Cooperative flight system based on magnetic attraction grabbing and collision docking of spherical unmanned aerial vehicle

Abstract

The invention discloses a cooperative flight system based on spherical unmanned aerial vehicle magnetic attraction grabbing and collision docking, and belongs to the technical field of unmanned aerial vehicles, the system comprises an unmanned aerial vehicle and connecting arms arranged on the unmanned aerial vehicle, a control module of the unmanned aerial vehicle, and external force is sensed through collision detection so as to adjust the attitude of the unmanned aerial vehicle for docking of the connecting arms; a threaded connecting piece matched with the other unmanned aerial vehicle is arranged on the connecting arm, and the control module controls the unmanned aerial vehicle to axially rotate with the connecting arm as the axis so as to complete fixed connection between the unmanned aerial vehicles; the unmanned aerial vehicle is provided with a spherical shell, a through hole is formed in the shell, a magnetic attraction assembly is arranged on the portion, on the outer side of the through hole, of the spherical shell, preliminary butt joint of the unmanned aerial vehicle is completed through magnetic attraction, the connecting arm is a telescopic bent L-shaped cantilever, and after threaded connection is conducted through the cantilever in the vertical state, the cantilever continues to stretch so that preliminary butt joint stripping of the magnetic attraction assembly can be achieved; and the other unmanned aerial vehicle rolls horizontally, so that the two unmanned aerial vehicles are parallel, and the bottoms of the two unmanned aerial vehicles are connected through a cantilever bent into an L shape.

Description

Technical Field

[0001] The invention belongs to the technical field of unmanned aerial vehicles, and in particular relates to a cooperative flight system based on magnetic capture and collision docking of a spherical unmanned aerial vehicle. Background Art

[0002] In the existing UAV technology, although UAVs have been widely used in flight control and object grasping, they still face many challenges in multi-UAV collaboration, precise docking and autonomous operation. The traditional multi-UAV collaborative operation method relies on highly accurate sensors and complex control algorithms, which makes it less adaptable in complex environments and more expensive. Especially in the precision operations such as docking and handling of UAVs, the existing technology often needs to rely on cumbersome mechanical structures and complex perception systems, which increases the weight and energy consumption of the system. For example: when docking UAVs, a visual perception system is usually used, but due to the complex aerial docking environment, the visual sensor has a certain blind spot during docking, which affects the docking. At the same time, the visual sensor and complex mechanical docking structure increase the load of the UAV, which is not conducive to the balance and endurance of the UAV. Therefore, it is of great significance to develop a more concise, efficient and stable UAV system that can perform tasks in complex environments. Summary of the invention

[0003] In order to solve the shortcomings of the prior art, improve the reliability of drone docking, and reduce the complexity and cost of drone docking, the present invention adopts the following technical solutions:

[0004] A collaborative flight system based on collision docking includes a drone and a connecting arm arranged thereon. The control module of the drone senses external force through collision detection to adjust the drone's posture for docking with the connecting arm. This perception ability can be approximately regarded as a kind of tactile perception, that is, dynamic perception, which can sense the size and direction of external interference force. The connecting arm is provided with a threaded connector that is arranged to cooperate with the end of the connecting arm of another drone. The control module controls the drone to rotate axially with the connecting arm as the axis to complete the fixed connection between the drones. Compared with the docking based on vision, this docking method is more accurate and stable, and reduces the load of the drone.

[0005] Furthermore, the connecting arm is transversely arranged on one side of the UAV, and the UAV is vertically flipped with the connecting arm as an axis so that the threaded connector on the connecting arm is threadedly connected to another UAV.

[0006] Furthermore, a threaded connector is provided on the connecting arm, so that the threaded connector is matched with a threaded connector on another drone connecting arm for threaded connection.

[0007] Furthermore, the connecting arm is a bendable L-shaped cantilever arranged at the bottom of the drone. When docking, the cantilever is in a vertical downward state, and is connected to the cantilever with the value pointing upward at the bottom of another inverted flying drone through a threaded connection. After docking, the other drone is rolled horizontally to form two drones in parallel, and the bottom is connected by the cantilever bent into an L shape.

[0008] Furthermore, the UAV is provided with a spherical shell with a through hole on the shell, and the connecting arm is a retractable connecting arm. The threaded connector of the connecting arm contacts a threaded connector provided on another UAV through the through hole inside the spherical shell and rotates around the connecting arm to complete a fixed connection. The spherical shell can not only achieve stable balance control through the cooperation of sensors and flywheel systems, but also protect the internal UAV mechanical structure and connecting arm, and is also suitable for tactile collision detection to better complete dynamic tactile perception.

[0009] Furthermore, a magnetic attraction component is provided on the spherical shell outside the through hole, and the magnetic attraction component is coordinated with the magnetic attraction component outside the through hole of another drone, and the initial docking of the drones is completed by magnetic attraction, and then fixedly connected by the connecting arm.

[0010] Furthermore, the magnetic suction component is an annular structure surrounding the through hole, so that when another drone rotates relative to the current drone with the connecting arm as the axis, the current drone and the other drone can still maintain preliminary docking, avoiding initial docking and peeling due to external forces received from the flight environment or changes in posture caused by the rotation of the drone.

[0011] Furthermore, during the separation process of the system, the drone rotates in the opposite direction with the connecting arm as the axis to separate the fixed connection between the drones; the drones fly in the opposite direction based on the magnetic connection to peel off the initial docking of the magnetic components. It is necessary to consider that the magnetic force is less than the lift of the drone or the driving force of the two drones flying in opposite directions.

