Unmanned aerial vehicle autonomous landing control method for fixed-wing unmanned aerial vehicle facing moving base platform

Abstract

This invention proposes an autonomous air-to-ground landing control method for fixed-wing UAVs on a moving base platform. First, a dynamic model of the fixed-wing UAV and a relative motion model between the UAV and the moving base are established. Then, considering the effects of ground effect, the wake of the moving base, and the motion of the moving base on landing accuracy, disturbance models are introduced respectively. Next, a trajectory control law based on a nonlinear L1 control law is designed, which generates real-time trajectory acceleration commands to control the UAV to glide along the optimal trajectory to the target point. A hierarchical attitude control method based on incremental nonlinear dynamic inverse is designed. The outer-loop attitude angle uses a nonlinear dynamic inverse control method to design the angular velocity control law, while the inner-loop angular velocity uses an incremental nonlinear dynamic inverse method to obtain the angular acceleration control law. Finally, incremental power-compensated throttle control is designed, introducing speed error feedback to achieve dynamic adjustment of thrust commands. This invention solves the problem of strong coupling interference during the landing process of UAVs on a dynamic base platform.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) control technology, and more specifically to a method for autonomous air-to-ground landing control of a fixed-wing UAV oriented towards a moving base platform. Background Technology

[0002] With the rapid development of drone technology, drones are being used more and more widely in various fields. Especially in logistics, rescue and other fields, the flexibility and efficiency of drones have brought enormous potential to mission execution.

[0003] In practical applications, to overcome their limited payload and endurance, drones need to take off and land on moving platforms such as vehicles, ships, and offshore oil drilling platforms. Offshore platforms, in particular, present significant challenges due to the complex marine and atmospheric environments they face, requiring stable flight and landing, such as strong winds, large waves, and sea fog. The dynamic and unstable nature of these environments poses considerable challenges to drone takeoff and landing on moving platforms. Therefore, designing docking control methods for drones on moving platforms is crucial. This allows for precise docking in dynamic environments, thereby improving the application range and efficiency of drones, and enhancing their safety and stability.

[0004] However, existing UAV systems still face some challenges in docking and controlling the UAVs with the moving base platform during missions in complex sea and air conditions. These challenges mainly include high coupling of multi-source disturbances, difficulty in tracking time-varying trajectories, and poor terminal landing accuracy.

[0005] First, the UAV's own dynamics and the six-DOF motion of the moving base, especially the pitch, yaw, and heave motions, create strong coupling interference, which is difficult to decouple using traditional single-channel control methods. Second, the continuous motion of the moving base causes the landing point position, velocity, and acceleration to exhibit time-varying characteristics, requiring the UAV to track its non-periodic trajectory in real time. Finally, during the low-speed, high-angle-of-attack phase of landing, the UAV's aerodynamic forces exhibit strong nonlinearity, placing higher demands on flight attitude control.

[0006] Currently, PID control and nonlinear dynamic inverse control are the most widely used methods. However, UAVs encounter various strong disturbances during landing on a moving platform. Traditional PID control methods have poor robustness to model uncertainties and disturbances, easily leading to large tracking errors and severe overshoot. Furthermore, when faced with combined disturbances such as wind field disturbances and platform undulations, traditional PID control has limited adjustment speed, exhibiting slow attitude response and a tendency to chatter. Meanwhile, traditional nonlinear dynamic inverse control is highly dependent on accurate models and has poor compensation capabilities for model uncertainties and external disturbances. When facing complex disturbances, traditional nonlinear dynamic inverse control is less sensitive to disturbances, responding more slowly to disturbances such as wake turbulence and deck motion, resulting in poorer adaptability. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention proposes an autonomous air-to-ground landing control method for fixed-wing UAVs on dynamic base platforms. It employs an adaptive composite control scheme that includes a nonlinear L1 control law and an incremental nonlinear dynamic inverse to solve the problem of strong coupling interference during the UAV's landing process on a dynamic base platform. This method has low dependence on precise dynamic models, strong anti-interference capabilities, and faster response, achieving robust full-modal control of fixed-wing UAVs and greatly improving the flexibility and efficiency of fixed-wing UAV missions.

[0008] The technical solution of this invention is as follows:

[0009] A method for autonomous air-to-ground landing control of a fixed-wing unmanned aerial vehicle (UAV) on a moving base platform includes the following steps:

[0010] Step 1: Treat the drone as a rigid body whose volume and shape do not change, and establish a dynamic model of the fixed-wing drone and a model of the relative motion between the drone and the moving base.

[0011] Step 2: Considering the effects of ground effect, dynamic base wake, and dynamic base motion on landing accuracy, introduce disturbance models for each.

[0012] Step 3: Based on the dynamic model of the fixed-wing UAV established in Step 1 and the relative motion model between the UAV and the moving base, as well as the factors affecting the landing accuracy of the ground effect, the wake of the moving base and the motion of the moving base obtained in Step 2, a trajectory control law based on the nonlinear L1 control law is designed. By generating trajectory acceleration commands in real time, the UAV is controlled to glide along the optimal trajectory to the target point.

[0013] Step 4: Design a hierarchical attitude control method based on incremental nonlinear dynamic inverse. According to the time scale separation principle, a dual closed-loop architecture is adopted. The outer loop attitude angle adopts the nonlinear dynamic inverse control method, and the angular velocity control law is designed through Lyapunov stability theory. The inner loop angular velocity adopts the incremental nonlinear dynamic inverse method to obtain the angular acceleration control law, and an enhanced differential filtering method is designed to estimate the angular acceleration.

[0014] Step 5: Based on Step 3 and Step 4, design incremental power compensation throttle control, introduce speed error feedback, and realize dynamic adjustment of thrust command.

[0015] Furthermore, in step 1, the dynamic model of the fixed-wing UAV includes the dynamic equations for the movement of the center of mass and the dynamic equations for the rotation about the center of mass:

[0016] The dynamic equation for the movement of the center of mass is:

[0017]

[0018] in, The following are the unmanned aerial vehicle body axis systems respectively Triaxial forces, For the quality of drones, The values, in order, are the forward velocity, lateral velocity, and vertical velocity along the unmanned aerial vehicle's axis. These are, in order, the unmanned aerial vehicle's roll rate, pitch rate, and yaw rate along its axis.

