An image servo hose-drag automatic aerial refueling docking control method under a speed control mode

Abstract

The application provides a kind of image servo hose type automatic aerial refueling docking control method under speed control mode, comprising the following: one: complete the hose type automatic aerial refueling docking control problem description under speed control mode;Two: image servo model establishment under speed mode, camera motion model establishment;Three: based on image servo model and camera motion model under speed mode, design the expected speed of receiver relative to ground coordinate system;Four: according to the guidance model, design dynamic backstepping speed tracking controller based on additive decomposition.The application solves the control problem of hose type automatic aerial refueling docking under speed control mode based on image servo method under the condition that existing bottom controller cannot be modified or throttle and rudder surface cannot be directly controlled for safety, meets safety, reliability, robustness, so that hose type automatic aerial refueling docking can be efficiently and safely completed, and is closer to actual application scene.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of aerial refueling control, and particularly relates to a method for image servo hose-type automatic aerial refueling docking control in a speed control mode. BACKGROUND

[0002] Automatic aerial refueling is the crown of automatic flight control, because on the one hand, aerial refueling can greatly expand the flight distance and combat radius of an aircraft and improve combat capability, and on the other hand, automatic aerial refueling has high control difficulty due to many kinds of disturbances in the docking process and high docking accuracy requirements. Hose-type aerial refueling has the advantages of flexibility and simultaneous refueling of multiple aircrafts, but is greatly affected by wake, turbulence and head wave effect, so it is of great significance to research a robust hose-type automatic aerial refueling control method.

[0003] Existing automatic aerial refueling control methods include navigation based on a differential GPS system, an optical-electric system, an inertial navigation system, a radar or vision. However, the above methods have many problems and are difficult to apply in practice, for example, a docking control method based on position servo, according to which there are many errors in calculating the position of a cone sleeve relative to a camera, thereby leading to docking failure. A docking control method based on image servo directly obtains a control instruction according to image information and has strong robustness. In addition, almost all current researches on automatic aerial refueling are directly controlling actuators such as a throttle, an elevator, an aileron and a rudder. However, in practice, most cases are that a receiver aircraft has an encapsulated bottom controller, and direct control to the rudder surface is not allowed due to safety considerations. This is also the reason why current related researches are limited to simulation and cannot be flown in practice. Generally, the role of a top controller is to control the position of an aircraft or make the aircraft track a trajectory. The role of the top controller of automatic aerial refueling is to control the position of a cone pipe of a receiver aircraft, and since the position of the cone pipe is related to the position, speed, attitude and angular velocity of the receiver aircraft, the design difficulty of the top controller is high. In addition, generally, safety is the primary requirement of a bottom controller, thereby sacrificing maneuverability, which means that the response of the receiver aircraft to a speed instruction is not accurate and fast enough, and it is likely that the high dynamic and high precision docking control requirements cannot be met. Therefore, it is of great significance to research a hose-type automatic aerial refueling docking control method based on image servo in a speed control mode for the practice of aerial refueling in military and civilian fields. SUMMARY

[0004] The application provides a hose type automatic aerial refueling docking control method based on image servo in a speed control mode, which has the characteristics of reliability and closeness to practice.

[0005] In order to achieve the above-mentioned purpose, the application provides a hose type automatic aerial refueling docking control method based on image servo in a speed control mode, and the implementation steps are as follows:

[0006] Step one: complete the hose type automatic aerial refueling docking control problem description in a speed control mode, including coordinate system establishment, receiver aircraft six degree of freedom model establishment, receiver aircraft guidance model establishment in a speed control mode, and control problem summary, which are as follows:

[0007] Step 1.1: according to the motion relationship of each object in the hose type automatic aerial refueling docking process in a speed control mode, all coordinate systems required for describing the docking control problem are established.

[0008] In order to establish the system model in the docking stage, five coordinate systems need to be defined: ground coordinate system o g x g y g z g , receiver aircraft coordinate system o r x r y r z r , tanker aircraft coordinate system o t x t y t z t , camera coordinate system o c x c y c z c and relative coordinate system o re x re y re z re , as shown in Figure 1 . represents a certain state of object i in j coordinate system. R i / j (i,j∈{g,r,t,c,re}) represents the rotation matrix from j coordinate system to i coordinate system. For example, V t g is the forward flight speed of the tanker aircraft in the ground coordinate system, and R r / t represents the rotation matrix for describing the rotation of the receiver aircraft coordinate system "r" relative to the tanker aircraft coordinate system "t".

