem, the pose information of the load, cable, and UAV is obtained, further improving the accuracy of this individual pose information.

[0051] Finally, the acquired pose information of the load, rope, and drone is processed by the load subsystem, rope subsystem, and drone subsystem respectively to obtain the tracking error of the transportation system. Based on the feedback information, the desired pose information of the load subsystem, rope subsystem, and drone subsystem are adjusted until the tracking error of the transportation system converges, thereby improving the tracking and control performance of the drone-borne load transportation system.

[0052] According to embodiments of this disclosure, processing the acquired load desired pose information using the load subsystem to obtain the load virtual desired control force and load virtual desired control torque includes: processing the load desired attitude rotation matrix and load desired angular velocity using the load attitude controller to obtain the load virtual desired control torque; and processing the load desired position, load desired velocity, and load desired acceleration using the load position controller to obtain the load virtual desired control force.

[0053] According to embodiments of this disclosure, the load subsystem includes a load attitude controller and a load position controller, and the load desired pose information includes a load desired attitude rotation matrix, a load desired angular velocity, a load desired position, a load desired velocity, and a load desired acceleration.

[0054] According to embodiments of this disclosure, the coordinate system definition and dynamic model of a multi-UAV sling transport system are determined; a system model considering n quadcopter UAVs slinging a rigid body load is provided. The definitions of relevant system state variables, some constants, and some special symbols are detailed in Table 1. Among them, the special orthogonal spaces SO(3) and S... 2 Defined respectively The definitions of system-related state variables, some constants, and some special symbols are detailed in Table 1. Among them, the special orthogonal spaces SO(3) and S 2 Defined respectively and Let represent the 1-dimensional real number space, the 3×1-dimensional real number vector space, and the 3×3-dimensional real number matrix space, respectively.

[0055] Table 1: Symbol Definitions

[0056]

[0057] Figure 2 A schematic diagram of a multi-UAV sling transport system structure is shown according to an embodiment of the present disclosure.

[0058] A simplified structural diagram of the model is shown below. Figure 2 As shown. First, define an inertial coordinate system. and n+1 individual coordinate system The center of the coordinate system of the 0th individual is established at the center of mass of the rigid body load, and the center of the coordinate system of the i-th individual is established at the center of mass of the i-th UAV.

[0059] The kinematic model of the system is shown below:

[0060]

[0061]

[0062] Where, matrix S(a) represents the vector An expanded oblique-symmetric matrix.

[0063] Assuming the rope is rigid, in an inertial coordinate system In the above, the positional relationship between the payload and the i-th quadcopter UAV is as follows:

[0064] p i =p L +R L η i -l i ξ i (3)

[0065] In inertial coordinate system In the above, the thrust that the i-th quadcopter UAV can generate is This represents the magnitude of the total thrust, e3 = [0, 0, 1] T In the body coordinate system Below, the torque generated by the i-th quadcopter UAV is Therefore {f i , τi}, i∈{0,1,…n} can represent the control input of the system.

[0066] Taking into account the modeling uncertainties of the sling load and the unknown external disturbances affecting the system, the aerodynamic effects of the quadcopter UAV's attitude loop are further considered. The complete dynamic model of the multi-UAV sling transport system can be obtained as follows:

[0067]

[0068]

[0069]

[0070]

[0071] in, Let Δ represent the air damping coefficient matrix of the i-th quadcopter UAV. p (t), Δ R (t), These represent unknown external disturbances acting on the load position, attitude, and UAV attitude, respectively. This represents the modeling uncertainty of the load attitude. The lift vector generated by the i-th UAV is... Decomposed into lift components f parallel to the direction of the rope i (p) (t) and the lift component f perpendicular to the rope direction i (n) (t), which can be obtained respectively

[0072]

[0073]

[0074] in, Representing the identity matrix, an auxiliary variable that simplifies model representation. The expression is as follows:

[0075]

[0076] According to embodiments of this disclosure, the control objectives and overall control logic of a multi-UAV sling transport system are determined;

[0077] Considering the dynamics of the multi-quadrotor UAV sling transport system given in equations (4) and (10), define... and Define the desired position and orientation of the load. Let be the desired attitude of the i-th quadcopter UAV. Therefore, the tracking error signal sets of the payload subsystem and the UAV attitude subsystem can be defined as follows:

[0078]

[0079] Among them, load position tracking error Load speed tracking error Load attitude tracking error Load attitude angular velocity tracking error Cable attitude tracking error Cable attitude angular velocity tracking error UAV attitude tracking error UAV attitude angular velocity tracking error The detailed definition of the desired trajectory will be given in the subsequent controller design section.

[0080] The system dynamics described by equations (4)-(10) is a cascaded structure, in which the load subsystem and the UAV attitude subsystem are connected by a rotation matrix R. i Establish contact.

