CN106292294B - Shipborne UAV auto landing on deck control device based on model reference self-adapting control - Google Patents

Abstract

The invention discloses an automatic landing control device for a ship-borne unmanned aerial vehicle based on model reference adaptive control, and belongs to the technical field of aviation vehicle control. The device of the present invention includes: a landing command and a glide reference trajectory generating module, which is used to generate a three-dimensional glide reference trajectory signal and a speed command signal according to the relative position and absolute position information of the ship and the ship-borne UAV; the model reference adaptive flight The control module uses the model reference adaptive control algorithm to generate the flight control signal of the shipboard UAV, so that the actual flight trajectory and speed of the shipboard UAV track the landing command and the three-dimensional glide reference trajectory generated by the glide reference trajectory generation module signal and speed command. The invention is applicable to fixed-wing manned carrier-based aircraft and unmanned carrier-based aircraft. Compared with the prior art, the present invention does not need the calculation of the guidance control law, the control of landing is more accurate, the control system is simpler, and the control real-time performance is good.

Description

Automatic landing control device for shipborne UAV based on model reference adaptive control

technical field

The invention relates to an automatic landing control device for a ship-borne unmanned aerial vehicle based on model reference adaptive control, and belongs to the technical field of aerospace control.

Background technique

The "landing" in the present invention includes landing methods such as runway arrest landing, net collision recovery, etc., and has universality in the principle of the control method.

As an important weapon force of the aircraft carrier, the key technology of carrier-based aircraft is how to ensure safe and accurate landing in a very harsh environment. Due to the harsh landing environment, disturbances such as mothership movement and ship tail airflow will have a great impact on UAV landing, which greatly increases the difficulty of carrier-based aircraft landing and seriously affects the safety of landing. During the sailing process of the ship, due to the influence of waves, sea surges and wind, the hull will produce deck motions in the form of pitch, yaw, roll, up and down, etc., resulting in the landing point on the ship as The three-degree-of-freedom active point seriously affects the difficulty and safety of landing. In the changing environment at sea, when a carrier-based aircraft lands on a ship, the disturbance of the airflow at the tail of the ship is also an important factor affecting its landing performance. In the approach and landing section, as the flight speed decreases, the flight angle of attack generally exceeds the critical angle of attack, and is in the region of unstable speed, making it very difficult to maintain the flight trajectory. At the same time, the carrier-based UAV itself is a complex control object with characteristics such as nonlinearity, uncertainty, multivariability, and strong coupling. The interference of complex environmental factors, the change of flight altitude and state, and modeling errors together constitute the uncertain factors of the shipboard UAV system.

The current carrier-based automatic landing system (ACLS) usually consists of two parts: on-board equipment and on-board equipment. The part on the ship has tracking radar, stable platform, high-speed computer, display device, data link coder/transmitter, data link monitor, flight track recorder, etc. The on-board part has a data link receiver, a receiver decoder, an autopilot coupler, an automatic flight control system, an automatic throttle controller, and a radar booster. The automatic landing control method usually uses the trajectory control loop as the outer loop, and the attitude control loop and speed control as the inner loop. The trajectory control loop is based on the trajectory tracking error information, combined with the deck motion prediction and compensation information, and generates the attitude and The speed command signal is sent to the flight control system, and the flight control system requires tracking of these command signals to obtain the desired trajectory, attitude and speed. The design of the inner and outer loop control laws is based on the traditional single-loop design method, such as the PID control method. However, there is no public application report on the automatic landing system (ACLS) applicable to UAVs.

To sum up, it can be seen that the existing automatic landing control technology for carrier-based aircraft generally has defects such as complex system, poor real-time performance, and high hardware requirements.

Contents of the invention

The technical problem to be solved by the present invention is to overcome the deficiencies in the prior art, and provide a shipboard unmanned aerial vehicle automatic landing control device based on model reference adaptive control, which does not need the calculation of the guidance control law, and the landing is more accurate and the control system Simpler, better real-time control.

The present invention specifically adopts the following technical solutions to solve the above-mentioned technical problems:

An automatic landing control device for shipboard unmanned aerial vehicles based on model reference adaptive control, the device includes: a landing command and a glide reference trajectory generation module, which is used to base on the relative position and absolute Position information, generate three-dimensional sliding reference trajectory signal and speed command signal;

The model reference adaptive flight control module uses the model reference adaptive control algorithm to generate the flight control signal of the shipboard UAV, so that the actual flight trajectory and speed of the shipboard UAV follow the landing command and the glide reference trajectory generation module. The three-dimensional glide reference trajectory signal and speed command.

