Oil receiver robust flight control method based on adaptive extended state observer
Through the adaptive extended state observer and reverse step control strategy, combined with the command filter, the problem of interference of the oil receiver during air refueling is solved, and the estimation compensation for the induced pneumatic tail vortex effect and external disturbance is achieved, thereby improving the oil receiver The robust flight control performance of the aircraft ensures the safety and stability of aerial refueling missions.
Patent Information
- Application Number
- CN202410876730.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-02
- Publication Date
- 2024-11-15
AI Technical Summary
During the aerial refueling process, the oil-received engine is disturbed by unmodeled dynamics and external meteorological environment, which increases the difficulty of control. The existing technology is difficult to effectively suppress these interferences and ensure the stable, reliable and robust control of the oil-received engine.
Using a robust flight control method based on an adaptive extended state observer, an inverse step control strategy is designed by estimating the interference values in each state channel, and combining the command filter and the extended state observer to achieve the effect of induced aerodynamic tail vortex. Model the estimation compensation of dynamic impact and external meteorological disturbances, optimize the internal and external ring controllers, and ensure the transient and steady-state response performance of the oil-received engine.
It effectively reduces the system's steady-state error and transient overshoot, improves the safety and control accuracy of aerial refueling tasks, and ensures the robust flight performance of the oil-receiving engine in tight formations.
Smart Images

Figure CN118963401A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of aircraft attitude control, and in particular to a robust flight control method for a receiving aircraft based on an adaptive extended state observer. Background Art
[0002] Aerial refueling can increase the range and flight time of aircraft in the aviation field, and has research value in both military and civilian fields. In the military field, aerial refueling can improve the endurance and mounting capacity of fighter jets, extend the time in the air, and increase the suddenness of attacks. In the civilian field, it can increase the payload of aircraft, reduce operating costs, and effectively improve the economy and safety of civil flights. Autonomous aerial refueling can effectively reduce the dependence of the refueling process on pilot operations, and has attracted the attention of more and more researchers in recent years.
[0003] The autonomous aerial refueling mission of fixed-wing aircraft can be divided into five links: rendezvous, formation, docking, refueling, and separation. During the docking and refueling process, the distance between the tanker and the receiver is relatively close, which belongs to a close formation. At this time, the receiver will fly in the tail vortex of the tanker and be affected by the induced aerodynamic disturbance generated by the tail vortex of the tanker. In addition, the receiver will also be disturbed by unmodeled dynamics and external meteorological environment. It is worth noting that these lumped disturbances composed of unmodeled dynamics, induced aerodynamic disturbances and various external disturbances will act on each state channel of the receiver, including both matching disturbances that appear in the same channel as the control input and unmatched disturbances in other state channels. How to suppress these disturbances and achieve stable and reliable robust control of the receiver is a key problem that needs to be solved in autonomous aerial refueling.
[0004] At present, for the problem of aerial refueling in close formation, a close formation kinematic and dynamic model based on wind shear term is often used, which introduces the influence of the induced aerodynamic force and torque generated by the leader on the follower. For the problem of the aerodynamic effect model of the trailing vortex, a close formation aerodynamic model based on lift line theory is commonly used. This model characterizes the aerodynamic induced effect caused by the leader's trailing vortex through the position function and direction function between the leader and the follower. On this basis, since backstepping control has unique performance advantages for disturbances that do not meet the matching conditions, there is a backstepping control strategy that combines the uncertainty and disturbance estimator (UDE) with the command filter (CF) for the robust flight control of close formation UAVs. Among them, the UDE used is a disturbance observer based on low-pass filtering, and its observation performance is determined by the constant parameters adjusted by the designer. It is particularly worth noting that when using it for disturbance estimation, the initial conditions must be strictly set, otherwise the initial estimation error will produce a peak phenomenon, resulting in a larger overshoot in the transient response of the system, and even control divergence. Similarly, the extended state observer (ESO) and extended high-gain observer also have such problems. In order to ensure the safety of aerial refueling, more stringent requirements are put forward for the transient control performance and steady-state control performance of the receiving aircraft, which is particularly important in the aerial docking and refueling maintenance phase. In the docking phase, if the transient performance is poor and a large overshoot occurs, it may cause the receiving aircraft to collide with the refueling aircraft. In the refueling maintenance phase, if the steady-state performance is poor, a large steady-state error occurs or the position fluctuates greatly, it may cause the refueling soft / hard pipe to break. Summary of the invention
[0005] In order to overcome the problems of the above controller, the present invention proposes a robust flight control method for a receiving aircraft based on an adaptive extended state observer, so as to take into account the transient and steady-state response performance of the inner and outer loop attitude control systems.
