A rescue unmanned aerial vehicle based on a compound wing vertical take-off and landing platform and a control method thereof

Abstract

The application relates to the technical field of unmanned planes, and discloses a rescue unmanned plane based on a compound-wing vertical take-off and landing platform and a control method thereof. The method monitors the tension of a sling through a high-precision tension sensor, estimates and compensates for system disturbance caused by cable swing and crosswind by adopting a compound control strategy combining an extended state observer and a recursive sliding mode controller, dynamically adjusts the thrust distribution of a rotor and the included angle of a rotating plane, effectively suppresses the "nodding phenomenon" and cargo swing when the load suddenly increases, and significantly improves the flight stability and safety of the unmanned plane during the hoisting and rescue of materials.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of compound wing unmanned aerial vehicle, and particularly relates to a rescue unmanned aerial vehicle based on a compound wing vertical take-off and landing platform and a control method thereof. BACKGROUND

[0002] The compound wing vertical take-off and landing unmanned aerial vehicle is a new unmanned aerial vehicle combining multi-rotor and fixed wing technologies, and realizes the dual advantages of vertical take-off and high-speed cruising through the cooperation of two power systems. The core design adopts multiple rotors (usually four to eight motors) to provide vertical lift, so that the unmanned aerial vehicle can take off and land in place without a runway. At the same time, the fixed wing and the aerodynamic layout are provided, and the forward flight power is provided by the propulsion motor or the propeller in the flat flight, and the lift generated by the wing is used to maintain efficient cruising. This configuration solves the field limitation of the traditional fixed wing unmanned aerial vehicle relying on the launch / slide take-off, and overcomes the defects of short flight time and low speed of the multi-rotor unmanned aerial vehicle. This unmanned aerial vehicle is widely used in logistics transportation, power inspection, disaster rescue and other fields, and becomes an important development direction of the industrial unmanned aerial vehicle market. In the future, with the progress of battery energy density and flight control system, the compound wing unmanned aerial vehicle will further evolve towards long flight time and intelligence;

[0003] However, when the rescue compound wing unmanned aerial vehicle suspends heavy objects such as rescue materials for take-off and landing, the dynamic change of the center of gravity is the core factor causing flight risk, because the rescue materials are connected with the unmanned aerial vehicle through hooks or slings, the overall center of gravity of the unmanned aerial vehicle will immediately deviate to the suspension point direction, when the weight of the rescue materials is large, the center of gravity deviation may exceed the design safety threshold, therefore, it is necessary to adjust the flight attitude of the unmanned aerial vehicle when the heavy object is pulled up and put down, so as to ensure that the compound wing unmanned aerial vehicle remains stable during operation and does not occur side tilt and crash. SUMMARY

[0004] The present application aims at solving the problems in the prior art, and provides a rescue unmanned aerial vehicle based on a compound wing vertical take-off and landing platform and a control method thereof.

[0005] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme:

[0006] A control method of a rescue unmanned aerial vehicle based on a compound wing vertical take-off and landing platform, comprising the following steps:

[0007] S1: in the vertical take-off stage, the change of the sling tension is monitored in real time by a high-precision tension sensor, and an extended state observer is used to estimate the total disturbance of the system caused by cable swing and crosswind, and the state equation of the extended state observer is as follows:

[0008]

[0009] Wherein, x1 is the system state estimation, x2 is the state differential estimation, x3 is the total disturbance estimation, y is the system output, u is the control input, β1, β2, β3 are observer gains, b is the control efficiency coefficient;

[0010] The recursive sliding mode controller generates the control input u=k·e+δ based on the attitude error e and the disturbance estimation value δ, wherein k is the control gain, and the disturbance estimation value δ is used to maintain the stable attitude of the unmanned aerial vehicle by adjusting the speed difference of the four rotors;

[0011] S2: Load surge response stage, when detecting the cable tension surge ΔT, calculating the required additional lift according to the formula ΔF_rotor=(m_load·g+ΔT) / 4, wherein m_load is the load mass, g is the gravity acceleration, and adjusting the thrust distribution of the four rotors in real time through the recursive sliding mode control;

[0012] S3: Steady flight stage, continuously monitoring the cable tension change, and suppressing the weight swing through rotor thrust distribution and tail deflection angle adjustment.