[0009] Up to now, all coordinate systems required for describing the docking control problem are established.

[0010] Step 1.2: According to the scenario of hose automatic aerial refueling docking problem under speed control mode, the six degrees of freedom model of receiver aircraft is established.

[0011] In the docking phase, the pitch angle of receiver aircraft changes in a small range, and the reference state is constant speed straight flight, whose attitude satisfies the coordinated turn condition, that is, the roll angle and sideslip angle satisfy Φ = β = 0, and the angle of attack α, the track angle γ and the pitch angle θ satisfy θ = γ + α. In this case, the motion equation of receiver aircraft in the tanker coordinate system can be decoupled into lateral and longitudinal channel motion. The motion equation of longitudinal channel is:

[0012]

[0013] The motion equation of lateral channel is:

[0014]

[0015] where, respectively represent the coordinates of the center of gravity of the receiver aircraft in the ground coordinate system, (φ, θ, ψ) respectively represent the Euler angles of the three axes, V r g represents the speed of the receiver aircraft relative to the ground, p, q, r respectively represent the angular velocities of the three axes, respectively represent the lift, drag, engine thrust, moment of three channels of pitch, roll and yaw, and side force, c1-c9 respectively represent the nine elements of the moment of inertia matrix. Equations (1), (2) can be combined and simplified as:

[0016]

[0017] where, represents the state of the receiver,

[0018] vector u r = [δ t δ e δ a δ r ] T represents the control input, including throttle, elevator, aileron and rudder.

[0019] So far, the six degrees of freedom model of receiver aircraft is established.

[0020] Step 1.3: According to the scenario of hose automatic aerial refueling docking problem under speed control mode, the guidance model of receiver aircraft under speed control mode is established.

[0021] The underlying controller implementing the speed control mode of the receiver aircraft can be implemented in various ways, such as through a linear quadratic regulator (LQR), a model predictive controller (MPC), a backstepping controller, or a differential flatness controller. Taking the underlying controller implementing the speed control mode through a model predictive controller as an example, the guidance model of the receiver aircraft under the action of the underlying controller can be described by equation (4):

[0022]

[0023] wherein, represents the position of the receiver aircraft in the ground coordinate system, represents the speed of the receiver aircraft in the ground coordinate system, represents the reference speed of the receiver aircraft in the ground coordinate system, G x y z is a transfer function from to . The transfer function G x y z can be obtained through system identification.

[0024] At this point, the guidance model of the receiver aircraft in the speed control mode is established.

[0025] Step 1.4: According to the scenario of the hose-type automatic mid-air refueling docking problem in the speed control mode, complete the description of the control problem.

[0026] To simplify the coordinate system conversion, it is assumed that the camera is installed at the top end of the cone. As shown in Figure 2 , define the projection of the cone sleeve center in the image coordinate system as The image tracking error e is defined as:

[0027]

[0028] wherein, is the origin of the image coordinate system. As shown in Figure 3 , define as the depth, i.e., the distance between the camera and the cone sleeve center plane in the camera coordinate system along the z c axis.

[0029] The following assumptions are made:

[0030] Assumption one: the vision recognition module can accurately and in real time return the center coordinates of the cone sleeve and can roughly estimate the depth.

[0031] Assumption two: the tanker aircraft is in a constant speed and level flight state during the entire docking stage, and the initial state of the receiver aircraft is the same as that of the tanker aircraft. ​​​​

[0032] Assumption 3: The receiving engine can track the reference speed command without static error under the action of the underlying controller in the absence of disturbances.

[0033] The control problem can be summarized as: designing a suitable reference speed for system (4) As the system input, the image tracking error and depth error tend to zero over time. And make the docking error tend to zero over time

[0034] At this point, the control problem is described.

[0035] Step 2: Establish the image servo model and camera motion model in speed mode, as follows:

[0036] Step 2.1: Based on the scenario of the hose-type automatic aerial refueling docking problem in speed control mode, establish the corresponding image servo model.