[0006] The technical solution adopted by the present invention is:
[0007] A robust flight control method for a receiving aircraft based on an adaptive extended state observer comprises the following steps:
[0008] Step 1, based on the position coordinates, speed, track inclination angle, track azimuth of the tanker in the inertial coordinate system and the optimal relative distance between the tanker and the tanker, the expected position, expected track inclination angle and track azimuth of the tanker in the inertial coordinate system are obtained; the expected position derivative of the tanker is estimated according to the three-degree-of-freedom kinematic model and attitude dynamics model of the tanker and the command filter (CF);
[0009] Step 2, based on the adaptive extended state observer (AESO), obtain the interference estimation value in each state channel;
[0010] Step 3, design the outer loop controller, using the backstepping control strategy: design the outer loop controller according to the position of the tanker and the desired speed of the receiving aircraft, the desired track inclination angle and the track azimuth angle, and the outer loop controller generates the desired thrust command, the desired track roll angle and the desired angle of attack as the input of the inner loop controller;
[0011] The actual position, speed, track inclination angle and track azimuth of the receiving aircraft are respectively input into the adaptive extended state observer, and based on the output thereof, the position channel interference estimation value, the speed channel interference estimation value, the track inclination angle channel interference estimation value and the track azimuth angle channel interference estimation value are obtained;
[0012] Write the dynamic equation of position error based on the obtained position interference estimate, receiver aircraft speed, track inclination angle and track azimuth;
[0013] According to the position error dynamic equation, the expected speed, expected track inclination angle and expected track azimuth of the receiving aircraft are designed;
[0014] According to the state equation of the receiving aircraft, the error dynamic equations of the receiving aircraft speed, track inclination angle and track azimuth can be obtained;
[0015] According to the error dynamic equations of the receiving aircraft speed, track inclination angle and track azimuth angle, a virtual intermediate control input is defined, thereby the desired control input can be designed;
[0016] According to the designed desired control input and the error dynamic equations of the receiving aircraft speed, track inclination angle and track azimuth angle, the desired output of the outer loop controller can be obtained, including: the desired thrust command, the desired track roll angle and the desired angle of attack;
[0017] Step 4, design the inner loop controller, adopt the backstepping control strategy: take the output of the outer loop controller as the expected input of the inner loop controller, and solve the inner loop attitude stability and controller to obtain the rudder deflection control input of the receiving aircraft, which includes: aileron deflection angle, elevator deflection angle and rudder deflection angle;
[0018] Inputting the desired track roll angle and the desired angle of attack obtained by the outer loop controller into the command filter, and obtaining derivative estimates of the desired track roll angle and the desired angle of attack based on the output;
[0019] According to the state equation of the receiving aircraft, the error dynamic equations of the receiving aircraft's track roll angle, angle of attack and sideslip angle are defined;
[0020] The intermediate virtual control quantity is introduced to obtain the error dynamic equations of the receiving aircraft's track roll angle, angle of attack and sideslip angle, in which the estimated value of the disturbance term is given by the adaptive extended state observer.
[0021] The desired control input is constructed based on the error dynamic equation, thereby inferring the desired angular velocity signal;
[0022] The obtained desired angular velocity signal is input into a second-order command filter to obtain a derivative estimation value of the desired angular velocity;
[0023] The error equation of the angular velocity signal of the receiving engine can be obtained according to the state equation of the receiving engine. The error dynamic equation of the angular velocity signal is obtained by introducing the intermediate virtual control quantity.
[0024] According to the error dynamic equation of the angular velocity signal, the expected control input is designed to obtain the rudder deflection control input of the receiving aircraft, including: aileron deflection angle, elevator deflection angle and rudder deflection angle, that is, the deflection amount of the operating control surface is obtained.
[0025] Furthermore, the three-degree-of-freedom kinematic and dynamic models of the fixed-wing aircraft constructed by the present invention are specifically:
[0026]
[0027] Among them, x f ,y f ,z f V is the position of the receiving aircraft in the inertial coordinate system (such as the North East Earth (NED) coordinate system), in meters. f ,γ f ,x f are the airspeed, track inclination angle and track azimuth of the receiving aircraft, in m / s, rad; D, Y, L are the initial aerodynamic drag, side force and lift of the receiving aircraft, in N; T is the throttle thrust, m f is the mass of the receiving machine, μ f ,α f ,β f are the rolling angle, angle of attack and sideslip angle of the receiving aircraft, respectively, in rad; p, q, r are the components of the angular velocity of the receiving aircraft relative to the inertial coordinate system in the x, y, and z axes of the body coordinate system, in rad / s; l, M, N are the original rolling, pitching, and yaw moments acting on the receiving aircraft, in N·m; I x ,I y ,I z ,I xz are the moment of inertia and product of inertia around the x, y, and z axes of the aircraft, respectively, and g represents the acceleration due to gravity. In addition, the induced wake velocity ω x ,ω y ,ω z, induced aerodynamic drag ΔD, side force ΔY, lift ΔL and induced rolling moment Δl, pitch moment ΔM, yaw moment ΔN, etc. are regarded as the interference caused by the tail vortex of the tanker on the receiving aircraft, and the units are m / s, N and N·m respectively. The uncertainty and interference terms in the lumped form of the above system can be obtained as follows:
[0028]
[0029] When designing the trajectory tracking controller of the receiving aircraft, the present invention first adopts the traditional inner and outer loop feedback loop structure to solve the under-driven characteristic problem of the receiving aircraft. The outer loop is the guidance loop, which is mainly used to stabilize and control the aircraft's motion trajectory, while the inner loop is the stabilization loop, which mainly plays a role in stabilizing and controlling the aircraft's attitude. The outer loop controller generates the desired thrust command T c , the desired track roll angle μ d and the desired angle of attack α d Since the present invention adopts the tilt turning control method, the desired sideslip angle β d = 0. The expected value μ d and α d , β d As the input of the inner loop, the control input of the aircraft is obtained by the inner loop attitude stability and controller solution, that is, the rudder command (i.e., rudder deflection control input), including the aileron deflection angle δ a , elevator deflection angle δ e and the rudder deflection angle δ r . This allows the receiving aircraft to track its attitude to the desired attitude angle μ d ,α d , and then complete the trajectory tracking control. Secondly, the inner and outer loops are composed of two sub-loops each. Since the backstepping control strategy has unique performance advantages for non-matching disturbances, the backstepping method is used to gradually design the controllers in the inner and outer loops and their sub-loops. At the same time, in order to solve the derivation problem of the intermediate virtual control quantity involved in the backstepping method, the second-order command filter of step 1 is adopted. In addition, in order to take into account the transient performance and steady-state performance of the trajectory tracking of the receiving machine, AESO is used to estimate the interference in each state channel and compensate it in the controller. That is, each loop controller contains two parts: the nominal controller part for the undisturbed nominal system and the compensation part for the interference.