[0019] The dynamic equation for the rotation about the center of mass is:

[0020]

[0021] in, The following are the axes around the unmanned aerial vehicle body, respectively. The resultant torque of the three-axis rotation; In order, they are in the unmanned aerial vehicle body axis system Moment of inertia in the three axes; In order, they are in the unmanned aerial vehicle body axis system Inertial product in the three axes.

[0022] Furthermore, in step 1, the relative motion model between the UAV and the moving base includes a relative angular motion model and a relative linear motion model;

[0023] The relative angular motion model includes:

[0024] Relative attitude angle is

[0025]

[0026] Relative angular velocity is

[0027] [ p as s q as s r as s ] = S yes ( [ p a e q a e r a e ] − [ p s e q s e r s e ] )

[0028] Relative angular acceleration is

[0029] [ p ˙ as s q ˙ as s r ˙ as s ] = S yes ( [ p ˙ a e q ˙ a e r ˙ a e ] − [ p ˙ s e q ˙ s e r ˙ s e ] ) + S ˙ yes ( [ p a e q a e r a e ] − [ p s e q s e r s e ] ) − [ p s e q s e r s e ] × [ p as s q as s r as s ]

[0030] in [ f as s i as s P as s ] T , [ p as s q as s r as s ] T and [ p ˙ as s q ˙ as s r ˙ as s ] T These represent the relative attitude angle, relative angular velocity, and relative angular acceleration between the UAV and the moving base platform, respectively. The relative transformation matrix between the UAV and the mobile base platform The elements in; the Euler angles of the drone are [ f a e i a e P a e ] T angular velocity is Oh → a e = [ p a e q a e r a e ] T Angular acceleration is Oh → ˙ a e = [ p ˙ a e q ˙ a e r ˙ a e ] T The Euler angles of the moving base platform are [ f s e i s e P s e ] T angular velocity is Oh → s e = [ p s e q s e r s e ] T Angular acceleration is Oh → ˙ s e = [ p ˙ s e q ˙ s e r ˙ s e ] T ;

[0031] The relative linear motion model includes:

[0032] Relative position is

[0033] [ x as s y as s z as s ] = S yes ( [ x a e y a e z a e ] − [ x s e y s e z s e ] )

[0034] relative speed is

[0035] [ v asx s v easy s v asz s ] = S yes ( [ v axe e v ay e v az e ] − [ v sx e v s e v sz e ] ) + S ˙ yes ( [ x a e y a e z a e ] − [ x s e y s e z s e ] ) − [ p s e q s e r s e ] × [ x as s y as s z as s ]

[0036] Relative acceleration is

[0037] [ a asx s a easy s a asz s ] = S yes ( [ a axe e a ay e a az e ] − [ a sx e a s e a sz e ] ) − 2 S ˙ yes ( [ v axe e v ay e v az e ] − [ v sx e v s e v sz e ] ) + S ¨ yes ( [ x a e y a e z a e ] − [ x s e y s e z s e ] ) − [ p ˙ s e q ˙ s e r ˙ s e ] × [ x as s y as s z as s ] − [ p s e q s e r s e ] × [ p s e q s e r s e ] × [ x as s y as s z as s ] − 2 [ p s e q s e r s e ] × [ v asx s v easy s v asz s ]

[0038] in [ x as s y as s z as s ] T , [ v asx s v easy s v asz s ] T , [ a asx s a easy s a asz s ] T These represent the relative position, relative velocity, and relative acceleration of the UAV and the moving base platform, respectively; the position vector, velocity vector, and acceleration vector of the UAV in the ground axis system are respectively... R → a e = [ x a e y a e z a e ] T , V → a e = [ v axe e v ay e z az e ] T , A → a e = [ a axe e a ay e a az e ] T The position vector, velocity vector, and acceleration vector of the moving base platform in the ground axis system are respectively as follows: R → s e = [ x s e y s e z s e ] T , V → s e = [ v sx e v s e z sz e ] T , A → s e = [ a sx e a s e a sz e ] T .

[0039] Furthermore, in step 2, after considering the ground effect, the UAV body axis is... The forces acting on the three axes are as follows:

[0040]

[0041] In the formula, As resistance, For thrust, For lateral force, For lift, These are, in order, the pitch angle and roll angle of the UAV. This is the acceleration due to gravity.

[0042] Furthermore, in step 2, after considering the influence of the wake of the moving base, the forward velocity and vertical velocity of the UAV body axis are:

[0043]

[0044] in The initial forward velocity and initial vertical velocity along the axis of the UAV body are, in order. These represent the corresponding disturbance quantities of the steady-state component and the periodic component in the body axis system, respectively:

[0045]

[0046] in This is the transformation matrix from the ground axis system to the UAV body axis system;

[0047] These are the horizontal and vertical components of the steady-state wake of the moving base, respectively:

[0048]

[0049] in For the deck wind speed of the dynamic base platform, The distance between the drone and the center of the platform's pitch. Reference length for the moving base platform;

[0050] These are the horizontal and vertical components of the periodic components of the wake from the moving base, respectively:

[0051]

[0052] in This is the proportionality coefficient. To measure the pitch amplitude of the base platform, This refers to the horizontal distance between the actual horizontal position of the drone and the ideal landing point.

[0053] The velocity of the horizontal turbulent disturbance under the shaft system. The vertical turbulent disturbance velocity along the machine shaft is given by the formula:

[0054]

[0055] D V B 1 ( t ) = [ D u B 1 ( t ) 0 D w B 1 ( t ) ] , D V N 1 ( t ) = [ D u N 1 ( t ) 0 D w N 1 ( t ) ]

[0056] The calculation yielded, where The turbulent horizontal disturbance velocity under the ground axis system. The vertical turbulent disturbance velocity under the ground axis;

[0057]

[0058] Let be the first transfer function of the filter. This is the second transfer function of the filter. Given the first white noise, Given the second white noise.

[0059] Furthermore, the proportionality coefficient is based on the formula

[0060] C = c o s { w s [ t ( 1 + V x − V w i d 0 . 8 5 V w i d ) + x 0 . 8 5 V w i d ] + D }

[0061] The calculation yielded, where For time, The pitching frequency of the dynamic base platform, For random phase, The horizontal speed of the moving base platform.

