A flight and land control method of a coaxial amphibious unmanned aerial vehicle

CN122653259BActive Publication Date: 2026-10-27NANJING UNIV OF SCI & TECH
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202611161022.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-08-03
Publication Date
2026-10-27
Estimated Expiration
2046-08-03

AI Technical Summary

Technical Problem

[0004]本发明解决的技术问题是:解决现有两栖无人机飞行与陆行共用一套动力系统,在陆行模式下机身姿态难以稳定控制的问题,特别是在躺倒姿态下,旋翼的反扭矩容易导致机身发生非期望的自转,严重影响操控精度的技术问题

Benefits of technology

[0013]与现有技术相比,本发明具有以下技术优势:(1)本发明陆行模式动力传输控制的控制策略可以在执行直线前进时,所述控制系统控制机身左右两侧所述旋翼电机同速同向运转。进行后退行动时,则控制双侧所述旋翼电机同步反转。所述控制系统控制两侧所述旋翼电机产生大小不一的转速,使得无人机通过两侧所述行驶轮实现差速转弯。所述控制系统控制左右两侧所述旋翼电机以绝对值相等但方向相反的转速运行。所述无人机利用后部展开的所述脚架碳板以保持尾部平衡。位于所述脚架碳板末端的辅助轮存在与地面的作用力,其与所述脚架碳板的物理长度形成的力臂,共同构成一个抵抗机身自转的阻尼力矩,以防止非期望的原地打转现象的发生。(2)本发明陆行模式下的动力传输路径,利用桨叶推动陆行推杆带动行驶轮转动,实现了飞行电机与陆行动力的复用,显著降低了结构重量和系统复杂度。同时,针对共轴无人机躺倒后的自转问题,提出了利用后部脚架碳板及末端辅助轮形成阻尼力矩的解决方案,有效保证了陆行方向的稳定性。此外,本发明分别构建了飞行模式与陆行模式下的精确动力学模型,并设计了相应的控制策略,实现了两种模式下的高精度、高稳定性控制,极大地拓展了无人机的作业能力和应用场景。

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122653259B_ABST
    Figure CN122653259B_ABST
Patent Text Reader

Abstract

The application provides a flight and land control method of a coaxial amphibious unmanned aerial vehicle, comprising the following steps of: obtaining a desired space position quantity in a flight mode, and solving a posture angle signal and a lift signal by a position controller and a speed controller in combination with a state variable in a dynamic model of the unmanned aerial vehicle in the flight mode; inputting the posture angle signal and the state variable into an attitude control module to obtain a desired torque signal; giving control signals of upper and lower motors and a rudder motor according to the desired torque signal and the lift signal by using a control instruction distributor, and controlling an actuator to execute a flight task according to the control signals. The control strategy of land mode power transmission control can control the left and right rotors of the body to rotate at the same speed and in the same direction when executing straight forward movement. When performing a backward movement, the left and right rotors are controlled to be reversely rotated synchronously.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) control technology, and in particular to a method for controlling the flight and land movement of a coaxial amphibious UAV. Background Technology

[0002] In recent years, with the inclusion of the low-altitude economy in the national strategic emerging industry cluster development plan, the unmanned aerial vehicle (UAV) industry has ushered in unprecedented development opportunities. Amphibious UAVs, with their long endurance, strong obstacle-crossing capabilities, and adaptability to complex environments, have demonstrated enormous application potential in short-distance material transportation, resource exploration, disaster search and rescue, and obstacle removal in confined spaces. However, how to achieve efficient switching and stable control between flight and land-based modes remains a key technical challenge in this field.

[0003] In the prior art, patent CN113199915A discloses a deformable amphibious ducted unmanned aerial vehicle (UAV) and its operation mode. This UAV has a large frontal projection in flight mode to ensure flight capability, and employs a tracked structure in land mode to enhance obstacle-crossing ability. However, this solution achieves mode switching through mechanical deformation of the power structure, which not only increases structural complexity and reduces system robustness, but also results in poor turning flexibility due to the fixed tracked structure, limiting its maneuverability in confined spaces. Patent CN118907466A discloses a wheeled amphibious ducted UAV and its operation mode. This solution has a compact structure, sharing a single power system for flight and land travel, and offers a certain level of safety and hovering performance. However, its power source in land mode mainly relies on the air thrust generated by the low-speed rotation of the rotor, which has lower propulsion efficiency and higher energy consumption compared to mechanical transmission; simultaneously, its wheeled structure requires a high degree of ground flatness, limiting its obstacle-crossing performance. In addition, existing amphibious drones often face the problem of unstable control of their fuselage attitude in land mode. Especially in the lying position, the counter-torque of the rotor can easily cause the fuselage to rotate unintended, which seriously affects the control accuracy. Summary of the Invention

[0004] The technical problem solved by this invention is to address the issue that existing amphibious drones share a single power system for both flight and land travel, making it difficult to maintain stable control of the fuselage attitude in land travel mode. In particular, in the prone position, the counter-torque of the rotor can easily cause the fuselage to rotate undesirably, seriously affecting the control precision.