[0030] Furthermore, to facilitate the controller design, the generation of the desired trajectory is first introduced: based on the position coordinates, speed, track inclination angle, track azimuth of the tanker in the inertial coordinate system and the optimal relative distance between the tanker and the tanker, the desired speed, expected track inclination angle and expected track azimuth of the tanker in the inertial coordinate system are obtained. Let x l ,y l ,z lare the components of the x, y, and z axes of the tanker in the inertial coordinate system (referred to as the inertial system), V l ,γ l ,x l denote the tanker's airspeed, track inclination angle, and track azimuth, respectively. Assuming that the receiving aircraft can obtain the above information of the tanker, let l x ,l y ,l z are the components of the relative optimal distance vector between the receiving aircraft and the refueling aircraft in the inertial system on the x, y, and z axes, respectively. The first-order differential form of the expected trajectory of the receiving aircraft in the inertial system is as follows:
[0031]
[0032] Among them, x r =x l +l x ,y r =y l +l y ,z r =z l +l z They represent the expected positions of the receiving aircraft on the x, y, and z axes in the inertial system, respectively. r ,γ r ,x r They represent the desired speed, desired track inclination angle and desired track azimuth of the receiving aircraft respectively, and are calculated by the following formula:
[0033]
[0034] Since the expected relative optimal distance is generally expressed by the tanker body coordinate system or airflow coordinate system, It is related to the unknown rotational angular velocity of the refueling maneuver coordinates and cannot be directly obtained analytically. To solve this problem, use the command filter (CF) in step 1 to filter l x ,l y ,l z Filter to obtain the expected relative position signal l xc ,l yc ,l zc and its derivatives Observed values:
[0035]
[0036] in, By adjusting the filter damping ratio and bandwidth Get the target signal The command filter signal
[0037] Furthermore, in the backstepping control strategy, the model uncertainty and external interference in different channels will have an adverse effect on the trajectory tracking effect. In order to compensate for it, the present invention designs AESO to estimate and compensate for the interference such as the induced aerodynamic wake effect generated by the tanker, the unmodeled dynamic influence and the external meteorological disturbance suffered by the receiving aircraft, so that a robust flight control law considering the external lumped interference can be obtained.
[0038] Furthermore, the design process of the adaptive extended state observer in step 2 is as follows:
[0039] Based on the designed time varying extended state observer (TVESO), the estimation error is introduced into the TVESO to form AESO, which can effectively solve the peak phenomenon problem caused by the initial observation error and reduce the final steady-state error.
[0040] For a first-order disturbed system
[0041]
[0042] Among them, U x is the system input, that is, in formula (6), x represents the system state, Δx represents the system model uncertainty and interference term, and a general second-order linear extended state observer (LESO) system can be designed:
[0043]
[0044] in, are the estimated values of x and Δx in equation (6), is the observation error of x, l1,l2 are constant observation gains. Using the closed-loop observer bandwidth ω n , the second-order LESO gain l1, l2 can be designed as l1 = 2ω n , ω n >0 can ensure the stability of LESO. However, because LESO requires a higher gain to achieve its steady-state performance, its peak phenomenon at the initial moment is more serious, especially when there is an initial estimation error, the transient performance is poor. To solve this problem, the following TVESO can be established:
[0045]
[0046] Among them, l1(t), l2(t) are time-varying observation gains, which can be designed as:
[0047]
[0048] Among them, ξ is a constant damping ratio, ω n (t) is the closed-loop time-varying observer bandwidth.
[0049] The designed AESO introduces the estimation error e1 into TVESO based on TVESO, and introduces e1 as a parameter into the desired observer bandwidth ω nc Therefore, the gains l1(t) and l2(t) are adjusted in real time with e1.
[0050] Next, we use e1 as a parameter to design the desired observer bandwidth as follows:
[0051]
[0052] Among them, ω nc is the desired observer bandwidth, is the fixed gain of different interference items. For the total interference items in the system, the interference item identifier ω ncmax ,ω ncmin For the preset ω nc The upper and lower limits of e1 may be due to the change of ω nc The expected observer bandwidth ω obtained by equation (10) is nc (t) is input to the Butterworth filter to obtain the observer bandwidth ω n (t) and its higher-order differential terms. This makes the observer bandwidth ω n (t) can be increased from a relatively small initial bandwidth value ω as e1 converges. n (0), smoothly transition to the minimum expected bandwidth ω ncmin (t), and finally to the maximum expected bandwidth ω ncmax (t).
[0053] For a second-order system, a third-order Butterworth filter can be designed as:
[0054]
[0055] Its pole can be set to -ω B ,(-0.5±j0.866)ω B , through the parameter ω B Adjust ω n (t) response speed, so that it can smoothly transition to the desired bandwidth ω nc (t). This filter makes ω n (t) In bandwidth instruction ω nc(t), and thus obtain appropriate time-varying gains l1(t), l2(t).
[0056] Furthermore, the controller is designed. The nonlinear model of the receiving machine is decomposed into two parts: the inner and outer loop models. The outer loop kinematic and dynamic model consists of the first six state equations in formula (1), and the inner loop model consists of the last six state equations in formula (1).
[0057] First, the outer loop controller is designed. The controllers of the outer loop and its sub-loops are designed step by step using the backstepping method.