Preferably, the model reference adaptive flight control module is used to estimate the nominal control matrix of the reference model The adaptive update law for online update of K ₂ (t) is designed according to the following method:

make ω(t)＝[Δx ^T (t),Δr ^T (t)] ^T , Δr is the reference input signal, Δx is the state vector, then the output tracking error

e(t)=Δy(t) _-Δym (t),

In the formula, Δy and Δy _m are the output of the system and the output of the reference model, respectively;

Define a new error signal as

ε(t)=ξm(s)h(s)[e](t)+Ψ(t)ξ( _t ),

In the formula, h(s)=1/f(s), f(s) is a stable polynomial, Ψ(t) is the estimated value of Ψ ^* =K _p , K _p is the high-frequency gain matrix, ξ _m (s) is the interaction matrix of the reference model, ξ(t)=Θ ^T (t)ζ(t)-h(s)[Δu](t);

make

ζ(t)=h(s)[ω](t)

Then the new error signal transforms into

In the formula,

Therefore, the adaptive update law of the control matrix parameters is designed as:

In the formula, Γ=Γ ^T >0, S _p denotes a reversible constant matrix.

Preferably, the input signal of the model reference adaptive flight control module includes: four longitudinal state quantities of the carrier-based UAV—flight speed V, angle of attack α, pitch angle rate q, pitch angle θ; State quantity—sideslip angle β, roll angular rate p, yaw angular rate r, roll angle φ, yaw angle ψ; landing command and glide reference trajectory generation module output speed command V _c and glide reference trajectory signal X _EATDc (t), Y _EATDc (t), Z _EATDc (t);

The output signals of the model reference adaptive flight control module include: throttle opening Δδ _T , elevator deflection angle Δδ _e , aileron deflection angle δ _a , rudder deflection angle δ _r ;

The flight control laws in the model reference adaptive flight control module include longitudinal and lateral flight control laws, which are designed by the following methods:

The first step, based on the following longitudinal linear model

Determine the relative order l _i of the transfer function matrix, i=1,2, and calculate the high-frequency gain matrix

Guaranteed to be non-singular; where, A _lon , B _lon , C _lon are longitudinal linear system matrices described by variable symbols, and c _1,lon , c _2,lon are the first and second rows of C _lon respectively;

In the second step, according to the relative order l _i of the transfer function matrix, i=1,2, select the interaction matrix Among them, p _1i and p _2i are the expected poles of the longitudinal system, so the following reference model is designed

Δy _m,lon (t)=W _m,lon (s)[Δr _lon ](t)

In the formula, Δr _ion (t)=[0, ΔH _EATDc ] ^T ,

The third step is to calculate the longitudinal flight control law

in, K _2,lon (t) is the control matrix updated online;

The fourth step is based on the following lateral linear model

Determine the relative order l _i of the transfer function matrix, i=1,2, and calculate the high-frequency gain matrix

Guaranteed to be non-singular; where, A _lat , B _lat , C _lat are lateral linear system matrices, c _1,lat , c _2,lat are the first row and second row of C _lat respectively;

The fifth step is to select the interaction matrix according to the relative order l _i of the transfer function matrix, i=1,2 Among them, p _1i and p _2i are the expected poles of the lateral system, so the following reference model is designed

y _m,lat (t)=W _m,lat (s)[r _lat ](t)

In the formula, r _lat (t)=[0,Y _EATDc ] ^T ,

The sixth step is to calculate the lateral flight control law

in, K _2,lat (t) is the control matrix updated online.

Preferably, the input signals of the landing command and glide reference trajectory generation module include: the azimuth angle (ψ _S +λ _ac ) of the ship's runway or glideslope, where ψ _S is the azimuth angle of the ship, and λ _ac is the angled deck clip The output signals of the landing command and glide reference trajectory generation module include: speed command V _c and glide reference trajectory signals X _EATDc (t), Y _EATDc (t), Z _EATDc (t).

further. The landing command and glide reference trajectory generation module uses the following method to generate the velocity command V _c and glide reference trajectory signals X _EATDc (t), Y _EATDc (t), Z _EATDc (t):

After capturing the glideslope, calculate the landing time based on the known initial glide height -Z _EA0 , glide angle γ _c , and glide speed V _c

and glideslope length

Then calculate the three-dimensional glide reference trajectory in the ground coordinate system with the ideal landing point as the origin:

Compared with the prior art, the present invention has the following beneficial effects:

(1) According to the relative position and absolute position information of the ship and the shipboard UAV, the present invention calculates the shipboard command signal online, generates the glide reference trajectory of the shipboard UAV, and controls the shipboard UAV through the flight control system Track the reference trajectory; compared with the existing technology, it can improve the coordination between the ship-borne UAV and the ship.