[0081] According to embodiments of this disclosure, the method further includes: establishing a dynamic model of the load subsystem based on a virtual control vector. This is achieved by introducing a virtual control vector. The control design for this cascaded system is then performed. Therefore, the dynamic model of the system disclosed in this paper is written as follows:

[0082]

[0083] Among them, f Δ (u i R i , ε) represents the coupling term between the UAV attitude subsystem and the payload subsystem. It is a constant matrix.

[0084] In the load subsystem In the middle, referring to equations (8) and (9), the virtual control input u i (t) can be decomposed into a virtual control vector for the rope (control components parallel to the rope direction). And load virtual control vector (control component perpendicular to the rope direction) The details are as follows:

[0085]

[0086]

[0087] Therefore, the dynamic equations of the load subsystem can be reconstructed as follows:

[0088]

[0089]

[0090]

[0091] Where, vector This represents the control input to the load subsystem. Parallel control component. With ν i (t) has the following relationship:

[0092]

[0093] In the above formula, Let g be the acceleration due to the load, g be the acceleration due to gravity, and e3 be the unit vector, i.e., e3 = [0, 0, 1]. T Δ p This represents the unknown disturbance acting at the load location. S(.) denotes the skew-symmetric matrix. J L Ω represents the load inertia matrix. L η represents the load angular velocity. i R represents the connection point of the i-th rope on the load. L Represents the load attitude rotation matrix, Represents the transpose of the load attitude rotation matrix, Δ R ξ represents the unknown perturbation of the payload's attitude. t represents time, and n represents the number of tethers used by the drone to transport the payload. Each tether connects to one drone, and there are i drones in total, where i is a positive integer greater than or equal to 1. i Let l represent the rope direction vector from the center of mass of the i-th UAV to the load connection point. i This represents the length of the i-th rope. Let represent the mass of the i-th drone. This represents the virtual control quantity for the load. i This represents an auxiliary variable. Let represent the angular acceleration of the i-th rope.

[0094] According to embodiments of this disclosure, the dynamic model is suitable for solving the load virtual desired control force and load virtual desired control torque to obtain the load virtual control quantity. The obtained load virtual desired control force and load desired control torque are processed using a rope subsystem to obtain the rope virtual control vector.

[0095] The objective of this invention is to design a controller. and The tracking errors χ(t) and ε(t) are asymptotically converged to zero. Considering the system model described in equation (12), the load subsystem... The coupling term f can be temporarily disregarded.Δ The influence of the UAV attitude subsystem is defined as follows: The main controller design can be divided into the following three steps:

[0096] (1) For the load subsystem Design a robust controller u i (t), ensuring that the tracking error χ(t) asymptotically converges to 0.

[0097] (2) Attitude subsystem of quadcopter UAV Design a robust controller τ i (t), ensuring that the tracking error ε(t) asymptotically converges to 0.

[0098] (3) Consider the inclusion of coupling term f Δ (u i R i For a complete closed-loop system of χ(t) and ε(t), it is proved that the tracking errors χ(t) and ε(t) converge asymptotically to 0.

[0099] Controllers were designed for the attitude subsystem and payload subsystem of the quadcopter UAV. The controller design for the payload subsystem adopted a robust control strategy based on RISE and geometric control methods. This is used to compensate for modeling uncertainties in the load and unknown external disturbances. Furthermore, to compensate for unknown disturbances and aerodynamic effects, a combined strategy based on geometric control and sliding mode control is used in the attitude subsystem of the quadcopter UAV.

[0100] According to embodiments of this disclosure, the UAV subsystem calculates the acquired virtual control quantities into a desired UAV attitude rotation matrix, and processes the virtual control vector and the desired UAV attitude rotation matrix to obtain the UAV control torque and UAV thrust values, including: processing the acquired virtual control vector using a desired UAV attitude solver to obtain a third desired attitude vector; calculating the desired UAV attitude rotation matrix using a time-smoothing function and the third desired attitude vector using the desired UAV attitude solver; processing the desired UAV attitude rotation matrix using a UAV attitude controller to obtain the UAV control torque; and processing the virtual control vector using a UAV thrust value solver to obtain the UAV thrust value.

[0101] According to embodiments of this disclosure, the desired attitude rotation matrix of the UAV includes three desired attitude vectors. The third desired attitude vector represents the desired direction of the UAV on a preset axis, and the third desired attitude vector is the third column vector of the desired attitude rotation matrix of the UAV. The UAV subsystem includes a UAV desired attitude solver, a UAV attitude controller, and a UAV thrust value solver.

[0102] According to embodiments of this disclosure, the drone can be a quadcopter drone. Through the proposed hierarchical robust geometric control strategy, the tracking errors χ(t) and ε(t) of each subsystem will be ensured to converge asymptotically to zero. As for the drone attitude subsystem... The expected rotation matrix of the i-th quadrotor UAV Defined as in, It is the expected rotation matrix The three column vectors.

[0103] The UAV desired attitude solver solves the time smoothing function and the third desired attitude vector to obtain the UAV desired attitude rotation matrix as shown in Equations 19 and 20. The preset axis can be the i-th UAV. axis, It can be the third desired attitude vector.