[0058] In formula (2), ω P ,P∈{x,y,z}and d Λ ,Λ∈{V,γ,χ} is the interference term in the outer loop state equation. In order to compensate for it, the present invention uses AESO to estimate it. First, let Denote the interference term ω x ,ω y ,ω z The estimated value of , the position equation of the receiving machine can be rewritten as:
[0059]
[0060] in,
[0061]
[0062] Therefore, the position tracking error of the receiving machine is introduced as follows:
[0063]
[0064] Among them, e x 、e y 、e z are the components of the position tracking error of the receiving aircraft in the x, y, and z axes respectively;
[0065] Then, the first-order differential equation of position error is obtained by derivation of equation (14):
[0066]
[0067] in, In order to gradually reduce the error, the expected speed V is designed d , expected track tilt angle γ d and the expected track azimuth χ d for:
[0068]
[0069] Among them, K x >0,Kz >0 is the controller parameter gain, auxiliary amount V d ,γ d Input to the command filter (CF) to obtain the command expected V c ,γ c and its derivatives Next, define e V =V f -V c , e γ =γ f -γ c , e χ = x f -x d , then according to the state equation (1) of the receiving aircraft, the error dynamic equations of the receiving aircraft speed, track inclination angle and track azimuth angle are:
[0070]
[0071] Among them, L0 is the lift at zero angle of attack, L α is the aerodynamic derivative of lift with respect to angle of attack, and L = L0 + L α α f From this, we can see that by changing the thrust T and the angle of attack α d and roll angle μ d The size of the virtual intermediate control input u can be changed Λ ,Λ∈{V,γ,χ}, and then achieve the tracking control of the outer loop. The desired control input is designed as follows:
[0072]
[0073] in, is d Λ The estimated value of d Λ Input into AESO to obtain the baseline vacancy rate. Specifically:
[0074]
[0075] Among them, the auxiliary K V >0,K γ >0,K χ >0 is the control gain, c V ,c χ is a constant, ε x =e x -ξ x , ε z =e z -ξ zAn auxiliary signal is created to compensate for the error in estimating the desired signal and its derivative using the command filter (CF), and is:
[0076]
[0077] Finally, the desired output of the outer loop controller can be derived from the error dynamic equations (17) of the receiving aircraft speed, track inclination angle and track azimuth angle and the designed desired control input (19): that is, the desired thrust command T c , the desired angle of attack α d and the desired track roll angle μ d .
[0078]
[0079] In this way, the thrust command T is generated through the two sub-loop controllers of the outer loop. c and the desired angle of attack α required for the inner ring d and the desired track roll angle μ d .
[0080] Furthermore, the inner loop attitude controller is designed, and the controller of the inner loop and its sub-loops is gradually designed by backstepping method so that the aircraft angle of attack and track roll angle can quickly track the expected value α respectively. d ,μ d , while keeping the sideslip angle at 0.
[0081] First, the command expected value α of the inner loop is obtained through the command filter (CF) c ,μ c and its derivatives And let e α =α f -α d ,e μ =μ f -μ d ,e β =β f -0,e Θ =[e α ,e μ ,e β ] T ,Θ d =[μ d ,α d ,0] T ,Θ c =[μ c ,α c ,0] T ,Ψ=[γ f ,x f ] T ,Ω=[p,q,r] TAccording to the state equation (1) of the receiving aircraft, the error equations of the aircraft's track roll angle, angle of attack and sideslip angle can be obtained:
[0082]
[0083] in
[0084]
[0085] because Unmeasurable, but can be measured by observation In addition, the expected value obtained by the command filter (CF) and There is also an error between them, so the model uncertainty and interference on the final Θ channel can be expressed as:
[0086]
[0087] Furthermore, an intermediate virtual control quantity is introduced:
[0088]
[0089] The dynamic equations of the receiving aircraft's track roll angle, angle of attack and sideslip angle errors can be obtained:
[0090]
[0091] The expected control input is:
[0092]
[0093] Where K Θ =diag{K μ ,K α ,K β} is a positive definite matrix, and d Θ Input into AESO to get d Θ Estimated value By calculating from formula (25), we can know that the expected angular velocity signal is:
[0094]
[0095] Among them, Ω d =[p d ,q d ,r d ] T , Ω d Input to the second-order command filter (CF) to get the expected signal Ω c =[p c ,q c ,r c ] Tand its derivatives
[0096] Furthermore, define e Ω =Ω-Ω c , the error equation of the angular velocity signal of the receiving engine can be obtained from the receiving engine state equation (1):
[0097]
[0098] where τ = [l, M, N] T , Δτ=[Δl,ΔM,ΔN] T , Ι is the moment of inertia matrix of the aircraft. And the aerodynamic control surface of the aircraft (i.e., the rudder control input) δ u =[δ a ,δ e ,δ r ] T and τ=[l,M,N] T There are the following relationships between them:
[0099] τ=τ0+M τ δ u (30)
[0100] Among them, M τ is the control derivative matrix, τ0 is the rudder deviation δ u =[0,0,0] T The torque vector at .
[0101] Therefore, according to the error equation of the angular velocity signal of the receiving engine, the virtual control input (i.e., the intermediate virtual control amount) u can be defined Ω =[u p ,u q ,u r ] T for:
[0102] u Ω =-I -1 Ω×IΩ+I -1 (τ0+M τ δ u ) (31)
[0103] Among them, u p ,u q ,u r They are the virtual control quantities corresponding to the components p, q, and r of the angular velocity of the receiving aircraft relative to the inertial coordinate system in the aircraft body coordinate system's x, y, and z axes.
[0104] Based on the above expression, the error differential equation of the angular velocity signal of the receiving aircraft can be obtained:
[0105]
[0106] Among them, d Ω =[d p ,d q ,d r ] T =I -1 Δτ is the lumped interference term, and its specific form is shown in formula (2).