[0005] To address the aforementioned technical problems, this invention provides a flight and land-based control method for a coaxial amphibious unmanned aerial vehicle (UAV). In flight mode, the control method includes: Step S100: Obtain the desired spatial position quantity and solve it with the state variables in the dynamic model of the UAV flight mode through the position controller and speed controller to obtain the attitude angle signal and lift signal; Step S200: Input the attitude angle signal and state variables into the control system to obtain the desired torque signal; Step S300: Using the control command distributor, control signals are given to the upper motor, lower motor and servo motor according to the desired torque signal and lift signal, and the actuator is controlled to perform the flight mission according to the control signals. The actuator includes two sets of rotor motors and servo motors arranged on the fuselage. Control methods in land travel mode include: Step S400: Position the drone in a lying-down position on land. In step S500, the control system performs deviation processing on the desired pose and the actual pose to obtain the desired fuselage motion information, and calculates the target control torque of the left and right rotor motors based on the desired fuselage motion information. Step S600: The target control torque is converted into a motor drive signal and input to the left and right rotor motors respectively. The rotation of the two rotor motors drives the two driving wheels to rotate around the axis respectively, realizing the differential power output for land travel. In step S700, during the land travel process, the frictional force between the auxiliary wheels and the ground is used to generate a reverse damping torque to balance the rotation of the fuselage.

[0006] Furthermore, the expression for the UAV dynamics model in the flight mode is as follows: , , , , , , , , , , , , in, This represents the X-axis coordinate obtained by the UAV in the ground coordinate system. This represents the Y-axis coordinate obtained by the UAV in the ground coordinate system. This represents the Z-axis coordinate obtained by the UAV in the ground coordinate system. This represents the flight speed of the UAV along the X-axis as observed in the ground coordinate system. This represents the UAV's flight speed along the Y-axis as observed in the ground coordinate system. This represents the UAV's flight speed along the Z-axis as observed in the ground coordinate system. This indicates the lift signal of the drone. This represents the drag coefficient experienced by the drone during flight. i =1,2,3 represent the X-axis, Y-axis, and Z-axis, respectively. Indicates pitch angle information. Indicates roll angle information. Indicates yaw angle information. Indicates the roll rate. Indicates pitch rate, Indicates the yaw rate. This indicates the moment of inertia of the drone along the X-axis. This indicates the moment of inertia of the drone's axes. This provides the rotational inertia information for the UAV's Z-axis. This indicates the rolling torque of the drone. This indicates the pitch moment of the drone. This indicates the yaw moment of the drone. M For the quality of drones, g It is the acceleration due to gravity. , , , , , , , , , , , They are respectively , , , , , , , θ , , p , q , r The first derivative.

[0007] Furthermore, the attitude angle signal and the state variable are input into the control system to obtain the desired torque signal, the mathematical expression of which is: , , , in, Indicates the rotor speed. Indicates the rotational speed of the lower rotor. It is a coefficient determined by the aerodynamic characteristics of the blades. Indicates the yaw moment coefficient. This represents the length of the lever arm from the rotor center to the drone's center of mass. The control system includes an attitude loop and an angular velocity loop: The attitude loop receives the difference between the desired attitude angle and the current attitude angle, calculates and outputs the desired angular velocity command through an LSTM network that has been processed by Gaussian smoothing filtering; The angular velocity loop receives the error between the desired angular velocity command and the current angular velocity, and outputs the three-axis control torque, the mathematical expression of which is: , , , in, This represents the control torque along the X-axis. This indicates the control torque along the Y-axis. This represents the control torque along the Z-axis.

[0008] Furthermore, the three-axis control torque and throttle signal The inverse model of the pre-trained MLP neural network is mapped to generate yaw angle commands for two servo motors and two sets of rotor motor speed commands.

[0009] Furthermore, the mathematical expression for the land-based UAV dynamics model is: , , , , , , , , , , , , in, , These are the angular velocities of the left and right rotor motors, respectively. , These are the angular accelerations of the left and right rotor motors, respectively. , These represent the rotational speeds of the left and right rotor motors, respectively. , The actual driving torques of the left and right rotor motors are respectively , These represent the output electromagnetic torques of the left and right rotor motors, respectively. The transmission efficiency constant is... Acceleration of the drone line, The fuselage angular velocity, For the fuselage yaw angle acceleration, The radius of the traveling wheel, d It is half the wheelbase of the left and right wheels. For the weight of the drone fuselage, For the mass of a single traveling wheel, Let be the moment of inertia of a single-sided traveling wheel about its own axis of rotation. The longitudinal rolling resistance of the entire machine, This is the combined damping torque experienced by the system when it turns. The coefficient of rolling friction is The coefficient of sliding friction for the tail auxiliary wheel. The normal load shared by this wheel, This is the equivalent force arm length of the tire's contact surface with the ground. The resistance torque caused by the slight slippage of the main driving wheels The equivalent translational mass of the entire machine. This is the equivalent rotational inertia of the entire machine.