[0012] Furthermore, during the connection process of the system, after the retractable connecting arm is fixedly connected to another drone, it continues to extend to allow the initial docking and separation of the magnetic suction component. During the separation process of the system, the drone is first rotated in the opposite direction with the connecting arm as the axis to separate the fixed connection between the drones, and then the retractable connecting arm is retracted, so that there is no need to consider the separation problem of the adsorption of the magnetic suction component during the separation process.

[0013] Furthermore, the connecting arm is a retractable, bent L-shaped cantilever disposed at the bottom of the drone;

[0014] During the connection process of the system, the drone and another inverted drone are preliminarily docked through their respective bottom magnetic suction components, and after being threadedly connected through the cantilever in the vertical state, the cantilever is further extended to peel off the preliminary docking of the magnetic suction component, and the other drone is rolled horizontally to form two drones in parallel, and the bottom is connected by the cantilever bent into an L shape;

[0015] During the separation process of the system, the other drone is caused to roll vertically to directly separate the fixed connection between the L-shaped cantilevers.

[0016] The advantages and beneficial effects of the present invention are:

[0017] (1) Highly simplified system design: The combined design of magnetic collision and robotic arm docking simplifies the overall structure of the drone. The robotic arm only needs a few degrees of freedom to complete docking and assembly tasks. Compared with traditional complex docking mechanisms, it greatly reduces equipment complexity and energy consumption, and improves mission reliability.

[0018] (2) Operational convenience is greatly improved: The innovative method of spherical drones and magnetic collision docking eliminates the need for high-precision scanning equipment and complex visual recognition algorithms, greatly simplifying the operation process while significantly reducing technical requirements and costs. It is more suitable for scenarios that are difficult for traditional drones to complete, such as high altitudes and narrow spaces.

[0019] (3) Efficient collaboration and precise assembly: Using dynamic control and multi-machine coordination algorithms, the spherical drone can quickly achieve a relatively static state, achieve initial docking through magnetic attraction and collision, and complete assembly through precise adjustment of the robotic arm. The entire process is time-saving and highly precise, suitable for high-efficiency mission requirements, such as emergency rescue or high-altitude construction.

[0020] (4) Adaptability to various complex environments: The spherical UAV shell has impact resistance and flexible attitude control capabilities, and can work stably in complex environments such as strong winds, high temperatures, and low temperatures. In addition, the combined effect of the magnetic suction device and the mechanical arm can overcome the docking difficulties caused by complex environments, making the present invention still highly applicable under extreme conditions.

[0021] (5) Modular expansion and mission diversification: The modular structure achieved through collision docking provides the possibility for the expansion of UAV missions. For example, heavy-load platforms and building components can be assembled through aerial docking, or a UAV headquarters can be established for mission management. Modular capabilities give UAVs a wider range of mission adaptability and flexibility, and are also the necessary foundation for the realization of future UAV aerial headquarters.

[0022] (6) Remote operation and automation capabilities: The present invention supports remote operation, combined with pre-programmed path planning and docking algorithms, to complete complex docking or assembly tasks without human intervention. Even in an unattended environment, tasks can be completed efficiently, which is particularly suitable for tasks such as post-disaster rescue. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 Schematic diagram of the system structure of an embodiment of the present invention.

[0024] Figure 2 It is a structural schematic diagram of the upper and lower docking state of a drone in a system of an embodiment of the present invention.

[0025] Figure 3 It is a structural schematic diagram of the parallel docking state in the system of an embodiment of the present invention.

[0026] Figure 4 It is a flow chart of the docking process in an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The specific implementation of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the present invention, and is not used to limit the present invention.

[0028] like Figure 1 As shown, a collaborative flight system based on magnetic grabbing and collision docking of a spherical drone is used for aerial assembly tasks, including a spherical drone 1 and a mechanical arm arranged thereon. In this system, the drone adjusts its posture through collision, so that the magnetic suction components 4 below it attract each other and perform preliminary docking. When the magnetic suction docking is initially completed, the mechanical arm (cantilever) will perform precise docking through telescopic action. The control module completes the entire collaborative process through a simple flight algorithm, reducing the complex perception and control system required for traditional drone docking.

[0029] The spherical drone 1 includes a magnetic suction component 4, a control module, a power module, a communication module and a sensor module. In the control module, the flight control unit (FCU) plans the path according to the instructions to fly, and performs pitch and yaw movements, thereby controlling the insertion, rotation and docking of the mechanical arm; the power module provides the flight power of the drone, and adopts a rotor system composed of an electric motor and an electronic speed regulator to achieve flight and hovering operations; the communication module performs two-way communication with the control module to ensure real-time data exchange during flight;

[0030] The spherical drone 1 adopts a spherical shell 2, on which a plurality of strong magnetic suction components 4 for docking and fixing between drones are evenly distributed, so as to realize adsorption and docking between drones.

[0031] The control module is responsible for the coordinated control of the flight control, attitude adjustment and magnetic docking process of the entire drone, including the flight control module, collision detection and control unit, sensor fusion unit, visual recognition unit and main control unit.

[0032] Flight control module, including flight control unit (FCU), IMU (inertial measurement unit), GPS unit, barometer.

[0033] The flight control unit (FCU) is responsible for attitude control, heading control, and position control. It combines the data from the IMU, GPS, and barometer to execute the flight control algorithm, adjust the flight status of the drone in real time, and ensure the drone's precise positioning and stable flight in three-dimensional space. The IMU monitors the drone's attitude changes in real time, combines the barometer to measure the altitude, and the GPS unit provides precise position data.