[0107] The expected control input is designed based on the error differential equation of the receiving engine angular velocity signal as follows:
[0108]
[0109] Among them, K Ω =diag{K p ,K q ,K r} is a positive definite diagonal constant gain matrix, C Ω =diag{c p ,c q ,c r} is a constant parameter matrix, and d Ω Input into AESO to get d Ω Estimated value of ε Θ =[ε μ ,ε α ,ε β ] T =e Θ -ξ Θ is an auxiliary signal created to compensate for the error generated when using the command filter (CF), and has:
[0110]
[0111] Finally, the control amount of the operating control surface can be calculated by formula (31): the aileron deflection angle δ a 、Elevator deflection angle δ e and the rudder deflection angle δ r
[0112]
[0113] So that the attitude of the receiving aircraft can track the desired attitude angle μ d and α d , and then complete the trajectory tracking control.
[0114] The technical solution provided by the present invention brings at least the following beneficial effects:
[0115] (1) The tracking control law of the present invention can ensure the global asymptotic stability of the nominal nonlinear closed-loop system, and the output of the adaptive extended state observer can be used to compensate for system uncertainty and aerodynamic disturbances induced by the tanker wake vortex;
[0116] (2) The present invention can effectively estimate the unknown wake vortex-induced aerodynamic effects and unmodeled dynamic influences and external meteorological disturbances, and has the characteristics of small transient overshoot and small steady-state error, ensuring the stability of the closed-loop system and improving the transient response performance and steady-state performance of the closed-loop system, which is more in line with the safety requirements of aerial refueling missions.
[0117] (3) The controller provided by the present invention is simple in form, easy to calculate and verify, and convenient for engineering implementation. BRIEF DESCRIPTION OF THE DRAWINGS
[0118] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0119] Figure 1 A control schematic diagram of each controller of a robust flight control method for a receiving aircraft based on an adaptive extended state observer provided in an embodiment of the present invention.
[0120] Figure 2 is the wake induced disturbance in the x, y, and z directions.
[0121] Figure 3 The aerodynamic force induced by the wake.
[0122] Figure 4 is the wake-induced aerodynamic torque.
[0123] Figure 5 The trajectory tracking diagram.
[0124] Figure 6 This is a schematic diagram of the three-axis tracking error under the control law designed based on multiple estimators in numerical simulation, where (6-1), (6-2), and (6-3) correspond to the tracking error diagrams of the x, y, and z axes, respectively.
[0125] Figure 7 Schematic diagram of tracking errors of V, γ, and χ in numerical simulation, where (7-1), (7-2), and (7-3) correspond to V, γ, and χ, respectively.
[0126] Figure 8 Schematic diagram of the tracking errors of α, β, and μ in numerical simulation, where (8-1), (8-2), and (8-3) correspond to α, β, and μ, respectively.
[0127] Fig. 9 Schematic diagram of the tracking errors of p, q, and r in numerical simulation, where (9-1), (9-2), and (9-3) correspond to p, q, and r respectively.
[0128] Fig.10 In numerical simulation, ω x ,ω y ,ω z The estimated value effect diagram of , where (10-1), (10-2), and (10-3) correspond to ω x ,ω y ,ω z . DETAILED DESCRIPTION
[0129] To make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be described in detail and completely in conjunction with the drawings in the implementation of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings can be arranged and designed using different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the present application for protection, but merely represents selected embodiments of the present invention.
[0130] In order to suppress the adverse effects of lumped interference and take into account the transient and steady-state response performance of the receiving aircraft, the embodiment of the present invention observes the mismatch and match interference of the receiving aircraft based on the designed adaptive extended state observer (AESO), and combines the nominal controller designed by the backstepping control strategy to complete the robust flight controller design of the complete inner and outer loops of the receiving aircraft, thereby achieving its accurate tracking of the desired refueling position. By introducing the estimation error term in the designed AESO, the peak phenomenon problem caused by the initial estimation error can be effectively solved, while reducing the large overshoot and final steady-state error caused by the estimation error in the initial stage, ensuring the safety of the docking and refueling holding stages during the autonomous aerial refueling process.
[0131] In a robust flight control method for a receiving aircraft based on an adaptive extended state observer provided by an embodiment of the present invention, firstly, a mathematical model of a fixed-wing aircraft of the receiving aircraft is established, namely, a three-degree-of-freedom kinematic model and an attitude dynamic model; an inner-outer loop feedback loop structure is used to design a backstepping controller, and a command filter is introduced to solve the derivative order problem of the backstepping method, namely, a nominal controller is designed by combining a command filter with a backstepping control method to ensure that the nominal closed-loop system is asymptotically stable; an adaptive extended state observer is used to estimate and compensate for interferences such as the induced aerodynamic wake effect, unmodeled dynamic influence, and external meteorological disturbances generated by the tanker received by the receiving aircraft, so as to obtain a robust flight control law considering external lumped interference; based on the control input solved by the command filter backstepping control loop, combined with the estimated compensation considering interferences such as the induced aerodynamic wake effect, unmodeled dynamic influence, and external meteorological disturbances, and the constructed three-degree-of-freedom kinematic and dynamic model of the fixed-wing aircraft of the receiving aircraft, the control input of the fixed-wing aircraft is obtained, and it is applied to the receiving aircraft. The method provided in the embodiment of the present invention can effectively solve the control problems caused by the induced aerodynamic wake effect generated by the refueling aircraft in a close formation, the unmodeled dynamic influence and the external meteorological interference, effectively solve the peak phenomenon problem of the interference estimator, and at the same time take into account the transient overshoot and steady-state error performance of the closed-loop system.