[0104] The i-th drone The desired direction of the axis is used Represented as

[0105]

[0106] The desired rotation matrix can be obtained. for

[0107]

[0108] Where, ψ id It is a smooth function with respect to time.

[0109] According to embodiments of this disclosure, the tracking error of the transportation system also includes the tracking error of the UAV subsystem, which includes UAV attitude tracking error and UAV angular velocity tracking error.

[0110] According to embodiments of this disclosure, a robust controller for the attitude subsystem of a quadcopter unmanned aerial vehicle is designed.

[0111] Attitude subsystem for quadcopter UAVs Design a controller to ensure that the tracking error ε(t) asymptotically converges to zero.

[0112] The UAV attitude tracking error is obtained from the UAV's desired attitude rotation matrix and UAV attitude rotation matrix. The UAV angular velocity tracking error is obtained from the UAV's desired attitude rotation matrix, UAV attitude rotation matrix, UAV attitude angular velocity, and UAV attitude angular velocity.

[0113] According to equation (7), the attitude tracking error of the i-th quadcopter UAV is... and UAV angular velocity tracking error Can be defined as

[0114]

[0115]

[0116] in, The desired rotation matrix of the i-th quadcopter UAV. Let represent the desired angular velocity of the i-th quadcopter UAV's attitude. (·) ∨ This represents the inverse operation of S(·).

[0117] right Taking the derivative with respect to time, we can obtain...

[0118]

[0119] The auxiliary function E(·) is defined as follows:

[0120]

[0121] Where tr(·) denotes the trace of the matrix. Taking the derivative with respect to time, we can obtain...

[0122]

[0123] Define a filter error vector As shown below:

[0124]

[0125] in, It is a positive filter gain. Based on the UAV attitude tracking error, UAV angular velocity tracking error, and UAV inertial matrix, a UAV attitude controller is established. Control torque τ i (t) The design is as follows:

[0126]

[0127] in, Defined as

[0128]

[0129] Where sgn(·) represents the standard symbolic function, Indicates positive control gain. express The upper boundary.

[0130] A thrust solver for the UAV is established based on the UAV attitude rotation matrix and virtual control vector. The thrust magnitude is determined by u. i(t) in -R i The projection on e3 is calculated as follows:

[0131] f i =-u i ·R i e3. (29)

[0132] Substituting the attitude dynamics model of the quadcopter UAV in equation (7) into equation (25), we can obtain the following from equation (28): The error dynamics are as follows

[0133]

[0134] Multiply both sides of the above equation by the inertia matrix J. i And substituting τ i (t), we can obtain The closed-loop dynamic equations are as follows:

[0135]

[0136] According to embodiments of this disclosure, obtaining a rope virtual control vector by processing the acquired load virtual desired control force and load virtual desired control torque using a rope subsystem includes: processing the load virtual desired control force and load virtual desired control torque using a rope desired attitude solver to obtain the desired direction of rope swing; and processing the desired direction of rope swing using a rope virtual vector solver to obtain the rope virtual control vector.

[0137] According to embodiments of this disclosure, the rope subsystem includes a rope desired attitude solver and a rope virtual vector solver.

[0138] According to embodiments of this disclosure, the control logic for the suspended load subsystem is determined; a tracking control strategy is designed for the load subsystem. i (t), to ensure that the tracking error χ(t) asymptotically converges to zero.

[0139] Referring to equations (13) and (14), the virtual control input u i (t) can be decomposed into two mutually perpendicular components. Perpendicular components This is the control input for the i-th rope subsystem in equation (17). The parallel component... ν is the control input for the load position and attitude subsystems in equations (15) and (16). According to equation (18), ν i (t) is parallel to ξ i (t), which cannot be guaranteed in control design.

[0140] The virtual control input ν for this load i (t) is selected as the expected value. Parallel to ξ i The component in the (t) direction is specifically expressed as follows:

[0141]

[0142] Define the desired load control force and load desired control torque Substituting these values ​​into equations (15) and (16) respectively, the position mechanics and attitude dynamics of the load can be rewritten as follows:

[0143]

[0144]

[0145] Among them, Y p (t) and Y R (t) represents the coupling term between the load and the i-th rope of the connection, and its value is

[0146]

[0147]

[0148] Define an auxiliary matrix As shown below:

[0149]

[0150] Virtual control input of load The following options are available:

[0151]

[0152] Therefore, the control design of the load subsystem can be decomposed into three sub-tasks, τ des (t), F des (t) and The detailed design will be discussed in a later section.

[0153] According to embodiments of this disclosure, the tracking error of the transportation system includes the tracking error of the load subsystem, which includes the tracking error of the load attitude rotation matrix, the tracking error of the load attitude angular velocity, the tracking error of the load position, and the tracking error of the load speed.