(2) The present invention designs flight controller under the uncertain situation of model parameter and structure of carrier-borne unmanned aerial vehicle, guarantees theoretically the output signal of the linear system of model uncertainty asymptotically tracks the output signal of reference model, and then tracks parameter input The signal, that is, the height, track and speed of the carrier-based UAV can track the reference trajectory and speed, and finally realize the tracking of the glide trajectory, so that the landing mission can be accurately completed. Therefore, the present invention can adjust control parameters online, and has strong self-adaptability and robust performance. However, traditional automatic landing systems are designed using classical control methods and rely on accurate models of carrier-based aircraft. External disturbances lack adaptability.

(3) The present invention does not have the calculation of the guidance control law, and the flight control system directly reduces or eliminates the trajectory error by designing the adaptive control law, so that the design of the flight control system becomes simpler.

Description of drawings

Fig. 1 represents the functional block diagram of the shipborne unmanned aerial vehicle automatic landing control device based on model reference adaptive control of the present invention;

Figure 2 shows the effect diagram of the altitude trajectory tracking during the landing process of the shipborne UAV;

Fig. 3 shows the effect diagram of the speed control during the landing process of the shipborne UAV;

In the figure, the solid line represents the expected value curve, and the dotted line represents the actual value curve.

Detailed ways

The technical scheme of the present invention is described in detail below in conjunction with accompanying drawing:

The principle of the shipborne UAV automatic landing control device based on model reference adaptive control of the present invention is shown in Figure 1, which consists of two parts: a landing command and a glide reference trajectory generation module, and an adaptive flight control module.

Landing command and glide reference trajectory generation module

The input signal of this module includes: the azimuth angle of the ship's runway or glideslope (ψ _S +λ _ac ), where ψ _S is the ship's azimuth angle, and λ _ac is the included angle of the inclined deck.

The output signals of this module include three-dimensional sliding reference trajectory signals X _EATDc (t), Y _EATDc (t), Z _EATDc (t) and speed command signal V _c . Wherein, the glide reference trajectory signal and the speed command signal are output to the adaptive flight control module.

In the first step, the carrier-based aircraft captures the glideslope, and the initial glide height -Z _EA0 , glide angle γ _c , and glide speed V _c are known, and the landing time is calculated

and glideslope length

The second step is to calculate the three-dimensional glide reference trajectory in the ground coordinate system with the ideal landing point as the origin

Model Reference Adaptive Flight Control Module

The input signals of this module include: four longitudinal state quantities of the shipboard drone fed back by the sensor x=(V,α,β,p,q,r,φ,θ,ψ,X,Y,H) ^T —— Flight speed V, angle of attack α, pitch rate q, pitch angle θ; five lateral state quantities fed back by sensors—sideslip angle β, roll rate p, yaw rate r, roll angle φ, yaw rate Flight angle ψ; the landing command and the speed command V _c output by the glide reference trajectory generation module, and the glide reference trajectory signals X _EATDc (t), Y _EATDc (t), Z _EATDc (t).

The output signals of this module include: throttle opening Δδ _T , elevator deflection angle Δδ _e , aileron deflection angle δ _a , and rudder deflection angle δ _r . Send it to the executive agency to control the flight of the carrier aircraft.

The specific process is: first calculate the longitudinal flight control law (steps 1, 2 and 3), and then calculate the lateral flight control law (steps 4, 5 and 6).

The first step, based on the following longitudinal linear model

Determine the relative order l _i of the transfer function matrix, i=1,2, and calculate the high-frequency gain matrix

Guaranteed to be nonsingular. In the formula, A _lon , B _lon , and C _lon are longitudinal linear system matrices described by variable symbols, and c _1,lon , c _2,lon are the first and second rows of C _lon , respectively.