[0062] Furthermore, in step 3, the L1 control law is:

[0063] [ ϕ d i d ψ d ] = 1 T s s + 1 [ K ϕ 1 ⋅ 2 V 2 L 1 z sin or K i 1 ⋅ 2 V 2 L 1 z ( h − h d L 1 z + h ˙ V ) 0 ] + [ 0 0 K ψ 1 ( ψ ref − ψ ) ]

[0064] In the formula, To achieve the desired Euler angle, For the gain of each control channel, For the drone's flight path angle, Let L1 be the time constant. V is the desired heading angle; V is the velocity after considering the ground effect and the wake of the moving base in step 2. L1 represents the longitudinal distance between the UAV and the reference point; For reference height, The altitude change rate of the drone.

[0065] Furthermore, in step 4, the angular velocity control law is:

[0066] [ p d q d r d ] = 1 0 . 2 5 s + 1 [ 1 0 − s i n i 0 c o s ϕ s i n ϕ c o s i 0 − s i n ϕ c o s ϕ c o s i ] [ K ϕ 2 ( ϕ d − ϕ ) K i 2 ( i d − i ) K ψ 2 ( ψ d − ψ ) ]

[0067] in For the desired angular velocity, This refers to the gain of each control channel.

[0068] Furthermore, in step 4, the angular acceleration control law is:

[0069]

[0070] in To control the quantity, The control quantity is the control quantity at the previous moment. The inertia matrix of the aircraft. For wing reference area, For dynamic pressure, For wingspan, The average aerodynamic chord length is given.

[0071] Furthermore, in step 4, the enhanced differential filtering method is used to estimate the angular acceleration as follows:

[0072]

[0073] in These are the parameters of the low-pass filter; These are the proportional control parameters and integral control parameters for the compensation stage, respectively. Angular acceleration estimation signal , Angular acceleration estimation signal The estimated angular velocity signal obtained through the integration process This is the angular velocity signal from the flight control system.

[0074] Beneficial effects

[0075] This invention comprehensively considers three typical and complex disturbance factors: ground effect, wake turbulence, and deck motion, and establishes corresponding disturbance models for each. This improves the realism of the simulation environment and significantly enhances the simulation capabilities for landing accuracy and safety. Compared with traditional methods, it has higher environmental adaptability and reliability.

[0076] This invention proposes a trajectory control strategy based on a nonlinear L1 adaptive guidance law, which achieves high-precision trajectory tracking for UAVs by generating optimal trajectory acceleration commands in real time. Furthermore, this method effectively decouples trajectory tracking and disturbance suppression commands through frequency domain separation technology, greatly improving the system's anti-interference performance and ensuring the safety and accuracy of the landing process. Compared with existing guidance methods, it has superior anti-interference capabilities.

[0077] The hierarchical attitude control method based on incremental nonlinear dynamic inverse control designed in this invention employs classical nonlinear dynamic inverse technology supplemented by Lyapunov stability analysis in the outer loop attitude angle control to achieve global asymptotic stability. The inner loop, addressing the significant impact of high-frequency disturbances on angular rate, innovatively introduces an incremental nonlinear dynamic inverse method and designs enhanced differential filtering to estimate angular acceleration, effectively suppressing disturbances and improving control accuracy. This strategy of designing separate inner and outer loops significantly improves the system's robustness, disturbance rejection capability, and dynamic response speed, far exceeding the performance of existing single control strategies.

[0078] The proposed enhanced differential filtering method for estimating angular acceleration combines the advantages of real-time performance of differential filtering with the drawback of high noise levels through signal compensation, and is model-independent. Compared to other methods, the proposed estimation strategy fully meets the practical requirements of engineering applications for high signal-to-noise ratio and low delay time in angular acceleration signals.

[0079] The proposed incremental power compensation throttle control method achieves dynamic real-time adjustment of thrust commands by introducing speed error feedback, effectively compensating for energy deviations caused by environmental disturbances and UAV attitude changes, significantly enhancing the precision control of flight speed and altitude, and improving flight quality during autonomous landing.

[0080] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0081] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0082] Figure 1 : Schematic diagram of the relative motion relationship between the UAV and the moving base platform;

[0083] Figure 2 : Overall structural block diagram of the control system of this invention;

[0084] Figure 3 Comparison of height tracking curves under the influence of ground effect, wake turbulence, and deck motion interference in embodiments of the present invention;

[0085] Figure 4 Comparison of height error curves under the influence of ground effect, wake turbulence, and deck motion interference in embodiments of the present invention;

[0086] Figure 5 Comparison of pitch angle variation curves under the influence of ground effect, wake turbulence, and deck motion interference in embodiments of the present invention;

[0087] Figure 6Comparison of vertical velocity variation curves under the influence of ground effect, wake turbulence and deck motion interference in embodiments of the present invention. Detailed Implementation

[0088] The embodiments of the present invention are described in detail below. These embodiments are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0089] In this embodiment, to address the requirement for autonomous air-to-ground landing of fixed-wing UAVs on a moving base platform, a control method for autonomous air-to-ground landing of fixed-wing UAVs on a moving base platform is proposed. This method employs an adaptive composite control scheme that includes a nonlinear L1 control law and an incremental nonlinear dynamic inverse, specifically comprising the following steps:

[0090] Step 1: Treat the drone as a rigid body whose volume and shape do not change, and establish a dynamic model of the fixed-wing drone and a model of the relative motion between the drone and the moving base.

[0091] like Figure 1 As shown, this invention refers to a fixed-wing unmanned aerial vehicle (UAV), which is considered as a rigid body whose volume and shape do not change. Its motion has six degrees of freedom; therefore, the dynamic equations include equations for the movement of the center of mass and equations for rotation about the center of mass.

[0092]

[0093]

[0094] In the formula, The velocity vector in the unmanned aerial vehicle's body axis system. The angular velocity vector of the UAV body axis. The values, in order, are the forward velocity, lateral velocity, and vertical velocity along the unmanned aerial vehicle's axis. These are, in order, the roll rate, pitch rate, and yaw rate of the UAV's body axis. These are, in order, the unmanned aerial vehicle's body axis under the following conditions. Differentiating the above equation with respect to the unit vectors of the three axes yields the absolute acceleration of the UAV's center of mass, i.e.:

[0095]

[0096] According to Newton's second law, the dynamic equation for the movement of the center of mass is:

[0097]

[0098] in, The following are the unmanned aerial vehicle body axis systems respectively Triaxial forces, Let be the mass of the drone. The rotation of the drone about its center of mass is caused by the torque acting on it. In the body axis system, according to the theorem of angular momentum, we can obtain:

[0099]

[0100] In the formula, Let be the angular momentum at the center of mass of the UAV. The resultant external torque. The projection of the angular momentum onto the UAV's body axis is:

[0101]

[0102] In the formula, These are, in order, the angular momentum of the UAV rigid body about the UAV body axis. Components in the three axes In order, they are in the unmanned aerial vehicle body axis system Moment of inertia in the three axes; In order, they are in the unmanned aerial vehicle body axis system The product of inertia along the three axes, and the derivative of the angular momentum with respect to time, yield:

[0103]

[0104] Because drones have a longitudinal plane of symmetry, the inertial product... Therefore, the dynamic equation for the rotation about the center of mass is:

[0105]

[0106] In the above formula, The following are the axes around the unmanned aerial vehicle body, respectively. The resultant torque of the three-axis rotation.