[0013] Preferred: In S1, the speed difference of the four rotors is adjusted by the PD controller to generate roll, pitch and yaw moments, and the proportional coefficient K_p of the PD controller is 0.8-1.2 and the differential coefficient K_d is 0.1-0.3.

[0014] Preferred: In S1, the control rotor rotation plane is 1-2° from the horizontal plane, and the generated horizontal component force f_h=f_v·sinθ is used to enhance the attitude adjustment response speed, wherein θ is the inclination angle and f_v is the rotor vertical thrust.

[0015] Preferred: In S1, when the crosswind speed exceeds 8m / s, the unmanned aerial vehicle head is turned to the wind direction through the rudder and rotor differential speed cooperative control, and the aerodynamic characteristics are used to maintain stable take-off.

[0016] Preferred: In S2, when the attitude sensor detects that the roll angle exceeds 10°, the controller outputs a correction instruction, and the unbalanced side rotor thrust is preferentially increased, and the correction instruction satisfies f_h=f_v·sinθ, wherein θ is the current inclination angle.

[0017] Preferred: In S2, when the tension sensor detects that the load weight exceeds the safety threshold, the flight control system triggers the power output limiting program, and limits the maximum rotor thrust within 120% of the rated value.

[0018] Preferred: In S3, the rotor horizontal component force f_h=f_v·sinθ is used to suppress the weight swing in cooperation with the tail deflection, wherein the rotor inclination angle θ is dynamically adjusted according to the swing amplitude, and the adjustment range is 0-5°.

[0019] A rescue unmanned aerial vehicle based on a compound wing vertical take-off and landing platform uses the above method for vertical take-off and landing. The structure comprises: an unmanned aerial vehicle body adopting a compound wing configuration combining a fixed wing and a four-rotor wing; four independently driven rotors symmetrically arranged at the four corners of the fuselage, providing the required lift for vertical take-off and landing and hovering; a fixed wing connected to the fuselage, adopting a high lift-drag ratio airfoil, providing the main lift in the horizontal flight stage; an electric propeller arranged at the rear of the fuselage, providing the thrust for horizontal flight; a high-precision tension sensor installed at the connection between the sling and the fuselage, monitoring the real-time change of the sling tension; a flight control system comprising a processor and a memory, configured to execute the above control method.

[0020] The beneficial effects of the present application are:

[0021] The present application adopts a double-layer control architecture combining an extended state observer (ESO) and a recursive sliding mode control (RSMC), cooperates with a high-precision tension sensor to monitor the real-time change of the sling tension, effectively suppresses the "nodding phenomenon" and cargo swing problem of the traditional rescue unmanned aerial vehicle when lifting heavy objects, significantly improves the anti-disturbance ability and response speed of the system; through the cooperative control mechanism of dynamic adjustment of the rotor plane angle and deflection of the tail wing, multi-dimensional precise regulation and control of roll, pitch and yaw are realized, so that the unmanned aerial vehicle can still maintain a stable flight attitude under complex airflow conditions; at the same time, the thrust distribution algorithm based on tension change can quickly and adaptively adjust the rotor thrust when the load suddenly increases, avoiding the control instability caused by power lag in the traditional method, and comprehensively improving the flight stability and safety of the rescue unmanned aerial vehicle in the hoisting operation. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 is a structural schematic diagram of a control method of a rescue unmanned aerial vehicle based on a compound wing vertical take-off and landing platform proposed by the present application. DETAILED DESCRIPTION

[0023] The technical solutions of the present application will be further described in detail below in combination with specific embodiments.

[0024] The embodiments of the present application will be described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be understood as a limitation of the present application.

[0025] Embodiment 1:

[0026] The present application discloses a control method of a rescue unmanned aerial vehicle based on a compound wing vertical take-off and landing platform, as shown in Figure 1 The method comprises the following steps:

[0027] S1: During the vertical takeoff phase, the changes in sling tension are monitored in real time using a high-precision tension sensor. An extended state observer is used to estimate the total system disturbance caused by cable swaying and crosswinds. The state equation of the extended state observer is as follows:

[0028]

[0029] Where x1 is the system state estimate, x2 is the state differential estimate, x3 is the total disturbance estimate, y is the system output, u is the control input, β1, β2, and β3 are the observer gains, and b is the control efficiency coefficient;

[0030] The recursive sliding mode controller generates control input u=k·e+δ based on attitude error e and disturbance estimate δ, where k is the control gain and δ is the disturbance estimate. It maintains the stable attitude of the UAV by adjusting the speed difference of the four rotors.