[0034] The collision detection and control unit uses ultrasonic sensors, lidar and other equipment for real-time environmental perception, monitors the relative positions of drones, and automatically adjusts the flight path or stops flying to avoid collision when potential collision risks are detected.

[0035] The visual recognition unit, including a high-definition camera and an image processing unit, uses computer vision algorithms (such as convolutional neural networks, CNN) for environmental perception and target recognition. The visual recognition unit can identify ground targets, docking objects, and surrounding obstacles in real time, and assist the drone in making precise flight adjustments.

[0036] The main control unit, including an embedded processor (such as ARM Cortex series processors) and a real-time operating system (RTOS), receives data from various sensors and coordinates flight control, robotic arm control, target recognition and other functions according to preset control algorithms and mission requirements. The flight path planning generates the optimal path based on the A* algorithm:

[0037] f(n)=g(n)+h(n)

[0038] Among them, g(n) represents the cost from the starting point to node n, and h(n) represents the heuristic estimate from node n to the target.

[0039] The power module, including four independent motors and an electronic speed controller (ESC), achieves stable flight of the drone through the PID control algorithm. Each motor controls the speed of the rotor and achieves attitude control by adjusting the thrust. The dynamics of the drone are described as follows:

[0040]

[0041] in, represents the position vector of the drone, m represents the mass of the drone, g represents the acceleration of gravity, e z represents the motor speed, R represents the rotation matrix of the drone, T represents the total thrust vector, Represents the modulus of the drone rotation matrix, represents the rotor angular velocity.

[0042] The attitude control uses a PID controller, and the formula is:

[0043]

[0044] Among them, θ d represents the target attitude angle, θ represents the current attitude angle, t represents the unit time, τ represents the delay, K p ,K i ,K d represent the proportional, integral, and differential coefficients respectively.

[0045] The communication module, including the radio communication module and high-gain antenna, ensures stable data transmission with the ground control station and supports remote control and real-time monitoring.

[0046] The sensing module, including ultrasonic sensors, infrared sensors, lidar, etc., monitors the surrounding environment in real time, provides obstacle detection and precise spatial positioning.

[0047] Mechanical arm, each drone is equipped with an L-shaped telescopic mechanical arm for completing precise docking and grasping operations in aerial assembly tasks. The mechanical arm is driven by a servo motor and can be precisely bent, stretched and rotated. The mechanical arm used in the present invention is an L-shaped cantilever 3, which is also equipped with a magnetic suction device, which can adjust the posture through collision during the docking process, and complete the docking task through the telescopic and rotating actions of the L-shaped cantilever 3.

[0048] The robot arm is composed of multi-degree-of-freedom joints and servo motors, and each joint is equipped with a high-precision position sensor (such as a rotary encoder). Through closed-loop control, the robot arm can bend and straighten precisely to adapt to different task requirements.

[0049] The multi-degree-of-freedom motion of the robot arm can be described by DH parameters, and its end pose matrix is:

[0050]

[0051] Where n represents the number of assumed joint rods, T i represents the transformation matrix, θ i represents the joint angle, α i Indicates the connecting rod length, d i Indicates the connecting rod offset.

[0052] like Figure 2 , Figure 3 As shown, the magnetic suction component 4 is distributed on the spherical shell 2. For example, a lower through hole is provided at the bottom of the spherical shell of the first UAV, and a lower magnetic suction component is provided around the lower through hole. An upper through hole is provided at the top of the spherical shell of the second UAV, and an upper magnetic suction component is also provided around the upper through hole. The magnetic poles of the lower magnetic suction component correspond to those of the upper magnetic suction component. When the two UAVs are close to each other, the initial docking is completed through the adsorption of the magnetic suction component 4, but the magnetic suction force should not be too large, otherwise it will affect the separation of the UAVs, which results in the magnetic connection of the two UAVs being unstable and further fixed by a mechanical arm. Therefore, after the magnetic docking, the screws at the bottom of the retractable lower cantilever of the first UAV extend downward through the lower through hole, and the screw holes at the top of the retractable upper cantilever of the second UAV extend upward through the upper through hole, so that the screws and the screw holes are docked, and the tightening of the screws and the screw holes is completed by the horizontal rotation of the second UAV relative to the first UAV.

[0053] Furthermore, the magnetic attraction component 4 can be an annular structure surrounding the through hole, so that when the second UAV rotates horizontally relative to the first UAV, the initial docking between the first UAV and the second UAV can be maintained, avoiding initial docking separation due to posture changes caused by rotation.

[0054] Furthermore, the upper and lower cantilevers are retractable L-shaped cantilevers 3. As the upper and lower cantilevers continue to extend and the drones rotate relative to each other, the initial docking of the magnetic suction component 4 is peeled off while the screw holes are fixedly connected. The second drone flips vertically to one side to form a parallel state with the first drone. At this time, the upper and lower cantilevers are changed from the original 1-shaped vertical connection mode to an L-shaped mirror-symmetrical connection mode.

[0055] In another embodiment, the two drones are respectively provided with left and right retractable cantilevers, and are initially docked through the magnetic suction components 4 on the periphery of their respective side through holes. Then, by extending the left and right retractable cantilevers and combining the second drone with a vertical flip relative to the first drone, the initial connection of the magnetic suction component 4 is directly stripped, the screws and screw holes are tightened, and the two drones are placed in parallel.