In the second step, according to the relative order l _i of the transfer function matrix, i=1,2, select the interaction matrix Among them, p _1i and p _2i are the expected poles of the longitudinal system, so the following reference model is designed

Δy _m,lon (t)=W _m,lon (s)[Δr _lon ](t) (7)

In the formula, Δr _lon (t)=[0,ΔH _EATDc ] ^T ,

The third step is to calculate the longitudinal flight control law

in, K _2,lon (t) is the control matrix, which is updated online according to the reference model adaptive control algorithm.

The fourth step is based on the following lateral linear model

Determine the relative order l _i of the transfer function matrix, i=1,2, and calculate the high-frequency gain matrix

Guaranteed to be nonsingular. In the formula, A _lat , B _lat , C _lat are lateral linear system matrices, and c _1,lat , c _2,lat are the first row and second row of C _lat respectively.

The fifth step is to select the interaction matrix according to the relative order l _i of the transfer function matrix, i=1,2 Among them, p _1i and p _2i are the expected poles of the lateral system, so the following reference model is designed

y _m,lat (t)=W _m,lat (s)[r _lat ](t) (6)

In the formula, r _lat (t)=[0,Y _EATDc ] ^T ,

The sixth step is to calculate the lateral flight control law

in, K _2,lat (t) is the control matrix, which is updated online according to the reference model adaptive control algorithm.

Model Reference Adaptive Control Algorithm

For the following linear system

In the formula, Δx is the state vector, Δu is the control vector, Δy is the output vector, and A, B, C are the system matrices.

Build a reference model as

In the formula, ξ _m (s) is the interaction matrix.

The purpose of control is to expect the system output Δy to track the output Δy _m of the reference model, so the control law structure is constructed as

where Δr is the reference input signal, K ₂ (t) is the nominal control matrix estimated value.

When the model parameters are fully known, design the control matrix in the nominal control law Satisfy the following equality conditions

Then it can ensure that the system output Δy completely tracks the output Δy _m of the reference model. However, when the model parameters are uncertain, the nominal control matrix cannot be obtained Therefore only estimates Instead of K ₂ (t), the estimated value needs to be updated online using the following adaptive algorithm.

make ω(t)＝[Δx ^T (t),Δr ^T (t)] ^T , then the output tracking error

e(t)=Δy(t) _-Δym (t) (18)

Define a new error signal as

ε(t)=ξm(s)h(s)[e](t)+Ψ(t)ξ( _t ) (19)

In the formula, h(s)=1/f(s), f(s) is a stable polynomial, and Ψ(t) is the estimated value of Ψ ^* =K _p .

make

ζ(t)=h(s)[ω](t), ξ(t)=Θ ^T (t)ζ(t)-h(s)[Δu](t) (11)

Then the new error signal transforms into

In the formula,

Therefore, the adaptive update law of the control matrix parameters is designed as

In the formula, Γ=Γ ^T >0,

According to the relevant theoretical proof of the multivariable reference model adaptive control algorithm principle, it can be seen that the algorithm can guarantee the boundedness of each variable of the linear system, and the output can asymptotically track the output of the reference model.

In order to verify the effect of the automatic landing control device for ship-borne unmanned aerial vehicle proposed by the present invention, taking the longitudinal dynamics and kinematics model of a certain unmanned aerial vehicle as an example, the deck motion compensation is added to the reference trajectory in the last 10 seconds, and the main simulation parameters The settings are as shown in Table 1:

Table 1

Through the numerical simulation verification under the MATLAB software platform, the results show that the invented shipboard UAV automatic landing control device can enable the shipboard UAV to track the glide reference trajectory with high precision, thus successfully completing the landing task.

Claims (4)