[0010] Furthermore, the motion control in the land driving mode includes straight-line motion control, turning motion control, and U-turn control; The straight-line motion control includes controlling the rotor motors on the left and right sides of the fuselage to rotate at the same speed and in the same direction to achieve forward movement, or controlling the rotor motors on the left and right sides of the fuselage to rotate in opposite directions synchronously to achieve backward movement. The turning motion control includes controlling the rotation speed difference between the two rotor motors to achieve steering through the linear speed difference between the two driving wheels; The in-situ turn control includes controlling the left and right rotor motors to run at speeds that are equal in absolute value but opposite in direction, so that the fuselage generates a pure rotational torque around the geometric center.

[0011] Furthermore, step S500 specifically includes: Step S510: The control system receives the desired position and heading angle command of the UAV in the global coordinate system, and compares it with the actual position and heading angle fed back by the sensor to obtain the pose deviation. Step S520: The position controller calculates the desired fuselage linear velocity and desired fuselage heading angular velocity to drive the UAV toward the target based on the pose deviation. Step S530: Based on the desired fuselage linear velocity and desired fuselage angular velocity, perform inverse transformation using the chassis kinematics model to calculate the desired rotational angular velocities of the left and right driving wheels respectively. Step S540: Compare the calculated desired rotational angular velocity with the actual wheel speed fed back by the encoder, and generate a desired angular acceleration command for the left and right driving wheels through the speed controller; Step S550: Substitute the desired angular acceleration command and driving resistance into the chassis dynamics model for inverse calculation to calculate the target control torque required to achieve the motion on the left and right rotor motors respectively.

[0012] Furthermore, step S600 specifically includes: Step S610: The calculated target control torque is converted into a corresponding motor drive signal and sent to the left and right rotor motors through the electronic speed controller to drive the rotor motors to start rotating at low speed. In step S620, while driving the driving wheels to move, the control system outputs a holding torque to the servo motor to fix the position of its output shaft, thereby achieving the locked state of the servo motor. In step S630, the low-speed rotating rotor motor drives the propeller clip to rotate. During the rotation, the propeller clip continuously pushes the land push rod, and the generated thrust is transmitted to the bearing through the reinforced carbon plate structure, thereby driving the running wheels installed on both sides of the fuselage to rotate around the axis.

[0013] Compared with the prior art, the present invention has the following technical advantages: (1) The control strategy of the power transmission control of the land mode of the present invention can control the rotor motors on the left and right sides of the fuselage to run at the same speed and in the same direction when performing straight forward movement. When performing backward movement, the rotor motors on both sides are controlled to reverse synchronously. The control system controls the rotor motors on both sides to generate different speeds, so that the UAV can achieve differential turning through the driving wheels on both sides. The control system controls the rotor motors on the left and right sides to run at speeds with equal absolute values ​​but opposite directions. The UAV uses the carbon plate of the landing gear that unfolds at the rear to maintain tail balance. The auxiliary wheel at the end of the carbon plate of the landing gear has a force with the ground, and the lever arm formed by it and the physical length of the carbon plate of the landing gear together constitutes a damping torque to resist the rotation of the fuselage, so as to prevent the occurrence of undesirable spinning in place. (2) The power transmission path in the land mode of the present invention uses the propeller to drive the land push rod to drive the driving wheel to rotate, realizing the reuse of flight motor and land driving force, which significantly reduces structural weight and system complexity. Meanwhile, to address the issue of autorotation when the coaxial UAV lies down, a solution is proposed that utilizes the carbon plates on the rear landing gear and the auxiliary wheels at the end to generate a damping torque, effectively ensuring stability in the land-walking direction. Furthermore, this invention constructs precise dynamic models for both flight and land-walking modes and designs corresponding control strategies, achieving high-precision and high-stability control in both modes, greatly expanding the operational capabilities and application scenarios of the UAV. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the basic flow of the control method of the present invention.

[0015] Figure 2 This is a schematic diagram of the flight control system of the present invention.

[0016] Figure 3 This is a schematic diagram of the flight mode structure of the present invention.

[0017] Figure 4 This is a top-down view of the flight mode of the present invention.

[0018] Figure 5 This is a schematic diagram of the land-based control system of the present invention.

[0019] Figure 6 This is a schematic diagram of the land travel mode results of the present invention.

[0020] Reference numerals: 1. Propeller clip, 2. Propeller blade, 3. Rotor motor, 4. Travel wheel, 5. Land travel push rod, 6. Servo, 7. Control system, 8. Telescopic motor, 9. Auxiliary wheel, 10. Electronic speed controller, 11. Carbon fiber plate of the landing gear, 12. Fixed spring. Detailed Implementation

[0021] Reference Figures 1 to 6 This embodiment provides a flight and land-based control method for a coaxial amphibious unmanned aerial vehicle. In flight mode, the control method includes: S1, obtain the desired spatial position quantity, and solve it with the state variables in the dynamic model of the UAV flight mode through the position controller and velocity controller to obtain the attitude angle signal and lift signal. The UAV's fuselage is generally vertical during flight. In this embodiment, the desired spatial position refers to the target position coordinates of the UAV in three-dimensional space set by the operator or the automatic control system 7, used to guide the UAV from its current position to the target position. In this embodiment, the desired spatial position is sent by the ground control station or automatically generated by the onboard computer, input to the position controller, compared with the actual position fed back by the sensors, and after two levels of calculation by the position controller and the speed controller, finally output attitude angle signals and lift signals, which are transmitted to the attitude control module. The desired spatial position specifically includes the target longitude coordinates, target latitude coordinates, and target altitude coordinates.