[0154] According to embodiments of this disclosure, a control law for the attitude of a suspended load is designed. Considering the suspended load attitude model given in equation (16), to facilitate the subsequent controller design, the load attitude tracking error is first obtained based on the load desired attitude rotation matrix and the load attitude rotation matrix. The load attitude angular velocity tracking error is then obtained based on the load desired attitude rotation matrix, the load attitude rotation matrix, the load attitude desired angular velocity, and the load attitude angular velocity. A load attitude tracking error vector is defined. Load attitude tracking error Load attitude angular velocity tracking error The expressions are respectively

[0155]

[0156]

[0157] in Represents the load-desired attitude rotation matrix. This represents the expected angular velocity of the load attitude.

[0158] The first auxiliary filtering error is obtained based on the load attitude angular velocity tracking error and the load attitude rotation matrix tracking error. The first auxiliary filtering error is defined as follows: As shown below:

[0159]

[0160] in, This is a normal amount. and Taking the derivative with respect to time, we can obtain...

[0161]

[0162]

[0163] Taking the derivative of r1(t) with respect to time and substituting it into equations (34), (42), and (43), we can obtain...

[0164]

[0165] The auxiliary functions E(·) and Y(·) are defined as follows:

[0166]

[0167]

[0168] A load attitude controller is established based on the first auxiliary filter error and the load inertia matrix. The load attitude controller determines the desired control torque of the load. The design is as follows:

[0169]

[0170] Where r1(0) represents the initial value of r1(t), It is a positive gain.

[0171] According to an embodiment of this disclosure, a load position control law is designed; considering the load position model given in equation (15), the load position tracking error and load speed tracking error are obtained based on the load centroid position and the load desired centroid position.

[0172] Define load position tracking error Load speed tracking error and third auxiliary filter error The expressions are respectively

[0173]

[0174]

[0175]

[0176] in, This is a normal quantity. A load position controller is established based on the load mass, load position tracking error, and load speed tracking error. The load position controller contains the desired control force F. des (t) can be designed as

[0177]

[0178] in, and Positive control gain, Represents ||Δ p The upper bound of (t)||.

[0179] According to embodiments of this disclosure, the tracking error of the transportation system also includes the tracking error of the rope subsystem, which includes rope direction tracking error and rope angular velocity tracking error.

[0180] According to an embodiment of this disclosure, a control law for the swing of the suspended load is designed; considering the swing model of the suspended rope given in equation (17), the desired swing direction of the i-th rope is defined. for

[0181]

[0182] By using formulas (52) and (38), the desired direction of rope swing is processed using the rope virtual vector solver to obtain the rope virtual control vector.

[0183] The desired angular velocity of the rope is obtained based on the desired direction of rope swing. For a given desired direction of the i-th rope, the desired angular velocity of the i-th rope can be obtained as follows:

[0184]

[0185] The rope direction tracking error is obtained based on the desired direction and direction of rope swing. The rope angular velocity tracking error is obtained based on the rope angular velocity, desired angular velocity, desired direction of rope swing, and direction of rope swing. The rope direction tracking error for the i-th rope is... and rope angular velocity tracking error Defined respectively

[0186]

[0187] Based on the rope direction tracking error, the rope angular velocity tracking error, auxiliary variables, and the mass of the UAV, the rope virtual control vector is obtained.

[0188] Based on the dynamic model in equation (17), the control input for the vertical portion of the i-th rope can be designed as follows:

[0189]

[0190] in, The control gain is positive. Based on equations (18) and (55), the desired control input for each UAV, i.e., the outer loop subsystem, can be calculated. Virtual control input u i (t), is

[0191]

[0192] Based on the above hierarchical robust control strategy design, the following conclusion can be drawn using the Lyapunov-based stability analysis method: the control strategy designed in this invention can guarantee the asymptotic stability of the multi-quadrotor UAV sling transport system (χ, ε).

[0193] Figure 3 A schematic block diagram of the control structure of a multi-UAV sling transport system according to an embodiment of the present disclosure is shown.

[0194] like Figure 3 As shown, the transportation system includes a load subsystem, a rope subsystem, and an unmanned aerial vehicle (UAV) subsystem. The load subsystem includes a load position controller, a load attitude controller, and a load virtual vector solver.

[0195] The load attitude controller is used to process the load's desired attitude rotation matrix and the load's desired angular velocity to obtain the load's virtual desired control torque.

[0196] By using a load position controller to process the load's desired position, desired velocity, and desired acceleration, the virtual desired control force of the load is obtained.

[0197] The load virtual desired control force and load desired control torque are processed using the load virtual vector solver to obtain the rope virtual control vector.

[0198] The rope subsystem includes a rope desired attitude solver and a rope swing controller / rope virtual control vector solver.

[0199] The virtual control input of the load is processed by the rope desired attitude solver to obtain the virtual desired control force and the virtual desired control torque of the load, and thus the desired direction of rope swing.

[0200] The desired direction of rope swing is processed using a rope swing controller or a rope virtual vector solver to obtain the rope virtual control vector.

[0201] The UAV subsystem includes a UAV desired attitude solver, a UAV thrust value solver, and a UAV attitude controller.

