CN116203840A - Adaptive gain scheduling control method for reusable carrier - Google Patents

Abstract

The invention discloses an adaptive gain scheduling control method for a reusable carrier, based on the real-time estimation and compensation of uncertainty and disturbance by ESO, to realize the accurate RLV under the action of unmodeled dynamics, unknown external disturbance and channel coupling Attitude control; based on the estimator's classification compensation mechanism for disturbances, the adaptive scheduling of control gains under cross-domain maneuvers is realized; adaptive compensation is performed based on the nominal controller, and the scheduling range is designed based on prior information to ensure that the proposed controller stable performance. The invention improves the attitude control performance and anti-interference ability of the RLV cross-domain recovery, and has good adaptability and robust performance.

Description

Adaptive gain scheduling control method for reusable vehicle

Technical Field

The present invention relates to the field of aerospace vehicle control technology, and in particular to an estimation and compensation-based adaptive gain scheduling control method for a reusable vehicle.

Background Art

A reusable launch vehicle (RLV) is a type of aircraft that can perform a launch mission again after completing a scheduled launch mission, after maintenance and refueling. The vertical take-off and landing scheme inherits the configuration of traditional launch vehicles, and has advantages such as technical span and low R&D cost. However, the airspace and speed domain span of the RLV vertical return flight is large, the dynamic pressure changes dramatically, the flight environment is complex and changeable, the internal and external uncertainties and disturbances such as aerodynamic parameter deviation, structure and wind interference are strong, and each channel shows serious nonlinear coupling characteristics. Especially when the internal parameters of the system change or serious external disturbances occur, traditional control theory is difficult to meet the high-performance control requirements of modern launch vehicles under special maneuvers.

Precise attitude control during vertical landing is the key to achieving safe recovery of the vehicle. At present, most projects use gain scheduling methods to design the attitude of the aircraft, such as using classical control methods such as PID for control gain scheduling design. However, simple linear control methods cannot cope with strong nonlinear coupling, unknown disturbances and uncertainties in the system, and it is difficult to obtain satisfactory control effects. With the development of modern technology, advanced control theories such as robust control, sliding mode control, adaptive backstepping control, and control methods based on disturbance observers have been developed in RLV attitude control.

Robust control quantitatively analyzes uncertainty and designs controllers to suppress the effects of uncertainty. Based on the H _∞ robust control method, uncertainty problems caused by uncertain aerodynamic parameters and liquid fuel sloshing can be handled, thereby realizing the robust attitude control design of RLV. Sliding mode control designs the switching hyperplane of the system according to the desired dynamic characteristics of the system, thereby effectively dealing with parameter uncertainty and external disturbances. Since the traditional sliding mode control method based on the linear sliding surface is difficult to avoid the chattering problem of the control quantity, terminal sliding mode control has gradually developed, achieving rapid convergence of the control system and solving the much-discussed chattering problem. Non-singular terminal sliding mode control further solves the singularity problem of traditional terminal sliding mode and promotes the application of sliding mode control technology in aircraft control. In addition, adaptive backstepping control decomposes the nonlinear control system into several subsystems that do not exceed the system order, and then designs virtual controllers that make each subsystem stable, and uses Lyapunov function to gradually back-step, and finally obtains a controller that makes the entire system stable. Due to the multi-layer recursion of this method, the designed controller is extremely complex.

In contrast, the control concept based on disturbance estimation and compensation is more intuitive and has significant advantages in dealing with system disturbances and uncertainties, such as unknown input observer, disturbance observer, generalized proportional integral observer, uncertainty and disturbance estimator, extended state observer (ESO), etc. Among them, ESO design requires the least dynamic system information, and its structure is simple and easy to design. It can estimate the system state while estimating the disturbance, and has received widespread attention in the control field.

However, when a reusable vehicle returns to the atmosphere, the cross-domain maneuvers of the RLV in a large airspace make the vehicle's own characteristics and the external flight environment change with the airspace and speed domain, resulting in complex and variable total disturbance types in the ESO estimate. Different disturbance types have different effects on the system, and inappropriate compensation mechanisms will be counterproductive, affecting control performance and stability margin. At present, there is still a lack of research on relevant control methods specifically for cross-domain maneuvering characteristics.

Summary of the invention

The purpose of the present invention is to provide a reusable vehicle adaptive gain scheduling attitude control method based on a strategy combining an observer and an estimator to address the unmodeled dynamics, uncertainties and disturbances such as the wide airspace across which the RLV reentry flight crosses, the large changes in air density, flight speed and dynamic characteristics, the serious coupling effect between channels, the fuel consumption and large shaking, and the changes in the structure of the aircraft. The method is applied to the attitude control of the reusable vehicle re-entering the atmosphere to solve the control gain scheduling problem of the RLV cross-domain maneuvering process in a large airspace, thereby realizing the safe recovery of the RLV.