[0022] The dynamics model in UAV flight mode refers to a set of mathematical equations describing the force and motion relationships of the UAV during flight, used to predict and control the UAV's dynamic behavior. The dynamics model is stored in the flight control computer, serving as the foundation for the control algorithm. It receives state variable inputs and outputs predicted motion states. Specifically, the dynamics model in UAV flight mode includes position dynamics equations, attitude kinematics equations, and attitude dynamics equations.

[0023] S2, the attitude angle signal and state variables are input to the attitude control module 7 to obtain the desired torque signal. In this embodiment, the attitude control module refers to the algorithm module responsible for converting the desired attitude into control torque, used to achieve attitude stabilization and tracking of the UAV. The attitude control module receives the desired attitude angle signal from the outer loop (position / velocity loop), compares it with the actual attitude angle fed back by the inertial measurement unit (IMU), calculates the control law, and outputs the three-axis desired torque signal to the control command distributor.

[0024] S3 utilizes the control command distributor to provide control signals to the upper motor, lower motor, and servo motor based on the desired torque and lift signals. The actuators, including two sets of rotor motors 3 and servo motors 6 mounted on the fuselage, are then controlled to perform flight tasks according to these control signals. In this embodiment, the rotor motors in flight mode are used as the power source for land-walking mode. In land-walking mode, the UAV is in a reclining position. By controlling the motors to rotate at low speed, the propellers drive the land-walking push rod, which in turn rotates the travel wheels. This achieves power system reuse, eliminating the need for additional dedicated land-walking motors, significantly reducing fuselage weight, and improving flight endurance. Simultaneously, the carbon fiber plates of the rear-mounted landing gear and the auxiliary wheels at the end interact with the ground to generate damping torque, effectively solving the fuselage rotation problem caused by rotor counter-torque in the reclining state of the coaxial UAV, ensuring stability in the land-walking direction.

[0025] The mathematical expression for the UAV dynamics model in flight mode is: , , , , , , , , , , , , in, This represents the X-axis coordinate obtained by the UAV in the ground coordinate system. This represents the Y-axis coordinate obtained by the UAV in the ground coordinate system. This represents the Z-axis coordinate obtained by the UAV in the ground coordinate system. , and The speed of the drone, specifically This represents the flight speed of the UAV along the X-axis as observed in the ground coordinate system. This represents the UAV's flight speed along the Y-axis as observed in the ground coordinate system. This represents the UAV's flight speed along the Z-axis as observed in the ground coordinate system. This indicates the lift signal of the drone. This represents the drag coefficient experienced by the drone during flight.i =1,2,3 represent the X-axis, Y-axis, and Z-axis, respectively. , and The attitude angle signal of the UAV is specifically... Indicates pitch angle information. Indicates roll angle information. Indicates yaw angle information. , and This refers to the angular rate information of the drone, specifically... Indicates the roll rate. Indicates pitch rate, Indicates the yaw rate. This indicates the moment of inertia of the drone along the X-axis. This indicates the moment of inertia of the drone's axes. This provides the rotational inertia information for the UAV's Z-axis. This indicates the rolling torque of the drone. This indicates the pitch moment of the drone. This indicates the yaw moment of the drone. , , , , , , , , , , , They are respectively , , , , , , , θ , , p , q , r The first derivative; The drone's velocity is converted into roll angle information by the speed controller. and pitch rate Yaw angle information Fixed, input yaw angle information initial value signal and throttle signal .

[0026] The attitude angle signal and state variable are input into attitude control module 7 to obtain the desired torque signal; its mathematical expression is: , , , in, Indicates the rotor speed. Indicates the rotational speed of the lower rotor. It is a coefficient determined by the aerodynamic characteristics of the blades. This represents the yaw moment coefficient, which characterizes the counter-torque effect generated when the upper and lower rotors rotate in opposite directions. This indicates the length of the lever arm from the rotor center to the drone's center of mass.

[0027] The attitude control module 7 includes an attitude loop and an angular velocity loop. The attitude loop receives the difference between the desired attitude angle and the current attitude angle, calculates and outputs the desired angular velocity command through a Gaussian-smoothed LSTM network, and then outputs the desired angular velocity command. In this embodiment, the attitude loop refers to the outermost loop of the control loop, used to correct attitude angle errors. Specifically, the attitude loop includes an error calculation unit, a Gaussian smoothing filter unit, and an LSTM network control unit. The error calculation unit calculates the deviation between the desired and actual values; the Gaussian smoothing filter unit filters out high-frequency noise to prevent overfitting of the LSTM network; and the LSTM network control unit utilizes its memory characteristics to process the nonlinear system and outputs the desired angular velocity. The angular velocity loop is the inner loop of the control loop, used for rapid response to angular velocity commands. The angular velocity loop receives the error between the desired angular velocity command and the current angular velocity, and outputs the three-axis control torque through a PID or similar control algorithm. Specifically, the angular velocity loop includes an angular velocity error calculation unit and a torque calculation unit.