[0202] The virtual control vectors obtained are processed using the UAV desired attitude solver to obtain the third desired attitude vector. The UAV desired attitude solver then solves the time smoothing function and the third desired attitude vector to obtain the UAV desired attitude rotation matrix.

[0203] The desired attitude rotation matrix of the UAV is processed by the UAV attitude controller to obtain the UAV control torque.

[0204] The thrust value of the UAV is obtained by processing the virtual control vector using the UAV thrust value solver.

[0205] The invention will be described in detail below with reference to specific simulation examples and accompanying drawings.

[0206] I. Introduction to the Simulation Platform

[0207] Numerical simulations were performed to verify the effectiveness of the proposed control scheme.

[0208] Figure 4 This diagram illustrates a system simulation verification schematic of a multi-UAV sling transport system according to an embodiment of the present disclosure.

[0209] like Figure 4 As shown, three quadcopter drones transport a rectangular payload along an elliptical curve. The payload's length, width, and height are set to 1.0m, 0.8m, and 0.2m, respectively. The payload's mass is set to m. L =1.5kg. The parameters for the three quadcopter drones and the rope are set to... l i=1m, J i =diag[0.082,0.085,0.138]kg·m 2 .

[0210] The connection points of the three ropes to the load are: η1 = [0.5, 0, -0.1] T η2 = [-0.5, 0.4, -0.1] T η3 = [-0.5, -0.4, -0.1] T The expected load trajectory is shown below:

[0211] In the expected posture of the load The axis is tangent to the desired trajectory. With the axis parallel to e3, we can obtain...

[0212]

[0213] The desired attitude of the i-th drone Select [1, 0, 0] T .

[0214] Figure 5 This schematically illustrates a virtual narrow slit diagram used in the simulation verification of a multi-UAV sling transport system according to an embodiment of the present disclosure;

[0215] like Figure 5 As shown, assuming at the desired trajectory point At point 0.8m wide and 0.5m long, there exists a point parallel to the inertial coordinate system. A narrow slit along the y-axis. To avoid obstacles and pass through the slit, the system needs to shift the load along the volume coordinate system when passing through this point. The x-axis rolls 45 degrees. The initial conditions for the overhead conveyor system are set to p. L (0) = [1, 5, 0] T m,v L (0) = [0, 0, 0] T m / s, ξ i (0) = [0, 0, 1] T ω i (0) = [0, 0, 0] T rad / s, R i (0) = I3, Ω i (0) = [0, 0, 0] T rad / s.

[0216] The external disturbances and modeling uncertainties in the numerical simulation are set as follows:

[0217]

[0218] II. Simulation Verification and Analysis

[0219] The corresponding simulation results are shown below. Figure 7-14 . Figure 7 The trajectory tracking performance of the load is displayed. The position and attitude tracking errors of the load are shown as follows: Figure 8 and 9 As shown. Where, Ψ L (t) is defined as also, Figure 10 The attitude tracking curves of the load in the "ZYX" rotation sequence are displayed in Euler angle form. and These represent the desired yaw angle, pitch angle, and roll angle, respectively. It can be seen that at t=6s and t=8s, the required roll angle varies by 45 degrees because the air transport system needs to tilt to avoid obstacles.

[0220] Furthermore, based on the timing of the load avoidance maneuver, the numerical analysis is divided into three time periods. The specific time segments are shown in Tables 2 and 3. The first and third rows of Tables 2 and 3 list the maximum steady-state tracking error (MSSE) and root mean squared steady-state tracking error (RMSSE) for the load position and attitude, respectively. The second row of Tables 2 and 3 shows the maximum tracking error (ME) and root mean squared tracking error (RMSE) for the load position and attitude.

[0221] Table 2: Analysis of Load Location Tracking Results

[0222]

[0223] Table 3: Analysis of Load Attitude Tracking Results

[0224]

[0225]

[0226] according to Figure 6-9Before the obstacle avoidance maneuver, the load was able to track the required trajectory with a stable attitude. It can be seen that the load's position and attitude reached a stable state at t = 3.118s and t = 0.558s, respectively. When time reached t = 6s, the system initiated the obstacle avoidance maneuver while the load's trajectory tracking performance remained good. Then, the load's position and attitude reached a stable state again at t = 10.142s and t = 9.366s, respectively. Figure 10 and 11 The tracking errors for the rope swing motion and the attitude of the i-th quadcopter UAV are shown respectively. and They are defined as follows: and Figure 12-14 The position and control inputs of the i-th quadcopter UAV are displayed.

[0227] The numerical simulation results above show that the system can smoothly track the set desired trajectory throughout the flight. Furthermore, the air transport system can quickly react to avoid obstacles in narrow passages by adjusting the load's attitude. Figure 13 and 14 As can be seen, the control input of the i-th quadcopter UAV remains within a reasonable range. These results verify the effectiveness of the control strategy proposed in this invention.