[0132] See also Figure 1 The specific implementation of a robust flight control method for a receiving aircraft based on an adaptive extended state observer provided by an embodiment of the present invention includes the following steps:
[0133] 1. Establish the three-degree-of-freedom kinematic model and attitude dynamics model of the receiver fixed-wing aircraft;
[0134] For overdriven quadrotor systems (such as Figure 1 As shown), a specific mathematical model is established, as shown in formulas (1) to (2);
[0135] 2. Adopting an adaptive extended state observer, the receiver aircraft is estimated and compensated for the induced aerodynamic wake effect, unmodeled dynamic influence and external meteorological disturbances generated by the tanker, and the robust flight control law considering the external aggregate disturbance is obtained.
[0136] The adaptive extended state observer is used to estimate the interference amount in each channel, as shown in formulas (7) to (11), and it is used as compensation input to participate in the calculation, as shown in formulas (18), (27), and (33);
[0137] 3. For the constructed nonlinear kinematic and dynamic systems, the backstepping controller is designed using the internal and external feedback loop structure, and the command filter is introduced to solve the derivative order problem of the backstepping method;
[0138] The backstepping control strategy is adopted to obtain the thrust command of the outer loop and the operating surface command of the inner loop step by step from the outer loop to the inner loop, as shown in formulas (12) to (35);
[0139] In order to solve the problem of derivation of the intermediate virtual control quantity involved in the backstepping method, the command filter is used to estimate the required derivative quantity as the input quantity of the next step, as shown in formula (6);
[0140] 4. Based on the control input solved by the command filter backstepping control loop, combined with the estimated compensation of interference such as induced aerodynamic wake effect, unmodeled dynamic influence and external meteorological disturbance, and the three-degree-of-freedom kinematic and dynamic model of the receiver fixed-wing aircraft constructed in step 1, the control input of the fixed-wing aircraft is obtained and applied to the receiver.
[0141] In an embodiment of the present invention, an open source aircraft is used as an experimental object, and a three-degree-of-freedom kinematic and dynamic model of the fixed-wing aircraft is established based on its model parameters. The object exhibits strong nonlinear characteristics of the actual aircraft and can verify the effectiveness and rationality of the designed control method.
[0142] Example:
[0143] The MATLAB / SIMULINK simulation tool is used to perform attitude robust control on a fixed-wing aircraft system experimental platform to verify the desired trajectory tracking control performance of the AESO-based control law in the presence of interference conditions such as the induced aerodynamic wake effect generated by the tanker, unmodeled dynamic influences, and external meteorological disturbances.
[0144] In the simulation experiment, a high-fidelity model of the experimental object is used for the fixed-wing aircraft. In order to reflect the robustness and stability of the robust flight control method of the receiving aircraft proposed in the embodiment of the present invention, the wake induced interference ω in the x, y, and z directions is P ,P∈{x,y,z} is designed as a unified composite sinusoidal signal: ω P = sin(0.2t+4)+0.5cos(0.4t-2)+1.5sin(0.2t+2), such as Figure 2 As shown. The receiving aircraft is subject to the induced aerodynamic forces ΔD, ΔY, ΔL and moments Δl, ΔM, ΔN, whose values are as follows in the docking and refueling holding stages: Figure 3 , Figure 4 given in the form of .
[0145] This embodiment simulates the aerial refueling scenario, and the simulation conditions of the given tanker are as follows: l (0) = 45m, y l (0) = -15m, z l (0) = -5010m, V l (0) = 200 m / s, γl (0) = x l (0) = 0rad. And the refueling machine is V l (0) = 200m / s initial speed for constant altitude and constant speed level flight; start turning at 60s, The aircraft will hover at a constant altitude and speed at 240s; it will re-enter the constant altitude and constant speed level flight state at 300s; the mission will end at 300s. f (0) = x l (0)-120=-75m,y f (0) = y l (0)+80=65m, z f (0) = z l (0)+40=-4970m, V f (0) = 200 + 10 = 210 m / s, γ f (0) = x f (0) = μ f (0) = β f (0) = 0 rad, α f (0) = 0.0484 rad, p = q = r = 0 rad / s. Considering the actual close-range scenario of aerial refueling and the physical conditions of the soft / hard pipe, the mission requires the receiving aircraft to maintain a relative position of [-15, 9, 0] in the coordinate system of the tanker. T m. Note that it is necessary to transform the tanker body coordinate system to the inertial coordinate system to obtain [l x ,l y ,l z ] T The desired trajectory of the receiving aircraft is thus solved. The receiving aircraft needs to automatically approach the refueling machine under the control of the controller, and then stabilize at the best relative position behind the refueling machine to ensure the successful completion of the refueling task.
[0146] In this embodiment, the corresponding parameter settings of the command filter, the controller and the observer are shown in Tables 1 to 3:
[0147] Table 1 Command filter parameters
[0148]
[0149] Table 2 Controller parameters
[0150]
[0151] Table 3 Observer parameters (AESO)
[0152]
[0153] In the scenario corresponding to this embodiment, a stateless observer, a low-gain linear extended state observer LESO (ω n =ω n (0))、LESO(ω n =ω ncmin )、LESO(ω n =ω ncmax )、TVESO(ω nc =ω ncmax ), trajectory tracking control under AESO. The three fixed-gain LESOs use the ω in their respective channels. ncmax ,ω ncmin ,ω n (0) is used as a parameter for interference observation comparison experiments, as shown in Table 4, while TVESO uses ω in Table 4 ncmax As the input bandwidth ω nc Perform simulation.
[0154] Table 4 Observer bandwidth parameters
[0155]
[0156] In actual situations, the observer initial value cannot be guaranteed to be completely consistent with the actual state initial value, that is, The initial value error will lead to the strength of the peak phenomenon. In order to simulate the real situation, the observer initial value error is set in each state channel, and its size is 5% of the state signal fluctuation amount in each channel. Set the observer initial value respectively:
[0157]
[0158]
[0159] According to the flight trajectory of the tanker as described above, the expected trajectory of the receiving aircraft is calculated, and the tracking is as follows: Figure 5 For Figure 5 Without using an observer (Without disturbance observer, Without Do), it is difficult for the controller to track the tanker to perform the refueling task under the influence of the aerodynamic induced effect.