[0037] In the basic image servo control model, and The relationship between them is:

[0038]

[0039] in, is the relative linear velocity, is the relative angular velocity. R re / t =R c / t is the transformation matrix from the tanker coordinate system to the relative coordinate system, and we have,

[0040]

[0041] and are the velocities of the camera and the spinal canal in the relative coordinate system. It is called the Jacobian matrix. Combining equations (5) and (6) we can get:

[0042]

[0043]

[0044] At this point, the image servo model in speed mode is established.

[0045] Step 2.2: According to the scenario of the hose-type automatic aerial refueling docking problem in speed control mode, establish the corresponding camera motion model.

[0046] The receiver is regarded as a rigid body, and the linear velocity of the camera equals to the vector sum of the linear velocity of the mass center of the receiver and the velocity of the camera rotating around the mass center in the receiver coordinate system:

[0047]

[0048] wherein, is the velocity of the camera, is the linear velocity of the receiver,

[0049] is the angular velocity of the receiver, is the coordinate of the camera in the receiver coordinate system.

[0050] According to the definition of the coordinate system, we have,

[0051]

[0052] wherein, is the motion velocity of the camera in the relative coordinate system. Combining equation (10) and (11), we have,

[0053]

[0054] According to the relative motion relationship between the receiver and the receiver, we have,

[0055]

[0056] and assuming 3, equation (12) can be transformed into:

[0057]

[0058] Thus far, the camera motion model in the velocity mode is established.

[0059] Step three: based on the image servo model and the camera motion model in the velocity mode, the expected velocity of the receiver relative to the ground coordinate system is designed.

[0060] According to equation (8), the expected velocity is designed as:

[0061]

[0062] In this case, if then equation (8) becomes:

[0063]

[0064] wherein, is the simplified coefficient, k1 is the adjustable controller parameter and satisfies At this time, ζ1>0, according to the second Lyapunov method, we have

[0065] According to equation (9), the desired velocity of the camera is designed as:

[0066]

[0067] In this case, if equation (9) becomes:

[0068]

[0069] where, is the simplified coefficient, k2 is an adjustable controller parameter and satisfies At this time, ζ2>0, according to the second Lyapunov method, we have

[0070] The desired velocity of the camera is designed as:

[0071]

[0072] At this time, if and k3>0, we have In order to prevent the velocity of the oil receiver from being too fast to miss the cone sleeve, we add the term "-k4|e x |-k5|e y |" in equation (19) to adjust the velocity of the oil receiver, where |e x |,|e y | represent the absolute value of e x ,e y , respectively. If |e x |,|e y | is large, this term can slow down the velocity to avoid overshoot. Thus,

[0073]

[0074] where k3, k4, k5>0 are controller parameters.

[0075] Since is a small quantity and cannot be measured, we regard as a disturbance. Thus we get,

[0076]

[0077] where, is the angular velocity of the camera relative to the cone sleeve, is the angular velocity of the camera in the relative coordinate system, is the angular velocity of the cone sleeve in the relative coordinate system. Therefore, The expected speed can be expressed as:

[0078]

[0079] in,

[0080]

[0081] because because It is the movement of the cone sleeve, so we can only control rather than Therefore, is considered as a disturbance, thus obtaining According to formula (14), the expected speed of the receiving aircraft in the ground system is for:

[0082]

[0083] At this point, the design of the expected speed of the receiving aircraft relative to the ground coordinate system based on the image servo model and the camera motion model is completed.

[0084] Step 4: According to the guidance model, design a dynamic inversion velocity tracking controller based on additive decomposition.

[0085] G in formula (4) x ,G y ,G z It reflects the response of the actual speed of the receiving machine to the reference speed input. In some cases, the response is fast enough to meet the requirements of the docking control. In this case, we can directly use the expected speed Used as reference speed However, in most cases, the underlying speed controller sacrifices maneuverability for safety, resulting in the receiving aircraft's speed response often failing to meet docking requirements. Furthermore, some underlying controllers can suffer from static errors when subjected to turbulence or gusty winds. Therefore, a dynamic inversion speed tracking controller based on additive decomposition is needed to enable the receiving aircraft to track the desired speed more quickly and without static errors.