[0228] This invention proposes a novel hierarchical robust geometric control strategy for a multi-quadrotor UAV load-carrying transportation system. The system dynamics model further considers the uncertainty of the load and the influence of unknown external disturbances on the entire system. In the controller design, the control system is decoupled into two parts: (1) the load-carrying subsystem: used for the position and attitude control design of the load and the swing control design of the rope. (2) the quadrotor UAV attitude subsystem: used for the attitude control design of the i-th quadrotor UAV. For the load subsystem, this invention proposes a novel robust controller based on the combination of RISE control method and geometric control method. This controller can quickly track the desired position and attitude of the load, without needing to add high-order time derivative terms of the desired system state in the controller design. Furthermore, considering the influence of aerodynamic effects,

[0229] The features and beneficial effects of this invention are:

[0230] 1. This invention proposes a robust geometric control strategy for the sling load subsystem based on the RISE method and geometric control method, which fully considers the dynamic characteristics of the load in the multi-quadrotor UAV sling transport system. This method can compensate for uncertainties in sling load modeling, effectively suppress unknown external disturbances, and ensure fast and accurate tracking of the sling load's position and attitude;

[0231] 2. Compared with existing controller designs, the robust geometric controller proposed in this invention for the attitude of suspended loads does not require higher-order time derivative terms of the expected system state, making it easier to apply in practice;

[0232] 3. This invention provides a detailed analysis of the dynamic coupling between multiple drones, the load, and the rope in a multi-drone sling transport system. Furthermore, this invention utilizes a Lyapunov-based stability analysis method to demonstrate the stability of the complete closed-loop system.

[0233] 4. The proposed hierarchical robust geometric control strategy was verified through numerical simulation, and the results demonstrate the effectiveness of the control strategy in a multi-UAV sling transport system.

[0234] This disclosure also provides a control method and apparatus for a transport system in which multiple unmanned aerial vehicles (UAVs) carry loads.

[0235] The control method and apparatus for the multi-UAV load-carrying transportation system of this embodiment includes a first processing module, a second processing module, a calculation module, a third processing module, a fourth processing module, and an adjustment module.

[0236] The first processing module is used to process the load desired pose information obtained by the load subsystem to obtain the load virtual desired control force and load virtual desired control torque, and to solve the load virtual desired control force and load virtual desired control torque to obtain the load virtual control quantity. In one embodiment, the first processing module can be used to perform the operation S110 described above, which will not be repeated here.

[0237] The second processing module is used to process the obtained virtual load desired control force and load desired control torque using the rope subsystem to obtain the rope virtual control vector. In one embodiment, the second processing module can be used to perform the operation S120 described above, which will not be repeated here.

[0238] The calculation module is used to calculate the acquired virtual control quantities into the desired attitude rotation matrix of the UAV using the UAV subsystem, and to process the virtual control vector and the desired attitude rotation matrix of the UAV to obtain the UAV control torque and UAV thrust values. In one embodiment, the calculation module can be used to perform the operation S130 described above, which will not be repeated here.

[0239] The third processing module is used to process the UAV control torque and thrust values ​​obtained by the UAV subsystem to obtain the attitude information of the load, the cable, and the UAV. In one embodiment, the third processing module can be used to perform the operation S140 described above, which will not be repeated here.

[0240] The fourth processing module is used to process the acquired pose information of the load, rope, and drone respectively using the load subsystem, rope subsystem, and drone subsystem to obtain the tracking error of the transportation system. In one embodiment, the fourth processing module can be used to perform the operation S150 described above, which will not be repeated here.

[0241] The adjustment module is used to adjust the desired pose information of the load subsystem, rope subsystem, and UAV subsystem according to the tracking error of the transportation system until the tracking error of the transportation system converges. In one embodiment, the adjustment module can be used to perform the operation S160 described above, which will not be repeated here.

[0242] It should be noted that the control method device part of the multi-UAV load-carrying transportation system in the embodiments of this disclosure corresponds to the control method part of the multi-UAV load-carrying transportation system in the embodiments of this disclosure. For a detailed description of the control method device part of the multi-UAV load-carrying transportation system, please refer to the control method part of the multi-UAV load-carrying transportation system, which will not be repeated here.

[0243] Those skilled in the art will understand that the features described in the various embodiments and / or claims of this disclosure can be combined or combined in various ways, even if such combinations or combinations are not explicitly described in this disclosure. In particular, the features described in the various embodiments and / or claims of this disclosure can be combined or combined in various ways without departing from the spirit and teachings of this disclosure. All such combinations and / or combinations fall within the scope of this disclosure.

[0244] The embodiments of this disclosure have been described above. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of this disclosure. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. The scope of this disclosure is defined by the appended claims and their equivalents. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of this disclosure, and all such substitutions and modifications should fall within the scope of this disclosure.