[0107] Figure 1 The relative motion relationship between the UAV and the moving base platform is shown. The platform coordinate system is selected as the moving reference axis system, and the moving speed of the moving base platform is 2m / s. The relative angular motion model and the relative linear motion model are obtained by using rigid body kinematics, dynamics and relative motion theory.

[0108] The relative angular motion model defines the Euler angles of the UAV as follows: [ f a e i a e P a e ] T angular velocity is Oh → a e = [ p a e q a e r a e ] T Angular acceleration is Oh → ˙ a e = [ p ˙ a e q ˙ a e r ˙ a e ] T The Euler angles of the dynamic base platform are defined as follows: [ f s e i s e P s e ] T angular velocity is Oh → s e = [ p s e q s e r s e ] T Angular acceleration is Oh → ˙ s e = [ p ˙ s e q ˙ s e r ˙ s e ] T The relative transformation matrix between the UAV and the moving base platform is:

[0109]

[0110] in and These represent the transformation matrices from the ground coordinate system to the UAV coordinate system and from the ground coordinate system to the platform coordinate system, respectively.

[0111] Let matrix The middle element is , and Representing rows and columns, we have:

[0112] Relative attitude angle is

[0113]

[0114] Relative angular velocity is

[0115] [ p as s q as s r as s ] = S yes ( [ p a e q a e r a e ] − [ p s e q s e r s e ] )

[0116] Relative angular acceleration is

[0117] [ p ˙ as s q ˙ as s r ˙ as s ] = S yes ( [ p ˙ a e q ˙ a e r ˙ a e ] − [ p ˙ s e q ˙ s e r ˙ s e ] ) + S ˙ yes ( [ p a e q a e r a e ] − [ p s e q s e r s e ] ) − [ p s e q s e r s e ] × [ p as s q as s r as s ]

[0118] In the formula [ f as s i as s P as s ] T , [ p as s q as s r as s ] T and [ p ˙ as s q ˙ as s r ˙ as s ] T These represent the relative attitude angle, relative angular velocity, and relative angular acceleration of the UAV and the moving base platform, respectively.

[0119] The relative linear motion model defines the position vector, velocity vector, and acceleration vector of the UAV in the ground axis system as follows: R → a e = [ x a e y a e z a e ] T , V → a e = [ v axe e v ay e z az e ] T , A → a e = [ a axe e a ay e a az e ] T Define the position vector, velocity vector, and acceleration vector of the moving base platform in the ground axis system as follows: R → s e = [ x s e y s e z s e ] T , V → s e = [ v sx e v s e z sz e ] T , A → s e = [ a sx e a s e a sz e ] T Then there is

[0120] Relative position is

[0121] [ x as s y as s z as s ] = S yes ( [ x a e y a e z a e ] − [ x s e y s e z s e ] )

[0122] relative speed is

[0123] [ v asx s v easy s v asz s ] = S yes ( [ v axe e v ay e v az e ] − [ v sx e v s e v sz e ] ) + S ˙ yes ( [ x a e y a e z a e ] − [ x s e y s e z s e ] ) − [ p s e q s e r s e ] × [ x as s y as s z as s ]

[0124] Relative acceleration is

[0125] [ a asx s a easy s a asz s ] = S yes ( [ a axe e a ay e a az e ] − [ a sx e a s e a sz e ] ) − 2 S ˙ yes ( [ v axe e v ay e v az e ] − [ v sx e v s e v sz e ] ) + S ¨ yes ( [ x a e y a e z a e ] − [ x s e y s e z s e ] ) − [ p ˙ s e q ˙ s e r ˙ s e ] × [ x as s y as s z as s ] − [ p s e q s e r s e ] × [ p s e q s e r s e ] × [ x as s y as s z as s ] − 2 [ p s e q s e r s e ] × [ v asx s v easy s v asz s ]

[0126] In the formula [ x as s y as s z as s ] T , [ v asx s v easy s v asz s ] T , [ a asx s a easy s a asz s ] T These represent the relative position, relative velocity, and relative acceleration of the UAV and the moving base platform, respectively.

[0127] Step 2: For complex environmental scenarios, considering the impact of ground effect, dynamic base wake, and dynamic base motion on landing accuracy, disturbance models are introduced to simulate complex interference environments under real sea and air conditions.

[0128] The ground effect primarily alters the lift of a UAV during landing, thus affecting its longitudinal motion. This invention collects data on the change in lift coefficient of the UAV under actual ground effect conditions. Through field test flights of the UAV, the change in lift coefficient at different altitudes is measured to establish a curve relating lift coefficient to flight altitude within the ground effect range. This measured data is then substituted into the UAV's dynamic model for high-precision calculations. The additional lift generated by the ground effect... It can be expressed by the following formula:

[0129]

[0130] In the formula, Atmospheric density, For the drone's flight speed, For wing area, This represents the change in the lift coefficient data of the UAV.

[0131] Combining the dynamic equations of the center of mass movement and the influence of ground effects, the UAV body axis... The forces acting on the three axes are as follows:

[0132]

[0133] In the above formula, As resistance, For thrust, For lateral force, For lift, These are, in order, the pitch angle and roll angle of the UAV. This is the acceleration due to gravity.

[0134] Dynamic base wake disturbance model:

[0135] The wake disturbance of the moving base appears approximately 20 meters behind the moving base platform. For the wake model of the moving base in the longitudinal plane, let the resultant velocities of the horizontal and vertical disturbances be respectively... The wake disturbance of the moving base is divided into four parts: the horizontal and vertical components of free atmospheric turbulence. Steady-state components of the wake of the moving base in the horizontal and vertical directions Periodic components of the wake of the moving base in the horizontal and vertical directions And the random components of the wake of the moving base in the horizontal and vertical directions. The random component has a negligible impact and is therefore not considered.