[0086] For the formula (4) Transfer function G x It can be obtained through system identification methods, generally meeting the following form:

[0087]

[0088] Among them, k x >0,p x <0 are all transfer function parameters. Therefore, the dynamic inversion velocity tracking controller based on additive decomposition can be designed as:

[0089]

[0090] where, is the observed integrated error, including uncertainty, disturbance and input. The function Q x is a low-pass filter, which is chosen as where ∈ x is a tunable controller parameter satisfying (∈ x > 0), and can be tuned to achieve better tracking performance. x

[0091] However, for many underlying controllers, the speed response of the oil machine will exhibit non-minimum phase characteristics, especially and on the channel. This means that for , the transfer function can satisfy the following form:

[0092]

[0093] where, p y1 , p y2 is a negative constant corresponding to a pole on the left side of the imaginary axis, and z y is a normal number corresponding to a non-minimum phase zero on the right side of the imaginary axis, so for the transfer function like formula (27), the function is unstable. In this case, the zero-phase error tracking control theory can be used for the design of the dynamic inversion speed tracking controller based on additive decomposition, and the compensator B y can be designed as:

[0094]

[0095] At this time, the system satisfies G y B y ≈ 1 at low frequencies. The dynamic inversion speed tracking controller based on additive decomposition can be designed as:

[0096]

[0097] where, is the observed integrated error, including uncertainty, disturbance and input. The function Q y is a low-pass filter, which is chosen as where ∈ y is a tunable controller parameter satisfying (∈ y > 0), and can be tuned to balance the tracking performance and stability. y ​​

[0098] Similarly, for The dynamic inversion velocity tracking controller based on additive decomposition can be designed as:

[0099]

[0100] where B z is the compensator, is the observed integrated error, including uncertainty, disturbance and input. Function Q z is a low-pass filter, which is selected as where ∈ z is an adjustable controller parameter and satisfies (∈ z > 0), the tracking performance and stability can be balanced by adjusting ∈ z .

[0101] So far, the dynamic inversion velocity tracking controller based on additive decomposition is designed.

[0102] By integrating step two, step three and step four, the control block diagram can be obtained as Figure 4 .

[0103] So far, the image servo hose type automatic aerial refueling docking controller in the speed control mode is designed.

[0104] The advantages and effects of the present application are that: the present application solves the control problem of hose type automatic aerial refueling docking in the speed control mode of the receiver based on the image servo method under the condition that the existing bottom controller cannot be modified or the throttle and rudder surface are not allowed to be directly controlled for safety, and can resist disturbances including head wave effect, turbulence, gust and wake, while meeting safety, reliability and robustness, so that the hose type automatic aerial refueling docking can be efficiently and safely completed, which is closer to the actual application scene and can be applied to automatic aerial refueling practice of manned or unmanned aircraft in military or civilian fields. BRIEF DESCRIPTION OF DRAWINGS

[0105] Figure 1 is a schematic diagram of ground, relative, tanker, receiver, camera coordinate systems.

[0106] Figure 2 is a schematic diagram of an image coordinate system.

[0107] Figure 3 is a depth diagram.

[0108] Figure 4 is a controller structure block diagram of the present application.

[0109] Figure 5 is a docking view.

[0110] Figure 6 is a docking process screenshot under undisturbed conditions.

[0111] Figure 7 is the image error curve, the depth error curve and the docking error curve of the docking process under undisturbed conditions.

[0112] Figure 8 is the image error curve, the depth error curve and the docking error curve of the docking process under the conditions of adding I-class turbulence, head wave effect and interference of tanker wake.

[0113] Figure 9 is the image error curve, the depth error curve and the docking error curve of the docking process under the conditions of adding gust, I-class turbulence, head wave effect and interference of tanker wake.

[0114] Figure 10 is the image error curve, the depth error curve and the docking error curve of the docking process under the conditions of adding gust, II-class turbulence, head wave effect and interference of tanker wake. DETAILED DESCRIPTION

[0115] The application provides an image servo hose type automatic aerial refueling docking control method in a speed control mode. The specific embodiments are further described by taking the scene of hose type automatic aerial refueling docking with different interference as an example. Simulation and calculation are performed on MATLAB R2022b, Python 3.8 and RflySim3D under the Win10 professional edition operating system of a computer with a frequency of 3.60 GHz and a memory of 64 GB.

[0116] The specific operation realized by the application is as follows:

[0117] Step 1.1 is completed according to the coordinate system establishment.

[0118] Step 1.2 is completed according to the six-degree-of-freedom model establishment of F16: the tanker cruises along a straight line at v t = 120 m / s (393.72 ft / s) at h t = 3000 m (9843 ft) in the high altitude. Formula (3) is the six-degree-of-freedom model of the receiver.