The technical solution adopted to achieve the purpose of the present invention is:

A method for adaptive gain scheduling control of a reusable vehicle comprises the steps of:

S1. According to the trajectory change requirements of the RLV reentry flight process, the guidance law is designed for the RLV flight attitude to generate the corresponding attitude tracking instructions; the guidance instructions for the three channels of pitch, roll and yaw are:

为俯仰角指令，β_r为侧滑角指令，γ_r为滚转角指令；设定侧滑角指令β_r＝0，用于实现偏航角控制Ψ_r＝Ψ_v，Ψ_r为偏航角，ψ_v为弹道偏角；θ is the trajectory inclination angle, _αr is the angle of attack guidance law,

is the pitch angle command, β _r is the sideslip angle command, and γ _r is the roll angle command; the sideslip angle command β _r = 0 is set to achieve yaw angle control Ψ _r = Ψ _v , Ψ _r is the yaw angle, ψ _v is the trajectory deviation angle;

S2. Design a six-degree-of-freedom dynamic model for the RLV reentry flight process, taking into account the unmodeled dynamics, uncertainties and disturbances including air density, flight speed, changes in dynamic characteristics, coupling between channels, fuel consumption, large sway, and changes in aircraft structure during cross-domain maneuvers, and establish an RLV attitude control model under disturbance and uncertainty;

The six-degree-of-freedom dynamic model is:

The dynamic equations for the RLV center of mass motion are:

The dynamic equation for the RLV to rotate around the center of mass is:

Kinematic equations for the motion of the RLV center of mass:

Kinematic equations for the rotation of the RLV around its center of mass

ψ,γ分别为俯仰角、偏航角和滚转角，θ,ψ_V,γ_V分别为弹道倾角、弹道偏角和倾侧角，V为位移速度，x,y,z为位置坐标，ω_x,ω_y,ω_z为角速度，J_x,J_y,J_z为转动惯量，M_x,M_y,M_z为分别为外力矩矢量在弹体坐标系各轴上的分量，X,Y,Z分别为阻力、升力和侧力；Where m is the mass of the RLV, g is the acceleration of gravity, α and β are the angle of attack and sideslip respectively.

ψ,γ are the pitch angle, yaw angle and roll angle respectively, θ, ψ _V ,γ _V are the trajectory inclination angle, trajectory deviation angle and roll angle respectively, V is the displacement velocity, x, y, z are the position coordinates, ω _x ,ω _y ,ω _z are the angular velocities, J _x ,J _y ,J _z are the moments of inertia, M _x ,M _y ,M _z are the components of the external torque vector on each axis of the projectile coordinate system, X, Y, Z are the drag, lift and side force respectively;

为动压，ρ为RLV所处飞行高度的空气密度，S为RLV的特征面积，L_b,L_c分别为RLV的侧向和纵向参考长度，c_x,c_y,c_z分别为阻力系数、升力系数、侧力系数，m_x,m_y,m_z分别代表滚动力矩系数、偏航力矩系数、俯仰力矩系数；in,

is the dynamic pressure, ρ is the air density at the flight altitude of the RLV, S is the characteristic area of the RLV, L _b , L _c are the lateral and longitudinal reference lengths of the RLV respectively, c _x , c _y , c _z are the drag coefficient, lift coefficient, and side force coefficient respectively, m _x , my _y , m _z represent the rolling moment coefficient, yaw moment coefficient, and pitch moment coefficient respectively;

ψ,γ二次求导，建立如下姿态控制模型：Considering the RLV characteristics and various uncertainties and disturbances of the return flight,

The second derivative of ψ and γ is used to establish the following attitude control model:

b_f,b_p,b_r代表俯仰通道、偏航通道、滚转通道的控制增益，

分别为俯仰通道、偏航通道和滚转通道的力矩系数分量，

为静导数，δ_z,δ_y,δ_x为俯仰通道、偏航通道和滚转通道待设计的数学舵；in,

_bf , _bp , _br represent the control gains of the pitch channel, yaw channel, and roll channel.

are the moment coefficient components of the pitch channel, yaw channel and roll channel respectively,

is the static derivative, δ _z ,δ _y ,δ _x are the mathematical rudders to be designed for the pitch channel, yaw channel and roll channel;

The state information in the pitch channel, yaw channel and roll channel except the rudder effect is defined as disturbance f ₁ , f ₂ , f ₃ , and f ₁ = f ₀₁ + d ₁ , f ₂ = f ₀₂ + d ₂ , f ₃ = f ₀₃ + d ₃ , f ₀₁ , f ₀₂ , f ₀₃ are modeled dynamics, d ₁ , d ₂ , d ₃ include the remaining unmodeled dynamics, uncertainties and unknown external disturbances in the pitch channel, yaw channel and roll channel;