[0136] Free atmospheric turbulence is independent of the relative position between the UAV and the moving base platform and exhibits anisotropy. The resulting turbulence alters the relative airflow direction, thus affecting the UAV's axial velocity components. Free atmospheric turbulence is a stochastic process and can only be described by spectral functions or correlation functions. The spectral relationship of free atmospheric turbulence is as follows:

[0137]

[0138] In the above formula, the free atmospheric turbulence spectrum is defined by spatial frequency. For the spatial spectrum of the independent variable, These are the horizontal and vertical components of the spatial spectrum of free atmospheric turbulence.

[0139] Since only the longitudinal airflow disturbance effect is calculated, the corresponding spatial spectrum is... Converted to time spectrum We can obtain:

[0140]

[0141] In the formula: These represent the horizontal and vertical velocities of the UAV in the ground-based axis, respectively. For time frequency.

[0142] Given white noise And based on the time spectrum Establish a filter to obtain the turbulent disturbance velocity under the ground axis:

[0143]

[0144] in The turbulent horizontal disturbance velocity under the ground axis system. The vertical turbulent disturbance velocity is located below the ground axis. Let be the first transfer function of the filter. This is the second transfer function of the filter. Given the first white noise, Given the second white noise.

[0145] The transformation matrix from the ground axis to the UAV body axis is established as follows:

[0146] R N B = [ cos i cos ψ cos i sin ψ − sin i sin ϕ sin i cos ψ − cos ψ sin ψ sin ϕ sin i sin ψ + cos ϕ co ψ s sin ϕ cos i cos ϕ sin i cos ψ + sin ϕ sin ψ cos ϕ sin i sin ψ − sin ϕ cos ψ cos ϕ cos i ]

[0147] In the above formula This is the yaw angle of the drone.

[0148] The turbulent disturbance velocities described above under the ground axis are then transferred to the body axis:

[0149]

[0150] The perturbation velocity matrix under the body axis is obtained:

[0151] D V B 1 ( t ) = [ D u B 1 ( t ) 0 D w B 1 ( t ) ]

[0152] in D V N 1 ( t ) = [ D u N 1 ( t ) 0 D w N 1 ( t ) ] , The velocity of the horizontal turbulent disturbance under the shaft system. The vertical turbulent disturbance velocity is the velocity of the lower axis of the machine body.

[0153] Add the disturbance velocity to the body velocity component:

[0154]

[0155] in, The initial velocities along the x and z axes of the UAV's body axis are shown in sequence.

[0156] The steady-state component of the wake of the moving base platform is generated by the airflow passing over the flat tail protrusion of the platform during headwind movement, and is also known as the "rooster wake". In the vertical direction, the velocity direction of the steady-state component of the wake is related to the distance from the tail protrusion. Near the tail protrusion, it is a strong downwash, and after a certain distance from the tail protrusion, the airflow begins to wash upward. The steady-state component is the continuous wind field distortion generated by the moving base platform and propulsion system, which is equivalent to affecting the airflow velocity experienced by the UAV, that is, changing the velocity component under the UAV's axis.

[0157] The horizontal component of the steady-state component of the wake of the moving base and the deck wind speed ratio and vertical component and deck wind speed ratio The distance between the UAV and the platform's center of pitch (COP) has a functional relationship, specifically:

[0158]

[0159] Therefore, we get:

[0160]

[0161] in For the deck wind speed of the dynamic base platform, The distance of the drone from COP. The reference length for the moving base platform.

[0162] The periodic component of the wake from the moving base platform is a disturbance generated by the pitch and heave motions of the platform. The pitch, heave amplitude, and frequency of the moving base platform, the distance between the fixed-wing UAV and the moving base platform, and the magnitude and direction of the deck wind all affect the periodic component of the wake. Correspondingly, this periodic component disturbance alters the UAV's motion relative to the atmosphere, directly impacting the UAV's axial velocity component. Since the vortex formed at the tail of the moving base platform decays at 85% of its horizontal distance, the closer the wake component is to the platform, the greater its influence. The mathematical model is shown below.

[0163]

[0164] in

[0165] C = c o s { w s [ t ( 1 + V x − V w i d 0 . 8 5 V w i d ) + x 0 . 8 5 V w i d ] + D }

[0166] In the formula: This is the proportionality coefficient. For time, The pitching frequency of the dynamic base platform, To measure the pitch amplitude of the base platform, For random phase, The horizontal distance between the actual horizontal position of the drone and the ideal landing point. The horizontal speed of the moving base platform.

[0167] The aforementioned steady-state and periodic components are both in the ground-based axis system; converting them to the body-based axis system:

[0168]

[0169] in, These represent the corresponding disturbances in the body axis system for the steady-state component and the periodic component, respectively.

[0170] Substituting the above expression into the body velocity components, we get:

[0171]

[0172] The impact of moving base motion on landing accuracy:

[0173] Due to the influence of sea conditions such as waves and winds, the motion of a moving platform at sea can be divided into three-dimensional yaw: pitch, roll, and yaw, and three-dimensional oscillation: heave, yaw, and pitch. The amplitudes of the yaw and pitch motions in the moving platform's motion are very small and have little impact on UAV landing, so they can be ignored. This invention only considers longitudinal control design and ignores roll. Therefore, the deck motion model only considers pitch, yaw, and heave motions. Pitch and yaw represent the rotation of the moving platform around the horizontal and vertical axes, respectively. Even if the UAV's own axis angle remains unchanged during landing, its pitch and yaw angles relative to the platform will change. Similarly, heave affects the relative height between the UAV and the platform.

[0174] Under moderate sea states, when the moving-base platform travels at a speed of 15 m / s, the pitching of the moving-base platform... , bow rocking and drooping As shown in the following formula:

[0175]

[0176] in Indicates units of degrees. The measurement is in meters; the pitch and roll of the base platform will be measured. , bow rocking and drooping By incorporating this into the UAV's dynamic model, the corrected UAV pitch angle can be obtained. Yaw angle and drone altitude for:

[0177]

[0178] In the above formula, The altitude of the UAV in the ground axis system.