[0028] The angular velocity loop receives the error between the desired angular velocity command and the current angular velocity, and outputs the three-axis control torque, the mathematical expression of which is: , , , Among them, the three-axis control torque includes , and Specifically This represents the control torque along the X-axis. This indicates the control torque along the Y-axis. This represents the control torque along the Z-axis.

[0029] Three-axis control torque and throttle signal The inverse model of the pre-trained MLP neural network is mapped to generate yaw angle commands for two servo motors and speed commands for two rotor motors.

[0030] In this embodiment, an LSTM network is used to process attitude loop errors, which can effectively cope with nonlinear systems and improve the response speed and robustness of attitude control. A precise mapping from attitude control to specific torque output is achieved through the torque calculation formula. The control command distributor is an algorithm module that allocates the total control target (torque and lift) to specific actuators. The control command distributor receives the three-axis control torque and total lift signals, and calculates the control commands for the upper rotor speed, lower rotor speed, and servo deflection angle by solving a system of nonlinear equations or using a neural network model. The control command distributor specifically includes an input interface, an MLP neural network inverse model, and an output interface.

[0031] In this embodiment, during flight mode, the UAV's actuators mainly include two sets of rotor motors and servo motors 6 mounted on the fuselage. The servo motors 6 can rotate left and right, and the motors on the servo motors 6 rotate at high speed according to control signals, thereby causing the foldable propellers 2 to unfold under significant centrifugal force. The motors drive the propellers 2, which unfold under high-speed rotation, providing flight power for the UAV. The driving wheels 4 in flight mode are not connected to the foldable propellers 2 and therefore do not rotate, serving to protect the propellers 2.

[0032] In land travel mode: S4, put the drone into a lying-down posture. In this embodiment, the lying-down posture refers to the state in which the drone's fuselage axis is parallel or nearly parallel to the ground, used for ground movement in land-walking mode. When the control system 7 detects the land-walking mode trigger signal, it controls the motors to reduce their speed, adjusts the center of gravity, or uses a dedicated mechanism to change the drone from a vertical state to a horizontal lying-down state, providing working conditions for the land-walking power transmission mechanism. The lying-down posture specifically includes states such as the drone being placed horizontally, the wheels in contact with the ground, and the tail landing gear deployed.

[0033] S5, the control system 7 in land-walking mode performs a difference operation between the desired pose and the actual pose. In this embodiment, the control system 7 in land-walking mode refers to a hardware and software system specifically designed for handling ground movement tasks, including a pose controller, a speed controller, and a torque controller. The control system 7 receives the actual pose provided by a global positioning system (such as GPS / RTK), compares it with the preset desired trajectory, and generates a control deviation.

[0034] S6 converts the target torque into a motor drive signal, causing the motor to rotate at low speed. The rotating blades continuously push the land-going push rod, driving the traveling wheels on both sides of the fuselage to rotate around their axes to achieve power output. Attitude error processing specifically includes position deviation calculation and heading angle deviation calculation. The position deviation is used to generate linear velocity commands, and the heading angle deviation is used to generate angular velocity commands. Target torque calculation refers to the process of inversely solving for the required output torque of the motor based on the dynamic model. Target torque calculation is completed in the velocity loop and torque loop. First, the desired wheel speed is calculated using inverse kinematics based on the desired speed. Then, the motor torque required to overcome resistance is calculated based on the dynamic model. Target torque calculation specifically includes desired wheel speed calculation and electromechanical coupling dynamics inverse operation.

[0035] S7 utilizes the interaction force between the rear-deployed carbon fiber landing gear and the end-mounted auxiliary wheels and the ground to generate a damping torque, thus suppressing fuselage rotation. In this embodiment, motion control in land-walking mode includes straight-line motion control, turning motion control, and U-turn control. Straight-line motion control involves controlling the left and right rotor motors to operate at the same speed and in the same direction to achieve forward movement, or controlling the left and right rotor motors to synchronously reverse direction to achieve backward movement. Turning motion control involves controlling the rotor motors on both sides to generate a speed difference, achieving turning through the linear speed difference of the two wheels. U-turn control involves controlling the left and right rotor motors to operate at speeds with equal absolute values ​​but opposite directions, causing the fuselage to generate a pure rotational torque around its geometric center. Through differential speed control strategy, flexible maneuverability in land-walking mode is achieved, meeting the motion requirements of different ground scenarios. The damping torque generation mechanism refers to the physical structure used to balance the rotor's anti-torque. In this embodiment, when the UAV lies down, the rear landing gear automatically deploys, and the end-mounted auxiliary wheels contact the ground. The damping torque generation mechanism specifically includes the carbon fiber landing gear, the end-mounted auxiliary wheels, and the ground contact surface. The damping force rectangular principle refers to the physical mechanism of using frictional torque to balance the counter-torque. When the rotor motor rotates, it generates a counter-torque that acts on the fuselage, causing the fuselage to tend to rotate around its vertical axis.