S3. Use ESO to estimate the integrated total disturbance of the three channels of pitch, roll and yaw in real time and perform feedback compensation;

The angular velocity information ω _z , ω _y , ω _x of the three channels of pitch, yaw and roll are used to design ESO to estimate the total disturbance in real time. The ESO designed for the pitch channel, yaw channel and roll channel are

分别为俯仰通道、偏航通道和滚转通道角速度估计值，即

z_ψ1≈ω_y,z_γ1≈ω_x，

分别为俯仰通道、偏航通道和滚转通道的总扰动估计值，即

z_ψ2≈f₂+(b_p-b_p0)δ_y，z_γ2≈f₃+(b_r-b_r0)δ_x，b_f0,b_p0,b_r0为b_f,b_p,b_r的估计标称值；in,

are the estimated angular velocities of the pitch channel, yaw channel and roll channel, namely

z _ψ1 ≈ω _y ,z _γ1 ≈ω _x ,

are the total disturbance estimates of the pitch channel, yaw channel and roll channel, namely

z _ψ2 ≈f ₂ +(b _p -b _p0 )δ _y , z _γ2 ≈f ₃ +(b _r -b _r0 )δ _x , b _f0 ,b _p0 ,b _r0 are the estimated nominal values of b _f ,b _p ,b _r ;

The ESO poles are configured at -ω _oz , -ω _oy , -ω _ox , where ω _oz , ω _oy , ω _ox are the bandwidths of the pitch, yaw and roll channels respectively. The six observer gains l _z1 , l _z2 , l _y1 , l _y2 , l _x1 , l _x2 satisfy

S4. Considering the dynamic characteristics of RLV cross-domain maneuvers, the system gain and uncertainty caused by aerodynamic parameters are constantly changing, and the symbolic estimator is used to classify and identify the disturbance and uncertainty of RLV cross-domain maneuvers, and the respective compensation is performed based on the nominal controller to achieve precise attitude control of RLV reentry flight;

The symbolic estimator is used to analyze, estimate and compensate for disturbances and uncertainties. The loss function established is:

f_ψ,f_r为与舵效无关的扰动与不确定性；采用符号投影梯度策略求解所建立损失函数，得到：where δ _bf ,δ _bp ,δ _br are the estimated uncertainties of the pitch, yaw and roll channel system gains b _f ,b _p , _br respectively.

f _ψ , f _r are disturbances and uncertainties that are independent of the steering effect. The loss function established is solved using the symbolic projection gradient strategy, and we get:

f_ψ,f_γ分别为俯仰、偏航和滚转通道中与舵效无关扰动和不确定性的估计值，α₁,α₂,α₃＞0分别为俯仰、偏航和滚转通道的待设计更新常数；Where, δ _bf ,δ _bp ,δ _br are the estimated values of the uncertainty of the rudder effect in the pitch, yaw and roll channels, respectively.

f _ψ ,f _γ are the estimated values of disturbances and uncertainties unrelated to the rudder effect in the pitch, yaw and roll channels, respectively; α ₁ ,α ₂ ,α ₃ >0 are the update constants to be designed for the pitch, yaw and roll channels, respectively;

Based on the estimation results and the attitude angle and angular rate information of the three channels, the mathematical rudder control law of the pitch, yaw and roll channels is designed as follows:

Among them, k _pf ,k _pp ,k _pr are the proportional feedback gains of pitch, yaw and roll channels respectively, k _df ,k _dp ,k _dr are the differential feedback gains of the three channels;

The equivalent grid rudders for the three channels of pitch, yaw and roll are obtained from the mathematical rudders δ _x ,δ _y ,δ _z :

ψ＝ψ_r＝ψ_V(β＝0),γ＝_r＝0，进行RLV跨域机动的姿态控制。The three-channel equivalent grid rudder acts on the six-degree-of-freedom model of the RLV to achieve the precise tracking guidance law of the pitch angle, sideslip angle and roll angle, that is,

ψ＝ψ _r ＝ψ _V (β＝0), γ＝ _r ＝0, and attitude control of RLV cross-domain maneuvers is performed.

The present invention designs an adaptive gain scheduling control strategy for RLV, which estimates and separates disturbances such as strong coupling between channels, parameter and model uncertainty in cross-domain maneuvers in real time, and can effectively improve the attitude control accuracy of cross-domain maneuvers.

The present invention performs adaptive gain scheduling design based on the nominal controller, solves the control gain selection problem in the traditional ADRC controller, has a simple structure, is easy to design, ensures the stability of attitude control under cross-domain maneuvers, and improves the stability margin of the control system, and has stronger robustness and adaptability.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic diagram of an adaptive gain scheduling control method for a reusable vehicle according to the present invention.

FIG. 2 is a cross-domain maneuvering altitude profile of the RLV of the present invention.