[0031] S2: Load surge response phase. When a sudden increase in cable tension ΔT is detected, the required additional lift is calculated according to the formula ΔF_rotor=(m_load·g+ΔT) / 4, where m_load is the load mass and g is the gravitational acceleration. The thrust distribution of the four rotors is adjusted in real time through recursive sliding mode control.

[0032] S3: During the stable flight phase, the changes in sling tension are continuously monitored, and the swaying of the load is suppressed by adjusting the rotor thrust distribution and tail fin deflection angle.

[0033] In a further embodiment:

[0034] S1: Vertical Takeoff Phase. The UAV takes off vertically, suspending the rescue supplies via a bottom sling. During this process, the load remains on the ground. The UAV activates its outer rotors for takeoff, which must provide sufficient thrust to overcome the lift caused by its own weight and the tension of the sling. Roll and pitch adjustments are made through differential control of different rotor speeds. The combined effects of adaptive control algorithms, fuzzy neural networks, ESO, and RSMC work together to counteract sling tension and sway disturbances, maintain a stable attitude, and monitor load changes in real time to dynamically optimize the control strategy.

[0035] As the drone's altitude and speed change, the tension of the sling gradually decreases. A high-precision tension sensor, installed between the sling and the fuselage, monitors these changes. The sensor has a range of 0-200 kg and an accuracy of 0.1% of full scale. The controller within the fuselage uses data on altitude, drone ascent speed, wind speed, and sling tension to extend the state observer's real-time estimation of disturbances such as cable sway and crosswinds, compensating for these disturbances and preventing rotor speed saturation due to disturbances, thus reducing flutter and improving robustness.

[0036] The PD controller controls the four rotors mounted on the drone to rotate at different speeds. The speed difference between the four rotors causes the drone's fuselage to roll or pitch, generating a yaw moment to counteract and offset the impact of the sling tension on the drone's takeoff stability. Simultaneously, the rotors mounted at the four corners of the drone can rotate under the drive of motors, finely adjusting the angle of the rotor's rotation plane. Based on the current drone status, the rotor rotation is controlled to maintain an angle of 1-2° with the horizontal plane. This dynamically adjusted horizontal component enhances the roll and pitch effects, and more quickly counteracts the influence of sling tension on the drone's attitude.

[0037] The rudder at the tail of the drone is adjusted synchronously. In crosswinds, the rudder deflection works in conjunction with the rotor's anti-torque to reduce the risk of sideslip. When the crosswind is too strong, the drone uses the rudder in conjunction with the speed difference of the four rotors and adjusts the rotation plane angle of the rotors to quickly rotate the drone horizontally, bringing the nose into the wind. As the drone ascends, the sling is straightened, reducing the lateral traction force and stabilizing the drone's attitude.

[0038] This embodiment realizes a complete three-stage control process. Under a 15kg load, vertical takeoff stability is improved by 67%, and wind resistance reaches 10m / s, effectively solving the control instability problem caused by power lag in traditional methods.

[0039] Example 2:

[0040] During vertical takeoff, the PD controller adjusts the speed difference of the four rotors to generate roll, pitch, and yaw moments. The proportional gain K_p of the PD controller is set to 1.0, and the differential gain K_d is set to 0.2, with a range of K_p = 0.8-1.2 and K_d = 0.1-0.3. The rotor rotation plane is controlled to form a 1.5° angle with the horizontal plane, with an angle range of 1-2°. The generated horizontal component force f_h = f_v·sinθ is used to enhance the attitude adjustment response speed, where θ is the tilt angle and f_v is the rotor vertical thrust. When the vertical thrust of a single rotor f_v = 200N, the generated horizontal component force f_h = 200 × sin(1.5°) ≈ 5.24N. When a 2° roll angle deviation of the UAV is detected, the PD controller calculates the control output and adjusts the speed difference of the four rotors, correcting the roll angle deviation to within 0.5° within 0.3s. By precisely configuring the parameters of the PD controller and adjusting the rotor angle, a rapid attitude response was achieved. The roll response time was reduced from 0.5s in the traditional method to 0.3s, the control accuracy was improved by 40%, and the attitude oscillation during takeoff was effectively suppressed.

[0041] Example 3:

[0042] When the crosswind speed exceeds 8 m / s, the UAV's nose is turned into the wind direction through coordinated control of the rudder and rotor differential speeds. Specifically: a rudder deflection of 25° generates a yaw moment; the left front and right rear rotors increase their rpm by 150 rpm, while the right front and left rear rotors decrease their rpm by 150 rpm; the yaw moment generated by the rotor speed difference is 8 N·m, and the rudder generates a yaw moment of 5 N·m. The total yaw moment is 13 N·m, enabling the UAV to complete a 60° nose turn within 3 seconds, achieving wind-facing flight. The sling generates a traction force in the same direction as the wind due to the wind's movement; the fuselage counteracts this traction force by controlling the pitch angle, effectively controlling the fuselage attitude and achieving stable vertical takeoff. The coordinated control of the rudder and rotor differential speeds allows the UAV to remain stable under 10 m / s crosswind conditions, with attitude angle deviation controlled within 3°, improving wind resistance by 67% compared to traditional methods.

[0043] Example 4:

[0044] The control methods for the lifting and airlifting phase of materials specifically include the following steps:

[0045] S2: Lifting and Ascent Phase. Before lifting, the sling is slack. The moment the rescue supplies or other heavy objects are lifted off the ground, the sling suddenly tightens, generating an impact load. This sudden tension is transmitted to the drone through the attachment point, causing high-frequency oscillations in the drone's body, commonly known as the "nose-nodding" phenomenon. The drone needs to provide additional lift instantaneously to offset the entire weight of the suspended object. At this point, the following condition must be met: F_rotor ≥ m_drone × g + m_load × g, where m_load is the mass of the suspended object and m_drone is the mass of the drone. Thrust distribution must be completed instantaneously.

[0046] S2.1: Vertical Shock Resistance Control. High-precision strain gauge sensors are used to monitor load weight changes in real time, and recursive sliding mode control (RSMC) is combined with ESO disturbance compensation. ESO estimates disturbances in real time, and when a sudden increase in load is detected, the flight control system immediately triggers the thrust distribution algorithm.

[0047] ΔF_rotor = (m_load × g + ΔT) / n

[0048] Where ΔT is the sudden increase in tension, and n is the number of rotors (n=4).

[0049] S2.2: Swinging Torque Compensation Control. A swinging torque will occur when the sling connection point is not perfectly aligned with the UAV's center of gravity. An extended state observer estimates disturbances such as sling swing and crosswinds in real time and compensates for these disturbances in the controller.

[0050] u = k × e + δ_hat

[0051] Where u is the control input, e is the error, and δ_hat is the disturbance estimate.

[0052] The correction angle is calculated based on the attitude sensor (IMU) data:

[0053] θ_correction = k × θ_swing

[0054] Where θ_correction is the correction angle, θ_swing is the sway angle of the weight, and k is the gain coefficient (taken as 0.8-1.2).

[0055] Real-time adjustment of rotor speed difference to counteract the torque caused by cable tension. For example, when the roll angle exceeds the threshold of 10°, the horizontal component force is calculated when the propeller axis tilts by 10°:

[0056] f_h = f_v × sin(10°) ≈ 0.17f_v

[0057] Increase the thrust of the right rotor to restore balance.

[0058] Rapid pitch response is achieved by adjusting the rotor's rotation plane angle, while roll response is adjusted synchronously using the rudder. A fuzzy neural network is used to adaptively adjust the controller gain based on load changes, avoiding overshoot or response lag and optimizing dynamic response parameters.

[0059] Safety protection mechanism: When the load sensor detects that the load is close to the safety threshold (set to 90% of the maximum load), it triggers an automatic power output limit or emergency descent program to avoid motor overload. If the high-precision strain gauge sensor detects a sudden and significant change in the sling tension (such as a drop of more than 50%), it indicates that the sling has broken or the load has lost weight. The descent program is then activated, the rotor decelerates rapidly, and the lift is reduced to minimize the risk of a fall. At the same time, GPS return-to-home or manual takeover mode is activated.

[0060] This embodiment achieves a rapid and safe response to sudden load increases. When a 15kg load increase is detected, the response time is reduced from 0.8s in the traditional method to 0.2s, an improvement of 75%; the amplitude of the head-nodding phenomenon is reduced by 70%; and the attitude recovery time is reduced by 60% in the case of sling connection point misalignment, significantly improving the safety and reliability of the system.

[0061] Example 5:

[0062] In Phase S3, the drone suspends a load and operates in normal flight mode. During stable flight, high-precision load cells or force sensors are used to monitor changes in tension on the cable in real time, accurately understanding the state of the suspended load and monitoring its swaying in the air.

[0063] Through a sensor feedback system, the flight control system acquires real-time information on the weight of the load and its changes, dynamically adjusting the rotor thrust. In the air, the yaw moment caused by the load's swaying is automatically counteracted by f_h = f_v * sinθ, with the horizontal component providing additional damping. The rotor tilt angle θ is dynamically adjusted according to the sway amplitude, with an adjustment range of 0-5°.

[0064] When the swing amplitude is less than 5°, θ = 1°;

[0065] When the swing amplitude is 5°-10°, θ=3°;

[0066] When the swing amplitude is greater than 10°, θ = 5°;

[0067] The tail fin, in conjunction with the rotator-adjustable rotor, prioritizes the suppression of sudden swaying, forming a horizontal force damping mechanism and a tail fin-rotor angle-priority suppression strategy. Through dynamic rotor angle adjustment, the sway decay time of the heavy object is shortened from 3.2s to 1.5s, the sway amplitude is reduced by 60%, and flight stability is significantly improved.

[0068] Example 6:

[0069] The Extended State Observer (ESO-RSMC) estimates disturbances such as cable sway and crosswinds in real time and compensates them for the controller. The state equations of the ESO are as follows:

[0070]

[0071] in:

[0072] x1 represents the system state estimate; x2 represents the state differential estimate; x3 represents the total disturbance estimate; y represents the system output (UAV attitude angle); u represents the control input (rotor thrust command); β1, β2, and β3 represent the observer gains, which are set to 100, 300, and 1000 respectively; b represents the control efficiency coefficient, which is set to 0.8; the control input u = k*e + δ, where u is the control input, e is the error, and δ is the disturbance estimate. This control architecture enables real-time estimation and compensation of disturbances when the sling connection point is not completely aligned with the UAV's center of gravity, resulting in a swinging torque. ESO achieves a disturbance estimation accuracy of 92%, enabling the control system to compensate for 85% of external disturbances in the feedforward stage, improving system robustness by 112% compared to traditional methods.

[0073] This method combines dynamic adjustment of the rotor rotation plane angle with active rudder control, achieving precise multi-dimensional response to roll, pitch, and yaw. It not only enhances the sensitivity of roll / pitch control, but also significantly reduces the risk of sideslip through the synergistic effect of rudder deflection and rotor anti-torque, overcoming the limitations of traditional methods that rely solely on rotor differential control and improving stability under complex airflow conditions.

[0074] By deeply coupling high-precision strain gauge sensors with RSMC, active prediction and real-time compensation for impact loads are achieved. When a sudden increase in cable tension is detected, the flight control system dynamically allocates rotor thrust and recursively estimates the disturbance using ESO, thereby completing thrust adjustment before the motor response lags. This effectively suppresses the "nodding phenomenon" caused by insufficient power reserve in traditional methods and avoids the fluttering problem caused by rotor speed saturation.

[0075] To address the periodic disturbances caused by the sling swing, a horizontal component damping mechanism and a tail rotor angle priority suppression strategy are employed. In the initial stage of the load swing, the yaw torque is directly offset by the horizontal component of the rotor. Furthermore, the rapid response characteristics of the tail adjustable rotor are utilized to prioritize the suppression of sudden swings. Compared to the traditional passive suppression method that relies solely on PID control, this significantly improves the swing attenuation efficiency and reduces the interference of attitude oscillations on the flight trajectory.

[0076] In another embodiment, a rescue drone based on a compound wing vertical takeoff and landing platform uses the above-described method for vertical takeoff and landing. Its structure includes: a drone body employing a compound wing configuration combining a fixed wing and a quadcopter; four independently driven rotors symmetrically arranged at the four corners of the fuselage, providing lift required for vertical takeoff, landing, and hovering; a fixed wing connected to the fuselage, employing a high lift-to-drag ratio airfoil, providing primary lift during horizontal flight; an electric propeller located at the rear of the fuselage, providing thrust for horizontal flight; a high-precision tension sensor installed at the connection point between the sling and the fuselage, monitoring changes in sling tension in real time; and a flight control system, including a processor and memory, configured to execute the above-described control method.

[0077] The above description represents a preferred embodiment of the present invention. The scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention, combined with existing technology or common knowledge, and within the spirit and principles of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A control method for a rescue unmanned aerial vehicle based on a compound wing vertical takeoff and landing platform, characterized in that, Includes the following steps: S1: During the vertical takeoff phase, the changes in sling tension are monitored in real time using a high-precision tension sensor. An extended state observer is used to estimate the total system disturbance caused by cable swaying and crosswinds. The state equation of the extended state observer is as follows: ； ； ； Where x1 is the system state estimate, x2 is the state differential estimate, x3 is the total disturbance estimate, y is the system output, and u is the control input. , , b is the observer gain, and b is the control efficiency coefficient; The recursive sliding mode controller generates control input u=k·e+δ based on attitude error e and disturbance estimate δ, where k is the control gain, and maintains the stable attitude of the UAV by adjusting the speed difference of the four rotors. S2: Load surge response phase. When a sudden increase in sling tension ΔT is detected, the required additional lift is calculated according to the formula ΔF_rotor=(m_load·g+ΔT) / 4, where m_load is the load mass and g is the gravitational acceleration. The thrust distribution of the four rotors is adjusted in real time through recursive sliding mode control. S3: During the stable flight phase, continuously monitor changes in sling tension and suppress load swaying by adjusting rotor thrust distribution and tail fin deflection angle.

2. The control method for a rescue UAV based on a compound wing vertical takeoff and landing platform according to claim 1, characterized in that, In S1, the rotational speed difference of the four rotors is adjusted by the PD controller to generate roll, pitch and yaw torques. The proportional coefficient of the PD controller is K_p=0.8-1.2 and the differential coefficient is K_d=0.1-0.

3.

3. The control method for a rescue UAV based on a compound wing vertical takeoff and landing platform according to claim 1, characterized in that, In S1, the rotor rotation plane is controlled to form an angle of 1-2° with the horizontal plane, and the generated horizontal component force f_h=f_v·sinθ is used to enhance the attitude adjustment response speed, where θ is the tilt angle.

4. The control method for a rescue UAV based on a compound wing vertical takeoff and landing platform according to claim 1, characterized in that, In S1, when the crosswind speed exceeds 8m / s, the nose of the UAV is turned towards the wind direction through the coordinated control of the rudder and the rotor differential speed.

5. The control method for a rescue UAV based on a compound wing vertical takeoff and landing platform according to claim 1, characterized in that, In S2, when the attitude sensor detects that the roll angle exceeds 10°, the controller outputs a correction command, which prioritizes increasing the thrust of the unbalanced rotor. The correction command satisfies f_h=f_v·sinθ, where θ is the current tilt angle, f_v is the vertical thrust of the rotor, and f_h is the horizontal component of the rotor force.

6. The control method for a rescue UAV based on a compound wing vertical takeoff and landing platform according to claim 1, characterized in that, In S2, when the tension sensor detects that the load weight exceeds the safety threshold, the flight control system triggers the power output limiting program to limit the maximum rotor thrust to within 120% of the rated value.

7. The control method for a rescue UAV based on a compound wing vertical takeoff and landing platform according to claim 1, characterized in that, In S3, the swaying of the heavy object is suppressed by the combined action of the horizontal component of the rotor force f_h=f_v·sinθ and the deflection of the tail fin. The rotor tilt angle θ is dynamically adjusted according to the sway amplitude, with an adjustment range of 0-5°. f_v is the vertical thrust of the rotor, and f_h is the horizontal component of the rotor force.

8. A rescue drone based on a compound wing vertical takeoff and landing platform, characterized in that, include: The drone body adopts a compound wing configuration that combines fixed wings and quadcopters; Four independently driven rotors symmetrically arranged at the four corners of the fuselage; A fixed wing connected to the fuselage; an electric propeller located at the rear of the fuselage; a high-precision tension sensor installed at the connection point between the sling and the fuselage; and a flight control system comprising a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to implement the control method as described in any one of claims 1-7.

Images