[0036] The motion control in land mode includes straight-line motion control, turning motion control, and U-turn control. Straight-line motion control includes controlling the left and right rotor motors 3 on both sides of the fuselage to run at the same speed and in the same direction to achieve forward movement, or controlling the left and right rotor motors 3 on both sides of the fuselage to reverse synchronously to achieve backward movement. Turning motion control includes controlling the rotor motors 3 on both sides to generate a speed difference, and using the linear speed difference of the two wheels to achieve turning. U-turn control includes controlling the left and right rotor motors 3 on both sides to run at speeds with equal absolute values ​​but opposite directions, so that the fuselage generates a pure rotational torque around the geometric center.

[0037] The control strategy of the control system 7 in land driving mode includes: (1) The control system 7 receives the desired position and heading angle command of the UAV in the global coordinate system and compares it with the actual position and heading angle fed back by the sensor; (2) Using the position controller, the desired fuselage linear velocity and desired yaw rate that drive the UAV to move toward the target are calculated based on the deviation between the desired and actual poses. (3) Based on the desired fuselage linear velocity and yaw rate, the chassis kinematics model is used to perform inverse transformation to calculate the desired rotational angular velocities of the left and right driving wheels 4 respectively. (4) Compare the calculated desired wheel speed with the actual wheel speed fed back by the encoder, and generate the desired angular acceleration command for the left and right wheels through the speed controller; (5) Substitute the desired angular acceleration and driving resistance into the electromechanical coupling dynamics model for inverse calculation, and calculate the target electromagnetic torque required to achieve the motion on the left and right drive motors respectively. (6) The calculated target torque is converted into a corresponding motor drive signal and sent to the left and right rotor motors 3 through the electronic speed controller to drive the rotor motors 3 to start rotating at low speed. (7) While the drive wheel is moving, the control system 7 outputs a holding torque to the servo motor 6 to fix the position of its output shaft, thereby achieving the locking state of the two servo motors 6; (8) The low-speed rotating motor drives the propeller clip 1 to rotate. During the rotation, the propeller clip 1 continuously pushes the land push rod 5 to move. The resulting thrust is transmitted to the head bearing through the reinforced carbon plate structure, driving the running wheels 4 installed on both sides of the fuselage to rotate around the axis.

[0038] Control system 7 continuously monitors the system. When the speed of motor 3 is greater than the speed of the driving wheel 4, the transmission control link is effective; otherwise, the control effect is ineffective. The output signal of control system 7 and the actual response of the system together constitute closed-loop control.

[0039] The mathematical expression for the dynamic model of the land-based UAV is: , , , , , , , , , , , , in, , The angular velocities of the left and right rotor motors (3) are respectively. , These are the angular accelerations of the left and right rotor motors (3), respectively. , The rotational speeds of the left and right rotor motors (3) are respectively. , The actual driving torques of the left and right rotor motors (3) are respectively , The output electromagnetic torques of the left and right rotor motors (3) are respectively. The transmission efficiency constant is... Acceleration of the drone line, The fuselage angular velocity, For the fuselage yaw angle acceleration, The radius of the traveling wheel (4) d It is half the wheelbase of the left and right wheels. For the weight of the drone fuselage, For the mass of the single-sided traveling wheel (4), The moment of inertia of the single-sided traveling wheel (4) about its own axis of rotation. The longitudinal rolling resistance of the entire machine, This is the combined damping torque experienced by the system when it turns. The coefficient of rolling friction is The coefficient of sliding friction of the tail auxiliary wheel (9) is given. The normal load shared by this wheel, This is the equivalent force arm length of the tire's contact surface with the ground. The resistance torque caused by the slight slippage of the main driving wheels The equivalent translational mass of the entire machine. The function is the equivalent rotational inertia of the entire machine. It is a symbolic function.

[0040] In this embodiment, in land-walking mode, the UAV's actuators mainly include two sets of rotor motors, servo motors 6, and two sets of telescopic motors 8 arranged coaxially on the fuselage. In land-walking mode, the two servo motors 6 remain locked. The telescopic motors 8 on the hub push the land-walking push rod 5 to collide with the propeller clamp 1, forming a non-rigid connection with the propeller blades 2, thereby propelling the driving wheels 4 forward. During land-walking, the foldable propeller blades 2 remain retracted because the tension of the springs they are attached to is greater than the centrifugal force caused by the rotational speed. The UAV fuselage is perpendicular to the direction of travel, and the two driving wheels 4 operate symmetrically from left to right. The block diagram of the control system 7 in land-walking mode is shown below. Figure 4 As shown, it consists of three parts: position loop, velocity loop, and torque loop.

[0041] The control strategy for power transmission control in land-based mode allows the control system 7 to control the left and right rotor motors 3 to operate at the same speed and in the same direction when moving forward in a straight line. When moving backward, the control system controls the rotor motors 3 to rotate in opposite directions synchronously. The control system 7 controls the rotor motors 3 on both sides to generate different speeds, enabling the UAV to achieve differential turning via the side wheels 4. The control system 7 controls the left and right rotor motors 3 to operate at speeds with equal absolute values ​​but opposite directions. The UAV uses the carbon fiber landing gear plates deployed at the rear to maintain tail balance. The auxiliary wheel 9 located at the end of the carbon fiber landing gear plates exerts a force on the ground, and the lever arm formed by this force and the physical length of the carbon fiber landing gear plates 11 together constitutes a damping torque to resist the rotation of the fuselage, preventing undesirable spinning in place.