FIG. 3 is a speed variation curve of the RLV cross-domain maneuver of the present invention.

FIG. 4 is a three-channel attitude angle tracking curve of the present invention.

FIG. 5 is a three-channel attitude angular velocity variation curve of the present invention.

FIG. 6 is a control torque curve of three mathematical rudders of the present invention.

DETAILED DESCRIPTION

The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

The present invention is a reusable vehicle adaptive gain scheduling control method, which can realize precise attitude control of RLV under unmodeled dynamics, unknown external disturbances and channel coupling based on the real-time estimation and compensation of ESO for uncertainty and disturbance. Based on the classification compensation mechanism of the estimator for disturbances, adaptive scheduling of control gain under cross-domain maneuvers can be realized. In order to ensure the stable performance of the proposed controller, adaptive compensation is performed based on the nominal controller, and the scheduling range is designed according to prior information. The present invention improves the attitude control performance and anti-interference ability of RLV cross-domain recovery, and has good adaptability and robust performance.

As shown in FIG1 , the adaptive gain scheduling control method for a reusable vehicle according to an embodiment of the present invention comprises the following steps:

Step 1: According to the trajectory change requirements of the RLV reentry flight process, design a suitable guidance law for the RLV flight attitude to generate the corresponding attitude tracking instructions;

Step (I) During the RLV reentry flight, its flight altitude and Mach number gradually decrease. Therefore, the longitudinal altitude profile is designed to be h _r ,

Among them, t ₀ , t ₁ , t ₂ , h ₀₁ , h ₀₂ , a ₁ are constants to be designed, t ₀ is the initial time value, t ₁ , t ₂ are function segmentation times, h ₀₁ is the initial altitude of the reentry flight, h ₀₂ is the terminal altitude of the reentry flight, a ₁ is the piecewise function coefficient,

For the pitch channel, according to the current true altitude h of the RLV reentry flight process, the angle of attack guidance law _αr is designed as:

i＝1,2，e_h＝h_r-h为高度跟踪误差，h_i,k_pα,k_dα为待设计常数，h_i为高度误差，k_pα为比例系数，k_dα为微分系数，α₀为攻角初值。The saturation function is defined as

i＝1,2, e _h = _hr -h is the height tracking error, _hi , k _pα , k _dα are constants to be designed, _hi is the height error, k _pα is the proportional coefficient, k _dα is the differential coefficient, and _α0 is the initial value of the angle of attack.

和α的时间常数，根据几何关系

将攻角制导律α_r转换为近似等效的俯仰角指令

Considering that the inertial time constant of the trajectory inclination angle θ is much larger than

and the time constant of α, according to the geometric relationship

Convert the angle of attack guidance law α _r into an approximately equivalent pitch angle command

For the yaw channel, the yaw command is set to ψ _r = ψ _v , that is, the yaw angle tracks the trajectory deviation angle. When high-precision command tracking is performed, β control can basically be achieved, so the sideslip angle command is set to β _r = 0 to achieve yaw angle control ψ _r = ψ _v .

For the roll channel, the roll angle tracking is set to γ _r =0.

Therefore, the guidance instructions for the three channels are set as follows:

Step 2: Design a six-degree-of-freedom dynamic model of the RLV reentry flight process, taking into account the changes in air density, flight speed, dynamic characteristics, strong coupling between channels, fuel consumption, large sway, and vehicle structure changes, and other unmodeled dynamics, uncertainties, and disturbances, and establish an RLV attitude control model under disturbances and uncertainties;

Among them, the established RLV six-degree-of-freedom model is:

The dynamic equations for the RLV center of mass motion are:

The dynamic equation for the RLV to rotate around the center of mass is:

Kinematic equations for the motion of the RLV center of mass:

Kinematic equations for the RLV rotation around the center of mass:

ψ，γ分别为俯仰角、偏航角和滚转角，θ,ψ_V,γ_V分别为弹道倾角、弹道偏角和倾侧角，V为位移速度，x,y,z为位置坐标，ω_x,ω_y,ω_z为角速度，J_x,J_y,J_z为转动惯量，M_x,M_y,M_z为分别为外力矩矢量在弹体坐标系各轴上的分量，X,Y,Z分别为阻力、升力和侧力。采用公开Winged-Cone模型的气动系数公式，有：Where m is the mass of the RLV, g is the acceleration of gravity, α and β are the angle of attack and sideslip respectively.