[0056] Regarding the separation of the drones, the preliminary connection of the magnetic suction component 4 can be restored by retracting the cantilever first, and then the fixed connection between the cantilevers can be separated by relatively rotating the drones. Since the magnetic suction force is not set large, for example, the magnetic suction force is set to be smaller than the lift of the drone, the preliminary connection of the magnetic suction component 4 can be peeled off by lifting the first drone upward or the second drone descending, or the two drones flying in opposite directions. On the other hand, the fixed connection of the cantilever can be directly separated by relatively rotating the drones first, without considering the magnetic suction problem.

[0057] The magnetic attraction component 4 can also use an electromagnet, but the electromagnet is too bulky and consumes a lot of energy, which will bring too much load and energy consumption to the drone.

[0058] The formula for magnetic adsorption force is as follows:

[0059]

[0060] Among them, μ0 is the magnetic permeability of vacuum, m1 and m2 are the magnetic moments of the UAV and the aerial station respectively, and r is the distance between them.

[0061] like Figure 4 As shown, the aerial assembly operation based on the collision docking control of the UAV includes the following steps:

[0062] Step 1: Preliminary mission planning and path generation;

[0063] Before the mission is launched, the ground control station sets the take-off position, target position and the target of the required operation (such as object grabbing, assembly position, etc.) of the drone through the mission planning system. The ground control station starts the mission by transmitting control instructions.

[0064] Path planning: Using classic path planning algorithms (such as A*, Dijkstra or RRT), the ground control station calculates the optimal flight path based on the mission requirements and the current position of the drone. Path planning takes into account factors such as obstacles and flight airspace restrictions to ensure safety during flight.

[0065] Step 2: UAV take-off and approach phase;

[0066] According to the mission requirements, the UAV takes off and starts flying to the target location. During the flight, the control module adjusts the attitude through the flight control unit to ensure a smooth flight.

[0067] Position adjustment and hovering,After reaching the target position, the drone uses the PID control algorithm to adjust its attitude, maintain the hovering state, ensure stability in the air, and prepare for precise docking.

[0068] After the drone takes off, it adjusts to the target position, and the hovering stability is ensured by the following formula:

[0069]

[0070] Among them, u represents the control quantity, x d Represents the horizontal coordinate of the connecting rod, y d represents the vertical coordinate of the connecting rod, x represents the horizontal coordinate of the drone, y represents the vertical coordinate of the drone, t represents the unit time, K p ,K i ,K d represent the proportional, integral, and differential coefficients respectively.

[0071] Step 3: Collision docking and robot arm operation, including the following steps:

[0072] Step 3.1: Initial docking (magnetic attraction stage);

[0073] When the two spherical drones 1 approach each other, they detect each other's positions through radar or sensors and adjust their attitudes so that the relative speed between the two approaches zero. At this time, the magnetic attraction device between the two drones starts to work to ensure the initial docking stability.

[0074] Magnetic attraction formula:

[0075]

[0076] Among them, μ0 is the magnetic permeability of vacuum, m1 and m2 are the magnetic moments of the UAV and the aerial station respectively, and r is the distance between them.

[0077] Step 3.2: Precise docking of the robotic arm;

[0078] After the initial magnetic docking, the cantilever begins to extend and approach the precise docking process. At this stage, the screws at one end of the cantilever dock with the screw holes at the other end, and a more solid physical connection is achieved through precise robotic arm control.

[0079] Threaded joint mechanics: Assuming the outer diameter of the screw is d, the pitch is p, and the torque is T, the axial force F generated by the screw during rotation axial It can be estimated by the following formula:

[0080]

[0081] Among them, r is the radius of the screw when it rotates, and T is the torque applied. In order to complete the docking, the cantilever needs to accurately adjust the torque through the rotation action to ensure that the screw is firmly screwed into the screw hole.

[0082] Robotic arm position control: Using kinematics formula:

[0083] q=q0+J·Δx

[0084] Among them, q represents the joint coordinates of the robot arm, q0 represents the joint coordinates of the drone, J represents the Jacobian matrix, and Δx represents the displacement increment of the target position.

[0085] Attitude adjustment: After the precise docking of the robotic arms is completed, the cantilever will return from a straight shape to an L shape, and the attitudes of the two drones will begin to adjust so that they are on the same horizontal plane. Through this process, the relative attitude of the drones is ensured to be stable, forming the final cooperative flight attitude.

[0086] Attitude adjustment formula: The attitude adjustment process involves the application of rotation matrix. Assuming the rotation matrix of the drone is R and its control input is angular velocity ω, the attitude adjustment process of the drone can be described as:

[0087]

[0088] in, is the antisymmetric matrix of angular velocity to represent the dynamic changes during rotation.

[0089] Stability control: The stability of the attitude can be achieved through a PID controller to ensure that the drone can quickly and stably adjust to a uniform attitude. The control signal u is:

[0090]

[0091] Among them, K p ,K i ,K d They represent the proportional, integral and differential coefficients respectively, and e represents the deviation between the current posture and the target posture.

[0092] Implementation Case 1: Four drones collide and dock in the air to merge and carry heavy objects

[0093] Background: This embodiment describes in detail the process of four drones performing collision docking and merging in the air and performing a heavy object handling task. The process includes mathematical analysis and algorithm details in flight path optimization, magnetic suction docking, mechanical arm precision docking, sensor fusion, relative position coordination, and collaborative control.