[0042] When switching from flight mode to land mode, when the drone needs to transition from flight to land mode, control system 7 first reduces its flight altitude and detects the landing terrain. After landing, the system sends a "lie down" command, adjusting the fuselage attitude to a horizontal lying position. Subsequently, the rear landing gear deploys, and the auxiliary wheels touch the ground. Control system 7 locks the servo positions and establishes the power transmission link. At this point, the flight control loop is disconnected, the land control loop is activated, and the system enters land standby mode.

[0043] In hill-climbing scenarios in land-walking mode, when the drone encounters a slope, the pose controller detects an increased deviation between the actual and desired pose (e.g., decreased forward speed). The speed controller increases the target torque output based on this deviation. The torque loop increases the motor drive current, thereby increasing the motor output torque. At this time, the force exerted by the propellers on the push rod increases, and the driving torque received by the travel wheels increases, overcoming the gravitational component to achieve hill climbing. If the motor torque exceeds a threshold, the control system triggers overload protection, stops output, and issues an alarm.

[0044] During the transmission link failure detection process, the control system 7 continuously monitors the motor speed and the wheel speed (obtained via an encoder) during land travel. If the motor speed is greater than the wheel speed and the difference is within the normal range, the transmission link is deemed valid, and control continues. If the motor speed is less than or equal to the wheel speed (abnormal situation, such as motor stall or slippage), or the difference is too large, the control effect is deemed invalid. The system immediately cuts off the motor output, activates the braking mechanism, and issues a fault alarm to prevent damage to the mechanism.

[0045] By constructing a dual-mode control architecture for both flight and land travel, and designing a unique power reuse mechanism and rotation suppression scheme for land travel mode, stable ground navigation of a coaxial amphibious UAV was successfully achieved. The complex control allocation problem of the coaxial UAV was solved using an MLP neural network inverse model, and a three-level closed-loop control strategy ensured the accuracy of land travel trajectory tracking. In particular, the damping torque generated by the tail auxiliary wheel cleverly solved the problem of fuselage rotation in the lying-down state, demonstrating significant engineering application value.

Claims

1. A method for flight and land-based control of a coaxial amphibious unmanned aerial vehicle, characterized in that, Control methods in flight mode include: Step S100: Obtain the desired spatial position quantity and solve it with the state variables in the dynamic model of the UAV flight mode through the position controller and speed controller to obtain the attitude angle signal and lift signal; Step S200: Input the attitude angle signal and state variable into the control system (7) to obtain the desired torque signal; Step S300: The control command distributor uses the desired torque signal and lift signal to give control signals to the upper motor, lower motor and servo motor, and controls the actuator to perform the flight mission according to the control signals. The actuator includes two sets of rotor motors (3) and servo motors (6) arranged on the fuselage. Control methods in land travel mode include: Step S400: Position the drone in a lying-down position on land. In step S500, the control system (7) performs deviation processing on the desired pose and the actual pose to obtain the desired fuselage motion information, and calculates the target control torque of the left and right rotor motors (3) based on the desired fuselage motion information. Step S600: The target control torque is converted into a motor drive signal and input to the left and right rotor motors (3) respectively. The rotation of the two rotor motors (3) drives the two driving wheels (4) to rotate around the axis respectively, thereby realizing the differential power output for land travel. Step S700: During the land travel process, the frictional force between the auxiliary wheels and the ground is used to generate a reverse damping torque to balance the rotation of the fuselage. Step S500 specifically includes: Step S510, the control system (7) receives the desired position and heading angle command of the UAV in the global coordinate system, and compares it with the actual position and heading angle fed back by the sensor to obtain the pose deviation; Step S520: The position controller calculates the desired fuselage linear velocity and desired fuselage heading angular velocity to drive the UAV toward the target based on the pose deviation. Step S530: Based on the desired fuselage linear velocity and desired fuselage heading angular velocity, the chassis kinematics model is used to perform inverse transformation to calculate the desired rotational angular velocities of the left and right driving wheels (4) respectively. Step S540: Compare the calculated desired rotational angular velocity with the actual wheel speed fed back by the encoder, and generate a desired angular acceleration command for the left and right driving wheels (4) through the speed controller; Step S550: Substitute the desired angular acceleration command and driving resistance into the chassis dynamics model for inverse calculation, and calculate the target control torque required to achieve the motion applied to the left and right rotor motors (3). Step S600 specifically includes: Step S610: The calculated target control torque is converted into a corresponding motor drive signal and sent to the rotor motors (3) on the left and right sides through the electronic speed controller, driving the rotor motors (3) to start rotating at low speed. In step S620, while driving the driving wheel (4) to move, the control system (7) outputs a holding torque to the servo motor (6) to fix the position of its output shaft, thereby achieving the locked state of the servo motor (6); In step S630, the low-speed rotating rotor motor (3) drives the propeller clip to rotate. During the rotation, the propeller clip continuously pushes the land push rod, and the generated thrust is transmitted to the bearing through the reinforced carbon plate structure, thereby driving the running wheels (4) installed on both sides of the fuselage to rotate around the axis.

2. The method according to claim 1, characterized in that, The expression for the UAV dynamics model under the specified flight mode is: , , , , , , , , , , , , in, This represents the X-axis coordinate obtained by the UAV in the ground coordinate system. This represents the Y-axis coordinate obtained by the UAV in the ground coordinate system. This represents the Z-axis coordinate obtained by the UAV in the ground coordinate system. This represents the flight speed of the UAV along the X-axis as observed in the ground coordinate system. This represents the UAV's flight speed along the Y-axis as observed in the ground coordinate system. This represents the UAV's flight speed along the Z-axis as observed in the ground coordinate system. This indicates the lift signal of the drone. This represents the drag coefficient experienced by the drone during flight. i =1,2,3 represent the X-axis, Y-axis, and Z-axis, respectively. Indicates pitch angle information. Indicates roll angle information. Indicates yaw angle information. Indicates the roll rate. Indicates pitch rate, Indicates the yaw rate. This indicates the moment of inertia of the drone along the X-axis. This indicates the moment of inertia of the drone's axes. This provides the rotational inertia information for the UAV's Z-axis. This indicates the rolling torque of the drone. This indicates the pitch moment of the drone. This indicates the yaw moment of the drone. M For the quality of drones, g It is the acceleration due to gravity. , , , , , , , , , , , They are respectively , , , , , , , θ , , p , q , r The first derivative.

3. The method according to claim 2, characterized in that, The attitude angle signal and the state variable are input into the control system (7) to obtain the desired torque signal, the mathematical expression of which is: , , , in, Indicates the rotor speed. Indicates the rotational speed of the lower rotor. It is a coefficient determined by the aerodynamic characteristics of the blades. Indicates the yaw moment coefficient. This represents the length of the lever arm from the rotor center to the drone's center of mass. The control system (7) includes an attitude loop and an angular velocity loop: The attitude loop receives the difference between the desired attitude angle and the current attitude angle, calculates and outputs the desired angular velocity command through an LSTM network that has been processed by Gaussian smoothing filtering; The angular velocity loop receives the error between the desired angular velocity command and the current angular velocity, and outputs the three-axis control torque, the mathematical expression of which is: , , , in, This represents the control torque along the X-axis. This indicates the control torque along the Y-axis. This represents the control torque along the Z-axis.

4. The method according to claim 3, characterized in that, The three-axis control torque and throttle signal The inverse model of the pre-trained MLP neural network is mapped to generate yaw angle commands for two servo motors and speed commands for two sets of rotor motors (3).

5. The method according to claim 1, characterized in that, The mathematical expression for the dynamic model of the land-based UAV is: , , , , , , , , , , , , in, , The angular velocities of the left and right rotor motors (3) are respectively. , These are the angular accelerations of the left and right rotor motors (3), respectively. , The rotational speeds of the left and right rotor motors (3) are respectively. , The actual driving torques of the left and right rotor motors (3) are respectively , The output electromagnetic torques of the left and right rotor motors (3) are respectively. The transmission efficiency constant is... Acceleration for drone lines, The fuselage angular velocity, For the fuselage yaw angle acceleration, The radius of the traveling wheel (4) d It is half the wheelbase of the left and right wheels. For the weight of the drone fuselage, For the mass of the single-sided traveling wheel (4), The moment of inertia of the single-sided traveling wheel (4) about its own axis of rotation. The longitudinal rolling resistance of the entire machine, This is the combined damping torque experienced by the system when it turns. The coefficient of rolling friction is The coefficient of sliding friction of the tail auxiliary wheel (9) is given. The normal load shared by this wheel, This is the equivalent force arm length of the tire's contact surface with the ground. The resistance torque caused by the slight slippage of the main driving wheels The equivalent translational mass of the entire machine. This is the equivalent rotational inertia of the entire machine.

6. The method according to claim 5, characterized in that, The motion control in the land driving mode includes straight-line motion control, turning motion control, and U-turn control. The straight-line motion control includes controlling the rotor motors (3) on the left and right sides of the fuselage to operate at the same speed and in the same direction to achieve forward movement, or controlling the rotor motors (3) on the left and right sides of the fuselage to rotate in opposite directions synchronously to achieve backward movement; The turning motion control includes controlling the rotation speed difference between the two rotor motors (3) to achieve steering through the linear speed difference between the two driving wheels; The in-situ turn control includes controlling the left and right rotor motors (3) to run at speeds with equal absolute values ​​but opposite directions, so that the fuselage generates a pure rotational torque around the geometric center.

Citation Information

Patent Citations

  • Air-ground amphibious unmanned aerial vehicle and flight control method thereof

    CN113199915A

  • Air-ground amphibious coaxial dual-rotor ducted unmanned aerial vehicle and motion control method thereof

    CN118907466A

  • Coaxial water air dual power unmanned aerial vehicle control system

    CN110515317A

  • Control system of power reuse type land-air amphibious unmanned aerial vehicle

    CN120779845A