Claims

1. A control method for a multi-UAV suspended load transport system, the transport system comprising: The method includes a load subsystem, a rope subsystem, and a drone subsystem, comprising: The load desired pose information obtained by the load subsystem is used to obtain the load virtual desired control force and load virtual desired control torque. The load virtual desired control force and load virtual desired control torque are then calculated to obtain the load virtual control quantity. Using the virtual expected control force and the virtual expected control torque of the load obtained by the rope subsystem, a rope virtual control vector is obtained, which is the sum of the virtual control quantity of the load and the rope virtual control vector. The virtual control vectors are obtained by using the UAV subsystem to calculate the desired attitude rotation matrix of the UAV, and the virtual control vectors and the desired attitude rotation matrix of the UAV are processed to obtain the UAV control torque and UAV thrust values. By using the drone subsystem to process and obtain the drone control torque and drone thrust value, the pose information of the load, the rope, and the drone is obtained, and the pose information includes attitude information. The load subsystem, the rope subsystem, and the UAV subsystem respectively process the obtained pose information of the load, the rope, and the UAV to obtain the tracking error of the transportation system. The tracking error of the transportation system represents the error between the pose information of the load, the rope, and the UAV and their respective expected pose information. Based on the tracking error of the transportation system and the obtained desired pose information of the load subsystem, the desired pose information of the rope subsystem and the UAV subsystem are adjusted until the tracking error of the transportation system converges, wherein the UAV control torque... The design is as follows: in, Defined as The tracking error of the transportation system includes the tracking error of the UAV subsystem, which includes UAV attitude tracking error and angular velocity tracking error. Let be the drone attitude tracking error of the i-th drone, where i is a positive integer greater than or equal to 1. Let be the angular velocity tracking error of the i-th UAV. Indicates positive control gain. Let be the angular velocity of the i-th drone. Represents the desired angular velocity of the i-th UAV. Let be the inertial matrix of the i-th UAV. Let the matrix represent the air damping coefficient of the i-th UAV. Represents the filtering error vector. Represents standard symbolic functions, express The upper realm, Represents an oblique symmetric matrix. Let the rotation matrix of the i-th UAV be represented. Let be the desired attitude of the i-th drone.

2. The method according to claim 1, wherein, The load subsystem includes a load attitude controller and a load position controller. The load desired pose information includes the load desired attitude rotation matrix, the load desired angular velocity, the load desired position, the load desired velocity, and the load desired acceleration. The step of using the load desired pose information obtained by the load subsystem to obtain the load virtual desired control force and load virtual desired control torque includes: The load attitude controller processes the load desired attitude rotation matrix and the load desired angular velocity to obtain the load virtual desired control torque; The load position controller processes the load's desired position, the load's desired velocity, and the load's desired acceleration to obtain the load's virtual desired control force.

3. The method according to claim 2, wherein, The tracking error of the transportation system includes the tracking error of the load subsystem, which includes the tracking error of the load attitude, the tracking error of the load attitude angular velocity, the tracking error of the load position, and the tracking error of the load speed. The method further includes: The load attitude tracking error is obtained based on the load desired attitude rotation matrix and the load attitude rotation matrix. The load attitude angular velocity tracking error is obtained based on the load attitude rotation matrix, the load attitude rotation matrix, the load attitude angular velocity, and the load attitude angular velocity. The first auxiliary filtering error is obtained based on the load attitude angular velocity tracking error and the load attitude tracking error; The load attitude controller is established based on the first auxiliary filter error and the load inertia matrix; Based on the load centroid position and the desired load centroid position, the load position tracking error and the load velocity tracking error are obtained. The load position controller is established based on the load mass, the load position tracking error, and the load speed tracking error.

4. The method according to claim 1, wherein, The rope subsystem includes a rope desired attitude solver and a rope virtual vector solver. The process of obtaining the rope virtual control vector using the load virtual desired control force and the load virtual desired control torque obtained through the rope subsystem includes: The desired direction of the rope swing is obtained by processing the load virtual desired control force and the load virtual desired control torque using the rope desired attitude solver. The desired direction of the rope swing is processed using the rope virtual vector solver to obtain the rope virtual control vector.

5. The method according to claim 4, wherein, The tracking error of the transportation system also includes the tracking error of the rope subsystem, which includes rope direction tracking error and rope angular velocity tracking error; The process of using the rope virtual vector solver to process the desired direction of the rope swing and obtain the rope virtual control vector includes: Based on the desired direction of the rope's swing, the desired angular velocity of the rope is obtained; The rope direction tracking error is obtained based on the desired direction of rope swing and the direction of rope swing. The rope angular velocity tracking error is obtained based on the rope angular velocity, the desired rope angular velocity, the desired direction of rope swing, and the direction of rope swing. The virtual control vector for the rope is obtained based on the rope direction tracking error, the rope angular velocity tracking error, auxiliary variables, and the mass of the UAV.