ψ, γ are pitch angle, yaw angle and roll angle respectively, θ, ψ _V , γ _V are trajectory inclination angle, trajectory deviation angle and roll angle respectively, V is displacement velocity, x, y, z are position coordinates, ω _x , ω _y , ω _z are angular velocities, J _x , J _y , J _z are moments of inertia, M _x , _My , M _z are components of external torque vector on each axis of the missile body coordinate system respectively, X, Y, Z are drag, lift and side force respectively. The aerodynamic coefficient formula of the public Winged-Cone model is adopted, which is:

为动压(ρ为RLV所处飞行高度的空气密度)，S为RLV的特征面积，L_b,L_c分别为RLV的侧向和纵向参考长度。c_x,c_y,c_z分别为阻力系数、升力系数、侧力系数，m_x,m_y,m_z分别代表滚动力矩系数、偏航力矩系数、俯仰力矩系数。此6个气动系数，均可由公开Winged-Cone模型数据进行实时计算。in,

is the dynamic pressure (ρ is the air density at the flight altitude of the RLV), S is the characteristic area of the RLV, L _b and L _c are the lateral and longitudinal reference lengths of the RLV respectively. _{c x} , c _y , c _z are the drag coefficient, lift coefficient, and side force coefficient respectively, and m _x , my _y , m _z represent the rolling moment coefficient, yaw moment coefficient, and pitch moment coefficient respectively. These six aerodynamic coefficients can all be calculated in real time using the public Winged-Cone model data.

进行二次求导，建立如下姿态控制模型：Considering the inherent characteristics of RLV and various uncertainties and disturbances of return flight,

Perform secondary derivation and establish the following attitude control model:

b_f,b_p,b_r代表俯仰通道、偏航通道、滚转通道的控制增益，

分别为俯仰通道、偏航通道和滚转通道的主要力矩系数分量，

为静导数，和攻角、马赫数相关，δ_z,δ_y,δ_x为俯仰通道、偏航通道和滚转通道待设计的数学舵。此外，将俯仰通道、偏航通道和滚转通道中除舵效外的状态信息定义为扰动f₁,f₂,f₃，且f₁=f₀₁+d₁,f₂=f₀₂+d₂,f₃=f₀₃+d₃，f₀₁,f₀₂,f₀₃为已建模动态，d₁,d₂,d₃包含了俯仰通道、偏航通道和滚转通道中剩余未建模动态、不确定性和未知外部扰动：in,

_bf , _bp , _br represent the control gains of the pitch channel, yaw channel, and roll channel.

are the main moment coefficient components of the pitch channel, yaw channel and roll channel respectively,

is the static derivative, which is related to the angle of attack and the Mach number. δ _z ,δ _y ,δ _x are the mathematical rudders to be designed for the pitch channel, yaw channel and roll channel. In addition, the state information in the pitch channel, yaw channel and roll channel except the rudder effect is defined as disturbance f ₁ ,f ₂ ,f ₃ , and f ₁ =f ₀₁ +d ₁ ,f ₂ =f ₀₂ +d ₂ ,f ₃ =f ₀₃ +d ₃ , f ₀₁ ,f ₀₂ ,f ₀₃ are the modeled dynamics, and d ₁ ,d ₂ ,d ₃ include the remaining unmodeled dynamics, uncertainties and unknown external disturbances in the pitch channel, yaw channel and roll channel:

It can be seen that the modelable disturbances f ₀₁ , f ₀₂ , f ₀₃ of the three channels of pitch, yaw and roll include inter-channel coupling, aerodynamic parameter uncertainty, etc. The uncertain disturbances f ₁ , f ₂ , f ₃ and system gains b _f , b _p , b _r are the key issues to be solved in the three-channel attitude control of RLV.

Step 3: To effectively handle the inter-channel coupling, unmodeled dynamics and unknown external disturbances in RLV attitude control, ESO is used to estimate the integrated total disturbance of pitch, roll and yaw channels in real time and perform feedback compensation;

In order to reduce the coupling effects between channels (aerodynamic coupling, inertial coupling, control coupling), parameter and model uncertainties, and other unknown external disturbances in f ₁ , f ₂ , f ₃ , the angular velocity information ω _z , ω _y , ω _x of the pitch, yaw and roll channels are used to design ESOs to estimate the total disturbance in real time. Based on (7), the ESOs designed for the pitch channel, yaw channel and roll channel are:

分别为俯仰通道、偏航通道和滚转通道角速度估计值，即

z_ψ1≈ω_y,z_γ1≈ω_x，

分别为俯仰通道、偏航通道和滚转通道的总扰动估计值，即

z_ψ2≈f₂+(b_p-b_p0)δ_y，z_γ2≈f₃+(b_r-b_r0)δ_x。b_f0,b_p0,b_r0为b_f,b_p,b_r的估计标称值。in,

are the estimated angular velocities of the pitch channel, yaw channel and roll channel, namely

z _ψ1 ≈ω _y ,z _γ1 ≈ω _x ,

are the total disturbance estimates of the pitch channel, yaw channel and roll channel, namely

z _ψ2 ≈f ₂ +(b _p -b _p0 )δ _y , z _γ2 ≈f ₃ +(b _r -b _r0 )δ _x . _{b f0} ,b _p0 ,b _r0 are the estimated nominal values of b _f ,b _p ,b _r .