[0119] Step 1.3 is completed according to the establishment of the guidance model of the receiver in the speed control mode: the bottom controller adopts the model predictive control method, at this time, the guidance model of the receiver can be represented by formula (4), and after system identification, the specific guidance model is:

[0120]

[0121] Step 1.4 is completed according to the control problem description: the control problem is to design a suitable reference speed for the system (4) As system input, make the image tracking error and the depth error grow with time both tend to zero And make the docking error grow with time tend to zero

[0122] According to step 2.1, the image servo model in speed mode is established: the image servo model is represented by formula (8), (9).

[0123] According to step 2.2, the camera motion model in speed mode is established: the camera motion model is represented by formula (14).

[0124] According to step three, the expected speed of the refueling machine relative to the ground coordinate system is designed: the expected speed is represented by formula (24), wherein the parameters are: k1=3, k2=3, k3=1, k4=3, k5=3.

[0125] According to step four, the dynamic inversion speed tracking controller based on additive decomposition is designed: the controller is represented by (26), (29), (30), wherein the parameters are: ∈ x =0.1, ∈ y =1.5, ∈ z =0.1.

[0126] Next, simulation experiments are carried out in the simulation platform to verify the effectiveness, reliability and robustness of the controller designed by the application. The visual display effect of the simulation platform is shown in the following figure, including the third person perspective and the camera perspective of the refueling machine. Various disturbances can be added in the simulation platform, including turbulence, refueling machine wake, head wave effect and gust, etc. The identification of the cone sleeve in the simulation platform is realized by using Yolo V5 neural network, which can accurately identify the cone sleeve and return the center position of the cone sleeve and the estimated depth. The specific simulation results are analyzed as follows. Figure 5

[0127] is a screenshot of the docking process under the condition of no disturbance, Figure 6 is the corresponding image error curve, depth error curve and docking error curve. From Figure 7 it can be seen that the refueling machine is located in the far rear of the refueling machine at 1s, and then gradually approaches the cone sleeve under the action of the controller until 67s when the cone pipe is inserted into the cone sleeve, and the docking is successful. Figure 6 In the figure, it can be seen that the two curves of the image error, the depth error curve and the three curves of the docking error all converge to about 0 at 67s, which shows that under the condition of no disturbance, the control method proposed by the application can successfully realize docking. Figure 7

[0128] Figure 8 ​is the image error curve, the depth error curve and the docking error curve of the docking process under the condition of adding I-class turbulence, head wave effect and tanker wake interference. Figure 8 The two curves of image error, the depth error curve and the three curves of docking error all converge to about 0 at 45s, although there are some fluctuations in each curve due to the addition of I-class turbulence, head wave effect and tanker wake, but the control method proposed in the application can successfully realize docking under these disturbances, and has certain reliability and robustness.

[0129] Figure 9 is the image error curve, the depth error curve and the docking error curve of the docking process under the condition of adding gust, I-class turbulence, head wave effect and tanker wake interference. It can be seen that Figure 9 The two curves of image error, the depth error curve and the three curves of docking error all converge to about 0 at 42s, although there are greater fluctuations compared with Figure 8 due to the addition of gust, I-class turbulence, head wave effect and tanker wake, but the control method proposed in the application can successfully realize docking under these disturbances, and has better reliability and robustness.

[0130] Figure 10 is the image error curve, the depth error curve and the docking error curve of the docking process under the condition of adding gust, II-class turbulence, head wave effect and tanker wake interference. It can be seen that Figure 10 The two curves of image error, the depth error curve and the three curves of docking error all converge to about 0 at 39s, although there are greater fluctuations compared with Figure 8 due to the increase of the intensity of gust, but the control method proposed in the application can successfully realize docking under the interference of gust, II-class turbulence, head wave effect and tanker wake, and has higher safety, reliability and robustness.

[0131] In summary, the effectiveness of the image servo hose type automatic aerial refueling docking control method under the speed control mode proposed in the application can be proved.