6. The method according to claim 1, wherein, The desired attitude rotation matrix of the UAV includes three desired attitude vectors. The third desired attitude vector represents the desired direction of the UAV on a preset axis. The third desired attitude vector is the third column vector of the desired attitude rotation matrix of the UAV. The UAV subsystem includes a UAV desired attitude solver, a UAV attitude controller, and a UAV thrust value solver. The process of using the UAV subsystem to calculate the acquired virtual control quantities into a desired UAV attitude rotation matrix, and processing the virtual control vector and the desired UAV attitude rotation matrix to obtain the UAV control torque and UAV thrust values ​​includes: The virtual control vector obtained by the UAV desired attitude solver is processed to obtain the third desired attitude vector; The desired attitude of the UAV is obtained by solving the smooth function of time and the third desired attitude vector using the UAV desired attitude solver. The UAV attitude controller processes the desired attitude rotation matrix of the UAV to obtain the UAV control torque; The virtual control vector is processed using a UAV thrust value solver to obtain the UAV thrust value.

7. The method according to claim 6, further comprising: The UAV attitude tracking error is obtained based on the UAV desired attitude rotation matrix and the UAV attitude rotation matrix. The UAV angular velocity tracking error is obtained based on the UAV's desired attitude rotation matrix, the UAV attitude rotation matrix, the UAV attitude angular velocity, and the UAV attitude angular velocity.

8. The method according to claim 7, further comprising: The UAV attitude controller is established based on the UAV attitude tracking error, the UAV angular velocity tracking error, and the UAV inertial matrix. A UAV thrust solver is established based on the UAV attitude rotation matrix and the virtual control vector.

9. The method according to claim 1, further comprising: Based on the virtual control vector, a dynamic model of the load subsystem is established. This dynamic model is suitable for solving the virtual expected control force and the virtual expected control torque of the load to obtain the virtual control quantity of the load. Using the virtual expected control force and the virtual expected control torque of the load obtained through the processing of the rope subsystem, the rope virtual control vector is obtained. The formula of the dynamic model is as follows: vector Represents the input force of the load subsystem; rope virtual control vector and The following relationship exists: In the formula, Indicates load quality. The location of the load's center of mass. For load acceleration, It is the acceleration due to gravity. Represents a unit vector, i.e. ; This indicates an unknown disturbance acting at the load location; Represents a skew-symmetric matrix; Represents the load inertia matrix. Indicates the load angular velocity. This represents the connection point of the i-th rope on the load. Represents the load attitude rotation matrix, Represents the transpose of the load attitude rotation matrix, This represents an unknown disturbance indicating the load attitude. The time represents the time, and n represents the number of ropes used by the drones to transport the payload; each rope connects to one drone, and there are a total of i drones, where i is a positive integer greater than or equal to 1; Let represent the rope direction vector from the centroid of the i-th UAV to the load connection point; This represents the length of the i-th rope. Indicate the mass of the i-th drone; This represents the virtual control quantity of the load; Indicates auxiliary variables; Let represent the angular acceleration of the i-th rope.

10. A control device for a drone-borne load transport system, the transport system comprising: The device includes a load subsystem, a rope subsystem, and a drone subsystem, comprising: The first processing module is used to process the load expected pose information obtained by the load subsystem to obtain the load virtual expected control force and load virtual expected control torque, and to solve the load virtual expected control force and load virtual expected control torque to obtain the load virtual control quantity. The second processing module is used to process the load virtual expected control force and the load virtual expected control torque obtained by the rope subsystem to obtain the rope virtual control vector, wherein the virtual control vector is the sum of the load virtual control quantity and the rope virtual control vector; The calculation module is used to use the UAV subsystem to calculate the virtual control quantity obtained into the desired attitude rotation matrix of the UAV, and process the virtual control vector and the desired attitude rotation matrix of the UAV to obtain the UAV control torque and UAV thrust value. The third processing module is used to process the drone control torque and drone thrust value obtained by the drone subsystem to obtain the pose information of the load, the rope and the drone respectively. The fourth processing module is used to process the pose information of the load, the pose information of the rope and the drone respectively using the load subsystem, the rope subsystem and the drone subsystem to obtain the transportation system tracking error. The transportation system tracking error represents the error between the pose information of the load, the rope and the drone and their respective expected pose information. The adjustment module is used to adjust the desired pose information of the rope subsystem and the UAV subsystem respectively based on the tracking error of the transportation system and the desired pose information of the load subsystem, until the tracking error of the transportation system converges, wherein the UAV control torque... The design is as follows: in, Defined as The tracking error of the transportation system includes the tracking error of the UAV subsystem, which includes UAV attitude tracking error and angular velocity tracking error. Let be the drone attitude tracking error of the i-th drone, where i is a positive integer greater than or equal to 1. Let be the angular velocity tracking error of the i-th UAV. Indicates positive control gain. Let be the angular velocity of the i-th drone. Represents the desired angular velocity of the i-th UAV. Let be the inertial matrix of the i-th UAV. Let the matrix represent the air damping coefficient of the i-th UAV. Represents the filtering error vector. Represents standard symbolic functions, express The upper realm, Represents an oblique symmetric matrix. Let the rotation matrix of the i-th UAV be represented. Let be the desired attitude of the i-th drone.

Images