To ensure the convergence of the observer, the ESO poles are configured at -ω _oz , -ω _oy , -ω _ox , where ω _oz , ω _oy , ω _ox are the bandwidths of the pitch, yaw, and roll channels, respectively. At this time, the six observer gains l _z1 , l _z2 , l _y1 , l _y2 , l _x1 , l _x2 satisfy

Considering the continuous changes in system gain caused by the coupling effect between channels and uncertain aerodynamic parameters under RLV cross-domain maneuvers, it is difficult to obtain satisfactory attitude control effects with the traditional single control gain method. Although ADRC technology has strong robustness and anti-disturbance capabilities, the control gains of the three channels of pitch, yaw and roll _bf , _bp , _br are very critical parameters in ADRC, which not only affect the compensation accuracy of disturbances, but also directly affect the stability margin of the control system. Therefore, an adaptive gain scheduling strategy is adopted to compensate disturbances and gains in real time respectively.

Step 4: Considering the continuous changes in system gain and uncertainty caused by dynamic characteristics and aerodynamic parameters during RLV cross-domain maneuvers, a symbolic estimator is used to classify and identify the disturbances and uncertainties of RLV cross-domain maneuvers, and the respective compensations are performed based on the nominal controller to achieve precise attitude control of RLV reentry flight.

The symbolic estimator is used to analyze, estimate and compensate for disturbances and uncertainties, and the loss function is established as

f_ψ,f_r为与舵效无关的扰动与不确定性。为实现损失函数的最小化，采用符号投影梯度策略求解(11)-(13)，得到：where δ _bf ,δ _bp ,δ _br are the estimated uncertainties of the pitch, yaw and roll channel system gains b _f ,b _p , _br respectively.

f _ψ , f _r are disturbances and uncertainties that are independent of the steering effect. In order to minimize the loss function, the symbolic projection gradient strategy is used to solve (11)-(13), and we get:

f_ψ,f_γ分别为俯仰、偏航和滚转通道中与舵效无关扰动和不确定性的估计值，α₁,α₂,α₃＞0分别为俯仰、偏航和滚转通道的待设计更新常数。Where, δ _bf ,δ _bp ,δ _br are the estimated values of the uncertainty of the rudder effect in the pitch, yaw and roll channels, respectively.

f _ψ ,f _γ are the estimated values of disturbances and uncertainties unrelated to the rudder effect in the pitch, yaw and roll channels, respectively. α ₁ ,α ₂ ,α ₃ > 0 are the update constants to be designed for the pitch, yaw and roll channels, respectively.

Based on the estimation results (14)-(16) and the attitude angle and angular rate information of the three channels, the mathematical rudder control law of the pitch, yaw and roll channels is designed as follows:

Among them, k _pf ,k _pp ,k _pr are the proportional feedback gains of pitch, yaw and roll channels respectively, k _df ,k _dp ,k _dr are the differential feedback gains of the three channels. This design transforms the three channels into an easily controllable series integral form by estimating, compensating for disturbances and controlling gains respectively.

During the aerodynamic deceleration phase of the RLV, only the grid rudder works. At this time, the equivalent grid rudders of the pitch, yaw and roll channels obtained by the mathematical rudders δ _x ,δ _y ,δ _z are:

ψ＝ψ_r＝ψ_V(β＝0),γ＝γ_r＝0。Applying the three-channel equivalent grid rudder (20) to the six-degree-of-freedom model of the RLV can achieve accurate tracking guidance law of pitch angle, sideslip angle and roll angle, that is,

ψ=ψ _r =ψ _V (β=0), γ=γ _r =0.

Therefore, the RLV attitude control based on the adaptive gain scheduling control laws (17)-(19) of the pitch, yaw and roll channels can realize the adaptive scheduling of the system gain during the cross-domain maneuvering process, thereby improving the attitude control performance of the cross-domain maneuvering.

The present invention comprehensively considers the unmodeled dynamics, uncertainties and disturbances such as the wide cross-domain airspace and speed range, large changes in air density, flight speed and dynamic characteristics, serious coupling between channels, fuel consumption and large shaking, and changes in aircraft structure during RLV re-entry flight, and proposes an adaptive gain scheduling control strategy. Based on the active disturbance rejection controller, the signed projection gradient strategy is used to adaptively schedule the system gain. The strategy does not rely on RLV feature point extraction, has a simple structure, is easy to design, and can effectively handle the uncertainty and disturbance of the RLV cross-domain maneuvering process, thereby improving the attitude control accuracy.