[0160] The tracking error and estimation error effects of the robust flight control method for the receiving aircraft based on the adaptive extended state observer can be obtained, as well as the comparison of the tracking effect and estimation effect with no estimator (Without DO), the introduction of the linear extended state observer (LESO), and the time-varying extended state observer (TVESO), such as Figure 6 to Figure 9 As shown, Figure 6 to Figure 9 Respectively represent the tracking errors on x, y, z, V, γ, χ, μ, α, β, p, q, and r channels; and Fig.10 Indicates the wake induced disturbance ω P , the estimation error of P∈{x,y,z}.
[0161] The experimental results of this embodiment show that in the simulation environment, the tracking errors of the actual angles of the three axes are all within a small range, and the root mean square errors of the three-axis angles are also very small, indicating that the control law designed based on the AESO embodiment of the present invention is effective; under the simulation conditions of time-varying disturbances, the estimated values of the three-axis disturbances are basically consistent with the actual values; under the simulation conditions of constant disturbances, the three-axis disturbance estimation errors and root mean square errors are at a small level, indicating that the AESO of the embodiment of the present invention can accurately estimate various forms of disturbances; the errors and root mean square errors between the estimated values of the three-axis angular velocities and the actual angular velocities in the simulation are of small magnitude, indicating that the angular velocity estimation values can be accurately estimated from the actual values, that is, the AESO of the embodiment of the present invention can accurately estimate the angular velocity information.
[0162] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
[0163] The above are only some embodiments of the present invention. For those skilled in the art, several modifications and improvements can be made without departing from the creative concept of the present invention, which all belong to the protection scope of the present invention.
Claims
1. A robust flight control method for a receiving aircraft based on an adaptive extended state observer, characterized in that: The following steps are involved: Step 1, based on the position coordinates, speed, track inclination angle, track azimuth of the tanker in the inertial coordinate system and the optimal relative distance between the tanker and the tanker, the expected speed, expected track inclination angle and track azimuth of the tanker in the inertial coordinate system are obtained; the expected position derivative of the tanker is estimated according to the three-degree-of-freedom kinematic model and attitude dynamics model of the tanker and the command filter; Step 2, based on a preset adaptive extended state observer, obtain interference estimation values in each state channel, including position interference estimation value, receiving aircraft speed, track inclination angle and track azimuth; Step 3, according to the position of the tanker, and the desired speed, desired track inclination angle and track azimuth of the receiving aircraft, an outer loop controller is set by adopting a backstepping control strategy, and the outer loop controller is used to generate the desired output and serve as the input of the inner loop controller; The actual position, speed, track inclination angle and track azimuth of the receiving aircraft are respectively input into the adaptive extended state observer, and based on the output, the position channel interference estimation value, the speed channel interference estimation value, the track inclination angle channel interference estimation value and the track azimuth angle channel interference estimation value are obtained; and based on the obtained position interference estimation value, the speed of the receiving aircraft, the track inclination angle and the track azimuth angle, the dynamic equation of the position error is obtained; According to the position error dynamic equation, the expected speed, expected track inclination angle and expected track azimuth of the receiving aircraft are set; According to the state equation of the receiving aircraft, the error dynamic equations of the receiving aircraft speed, track inclination angle and track azimuth angle are obtained; According to the error dynamic equations of the receiving aircraft speed, track inclination angle and track azimuth angle, a virtual intermediate control input is defined to set the desired control input; According to the desired control input and the error dynamic equations of the receiving aircraft speed, track inclination angle and track azimuth angle, the desired output of the outer loop controller is obtained, including: the desired thrust command, the desired track roll angle and the desired angle of attack; Step 4, taking the expected output of the outer loop controller as the expected input of the inner loop controller, and obtaining the rudder control input of the receiving aircraft by solving the inner loop attitude stability and controller; Inputting the desired track roll angle and the desired angle of attack obtained by the outer loop controller into the command filter, and obtaining derivative estimates of the desired track roll angle and the desired angle of attack based on the output of the command filter; According to the state equation of the receiving aircraft, the error dynamic equations of the receiving aircraft's track roll angle, angle of attack and sideslip angle are defined; An intermediate virtual control variable is introduced to obtain the error dynamic equations of the receiving aircraft's track roll angle, angle of attack and sideslip angle, where the estimated value of the disturbance term is obtained based on an adaptive extended state observer. Constructing a desired control input according to the error dynamic equation to obtain a desired angular velocity signal; inputting the desired angular velocity signal into a second-order command filter to obtain a derivative estimate of the desired angular velocity; The error equation of the angular velocity signal of the receiving engine is obtained according to the state equation of the receiving engine, and the error dynamic equation of the angular velocity signal is obtained by introducing the intermediate virtual control quantity; According to the error dynamic equation of the angular velocity signal, the desired control input is set to obtain the rudder deflection control input of the receiving aircraft, including: aileron deflection angle, elevator deflection angle and rudder deflection angle.