[0179] Step 3: Based on the dynamic model of the fixed-wing UAV established in Step 1 and the relative motion model between the UAV and the moving base, as well as the factors affecting landing accuracy such as ground effect, moving base wake, and moving base motion obtained in Step 2, a trajectory control law based on nonlinear L1 control law is designed. By generating trajectory acceleration commands in real time, the UAV is controlled to glide along the optimal trajectory to the target point, transforming the trajectory tracking problem into an optimal control problem to be solved. In terms of anti-disturbance performance, this method uses frequency domain separation technology to decouple low-frequency trajectory tracking commands from high-frequency disturbance suppression commands, thereby improving the landing safety of the UAV.

[0180] The L1 control law is based on L1 adaptive control theory. It takes a reference point on the desired trajectory that is L1 distance away from the UAV, calculates the acceleration command based on the current speed of the UAV, and generates the attitude angle command of the UAV through the mathematical relationship between acceleration and attitude angle.

[0181] During the descent, assume a radius of [missing information] between the UAV and the reference point in the longitudinal plane. The circular motion is tangential to the velocity direction of the UAV. According to the L1 control law, the longitudinal L1 distance is defined as follows:

[0182]

[0183] L1 is the longitudinal distance between the UAV and the reference point. The angle between the L1 direction and the UAV velocity V is the longitudinal angle between the UAV direction and the UAV velocity V, which is the velocity after considering the ground effect and the wake of the moving base in step 2.

[0184] Longitudinal acceleration of drones This is the centripetal acceleration of the circular arc segment:

[0185]

[0186] Due to the included angle Very small, with

[0187]

[0188] in, The angle between the UAV's flight path angle and the L1 direction is... Let be the angle between the UAV's flight path angle and its velocity direction. Further, we obtain...

[0189]

[0190] For reference height, Let be the rate of change of altitude of the drone, which is also known as the rate of climb. Substituting the centripetal acceleration, we get...

[0191]

[0192] After rearranging, we obtain a second-order system:

[0193]

[0194] Applying the Laplace transform to the second-order system yields the transfer function of the high-tracking system:

[0195]

[0196] In the formula, the damping ratio The undamped natural frequency is 0.707. for , For complex frequencies, For the zero-point polynomial of the height tracking system, Let be the pole polynomial of the high-tracking system.

[0197] Desired pitch angle Through gravity compensation and acceleration commands Calculations show that, since the gravitational component of the UAV's longitudinal attitude affects its longitudinal acceleration, the target pitch angle is:

[0198]

[0199] Similarly, lateral acceleration That is, the centripetal acceleration of the circular arc segment. :

[0200]

[0201] in For the drone's flight path angle;

[0202] The target roll angle is then obtained.

[0203]

[0204] Through the above derivation, the complete L1 control law is finally obtained as follows:

[0205] [ ϕ d i d ψ d ] = 1 T s s + 1 [ K ϕ 1 ⋅ 2 V 2 L 1 z sin or K i 1 ⋅ 2 V 2 L 1 z ( h − h d L 1 z + h ˙ V ) 0 ] + [ 0 0 K ψ 1 ( ψ ref − ψ ) ]

[0206] In the formula, To achieve the desired Euler angle, For the gain of each control channel, For the drone's flight path angle, Let L1 be the time constant. The desired heading angle.

[0207] Step 4: Building upon Step 3, a hierarchical attitude control method based on incremental nonlinear dynamic inverse is further designed. According to the time-scale separation principle, a dual-closed-loop architecture is adopted. Since the outer loop attitude angle does not contain uncertain terms, a nonlinear dynamic inverse control method is used. The control law is designed using Lyapunov stability theory to ensure the global asymptotic stability of the control system. Since the angular rate is affected by disturbances first, the inner loop angular velocity adopts an incremental nonlinear dynamic inverse method. An enhanced differential filtering method is designed to estimate the angular acceleration. That is, a compensation signal is introduced into the differential filtering method to improve the quality of the estimated signal, improve the anti-disturbance capability and response speed, thereby enhancing the robustness and control accuracy of the system.

[0208] In the landing control of the moving base platform, the angular velocity control loop is the outer loop of the attitude control, and the kinematic relationship between attitude angle and angular velocity is as follows:

[0209]

[0210] Introducing desired attitude angular velocity and virtual input Rewrite the above formula as:

[0211]

[0212] in The desired angular velocity is generated by the attitude angle error via a proportional controller. Taking pitch control as an example, the formula is as follows:

[0213]

[0214] To avoid excessive control input, a first-order hysteresis filter needs to be introduced before the controller for command shaping.

[0215]

[0216] The complete angular velocity control law expression is:

[0217] [ p d q d r d ] = 1 0 . 2 5 s + 1 [ 1 0 − s i n i 0 c o s ϕ s i n ϕ c o s i 0 − s i n ϕ c o s ϕ c o s i ] [ K ϕ 2 ( ϕ d − ϕ ) K i 2 ( i d − i ) K ψ 2 ( ψ d − ψ ) ]

[0218] This represents the desired angular velocity.

[0219] The angular acceleration control loop is the inner loop of attitude control. It uses an incremental dynamic inverse method to accurately track the desired attitude angular velocity. This can be obtained by inversely solving the control surface increments.

[0220]

[0221] in, The inertia matrix of the aircraft. For the output function, To manipulate the performance matrix, virtual inputs are required in actual control. replace The virtual input is provided by the outer loop controller. Adding the increment of the control surface to the existing control surface deflection yields the complete expression:

[0222]

[0223] Combining the above equation, the basic control law of the angular acceleration control loop can be obtained as follows:

[0224]

[0225] in and To account for the state and control variables from the previous moment, the efficiency of the torque coefficient change caused by the control surface needs to be introduced. Calculate the increment of control surface deflection, and the torque acting on the aircraft. The torques generated by the aircraft under different flight and control conditions are synthesized from the torques produced by the aerodynamic forces. Aerodynamic forces produce corresponding torques, and the corresponding roll, pitch, and yaw moments acting on the aircraft can be expressed as... The relevant formulas are shown below.

[0226] M = [ l m n ] = [ Q b S C l ( α , β , d e , d a , d r ) Q c S C m ( α , β , d e , d a , d r ) Q b S C n ( α , β , d e , d a , d r ) ]

[0227] In the formula For wingspan, For the average aerodynamic chord length, For wing reference area, For dynamic pressure, Let be the roll, pitch, and yaw moment coefficients of the aircraft. The roll, pitch, and yaw moment coefficients in the above equation can be expressed as:

[0228]

[0229] In the formula This represents the distance between the reference position and the actual center of gravity. To calculate the control surface effectiveness, the roll, pitch, and yaw moment coefficients caused by control surface deflection are separated and calculated. about The law of change, taking the elevator pitch control effectiveness as an example, can be rewritten as follows:

[0230]

[0231] By combining the aerodynamic formulas of the UAV, the control surface effectiveness matrix of this aircraft can be determined. It can be calculated as:

[0232]

[0233] In addition to the control surface performance matrix, the angular acceleration control loop also needs to introduce a virtual angular acceleration input. Similar to the angular velocity control loop, the difference between the desired angular velocity and the attitude angular velocity is calculated and passed through a proportional controller to obtain the virtual angular acceleration signal. For example, taking pitch control, it can be shown in the following formula:

[0234]

[0235] At this point, the complete angular acceleration control loop control law has been obtained:

[0236]

[0237] In the above-mentioned angular acceleration control loop design, an estimated value of the angular acceleration signal needs to be introduced when calculating the control surface deflection increment. This invention proposes an enhanced differential filtering method, which introduces a compensation signal into the differential filtering method to improve the quality of the estimated signal:

[0238] (1) Pass the angular velocity signal through a low-pass filter and a differentiator to obtain the estimated value of the angular acceleration signal. ;

[0239] (2) Make the angular acceleration estimation signal The estimated angular velocity signal is obtained through integration. and pitch angular velocity signals from the flight control system Difference to obtain the estimation error ;

[0240] (3) This error is then compensated for by a proportional-integral controller to obtain the angular acceleration estimation signal. The compensation signal is: ;

[0241] (4) The final estimated angular acceleration signal expression is shown below:

[0242]

[0243] In the formula, These are the parameters of the low-pass filter; These are the proportional control parameters and integral control parameters for the compensation stage, respectively.

[0244] Step 5: Building upon Steps 3 and 4, incremental power-compensated throttle control is designed, incorporating speed error feedback to achieve dynamic adjustment of thrust commands and effectively compensate for energy deviations caused by environmental disturbances or attitude changes. This method not only improves the rapid response capability to flight speed but also enhances the precise control of flight altitude.

[0245] The throttle opening is determined by controlling the speed of the drone. :

[0246]

[0247] In the formula, The drag coefficient, For the trajectory angle, This represents the aircraft's maximum available thrust.

[0248] Determine the expected thrust:

[0249]

[0250] Because the dynamic characteristics of an engine vary with parameters such as flight altitude, speed, and flight attitude, flight simulations typically simplify the process from throttle input to thrust output. The mathematical model of the engine's dynamic characteristics is described as follows:

[0251]

[0252] The thrust of the drone engine is:

[0253]

[0254] Because the dynamic characteristics of an engine vary with parameters such as flight altitude, speed, and flight attitude, flight simulations typically simplify the process from throttle input to thrust output. The mathematical model of the engine's dynamic characteristics is described as follows:

[0255]

[0256] To achieve precise power compensation control, an incremental power compensation throttle control method is employed, adjusting engine thrust to compensate for energy changes during platform descent. Assuming the total engine thrust acts within the UAV's plane of symmetry, and due to the engine's installation angle, the total engine thrust typically does not pass through the UAV's center of mass. If the distance from the UAV's center of mass to the line of action of the engine thrust is... Furthermore, the thrust line of the engine is parallel to the engine shaft. The angle of deflection is Then in the body axis system Under these conditions, the engine thrust can be expressed as:

[0257] [ T x T y T z ] = [ T c o s α p T s i n α p 0 ]

[0258] The torque of the engine thrust can be expressed as:

[0259] [ M x M y M z ] = [ 0 0 − T e p ]

[0260] Example verification:

[0261] The following section verifies the autonomous landing of an unmanned aerial vehicle (UAV) on an air-to-ground movable base platform using the method of this invention under the influence of ground effect, wake turbulence, and deck motion.

[0262] Ground effect is introduced when the UAV is 2.5m above the platform, while wake and deck motion are introduced when the UAV is 20m above the platform. The UAV's initial altitude is 48m and its relative speed is 15m / s. The UAV begins its descent at an angle of 8.6° at 8.8s in the simulation and catches up with the moving platform after 21.2s. Due to complex sea and air interference, the actual landing point will be far from the ideal landing point. Under these conditions, the performance differences between the trajectory control based on nonlinear L1 control law and the attitude control based on incremental nonlinear dynamic inverse proposed in this invention and the traditional PID control method are compared. The comparison results are as follows: Figure 3~Figure 6 As shown.

[0263] from Figure 3 The comparison results show that in the initial stage of descent, both the traditional PID control and the method proposed in this invention have tracking errors. However, the control method adopted in this invention can correct the error more quickly and respond faster. At the end of the descent, the horizontal distance from the ideal landing point using the traditional method is 14.85m, while that using this method is 10.95m, a reduction of 26.3%, significantly improving landing accuracy. Figure 4 The comparison results show that the proposed method exhibits smaller fluctuations in altitude tracking error, higher stability, and faster response during descent, significantly improving landing safety.

[0264] During the landing process, the proposed control method showed significant advantages over the traditional PID control scheme, such as... Figure 5 As shown. In terms of pitch angle control, this method has a faster response speed and can track command signals more quickly, effectively reducing the delay in attitude adjustment; at the same time, the pitch angle change is significantly reduced in the terminal phase of descent, indicating that this control method has stronger stability and anti-interference ability, thus ensuring that the aircraft lands in a more stable attitude.

[0265] In vertical speed control, the proposed method not only offers a faster dynamic response and converges quickly to the target descent rate, but also significantly reduces speed fluctuations as the aircraft approaches the ground, avoiding the "overshoot" or "oscillation" phenomena commonly found in traditional PID control. This enables the aircraft to achieve a smoother and more precise landing, further enhancing landing safety and reliability.

[0266] In summary, this control method, by optimizing dynamic response and suppressing terminal fluctuations, outperforms traditional PID schemes in altitude tracking, attitude and speed control, providing a more precise solution for autonomous landing missions and ensuring the reliability and safety of UAV landing on movable base platforms.

[0267] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention without departing from the principles and spirit of the present invention.

Claims

1. A method for autonomous air-to-ground landing control of a fixed-wing unmanned aerial vehicle (UAV) on a moving base platform, characterized in that: Includes the following steps: Step 1: Treat the drone as a rigid body whose volume and shape do not change, and establish a dynamic model of the fixed-wing drone and a model of the relative motion between the drone and the moving base. Step 2: Considering the effects of ground effect, dynamic base wake, and dynamic base motion on landing accuracy, introduce disturbance models for each. Step 3: Based on the dynamic model of the fixed-wing UAV established in Step 1 and the relative motion model between the UAV and the moving base, as well as the factors affecting landing accuracy obtained in Step 2, including ground effect, moving base wake, and moving base motion, a trajectory control law based on a nonlinear L1 control law is designed. This law generates real-time trajectory acceleration commands to control the UAV to glide along the optimal trajectory to the target point. The L1 control law is as follows: In the formula, To achieve the desired Euler angle, For the gain of each control channel, For the drone's flight path angle, Let L1 be the time constant. V is the desired heading angle; V is the velocity after considering the ground effect and the wake of the moving base in step 2. L1 represents the longitudinal distance between the UAV and the reference point; For reference height, The rate of change of altitude of the drone; Step 4: Design a hierarchical attitude control method based on incremental nonlinear dynamic inverse. According to the time scale separation principle, a dual closed-loop architecture is adopted. The outer loop attitude angle adopts the nonlinear dynamic inverse control method, and the angular velocity control law is designed through Lyapunov stability theory. The inner loop angular velocity adopts the incremental nonlinear dynamic inverse method to obtain the angular acceleration control law, and an enhanced differential filtering method is designed to estimate the angular acceleration. Step 5: Based on Step 3 and Step 4, design incremental power compensation throttle control, introduce speed error feedback, and realize dynamic adjustment of thrust command.

2. The method for autonomous air-to-ground landing control of a fixed-wing UAV oriented towards a moving base platform according to claim 1, characterized in that: In step 1, the dynamic model of the fixed-wing UAV includes the dynamic equations for the movement of the center of mass and the dynamic equations for the rotation about the center of mass: The dynamic equation for the movement of the center of mass is: in, The following are the unmanned aerial vehicle body axis systems respectively Triaxial forces, For the quality of drones, The values, in order, are the forward velocity, lateral velocity, and vertical velocity along the unmanned aerial vehicle's axis. These are, in order, the unmanned aerial vehicle's roll rate, pitch rate, and yaw rate along its axis. The dynamic equation for the rotation about the center of mass is: in, The following are the axes around the unmanned aerial vehicle body, respectively. The resultant torque of the three-axis rotation; In order, they are in the unmanned aerial vehicle body axis system Moment of inertia in the three axes; In order, they are in the unmanned aerial vehicle body axis system Inertial product in the three axes.

3. The method for autonomous air-to-ground landing control of a fixed-wing UAV oriented towards a moving base platform according to claim 1, characterized in that: In step 1, the relative motion model between the UAV and the moving base includes a relative angular motion model and a relative linear motion model; The relative angular motion model includes: Relative attitude angle is Relative angular velocity is Relative angular acceleration is in , and These represent the relative attitude angle, relative angular velocity, and relative angular acceleration between the UAV and the moving base platform, respectively. The relative transformation matrix between the UAV and the mobile base platform The elements in; the Euler angles of the drone are angular velocity is Angular acceleration is The Euler angles of the moving base platform are angular velocity is Angular acceleration is ; This is the transformation matrix from the ground coordinate system to the platform coordinate system; The relative linear motion model includes: Relative position is Relative speed is Relative acceleration is in , , These represent the relative position, relative velocity, and relative acceleration of the UAV and the moving base platform, respectively; the position vector, velocity vector, and acceleration vector of the UAV in the ground axis system are respectively... , , The position vector, velocity vector, and acceleration vector of the moving base platform in the ground axis system are respectively as follows: , , .

4. The method for autonomous air-to-ground landing control of a fixed-wing UAV oriented towards a moving base platform according to claim 2, characterized in that: In step 2, after considering the ground effect, the UAV body axis is... The forces acting on the three axes are as follows: In the formula, As resistance, For thrust, For lateral force, For lift, These are, in order, the pitch angle and roll angle of the UAV. It is the acceleration due to gravity. Additional lift generated by the ground effect.

5. The method for autonomous air-to-ground landing control of a fixed-wing UAV oriented towards a moving base platform according to claim 1, characterized in that: In step 2, after considering the influence of the wake of the moving base, the forward velocity and vertical velocity of the UAV body axis are: in The initial forward velocity and initial vertical velocity along the axis of the UAV body are, in order. These represent the corresponding disturbance quantities of the steady-state component and the periodic component in the body axis system, respectively: in This is the transformation matrix from the ground axis system to the UAV body axis system; These are the horizontal and vertical components of the steady-state wake of the moving base, respectively: in For the deck wind speed of the dynamic base platform, The distance between the drone and the center of the platform's pitch. Reference length for the moving base platform; These are the horizontal and vertical components of the periodic components of the wake from the moving base, respectively: in This is the proportionality coefficient. To measure the pitch amplitude of the base platform, This refers to the horizontal distance between the actual horizontal position of the drone and the ideal landing point. The velocity of the horizontal turbulent disturbance under the shaft system. The vertical turbulent disturbance velocity along the machine shaft is given by the formula: ， The calculation yielded, where The turbulent horizontal disturbance velocity under the ground axis system. The vertical turbulent disturbance velocity under the ground axis system; Let be the first transfer function of the filter. This is the second transfer function of the filter. Given the first white noise, Given the second white noise.

6. The method for autonomous air-to-ground landing control of a fixed-wing UAV oriented towards a moving base platform according to claim 5, characterized in that: The proportionality coefficient is based on the formula The calculation yielded, where For time, The pitching frequency of the dynamic base platform, For random phase, The horizontal speed of the moving base platform.

7. The method for autonomous air-to-ground landing control of a fixed-wing UAV oriented towards a moving base platform according to claim 1, characterized in that: In step 4, the angular acceleration is estimated using the enhanced differential filtering method as follows: in These are the parameters of the low-pass filter; These are the proportional control parameters and integral control parameters for the compensation stage, respectively. Angular acceleration estimation signal , Angular acceleration estimation signal The estimated angular velocity signal obtained through the integration process This is the angular velocity signal from the flight control system.

Images