In order to verify the effectiveness of the adaptive gain scheduling control strategy proposed in the present invention, the RLV six-degree-of-freedom model was built using MATLAB software, and the adaptive gain scheduling control strategy was designed based on the attitude control model, thereby verifying the effectiveness of the present invention in dealing with disturbances and uncertainties, as well as its strong robustness and adaptability in improving attitude control accuracy. During simulation, the data of relevant parameters are as follows:

t ₀ = 0, t ₁ = 60, t ₂ = 120, h ₀₁ = 25000, h ₀₂ = 1000, altitude error h ₁ = 500, h ₂ = 50, proportional coefficient k _pα = 0.2, differential coefficient k _dα = 1, initial value of angle of attack α ₀ = 5. The total mass of the RLV aircraft is 136080 kg, the longitudinal reference length is 24.384 m, the lateral reference length is 18.288 m, the reference area is 334.73 m ² , the initial moment of inertia in the x direction is 1355818 kg·m ² , and the initial moments of inertia in the y and z directions are 13558180 kg·m ² .

f_ψ,f_γ的初值均设定为0，α₁,α₂,α₃＝0.001。k_pf＝42,k_df＝20，k_pp＝0.5,k_dp＝20，k_pr＝62,k_dr＝38。控制姿态角初始值为[5 0 1]°，位置初始值为[0 25000 0]m，速度初始值为4Ma，角速度初始值为0。ω _of =10, b _f0 =2.45, ω _op =10, b _p0 =0.5, ω _or =10, b _r0 =11.18. δ _bf , δ _bp , δ _br and

The initial values of f _ψ , f _γ are set to 0, α ₁ , α ₂ , α ₃ ＝ 0.001. _{k pf} ＝ 42, k _df ＝ 20, k _pp ＝ 0.5, k _dp ＝ 20, k _pr ＝ 62, k _dr = 38. The initial value of the control attitude angle is [5 0 1]°, the initial value of the position is [0 25000 0]m, the initial value of the speed is 4Ma, and the initial value of the angular velocity is 0.

The simulation results are shown in Figure 2-6 and are divided into two parts:

Part I: Guidance Laws

According to the longitudinal profile reference value of the RLV re-entry flight, the angle of attack is designed, so that the pitch angle tracking control is performed according to the geometric relationship. The longitudinal profile reference value and the actual value obtained are shown in Figure 2. It can be seen that the control strategy proposed in the present invention can track the longitudinal profile value well and has good tracking performance. At the same time, the RLV flight speed is shown in Figure 3, which realizes the attenuation of the flight speed during the RLV re-entry flight, further illustrating the effectiveness of the guidance law adopted by the present invention, and laying a good foundation for achieving precise attitude control.

Part 2: Posture Control

The present invention proposes to use an adaptive gain scheduling strategy to perform RLV cross-domain maneuvers. In the simulation results, Figure 4 shows the control results of the attitude angle, Figure 5 shows the control results of the three angular velocities, and Figure 6 is the control mathematical rudder designed based on the present invention. From the results, it can be seen that the present invention can effectively handle uncertainties such as coupling between channels, uncertainty in aerodynamic parameters, and changes in air density; it achieves stable and accurate tracking of the guidance law, proving the effectiveness, strong robustness, and adaptability of the proposed algorithm.

The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention; therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is limited by the attached claims rather than the above description, and it is intended to include all changes within the meaning and scope of the equivalent elements of the claims in the present invention.

In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims (1)

1. A method for adaptive gain scheduling control of a reusable vehicle, comprising the steps of: S1. According to the trajectory change requirements of the RLV reentry flight process, the guidance law is designed for the RLV flight attitude to generate the corresponding attitude tracking instructions; the guidance instructions for the three channels of pitch, roll and yaw are:

θ is the trajectory inclination angle, α _r is the angle of attack guidance law, θ _r is the pitch angle command, β _r is the sideslip angle command, and γ _r is the roll angle command; the sideslip angle command β _r = 0 is set to achieve yaw angle control ψ _r = ψ _v , ψ _r is the yaw angle, and ψ _v is the trajectory deviation angle; S2. Design a six-degree-of-freedom dynamic model for the RLV reentry flight process, taking into account the unmodeled dynamics, uncertainties and disturbances including air density, flight speed, changes in dynamic characteristics, coupling between channels, fuel consumption, large sway, and changes in aircraft structure during cross-domain maneuvers, and establish an RLV attitude control model under disturbance and uncertainty; The six-degree-of-freedom dynamic model is: The dynamic equations for the RLV center of mass motion are:

The dynamic equation for the RLV to rotate around the center of mass is:

Kinematic equations for the motion of the RLV center of mass:

Kinematic equations for the rotation of the RLV around its center of mass

Where m is the mass of the RLV, g is the acceleration of gravity, α and β are the angle of attack and sideslip, respectively.