2. The method according to claim 1, characterized in that The three-degree-of-freedom kinematic model and attitude dynamics model of the receiving aircraft are as follows: Among them, x f ,y f ,z f is the position of the receiving aircraft in the inertial coordinate system, V f ,γ f ,x f are the receiving aircraft's airspeed, track inclination angle and track azimuth, position and V f ,γ f ,x f The three-degree-of-freedom kinematic model of the receiving aircraft is constructed. D, Y, and L are the initial aerodynamic drag, side force, and lift of the receiving aircraft, respectively. T is the throttle thrust, m f is the mass of the receiving machine, μ f ,α f ,β f are the rolling angle, angle of attack and sideslip angle of the receiving aircraft respectively; p, q, r are the components of the angular velocity of the receiving aircraft relative to the inertial coordinate system in the x, y, z axis of the body coordinate system respectively; l, M, N represent the original rolling, pitching and yaw moments acting on the receiving aircraft respectively; I x ,I y ,I z ,I xz are the moment of inertia and product of inertia around the x, y, and z axes of the body, ω x ,ω y ,ω z is the component of the induced wake velocity in the inertial coordinate system in the x, y, and z axes of the body coordinate system. The parameter symbol "·" represents the first-order derivative; ω x ,ω y ,ω z , induced aerodynamic drag ΔD, side force ΔY, lift ΔL and induced rolling moment Δl, pitching moment ΔM, yaw moment ΔN are regarded as the interference caused by the tail vortex of the tanker on the receiving aircraft. The uncertainty and interference terms are: Among them, d V is the speed disturbance term, d γ is the track tilt angle interference term, d χ is the track azimuth interference term, d p , d q , d r They are the three components of the angular velocity disturbance term in the x, y, and z axes of the body coordinate system.
3. The method according to claim 2, characterized in that The instruction filter in step 1 is specifically: in, represents the input signal, l x ,l y ,l z They are the components of the optimal relative distance vector between the receiving aircraft and the refueling aircraft in the inertial coordinate system on the x, y, and z axes. By adjusting the filter damping ratio and bandwidth Get the target signal The command filter signal and its first-order derivative Second-order derivative 4. The method according to claim 2, characterized in that The adaptive extended state observer in step 2 is specifically: Establish the time-varying extended state observer TVESO: in, is the estimated value of the state quantity x1, x2, U x For system input, is the estimation error of x1, l1(t), l2(t) are the time-varying observation gains, and their expressions are: Among them, ξ is a constant damping ratio, ω n (t) is the closed-loop time-varying observer bandwidth, They are ω n The first and second derivatives of (t); Taking e1 as a parameter, the observation gains l1(t) and l2(t) are adjusted in real time with e1. The obtained observer is called the adaptive extended state observer AESO. The expected observer bandwidth of the AESO is: Among them, ω nc is the desired observer bandwidth, is the fixed gain of different interference items, interference item identifier ω ncmax ,ω ncmin For the preset ω nc The upper and lower limits of .
5. The method according to claim 4, characterized in that Use Butterworth filter to obtain the desired observer bandwidth ω nc : Design a third-order Butterworth filter as: in, They are the closed-loop time-varying observer bandwidth ω n (t) third-order derivative, second-order derivative and first-order derivative, ω nc (t) is ω nc The time-varying representation of B Adjust ω n (t) response speed, so that ω n (t) smoothly transitions to the desired bandwidth ω nc (t) up.
6. The method according to claim 5, characterized in that The poles of the third-order Butterworth filter are [-ω B ,(-0.5±j0.866)ω B ]。 7. The method according to claim 1, characterized in that In step 3, the expected output of the outer loop controller is: Among them, T c is the desired thrust command, α d is the desired angle of attack, μ d The track roll angle of the target, m f is the quality of the receiving machine, is the expected output of the outer loop controller corresponding to the speed, track inclination angle and track azimuth of the receiving aircraft, g represents the acceleration of gravity, D is the initial aerodynamic drag on the receiving aircraft, V f ,γ f ,x f are the airspeed, track inclination angle and track azimuth of the receiving aircraft respectively, T is the throttle thrust, μ f ,α f ,β f are the roll angle, angle of attack and sideslip angle of the receiving aircraft respectively. L0 is the lift at zero angle of attack, L α is the aerodynamic derivative of lift with respect to angle of attack, and lift L = L0 + L α α f .
8. The method according to claim 1, characterized in that In step 4, the rudder control input of the receiving aircraft is specifically: Among them, δ u is the rudder control input of the receiving aircraft, and δ u =[δ a ,δ e ,δ r ] T , δ a is the aileron deflection angle, δ e is the elevator deflection angle, δ r is the rudder deflection angle, M τ To manipulate the derivative matrix, the auxiliary quantity Δτ = [Δl, ΔM, ΔN] T , Δl, ΔM, ΔN represent the induced rolling moment, pitching moment and yaw moment respectively, Ι is the moment of inertia matrix of the aircraft, and τ=τ0+M τ δ u , auxiliary amount τ=[l,M,N] T , l, M, N represent the original rolling, pitching and yaw moments acting on the receiving aircraft, respectively, and τ0 is the rudder deflection δ u =[0,0,0] T The moment vector at ; Expected control input Among them, K Ω =diag{K p ,K q ,K r } is a positive definite diagonal constant gain matrix, K p ,K q ,K r They are the gains of the components p, q, and r of the angular velocity of the receiving aircraft relative to the inertial coordinate system in the body coordinate system x, y, and z axes, respectively. The constant parameter matrix C Ω =diag{c p ,c q ,c r },c p ,c q ,c r are constants corresponding to the components p, q, and r respectively. The angular velocity of the receiving aircraft relative to the inertial coordinate system is Ω = [p, q, r] T .
Citation Information
Cited By
Robust control method for aircraft attitude angle in predetermined time under unknown disturbance
CN119535990A
A predetermined time robust control method for aircraft attitude angle under unknown disturbance
CN119535990B
Miniature steering engine servo control method based on combination of variable gain expansion state observer and backstepping control
CN121918378A
Servo control method of micro rudder based on variable gain extended state observer and backstepping control
CN121918378B
Close formation anti-interference control method and system considering eddy current influence
CN122488802A