Claims

1. A method for controlling an image servo hose-type automatic aerial refueling docking in a speed control mode, characterized by: The steps include the following: Step one: complete the hose automatic aerial refueling docking control problem description in the speed control mode, including the establishment of coordinate system, the establishment of the six degrees of freedom model of the receiver, the establishment of the receiver guidance model in the speed control mode, and the summary of the control problem; Step two: establish the image servo model in the speed control mode and the camera motion model; Step three: based on the image servo model and the camera motion model in the speed control mode, design the expected speed of the receiver relative to the ground coordinate system; Step four: according to the guidance model, design a dynamic backstepping speed tracking controller based on additive decomposition; In step one, according to the motion relationship of each object in the hose automatic aerial refueling docking process in the speed control mode, all coordinate systems required to describe the docking control problem are established; In order to establish the system model in the docking stage, five coordinate systems need to be defined: ground coordinate system o g x g y g z g , receiver coordinate system o r x r y r z r , refueling machine coordinate system o t x t y t z t , camera coordinate system o c x c y c z c and relative coordinate system o re x re y re z re ; represents a certain state of object i in j coordinate system; R i / j , i, j ∈ {g, r, t, c, re} represents the rotation matrix from j coordinate system to i coordinate system; is the forward flight speed of the tanker in ground coordinate system, R r / t represents the rotation matrix used to describe the rotation of the receiver coordinate system "r" relative to the tanker coordinate system "t"; In step one, according to the scene of the hose automatic aerial refueling docking problem in the speed control mode, the six degrees of freedom model of the receiver is established; In the docking stage, the pitch angle of the receiver changes in a small range, the reference state is constant speed and straight flight, and the attitude satisfies the coordinated turn condition, that is, the roll angle and the sideslip angle satisfy Φ = β = 0, the angle of attack α, the flight path angle γ and the pitch angle θ satisfy θ = γ + α; In this case, the motion equation of the receiver in the tanker coordinate system is decoupled into the lateral channel motion and the longitudinal channel motion; The motion equation of the longitudinal channel is: The motion equation of the lateral channel is: where, respectively represent the coordinates of the center of gravity of the oil receiver in the ground coordinate system, and (φ, θ, ψ) respectively represent the Euler angles of the three axes, respectively represent the velocities of the oil receiver relative to the ground, p, q, r respectively represent the angular velocities of the three axes, L, D, T, N respectively represent the lift, the drag, the engine thrust, the moments of the three channels of pitch, roll and yaw, and the side force, and c1-c9 respectively represent the nine elements of the moment of inertia matrix; the combination of equations (1) and (2) is simplified as: wherein represents the state of the receiver, Vector u r = [δ t δ e δ a δ r ] T denotes control inputs, including throttle, elevator, aileron, and rudder, respectively; In step one, according to the scene of the hose automatic aerial refueling docking problem in the speed control mode, the receiver guidance model in the speed control mode is established; The bottom controller of the speed control mode is realized by model predictive control, and the guidance model of the receiver under the action of the bottom controller is described by formula (4): wherein, represents the position of the receiver in the ground coordinate system, represents the velocity of the receiver in the ground coordinate system, represents the reference velocity of the receiver in the ground coordinate system, G x ,G y ,G z is a transfer function from to ; the transfer function G x ,G y ,G z is obtained by means of system identification; In step one, according to the scene of the hose automatic aerial refueling docking problem in the speed control mode, the control problem is completed; Let the camera be mounted at the tip of the cone; define the projection of the center of the cone sleeve in the image coordinate system as The image tracking error e is defined as: wherein is the origin of the image coordinate system; define is the depth, i.e. the distance of the camera and the center plane of the cone along the z c axis in the camera coordinate system; make the following settings: Setting one: the vision recognition module accurately and in real time returns the center coordinates of the cone sleeve and estimates the depth; Setting two: the tanker is in constant speed and straight flight state during the entire docking stage, and the initial state of the receiver is the same as that of the tanker; Setting three: the receiver tracks the reference speed command without static error under the action of the bottom controller without disturbance; Designing a suitable reference velocity for formula (4) As a system input, so that the image tracking error and the depth error both tend to zero as time grows, i.e. e(t)→0, and so that the docking error tends to zero as time grows, i.e. In step two, according to the scene of the hose automatic aerial refueling docking problem in the speed control mode, the corresponding image servo model is established; In the image servo control model, and the relationship between them is: wherein is the relative linear velocity, is the relative angular velocity; R re / t = R c / t is the transformation matrix from the fueler coordinate system to the relative coordinate system, and has, and are the velocities of the camera and the spinal canal in the relative coordinate system; known as the Jacobian matrix, combining equations (5) and (6) gives:

2. The method of controlling the automatic aerial refueling docking of a probe-and-drogue system in a speed control mode as in claim 1, wherein: In step two, according to the scene of the hose automatic aerial refueling docking problem in the speed control mode, the corresponding camera motion model is established; The receiver is regarded as a rigid body, and in the tanker coordinate system, the linear velocity of the camera is equal to the vector sum of the linear velocity of the receiver mass center and the velocity of the camera rotating around the mass center: wherein, is the speed of the camera, is the linear speed of the oil receiver machine, is the angular velocity of the oil receiver, is the coordinates of the camera in the oil receiver coordinate system; According to the definition of coordinate system, we get, wherein, is the motion velocity of the camera in the relative coordinate system; combining equations (10) and (11), we get, According to the relative motion relationship between the receiver and the tanker: And setting 3, formula (12) is simplified as:

3. The image servo hose-type automatic aerial refueling docking control method in a speed control mode according to claim 2, characterized in that: In step three, the desired velocity is designed according to formula (8), is designed as: In this case, if then equation (8) becomes: wherein, are the simplified coefficients, k1is an adjustable controller parameter and satisfies Then, ζ1> 0, according to the Lyapunov second method, we have According to equation (9), The desired velocity of the is designed to be: In this case, if then equation (9) becomes: wherein, are the simplified coefficients, k2is an adjustable controller parameter and satisfies Since ζ2> 0, according to the Lyapunov second method, we have The desired speed of the vehicle 100 is designed to be: At this time, if and k3 > 0, then To prevent the speed of the oil receiver from being too fast and missing the cone sleeve, a term "-k4|e x |-k5|e y " is added to equation (19) to adjust the speed of the oil receiver, where |e x |,|e y | respectively represent the absolute value of e x ,e y ; if |e x |,|e y | is large, the term slows down the speed to avoid overshoot; thus, Wherein, k3, k4, k5 > 0 are controller parameters; Since is a small quantity and not measurable, we consider as a perturbation; thus we get, wherein, is the angular velocity of the camera relative to the cone, is the angular velocity of the camera in the relative coordinate system, is the angular velocity of the cone in the relative coordinate system; so that, the desired velocity of is expressed as: Wherein, Due to Because Since the motion of the cone is controlled Rather than Therefore, the is considered as a disturbance, so that According to equation (14), the desired speed of the oil receiver in the ground system is:

4. The method of claim 3, wherein: In step four, for formula (4) Transfer function G x Obtained by system identification method, satisfying the following form: where k x > 0, p x < 0 are transfer function parameters; then the dynamic inversion velocity tracking controller based on additive decomposition is designed as: where, is the observed integrated error, including uncertainty, disturbance and input; function Q x is a low-pass filter, which is chosen here as where ∈ x is an adjustable controller parameter and satisfies ∈ x > 0, and better tracking performance is achieved by adjusting ∈x.

5. The method of claim 4, wherein: For many bottoming controllers, the speed response of the oil- lubricated machine will exhibit non-minimum phase characteristics on the channel and For the case of a hydraulic actuator, the transfer function satisfies the following form: G(s) = Ks2+ s Among them, p y1 ,p y2 is a negative constant, corresponding to the pole on the left side of the imaginary axis, and z y is a positive constant, corresponding to the non-minimum phase zero point located on the right side of the imaginary axis. Therefore, for the transfer function of the form (27), the function is unstable; in this case, the zero phase error tracking control theory is used to design a dynamic backstepping velocity tracking controller based on additive decomposition, and the compensator B y Designed to: At this time the system in low frequency band to meet G y B y ≈1; based on the dynamic inversion of the decomposition of the velocity tracking controller design: where is the observed integrated error, including uncertainty, disturbances, and inputs; function Q y is a low-pass filter, chosen here as where ∈ y is an adjustable controller parameter and satisfies ∈ y > 0, balancing tracking performance and stability by adjusting ∈y. For The dynamic inversion velocity tracking controller based on additive decomposition is designed as: where B z is a compensator, is the observed integrated error, including uncertainty, disturbance, and input; function Q z is a low-pass filter, chosen here as where ∈zis an adjustable controller parameter and satisfies ∈ z > 0, balancing tracking performance and stability by adjusting ∈z.

Images