[0094] 1. Drone takeoff and flight path calculation

[0095] 1.1 Flight path optimization and positioning

[0096] After takeoff, the four drones fly to the predetermined target area for docking. To ensure the flight path is optimized and avoid collisions, the flight path of each drone needs to be accurately calculated.

[0097] Displacement calculation:

[0098] The initial position of each drone is P i =(x i ,y i ,z i ), the target position is T i =(x i ′ ,y i ′ ,z i ′ ), the displacement of drone i

[0099] ΔP i The calculation formula is:

[0100] ΔP i =T i -P i =(x i ′ -x i ,y i ′ -y i ,z i ′ -z i )

[0101] Each drone needs to adjust its flight path according to this displacement.

[0102] Flight Path Optimization:

[0103] The goal of flight path optimization is to ensure that the UAV flies smoothly and avoids sudden changes. Common optimization algorithms include Bezier curves and spline curves. Assume that the flight path of the UAV is composed of several control points P0, P1, P2, …, P n Defined, the flight path P(t) can be represented by the following spline curve:

[0104]

[0105] Among them, α i (t) is the weight function of the control point, P i is the coordinate of the control point, t is the path parameter. Weight function α i (t) The calculation usually relies on a specific interpolation algorithm, such as B-spline interpolation.

[0106] 1.2 Flight Control System

[0107] The flight control of a drone relies on real-time adjustments of its attitude, speed, and deviation from the target position.

[0108] Acceleration control of drone:

[0109] The acceleration of the drone is i Determined by the difference between thrust and gravity:

[0110]

[0111] Among them, F thrust is the thrust, m i is the mass of the drone, and g is the acceleration due to gravity.

[0112] Equations of motion:

[0113] According to Newton's second law, the equation of motion is:

[0114]

[0115] Among them, r i is the position of the drone, F drag is the air resistance, and the air resistance calculation formula is:

[0116]

[0117] Among them, C d is the drag coefficient, ρ is the air density, A is the windward area, v i is the speed of the drone.

[0118] PID control:

[0119] The flight control of UAV is usually carried out through PID control. The controller is based on the current error e(t) = x target -x current The output u(t) of the PID controller is:

[0120]

[0121] Among them, K p ,K i ,K d are the proportional, integral and derivative gains respectively.

[0122] 2. Collision docking and magnetic attraction calculation

[0123] 2.1 Magnetic attraction

[0124] During the collision docking process, each drone is equipped with a magnetic device for initial docking. The magnitude of the magnetic attraction is related to the distance between the two drones. The magnetic attraction F mag Estimated by the following formula:

[0125]

[0126] Among them, μ0 is the magnetic permeability of vacuum, m1 and m2 are the magnetic moments of the UAV and the aerial station respectively, and r is the distance between them.

[0127] 2.2 Magnetic docking process

[0128] When the drone approaches, the magnetic device attracts it through magnetic force. Assume the relative speed between drone i and drone j is v ij , the relative motion between them follows the following equation:

[0129]

[0130] Among them, Δrij is the relative position between UAVs i and j, is air resistance.

[0131] 2.3 Docking through magnetic attraction

[0132] When the magnetic attraction F m When it reaches a certain value, it will reduce the relative speed between the two drones and make them approach steadily and complete the preliminary docking. At this time, the drone's mechanical arm begins to extend for precise docking.

[0133] 3. Kinematic control and precise docking of the robot arm

[0134] 3.1 Robotic arm kinematics

[0135] Each drone is equipped with a multi-degree-of-freedom robotic arm. In order to accurately control the extension and rotation of the robotic arm, inverse kinematics is used. Assume that the end position of the robotic arm is P end =(x,y,z), the joint angles are θ1, θ2, θ3, and the relationship between the end position of the robot and the joint angles is:

[0136] T=A1(θ1)A2(θ2)A3(θ3)

[0137] Among them, A i (θ i ) is the transformation matrix of the i-th joint. Each transformation matrix describes the spatial position transformation of the robot from one joint to the next.

[0138] 3.2 Precise docking of robotic arms

[0139] During the precise docking process, the robot arm needs to adjust its posture to ensure that the end point completely coincides with the target joint point. Assume that the target position of the end point of the robot arm is P target , the calculation formula of the end position adjustment angle Δθ is:

[0140] Δθ=θ final -θ initial

[0141] During this process, the robot arm solves the position change through inverse kinematics and adjusts the joint angles to reach the target position.

[0142] 4. Sensor fusion and relative position coordination

[0143] 4.1 Sensor Fusion

[0144] The drone obtains real-time attitude, position and speed information through sensors (such as IMU, GPS, and visual sensors). This information is fused through Kalman filtering to improve positioning accuracy. The state update formula of Kalman filtering is:

[0145] x k+1 =Ax k +Bu k +w k

[0146] Among them, A is the state transfer matrix, B is the control matrix, and u k is the control input, w k is the process noise.

[0147] 4.2 Relative position adjustment

[0148] The drones need to adjust their relative positions to ensure that the docking process is accurate. Let the relative position between drones i and j be Δr ij , the relative position is adjusted by the following control equation:

[0149]

[0150] Among them, k is the control gain, which ensures the relative position between the drones is stable and avoids excessive oscillation.

[0151] 5. Merge and task execution

[0152] 5.1 Joint Control and Task Allocation

[0153] When the four drones are successfully merged into a consortium, they will perform subsequent tasks. Task scheduling is assigned based on the status of each sub-drone. The target task T is the object handling task. The consortium control system coordinates flight and task execution through a distributed control algorithm:

[0154]

[0155] Among them, T i Subtasks to be performed for each sub-drone.

[0156] 5.2 Coordinated flight and precise control

[0157] During the flight of the consortium, the speed and position of each sub-UAV are adjusted in real time through the distributed PID control algorithm to ensure that they coordinate and accurately complete the task. The control system needs to constantly monitor the status of each sub-UAV to ensure the stability of the coordinated flight. Through this series of mathematical models and control algorithms, the four UAVs can accurately perform mid-air collision docking and merging, and complete complex tasks such as heavy object handling.

[0158] Implementation Case 2: Aerial Building Structure Assembly System

[0159] Background and Problem

[0160] In some special construction scenarios, especially during the construction of high-rise buildings or temporary facilities, traditional ground and climbing equipment cannot meet the needs of efficient and safe structural installation. For example, in a small area in the city center or in a complex high-altitude environment, the use of traditional lifting equipment may be limited by space, or there may be major safety hazards in high-altitude operations. Using the collision docking technology of the spherical drone 1, fast and accurate structural splicing can be achieved in the air.

[0161] Core Technology

[0162] Collision docking: The spherical drone 1 docks in the air through collision and attitude adjustment, and is precisely positioned and assembled through a magnetic suction system or a robotic arm.

[0163] Modular construction: Building components are pre-designed into modules that can be quickly docked. Each module can independently carry a certain structural load and can be quickly installed in the air.

[0164] Robotic arm and automatic adjustment: The initial docking is completed by magnetic attraction, and the robotic arm is further used to accurately adjust, fix and complete the connection between components.

[0165] Detailed mathematical analysis and details

[0166] 1. Flight path and collision simulation

[0167] During the aerial assembly process, the drone needs to plan its flight path accurately based on the location and height of the building components. The optimization of the flight path is to ensure that the drone can complete the docking in the shortest time while avoiding collisions or errors.

[0168] Flight trajectory optimization: In order to ensure that the drone can fly to the designated location efficiently and smoothly, we need to calculate the shortest path of the flight trajectory, considering the following factors:

[0169] Aerodynamics: Adjust the flight path by solving the impact of airflow during flight.

[0170] Rotational torque: The rotation angle of the drone needs to be controlled during flight to avoid unstable posture due to angle error.

[0171] Dynamic constraints: Calculate the maximum acceleration, steering angle, etc. during flight to avoid exceeding the physical limits of the drone during flight.

[0172] The shortest path algorithm (such as Dijkstra or A*) is used to optimize the flight path of the drone to ensure that multiple drones can reach the target location quickly and efficiently when working together.

[0173] Mathematical model:

[0174]

[0175] Among them, r(t) represents the position of the UAV at time t, and the goal is to minimize the flight time from the starting point to the end point.

[0176] 2. Dynamic analysis of collision and magnetic docking

[0177] The collision and attitude adjustment of the spherical drone 1 is a crucial part of the entire assembly process. Whether it is the initial docking through magnetic attraction or the precise adjustment through the robotic arm, the dynamic characteristics of the drone need to be considered during the flight.

[0178] Collision modeling: We can use a rigid collision model to describe the collision process between the drone and the building module. Let the mass of the drone be m, the velocity be v, and the angular velocity after the collision be ω. Assuming the collision is elastic, the momentum and angular momentum are conserved.

[0179] Conservation of momentum: mv1+Mv2=mv ′ 1+Mv ′ 2

[0180] Conservation of angular momentum: I1ω1+I2ω2=I1ω ′ 1+I2ω ′ 2

[0181] Among them, m and M are the masses of the drone and building module respectively, v1 and v2 are the velocities before the collision, and ω1 and ω2 are the angular velocities before the collision.

[0182] Magnetic attraction model: When the drone is initially docked through magnetic attraction, it attracts the building module through the magnetic field. The magnitude of the magnetic attraction and the distribution of the magnetic field can be described by the following formula:

[0183]

[0184] Among them, F mag is the magnetic attraction, μ0 is the magnetic permeability of vacuum, m1 and m2 are the magnetic moments of the two objects, and r is the distance between the two objects.

[0185] When calculating the magnetic attraction force, the magnetic attraction layout on the surface of the drone and the building module needs to be considered to ensure that the magnetic force can accurately act on the contact surface between the drone and the building module.

[0186] 3. Error model and adjustment to ensure docking accuracy

[0187] For high-precision assembly, especially in the process of multi-UAV collaboration, the accumulation of errors must be considered. Errors may come from many aspects, such as navigation errors, deviations caused by collisions, and robot arm operation errors.

[0188] Error propagation model: In order to improve the overall accuracy of the system, a Kalman filter is used to fuse the data of multiple sensors to reduce the error propagation. Suppose there is a multi-sensor system, in which the measurement value of each sensor i is z i , whose noise is v i , the following filtering model can be established:

[0189] z i =H i x+v i

[0190] Where x is the system state vector, H i is the observation matrix, v i To measure noise.

[0191] The Kalman filter reduces the error by updating the estimate of the system state, as analyzed above.

[0192] Post-collision adjustment strategy: When using the robotic arm to make precise adjustments after a collision, consider the kinematic limitations of the robotic arm and the docking accuracy with the building module. For example, if the drone has a slight deviation during docking, the robotic arm can adjust through inverse kinematics to achieve the required accuracy:

[0193] q final =q initial +Δq

[0194] Among them, q initial and q final are the initial and target joint angles of the robot arm, respectively, and Δq is the fine-tuning amount.

[0195] 4. Structural stability analysis

[0196] After the modules are docked, a structural stability analysis must be performed to ensure that the connection between the modules will not loosen or shift due to external forces (such as wind, vibration, etc.).

[0197] Static analysis: Use the finite element method (FEM) to analyze the joints to ensure that the module connection will not become unstable under external forces. Construct the stiffness matrix of each connection point, calculate the stress distribution of the system, and verify the connection strength between modules:

[0198] F=Kd

[0199] Among them, F is the external force, K is the stiffness matrix, and d is the displacement.

[0200] Vibration and dynamic stability: In the external environment, the modules of the drone after docking may be affected by vibration or other external forces. Therefore, it is necessary to conduct vibration analysis on the connection between modules to ensure the stability of the structure during long-term use:

[0201]

[0202] Among them, m is the mass matrix, c is the damping matrix, k is the stiffness matrix, and F ext For external motivation.

[0203] 5. Specific steps

[0204] Same as Case 1, no further details will be given.

[0205] Implementation Case 3: Aerial UAV Station: Detailed Analysis and Implementation

[0206] Background and Problem

[0207] With the rapid development of drone technology, more and more drones are being used for various high-altitude tasks, such as construction, environmental monitoring, and rescue missions. In order to improve the collaboration efficiency of drone groups and their ability to cope with complex scenarios, the concept of an aerial drone hub (ADH) is proposed. This hub not only provides real-time control and data support for drone groups, but also provides energy supply, maintenance, monitoring, and efficient scheduling management for drones according to demand.

[0208] The main goal of the aerial drone master station system is to improve the scheduling and coordination efficiency of drone groups to a higher level while ensuring their stability and autonomy in complex environments.

[0209] Key technical points

[0210] Air flight control and command system: The UAV center allocates flight missions, plans routes and coordinates management through real-time data analysis, communication and control systems.

[0211] Energy supply and maintenance: Through wireless charging, energy management systems or material supply drones, ensure that the drone group can continue to perform tasks, especially in long-term missions, the aerial drone station has great advantages.

[0212] Automatic docking and repair of drones: Using collision docking technology and equipment such as robotic arms, the aerial drone station can assist in receiving, repairing, loading and redeploying drones to ensure mission continuity.

[0213] Autonomous navigation and obstacle avoidance capabilities: The aerial terminal can autonomously adjust the flight path in complex urban airspace or other special environments to avoid conflicts with other aerial objects.

[0214] Mathematical analysis and details of the aerial drone station

[0215] 1. Flight management and path planning

[0216] A core function of the aerial drone station is scheduling and flight path management. In order to ensure that multiple drones can work together efficiently in the air, avoid collisions, and respond quickly according to mission requirements, a flight path optimization algorithm is required.

[0217] Path optimization model: Utilizes the reinforcement learning algorithm in dynamic programming (DP) or artificial intelligence (AI) to calculate the optimal flight route for a group of drones based on environmental information, target points, flight speed, airspace restrictions, and other factors.

[0218] Assume that the state of the drone group is x(t), where x(t) = [x1(t), x2(t), …, x N (t)] T represents the position of each drone in the drone group. The goal is to optimize the mission objective by controlling the path of each drone.

[0219] Objective function (such as the shortest path problem):

[0220]

[0221] Among them, v i (t) represents the flight speed of the i-th UAV at time t.

[0222] Using A* or Dijkstra algorithm, a path is designed for each drone in the airspace to ensure that all drones can complete the task within the specified time and avoid collision.

[0223] 2. Air docking and resupply

[0224] When performing a mission, the drone may need to regularly replace batteries, transmit data, or update missions. An important function of the aerial drone station is to provide support for these tasks. In order to achieve autonomous docking, energy replenishment, and material replenishment of drones, we use the following technologies.

[0225] Collision docking and docking accuracy control:

[0226] When the drone approaches the air station, it uses collision docking and magnetic attraction systems to locate and dock. First, the position of the target drone is monitored using radar or visual sensors to calculate its relative position to the air station. Then, the flight path is adjusted to ensure docking accuracy.

[0227] The dynamic model of docking can be described by using the elastic collision model and the magnetic field effect:

[0228] Elastic collision model:

[0229] At the time of collision, assuming that the contact between the drone and the aerial station is rigid and elastic, the momentum conservation equation is:

[0230] m1v1+m2v2=m1v ′ 1+m2v ′ 2

[0231] Among them, v1 and v2 are the velocities of the drone and the aerial station before the collision, v ′ 1 and v ′ 2 is the velocity after the collision.

[0232] Magnetic docking:

[0233] The Aerial Terminal precisely adjusts the docking by controlling the magnetic attraction. The effect of the magnetic attraction is given by the following formula:

[0234]

[0235] Among them, μ0 is the magnetic permeability of vacuum, m1 and m2 are the magnetic moments of the UAV and the aerial station respectively, and r is the distance between them.

[0236] After the magnetic force provides the initial docking power, the robotic arm is used for precise positioning to ensure the stability of the docking structure.

[0237] 3. In-flight energy management and automatic charging

[0238] The aerial drone station not only needs to perform flight management and docking, but also needs to provide charging and material supply for drones. To achieve this goal, automatic charging equipment and supply systems need to be deployed on the station.

[0239] Wireless charging technology: Using electromagnetic induction or laser charging technology, wireless charging of drones can be realized in the air. By controlling the matching of charging frequency and drone receiving capacity, the drone can always maintain sufficient power during flight.

[0240] Power model of wireless charging:

[0241]

[0242] Among them, P received is the received power, P transmitted is the transmitted power, A receiver is the area of ​​the receiver and r is the transmission distance.

[0243] When the drone is connected to charging, the charging strategy is adjusted according to the real-time power demand and charging rate to ensure that charging is completed in a short time.

[0244] Material supply system: The aerial drone station can also carry supply drones (Supply Drone), which can carry the required materials (such as new sensors, batteries, tools, etc.) and dock and exchange with the operating drones in the air. The material supply operation can be completed through precise path planning and control systems to ensure that the mission is not interrupted.

[0245] 4. Autonomous obstacle avoidance and flight stability

[0246] Aerial drone stations must avoid collisions with other objects (such as buildings, other drones, birds, etc.) during flight, especially in complex aerial environments.

[0247] Real-time obstacle avoidance algorithm: It uses deep learning-based obstacle avoidance algorithms (such as YOLO, PointNet, etc.) combined with sensors such as lidar and cameras to analyze the surrounding environment in real time. When the system detects potential obstacles, it adjusts the flight path to avoid collisions.

[0248] Attitude control and stability: The aerial drone station must ensure that it always maintains a stable attitude during flight to avoid system instability caused by external interference (such as wind and airflow). By using a PID controller (proportional-integral-differential controller) or a model-based control method (such as LQR control), the aerial station can maintain stable flight in a dynamic environment.

[0249] Posture control model:

[0250]

[0251] Among them, q is the attitude vector, M is the inertia matrix, C is the Coriolis matrix, K is the stiffness matrix, and τ is the input.

[0252] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some or all of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A collaborative flight system based on collision docking, comprising a drone (1) and a connecting arm arranged thereon, characterized in that: The control module of the drone (1) senses external forces through collision detection to adjust the drone's posture for docking with a connecting arm, the connecting arm being provided with a threaded connector that is arranged to cooperate with another drone, and the control module controls the drone to perform axial rotation with the connecting arm as an axis to complete the fixed connection between the drones.

2. The collision docking-based cooperative flight system according to claim 1, characterized in that: The connecting arm is arranged transversely on one side of the drone (1), and the drone is flipped vertically with the connecting arm as an axis so that the threaded connection piece on the connecting arm is threadedly connected to another drone.

3. The collision docking-based cooperative flight system according to claim 2, characterized in that: The connecting arm is provided with a threaded connector, so that the threaded connector is matched with the threaded connector on another UAV connecting arm for threaded connection.

4. The collision docking-based cooperative flight system according to claim 3, characterized in that: The connecting arm is a bendable L-shaped cantilever (3) arranged at the bottom of the drone. When docking, the cantilever is in a vertical downward state, so as to be connected to the cantilever with the value of the bottom of another inverted flying drone facing upward through a thread. After docking, the other drone is caused to roll horizontally so that the two drones are parallel, and the bottoms are connected through the cantilever bent into an L shape.

5. A collaborative flight system based on collision docking of spherical drones supporting the collaborative flight system based on collision docking according to claim 1, characterized in that: The drone (1) is provided with a spherical shell (2) provided with a through hole, the connecting arm is a retractable connecting arm, the threaded connecting piece of the connecting arm contacts a threaded connecting piece provided in cooperation with another drone through the through hole from inside the spherical shell (2), and rotates with the connecting arm as an axis to complete a fixed connection.

6. A cooperative flight system based on magnetic capture and collision docking of spherical drones supporting the cooperative flight system based on collision docking of spherical drones as claimed in claim 5, characterized in that: A magnetic attraction component (4) is provided on the spherical shell (2) outside the through hole, and the magnetic attraction component (4) is arranged in cooperation with a magnetic attraction component outside the through hole of another drone, so that the drones are initially docked by magnetic attraction, and then fixedly connected by a connecting arm.

7. The cooperative flight system based on magnetic capture and collision docking of a spherical UAV according to claim 6 is characterized in that: The magnetic attraction component (4) is an annular structure surrounding the through hole.

8. The cooperative flight system based on collision and docking of spherical UAVs according to claim 6 is characterized in that: During the separation process of the system, the drones rotate in the opposite direction with the connecting arm as the axis to separate the fixed connection between the drones; the drones fly in the opposite direction based on the magnetic connection to peel off the initial docking of the magnetic suction component (4).

9. The cooperative flight system based on collision and docking of spherical UAVs according to claim 6 is characterized in that: During the connection process of the system, after the retractable connecting arm is fixedly connected to another drone, it continues to extend to allow the initial docking and separation of the magnetic suction component (4). During the separation process of the system, the drone is first rotated in the opposite direction with the connecting arm as the axis to separate the fixed connection between the drones, and then the retractable connecting arm is retracted.

10. The cooperative flight system based on magnetic capture and collision docking of a spherical UAV according to claim 6 is characterized in that: The connecting arm is a retractable, bent L-shaped cantilever (3) arranged at the bottom of the drone; During the connection process of the system, the drone (1) and another inverted drone are preliminarily docked via the respective bottom magnetic suction components (4), and after being threadedly connected via the cantilever in the vertical state, the cantilever is further extended to separate the preliminarily docked magnetic suction components (4), and the other drone is rolled horizontally so that the two drones are parallel, and the bottoms are connected via the cantilever bent into an L shape; During the separation process of the system, the other drone is caused to roll vertically to directly separate the fixed connection between the L-shaped cantilevers.

Images