are the pitch angle, yaw angle and roll angle respectively, θ, ψ _V , γ _V are the trajectory inclination angle, trajectory deviation angle and roll angle respectively, V is the displacement velocity, x, y, z are the position coordinates, ω _x , ω _y , ω _z are the angular velocities, J _x , J _y , J _z are the moments of inertia, M _x , _My , M _z are the components of the external torque vector on each axis of the projectile coordinate system, X, Y, Z are the drag, lift and side force respectively;

in,

is the dynamic pressure, ρ is the air density at the flight altitude of the RLV, S is the characteristic area of the RLV, L _b and L _c are the lateral and longitudinal reference lengths of the RLV respectively, c _x , c _y , c _z are the drag coefficient, lift coefficient and side force coefficient respectively, m _x , my _y , m _z represent the rolling moment coefficient, yaw moment coefficient and pitching moment coefficient respectively; Considering the RLV characteristics and various uncertainties and disturbances of the return flight,

The second derivative of Ψ and γ is used to establish the following attitude control model:

in,

_bf , _bp , _br represent the control gains of the pitch channel, yaw channel, and roll channel.

are the moment coefficient components of the pitch channel, yaw channel and roll channel respectively,

is the static derivative, δ _z , δ _y , δ _x are the mathematical rudders to be designed for the pitch channel, yaw channel and roll channel; The state information in the pitch channel, yaw channel and roll channel except the rudder effect is defined as disturbances f ₁ , f ₂ , f ₃ , and f ₁ = f ₀₁ + d ₁ , f ₂ = f ₀₂ + d ₂ , f ₃ = f ₀₃ + d ₃ , f ₀₁ , f ₀₂ , f ₀₃ are modeled dynamics, d ₁ , d ₂ , d ₃ include the remaining unmodeled dynamics, uncertainties and unknown external disturbances in the pitch channel, yaw channel and roll channel;

S3. Use ESO to estimate the integrated total disturbance of the three channels of pitch, roll and yaw in real time and perform feedback compensation; The angular velocity information ω _z , ω _y , ω _x of the three channels of pitch, yaw and roll are used to design ESO to estimate the total disturbance in real time. The ESO designed for the pitch channel, yaw channel and roll channel are

in,

are the estimated angular velocities of the pitch channel, yaw channel and roll channel, namely

are the total disturbance estimates of the pitch channel, yaw channel and roll channel, namely

z _ψ2 ≈f ₂ +(b _p -b _p0 )δ _y , z _γ2 ≈f ₃ +(b _r -b _r0 )δ _x , b _f0 , b _p0 , b _r0 are the estimated nominal values of b _f , b _p , b _r ; The ESO poles are configured at -ω _oz , -ω _oy , -ω _ox , where ω _oz , ω _oy , ω _ox are the bandwidths of the pitch, yaw and roll channels respectively. The six observer gains l _z1 , l _z2 , l _y1 , l _y2 , l _x1 , l _x2 satisfy l _z1 ＝2ω _oz ,

l _y1 ＝2ω _oy ，

l _x1 = 2ω _ox ,

S4. Considering the dynamic characteristics of RLV cross-domain maneuvers, the system gain and uncertainty caused by aerodynamic parameters are constantly changing, and the symbolic estimator is used to classify and identify the disturbance and uncertainty of RLV cross-domain maneuvers, and the respective compensation is performed based on the nominal controller to achieve precise attitude control of RLV reentry flight; The symbolic estimator is used to analyze, estimate and compensate for disturbances and uncertainties, and the loss function established is:

where δ _bf , δ _bp , δ _br are the estimated uncertainties of the pitch, yaw and roll channel system gains b _f , b _p , _{br ,} respectively.

is the disturbance and uncertainty that is unrelated to the rudder effect; the symbolic projection gradient strategy is used to solve the established loss function and we get:

Where, δ _bf , δ _bp , δ _br are the estimated values of the uncertainty of the rudder effect in the pitch, yaw and roll channels, respectively.

are the estimated values of disturbances and uncertainties unrelated to the rudder effect in the pitch, yaw and roll channels, respectively; α ₁ , α ₂ , α ₃ >o are the update constants to be designed for the pitch, yaw and roll channels, respectively; Based on the estimation results and the attitude angle and angular rate information of the three channels, the mathematical rudder control law of the pitch, yaw and roll channels is designed as follows:

Wherein, k _pf , k _pp , k _pr are the proportional feedback gains of the pitch, yaw and roll channels respectively, k _df , k _dp , k _dr are the differential feedback gains of the three channels; The equivalent grid rudders for the three channels of pitch, yaw and roll are obtained from the mathematical rudders δ _x , δ _y , δ _z as follows:

The three-channel equivalent grid rudder acts on the six-degree-of-freedom model of the RLV to achieve the precise tracking guidance law of the pitch angle, sideslip angle and roll angle, that is,

Perform attitude control for RLV cross-domain maneuvers.

Images