1. A ship-borne UAV automatic landing control device based on model reference adaptive control, characterized in that the device includes: a landing command and a glide reference trajectory generation module for The relative position and absolute position information of the aircraft are used to generate a three-dimensional sliding reference trajectory signal and a speed command signal; The model reference adaptive flight control module uses the model reference adaptive control algorithm to generate the flight control signal of the shipboard UAV, so that the actual flight trajectory and speed of the shipboard UAV follow the landing command and the glide reference trajectory generation module. The three-dimensional glide reference trajectory signal and speed command; the model reference adaptive flight control module is used to estimate the nominal control matrix of the reference model The adaptive update law for online update of K ₂ (t) is designed according to the following method: make ω(t)＝[Δx ^T (t),Δr ^T (t)] ^T , Δr is the reference input signal, Δx is the state vector, then the output tracking error e(t)=Δy(t) _-Δym (t), In the formula, Δy and Δy _m are the output of the system and the output of the reference model, respectively; Define a new error signal as ε(t)=ξm(s)h(s)e(t)+Ψ(t)ξ( _t ), In the formula, h(s)=1/f(s), f(s) is a stable polynomial, Ψ(t) is the estimated value of Ψ ^* =K _p , K _p is the high-frequency gain matrix, ξ _m (s) is the interaction matrix of the reference model, ξ(t)=Θ ^T (t)ζ(t)-h(s)Δu(t), and Δu(t) is the control vector; make ζ(t)=h(s)ω(t), Then the new error signal transforms into In the formula, Therefore, the adaptive update law of the control matrix parameters is designed as: In the formula, Γ=Γ ^T >0, S _p is a reversible constant matrix. 2. ship-borne unmanned aerial vehicle automatic landing control device as claimed in claim 1, is characterized in that, the input signal of model reference adaptive flight control module comprises: four longitudinal state quantities of ship-borne unmanned aerial vehicle---flight speed V, angle of attack α, pitch angle rate q, pitch angle θ; five lateral state quantities—sideslip angle β, roll angle rate p, yaw angle rate r, roll angle φ, yaw angle ψ; The speed command V _c and the glide reference trajectory signal X _EATDc (t), Y _EATDc (t), Z _EATDc (t) output by the ship command and the glide reference trajectory generation module; The output signals of the model reference adaptive flight control module include: throttle opening Δδ _T , elevator deflection angle Δδ _e , aileron deflection angle δ _a , rudder deflection angle δ _r ; The flight control laws in the model reference adaptive flight control module include longitudinal and lateral flight control laws, which are designed by the following methods: The first step, based on the following longitudinal linear model Determine the relative order l _i of the transfer function matrix, i=1,2, and calculate the high-frequency gain matrix Guaranteed to be non-singular; where, ΔV, Δα, Δq, Δθ, ΔH are the offsets of flight speed V, angle of attack α, pitch rate q, pitch angle θ, and flight height H respectively; A _lon , B _lon , C _lon is the longitudinal linear system matrix described by variable symbols, c _1,lon and c _2,lon are the first and second rows of C _lon respectively; In the second step, according to the relative order l _i of the transfer function matrix, i=1,2, select the interaction matrix Among them, p _1i and p _2i are the expected poles of the longitudinal system, so the following reference model is designed Δy _m,lon (t)=W _m,lon (s)Δr _lon (t) In the formula, Δr _lon (t)=[0,ΔH _EATDc ] ^T , ΔH _EATDc is the altitude offset of the glide reference trajectory; The third step is to calculate the longitudinal flight control law in, K _2,lon (t) is the control matrix updated online in the longitudinal flight control law; The fourth step is based on the following lateral linear model Determine the relative order l _i of the transfer function matrix, i=1,2, and calculate the high-frequency gain matrix Guaranteed to be non-singular; where, A _lat , B _lat , C _lat are lateral linear system matrices, c _1,lat , c _2,lat are the first and second rows of C _lat respectively, and y is the lateral Deviate; The fifth step is to select the interaction matrix according to the relative order l _i of the transfer function matrix, i=1,2 Among them, p _1i and p _2i are the expected poles of the lateral system, so the following reference model is designed y _m,lat (t)=W _m,lat (s)r _lat (t) In the formula, r _lat (t)=[0,Y _EATDc ] ^T , The sixth step is to calculate the lateral flight control law in, K _2,lat (t) is the control matrix updated online in the lateral flight control rate. 3. The shipborne unmanned aerial vehicle automatic landing control device as claimed in claim 1, is characterized in that, the input signal of landing command and glide reference trajectory generation module comprises: the azimuth angle (ψ _S + λ _ac ), where ψ _S is the azimuth angle of the ship, and λ _ac is the included angle of the inclined deck; the output signals of the landing command and the glide reference trajectory generation module include: speed command V _c and glide reference trajectory signal X _EATDc (t) , Y _EATDc (t), Z _EATDc (t). 4. ship-borne unmanned aerial vehicle automatic landing control device as claimed in claim 3, it is characterized in that, landing command and glide reference trajectory generation module use the following method to generate speed command V _c and glide reference trajectory signal X _EATDc (t) , Y _EATDc (t), Z _EATDc (t): After capturing the glideslope, calculate the landing time based on the known initial glide height -Z _EA0 , glide angle γ _c , and glide speed V _c and glideslope length Then calculate the three-dimensional glide reference trajectory in the ground coordinate system with the ideal landing point as the origin: