CN109597303B - Full-mode flight control method of combined type rotor craft - Google Patents

Abstract

The invention discloses a full-mode flight control method for a compound rotorcraft. First, a nonlinear system model linearization theory is used to quickly and effectively solve a dynamic inverse model controller in a multi-mode reference flight state; then, a neural network is used to compensate the multi-mode dynamic The error of the inverse model; finally, the adaptive terminal sliding mode is used to ensure the stability, rapidity and robustness of the full-mode flight control system. In the all-mode flight control method of the compound rotorcraft of the present invention, a dynamic inverse adaptive terminal sliding mode control method is designed in the attitude loop and the speed loop respectively, so that the compound rotorcraft can complete the system command within the set limited time. Tracking, good convergence, and good robustness of the flight system, can realize full-envelope full-mode flight of the aircraft, no need to switch controllers during flight, reduce the complexity of the flight system and improve the mode switching of the aircraft security.

Description

An all-mode flight control method for a composite rotorcraft

technical field

The invention belongs to the technical field of flight mechanics and flight simulation, and in particular relates to a full-mode flight control technology for a compound rotor aircraft.

Background technique

The composite rotorcraft has both a thrust device, a fixed wing and a rotor, which can not only realize the vertical take-off and landing and hovering flight of the helicopter flight mode, but also have the high-speed and long-range cruise flight capability of the fixed-wing aircraft. Institutions and researchers. At present, the research on the composite rotorcraft carrier-based UAV at home and abroad is still in its infancy. Due to technical confidentiality and other reasons, at this stage, there are few issues related to the multi-mode flight control of high-speed helicopters at home and abroad.

The composite rotorcraft flight control system is the key and guarantee for the safe and stable flight of the aircraft and the completion of various flight tasks. Designing the flight control system of a compound rotorcraft is not an easy task. It is necessary to consider the helicopter flight control problem, the fixed-wing flight control problem, the transition flight control problem, and the control switching when switching between different flight modes. question. When the flight mode of the aircraft changes, the change of aerodynamic characteristics is quite complicated. It is difficult to establish a complete and accurate mathematical model for the verification of the control system design. The designed flight control system must have good adaptability and robustness.

The design method of the flight control law for a specific flight state based on the small disturbance linearization model is a common practice at present. For the multi-mode flight control problem of the compound rotorcraft, this design method faces challenges, and it is difficult to adopt the traditional linear control method, use The single-parameter controller achieves the desired control goal. During the flight, there are mode conversions, the uncertainty of the system caused by the change of the inertia moment of the body, the fluctuation of the rotor speed, and the external interference, which make the aircraft system present the characteristics of nonlinearity, strong coupling and uncertainty. The change of flight mode determines the change of flight dynamics, and the change of dynamics means that the flight control system needs to be adjusted to ensure the smooth flight of the aircraft. The adjustment of the control system is not only the adjustment of the control parameters, but also the adjustment of the control structure may be required. This brings great difficulty to the control system design. The change of flight dynamics caused by the change of flight mode is abstracted as the uncertainty of the dynamic characteristics of the controlled object, and a set of adaptive and robust flight control system suitable for all flight modes is designed using nonlinear control theory. , is an ideal method to solve the flight control of compound rotorcraft.

Based on the typical flight state, the present invention designs a nonlinear intelligent flight control system that can compensate for the inverse error of the model online, so as to ensure that the compound rotorcraft has good control performance in different flight modes, so that the system has real-time, stable, robustness. Explore the nonlinear dynamic inverse model solution method to solve the rapidity, stability and accuracy of the inverse model controller algorithm.

SUMMARY OF THE INVENTION

In view of the above-mentioned shortcomings of the prior art, the present invention proposes a full-mode flight control technology for a composite rotorcraft, which enables the system to have self-adaptation and robustness, and avoids the problem of controller switching during multi-mode flight. All-mode stable flight within the full envelope of the rotorcraft.

For solving the above problems, the technical scheme adopted in the present invention is:

An all-mode flight control method for a composite rotorcraft, the specific steps are as follows:

Step 1. Select a compound _rotorcraft dynamics model as the research object, and obtain the desired speed tracking command of the flight system according to the compound rotorcraft flight mission. _c and the desired yaw speed w _c , and use the desired speed tracking command as the input of the controller;

期望滚转角ψ_c；姿态回路作为内回路，可使复合式旋翼飞行器各个通道解耦，增强系统的稳定性；Step 2: Design the speed loop and the attitude loop control structure respectively. The speed loop is used as the outer loop to provide the desired attitude control command for the attitude loop. The desired attitude control command includes: the desired pitch angle θ _c , the desired yaw angle

Desired roll angle ψ _c ; the attitude loop is used as an inner loop, which can decouple each channel of the compound rotorcraft and enhance the stability of the system;

同时，设计结合神经网络的自适应终端滑模控制器模块，得到速度增量U_a1+U_t1，并通过控制分配计算得到期望姿态控制指令信号

设计姿态回路动态逆自适应终端滑模控制方法，首先通过动态逆控制器得到复合式旋翼操纵变量δ₂＝[A_1s B_1s θ_wr θ_wl θ₁ θ₂]，并结合速度回路得到的操纵变量δ₁，作为复合式旋翼飞行器期望舵面操纵信号

同时，设计结合神经网络的自适应终端滑模控制器模块，得到姿态角增量U_a2+U_t2，并通过控制分配计算舵面操纵信号增量Δδ，将U＝Δδ+δ_d作为复合式旋翼飞行器的实际舵面操纵信号；Step 3: Design the dynamic inverse adaptive terminal sliding mode control method of the speed loop, and obtain the manipulated variables of the compound rotorcraft through the dynamic inverse controller

At the same time, an adaptive terminal sliding mode controller module combined with a neural network is designed to obtain the velocity increment U _a1 +U _t1 , and the desired attitude control command signal is obtained through the control distribution calculation

The dynamic inverse adaptive terminal sliding mode control method of the attitude loop is designed. First, the composite rotor manipulation variable δ ₂ = [A _1s B _1s θ _wr θ _wl θ ₁ θ ₂ ] is obtained through the dynamic inverse controller, and combined with the control obtained by the speed loop The variable δ ₁ , as the desired rudder control signal of the compound rotorcraft

At the same time, an adaptive terminal sliding mode controller module combined with a neural network is designed to obtain the attitude angle increment U _a2 +U _t2 , and the rudder control signal increment Δδ is calculated through the control distribution, and U=Δδ+δ _d is taken as the compound formula The actual rudder control signal of the rotorcraft;

为旋翼总距，A_1s为横向周期变距、B_1s为纵向周期变距、θ_wr为右襟副翼偏角、θ_wl为左襟副翼偏角、θ₁为推力矢量与机体坐标系下XOY面夹角，θ₂为推力矢量在水平面投影与X轴的夹角。Among them, T is the thrust vector of the ducted fan,

is the rotor collective pitch, A _1s is the lateral cyclic pitch, B _1s is the longitudinal cyclic pitch, θ _wr is the right flaperon declination angle, θ _wl is the left flaperon declination angle, and θ ₁ is the thrust vector and the body coordinate system The angle between the lower XOY plane, θ ₂ is the angle between the projection of the thrust vector on the horizontal plane and the X axis.

偏航角θ、滚转角ψ、俯仰角速度p、偏航角速度q和滚转角速度r，并重复步骤一至四。Step 4: Detect the flight status of the compound rotorcraft in real time, the flight status mainly includes: forward flight speed u, lateral speed v, vertical speed w, pitch angle

Yaw angle θ, roll angle ψ, pitch angular velocity p, yaw angular velocity q and roll angular velocity r, and repeat steps 1 to 4.

Further, both the attitude loop control and the speed loop control described in step 2 include an instruction filtering module, a dynamic inverse neural network module and an adaptive terminal sliding mode module, which specifically includes the following steps:

或期望姿态控制指令信号

通过指令滤波模块，限制输入量的幅值和频率，得到期望速度跟踪指令的一阶导数和期望姿态控制指令的二阶导数；Step 3.1. Desired speed tracking command signal

or desired attitude control command signal

Through the command filtering module, the amplitude and frequency of the input quantity are limited, and the first-order derivative of the desired speed tracking command and the second-order derivative of the desired attitude control command are obtained;

速度回路得到的一阶导数为

指令滤波模块采用二阶滤波器，表达式如下：Among them, the second derivative obtained by the attitude loop is

The first derivative of the velocity loop is

The instruction filtering module adopts a second-order filter, and the expression is as follows:

In the formula, the natural frequency _wn = 3 is selected, and the damping ratio ξ = 0.7.

Step 3.2, design a dynamic inverse controller according to the first derivative of the desired speed tracking command, the second derivative of the desired attitude control command and the flight state of the compound rotorcraft, and at the same time, design a neural network control method to compensate the system model error;

Step 3.3. In order to enable the flight system of the composite rotorcraft to complete the system command tracking within a limited time and have good robustness, design the adaptive terminal sliding mode compensation value, including the speed loop adaptive terminal sliding mode compensation value U _t1 and attitude loop adaptive terminal sliding mode compensation value U _t2 to improve the command tracking speed and stability of the flight system;

Further, the design steps of the dynamic inverse controller and the neural network control method described in step 3.2 include:

Step 3.2.1. Establish a nonlinear dynamic model of the compound rotor, select the equilibrium point, and use the quasi-Newton iteration method to perform dynamic trimming and small disturbance linearization analysis to obtain an approximate linearization model of the compound rotor;

Since the compound rotorcraft has three flight modes, including the helicopter flight mode, the transition flight mode and the fixed-wing flight mode, and the values of the manipulated variables in each flight mode are quite different, the present invention selects the front and rear flight modes of the helicopter respectively. Dynamic trimming and small-disturbance linearization analysis were carried out at a flying speed of 20m/s, forward flight speed of 70m/s in transition flight mode, and 100m/s in fixed-wing flight mode;

It is assumed that the nonlinear dynamic model of the composite rotor is expressed as:

表示x的一阶导数，δ表示复合式旋翼飞行器舵面操纵信号。In the formula,

Represents the first derivative of x, and δ represents the control signal of the rudder surface of the compound rotorcraft.

The approximate linearized model obtained by trimming can be expressed as:

表示复合式旋翼飞行器近似舵面操纵信号。In the formula,

Indicates the approximate rudder control signal of the compound rotorcraft.

并以此来设计动态逆控制器；Step 3.2.2, in order to meet the dynamic inverse control conditions, according to the approximate linearization model of the compound rotorcraft, inversely solve the virtual control command

And use this to design a dynamic inverse controller;

姿态回路动态逆控制器近似舵面操纵信号

That is, the speed loop dynamic inverse controller approximates the control signal of the rudder surface

Attitude Loop Dynamic Inverse Controller Approximate Rudder Surface Control Signal

Step 3.2.3. Since there is a deviation between the approximate linearized model and the actual model of the compound rotorcraft, the dynamic inverse controller designed by the approximate linearized model may lead to instability of the flight system, so the present invention designs a single-layer sigma-pi Neural network to compensate for model errors.

The compensation value of the speed loop neural network is:

U _a1 =W ₁ ^T β ₁ (4)

The compensation value of the attitude loop neural network is:

U _a2 =W ₂ ^T β ₂ (5)

Among them, β ₁ and β ₂ are basis function vectors, W ₁ and W ₂ are weight coefficient vectors, and the values of β ₁ and β ₂ are as follows:

C₃＝[uv w]，

kron()表示矩阵叉乘。In the formula, C ₁ =C ₁ '=[0.1 VV ² ],

C ₃ =[uv w],

kron() represents matrix cross product.

Further, the design steps of the adaptive terminal sliding mode module described in step 3.3 include:

Step 3.3.1. Set the compound rotorcraft to track the control command within a limited time T _d , and design the terminal sliding surface of the flight system;

Define the speed command tracking error as:

E ₁ (t)=x _r1 -x _c1 =[e ₁₁ , e ₂₁ , e ₃₁ ] ^T (7)

Similarly, the attitude command error can be expressed as:

E ₂ (t)=x _r2 -x _c2 =[e ₁₂ , e ₂₂ , e ₃₂ ] ^T (8)

为飞行系统姿态角跟踪指令，

θ_r和ψ_r分别表示俯仰角跟踪指令、偏航角跟踪指令和滚转速度跟踪指令，

为期望姿态控制指令，e₁₁、e₂₁和e₃₁分别表示前飞速度、升降速度和偏航速度指令跟踪误差，e₁₂、e₂₂和e₃₂分别表示俯仰角、偏航角和滚转角指令跟踪误差；In the formula, x _r1 =[ur , v _r , w _r ] ^T is the flight system speed tracking command, _{ur , v r and wr represent} the forward flight speed tracking command, the ascending and descending speed tracking command and the yaw speed tracking command respectively , x _c1 =[u _c , v _c , w _c ] ^T is the desired speed tracking command,

is the flight system attitude angle tracking command,

θ _r and ψ _r represent pitch angle tracking command, yaw angle tracking command and roll speed tracking command, respectively,

is the desired attitude control command, e ₁₁ , e ₂₁ and e ₃₁ represent the forward flight speed, lift speed and yaw speed command tracking error respectively, e ₁₂ , e ₂₂ and e ₃₂ represent the pitch angle, yaw angle and roll angle commands respectively tracking error;

The terminal sliding surface of the flight system is designed as:

S(x)=CE _j (t)-CP _j (t),j=1,2 (9)

The present invention selects C as a third-order unit matrix, and P _j (t)=[p _1j (t), p _2j (t), p _3j (t)] ^T represents a time-varying compensation function.

In the formula, i=1, 2, 3, j=1, 2, e _ij (0) represents the initial instruction tracking error when t=0.

设计速度回路自适应终端滑膜补偿值U_t1和姿态回路自适应终端滑膜补偿值U_t2，补偿系统期望控制指令x_r1和x_r2；Step 3.3.2, by constructing the Lyapunov function

Design the speed loop adaptive terminal synovial compensation value U _t1 and the attitude loop adaptive terminal synovial compensation value U _t2 , and the compensation system expects control commands x _r1 and x _r2 ;

关于时间的导数为：Among them, the Lyapunov function

The derivative with respect to time is:

u_v为复合式旋翼飞行器虚拟控制指令；In the formula, j=1, 2, K is a constant, Δ _j =E _j (t)-P _j (t), A _j and B _j represent the state matrix and the approximate linearization model of the compound rotorcraft respectively. control matrix,

u _v is the virtual control command of the compound rotorcraft;

The design speed loop adaptive terminal synovial compensation value is:

In the formula, A ₁ and B ₁ respectively represent the state matrix and control matrix corresponding to the approximate linearized model velocity state quantity of the compound rotorcraft.

Taking u _v =U _t1 , and bringing it into equation (11), the derivative of the Lyapunov function in the velocity loop with respect to time can be obtained as:

In the same way, the design attitude loop adaptive terminal synovial compensation value is:

In the formula, A ₂ and B ₂ respectively represent the state matrix and control matrix corresponding to the attitude angle of the approximate linearized model of the compound rotorcraft.

Taking u _v =U _t2 , and bringing it into equation (11), the derivative of the Lyapunov function in the attitude loop with respect to time can be obtained as:

负定，故可以确定运动状态可在有限时间内沿着滑模面运动到平衡点，确保了飞行系统的鲁棒性。Through analysis, it can be seen that the Lyapunov function V(x) is positive definite, and its derivative with respect to time

Negative definite, so it can be determined that the motion state can move to the equilibrium point along the sliding surface within a limited time, which ensures the robustness of the flight system.

The beneficial effects of the present invention are:

The present invention designs speed loop control and attitude loop control respectively, the attitude loop can decouple each channel of the compound rotorcraft, and enhance the stability of the flight system; the speed loop takes the speed amount as the design target, which improves the command tracking ability of the flight system and the stability of the flight system. Anti-interference ability.

In the present invention, a dynamic inverse adaptive terminal sliding mode control method is designed in the attitude loop and the speed loop, respectively, so that the composite rotorcraft can complete the system command tracking within a set limited time, and the convergence is good, and the flight system has a relatively high performance. good robustness.

In addition, the all-mode control method of the composite rotorcraft proposed by the present invention can realize the all-envelope and all-mode flight of the aircraft, and does not need to switch controllers during flight, which reduces the complexity of the flight system and improves the safety of aircraft mode switching.

Description of drawings

1 is a block diagram of an all-mode flight control structure for a composite rotorcraft of the present invention;

Fig. 2 is the instruction filtering structural block diagram of the present invention;

Fig. 3 is the dynamic inverse neural network control structure block diagram of the present invention;

4 is a graph showing the relationship between the speed variable and time of the composite rotorcraft under simulated full-mode flight in an embodiment of the present invention;

In Fig. 4, (a) is a graph showing the relationship between the forward flight speed and time of the composite rotorcraft under simulated full-mode flight in the example of the present invention; (b) is a diagram of the composite rotor under simulated full-mode flight in the example of the present invention. The relationship diagram between the yaw speed of the aircraft and the time; (c) is the relationship diagram between the lifting speed and the time of the composite rotorcraft under the simulated full-mode flight in the example of the present invention;

5 is a diagram showing the relationship between the attitude angle and attitude angular velocity of the composite rotorcraft under simulated full-mode flight and time in an embodiment of the present invention;

In Fig. 5, (a) is a graph showing the relationship between the pitch angle and time of the composite rotorcraft under simulated full-mode flight in the example of the present invention; (b) is the composite rotorcraft under simulated full-mode flight in the example of the present invention The relationship diagram between the yaw angle and time; (c) is the relationship diagram between the roll angle and time of the composite rotorcraft under simulated full-mode flight in the example of the present invention; (d) is in the example of the present invention, under the simulated full-mode flight The relationship diagram of the compound rotorcraft pitch angular velocity and time; (e) is the relationship diagram of the compound rotorcraft yaw angular velocity and time under the simulation full-mode flight in the example of the present invention; (f) is the example of the present invention In , the relationship between the roll angular velocity and time of the compound rotorcraft under full-mode flight is simulated;

Identification in the figure: deg-degree (angle unit), t-time, s-second (time unit); m-meter (length unit), m s ^-1 - meter per second (speed unit), deg s ^{- 1} - degrees per second (units of angular velocity)

Detailed ways

In order to facilitate the understanding of those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

The full-mode flight control structure block diagram of the composite rotorcraft shown in Figure 1 includes the following steps:

Step 1: Select the composite _rotorcraft dynamics model as the research object, and obtain the desired speed tracking command of the flight system according to the composite _rotorcraft flight mission. The desired yaw speed w _c , and the desired speed tracking command is used as the input of the controller;

期望滚转角ψ_c；姿态回路作为内回路，可使复合式旋翼飞行器各个通道解耦，增强系统的稳定性；Step 2: Design the speed loop and the attitude loop control structure respectively. The speed loop is used as the outer loop to provide the desired attitude control command for the attitude loop. The desired control command includes: the desired pitch angle θ _c , the desired yaw angle

Desired roll angle ψ _c ; the attitude loop is used as an inner loop, which can decouple each channel of the compound rotorcraft and enhance the stability of the system;

同时，设计结合神经网络的自适应终端滑模控制器模块，得到速度增量U_a1+U_t1，并通过控制分配计算得到期望姿态控制指令信号

设计姿态回路动态逆自适应终端滑模控制方法，首先通过动态逆控制器得到复合式旋翼操纵变量δ₂＝[A_1s B_1s θ_wr θ_wl θ₁ θ₂]，并结合速度回路得到的操纵变量δ₁，作为复合式旋翼飞行器期望舵面操纵信号

同时，设计结合神经网络的自适应终端滑模控制器模块，得到姿态角增量U_a2+U_t2，并通过控制分配计算舵面操纵信号增量Δδ，将U＝Δδ+δ_d作为复合式旋翼飞行器的实际舵面操纵信号；Step 3: Design the dynamic inverse adaptive terminal sliding mode control method of the speed loop, and obtain the manipulated variables of the compound rotorcraft through the dynamic inverse controller

At the same time, an adaptive terminal sliding mode controller module combined with a neural network is designed to obtain the velocity increment U _a1 +U _t1 , and the desired attitude control command signal is obtained through the control distribution calculation

The dynamic inverse adaptive terminal sliding mode control method of the attitude loop is designed. First, the composite rotor manipulation variable δ ₂ = [A _1s B _1s θ _wr θ _wl θ ₁ θ ₂ ] is obtained through the dynamic inverse controller, and combined with the control obtained by the speed loop The variable δ ₁ , as the desired rudder control signal of the compound rotorcraft

At the same time, an adaptive terminal sliding mode controller module combined with a neural network is designed to obtain the attitude angle increment U _a2 +U _t2 , and the rudder control signal increment Δδ is calculated through the control distribution, and U=Δδ+δ _d is taken as the compound formula The actual rudder control signal of the rotorcraft;

为旋翼总距，A_1s为横向周期变距、B_1s为纵向周期变距、θ_wr为右襟副翼偏角、θ_wl为左襟副翼偏角、θ₁为推力矢量与机体坐标系下XOY面夹角，θ₂为推力矢量在水平面投影与X轴的夹角。Among them, T is the thrust vector of the ducted fan,

is the rotor collective pitch, A _1s is the lateral cyclic pitch, B _1s is the longitudinal cyclic pitch, θ _wr is the right flaperon declination angle, θ _wl is the left flaperon declination angle, and θ ₁ is the thrust vector and the body coordinate system The angle between the lower XOY plane, θ ₂ is the angle between the projection of the thrust vector on the horizontal plane and the X axis.

The specific implementation of the dynamic inverse adaptive terminal sliding mode control includes the following steps:

Step 3.1, the block diagram of the instruction filtering structure shown in Figure 2,

或期望姿态控制指令信号

通过指令滤波模块，限制输入量的幅值和频率，得到期望速度跟踪指令的一阶导数和期望姿态控制指令的二阶导数；Desired speed tracking command signal

or desired attitude control command signal

Through the command filtering module, the amplitude and frequency of the input quantity are limited, and the first-order derivative of the desired speed tracking command and the second-order derivative of the desired attitude control command are obtained;

速度回路得到的一阶导数为

指令滤波模块采用二阶滤波器，表达式如下：Among them, the second derivative obtained by the attitude loop is

The first derivative of the velocity loop is

The instruction filtering module adopts a second-order filter, and the expression is as follows:

In the formula, the natural frequency _wn = 3 is selected, and the damping ratio ξ = 0.7.

Step 3.2, design a dynamic inverse controller according to the first derivative of the desired speed tracking command, the second derivative of the desired attitude control command and the flight state of the compound rotorcraft, and at the same time, design a neural network control method to compensate the system model error;

The block diagram of the dynamic inverse neural network shown in Figure 3 is implemented as follows:

Suppose the composite rotor dynamics model is expressed as:

表示x的一阶导数，δ表示复合式旋翼飞行器舵面操纵信号。In the formula,

Represents the first derivative of x, and δ represents the control signal of the rudder surface of the compound rotorcraft.

Due to the complexity of the composite rotorcraft model, δ cannot be directly obtained. The present invention obtains the model approximation function by linearizing the composite rotorcraft model:

表示复合式旋翼飞行器近似虚拟控制指令。In the formula,

Represents an approximate virtual control command for a compound rotorcraft.

并以此来设计动态逆控制器。其中，速度回路动态逆控制器近似舵面操纵信号

姿态回路动态逆控制器近似舵面操纵信号

According to the inverse solution of the approximation function

And use this to design a dynamic inverse controller. Among them, the speed loop dynamic inverse controller approximates the control signal of the rudder surface

Attitude Loop Dynamic Inverse Controller Approximate Rudder Surface Control Signal

And due to the deviation between the approximation function and the actual model of the compound rotorcraft, a single-layer sigma-pi neural network is designed to compensate for the model error.

The compensation value of the speed loop neural network is:

U _a1 =W ₁ ^T β ₁ (4)

The compensation value of the attitude loop neural network is:

U _a2 =W ₂ ^T β ₂ (5)

Among them, β ₁ and β ₂ are basis function vectors, W ₁ and W ₂ are weight coefficient vectors, and the values of β ₁ and β ₂ are as follows:

C₃'＝[u v w]，

kron()表示矩阵叉乘。In the formula, C ₁ =C ₁ '=[0.1 VV ² ],

C ₃ '=[uvw],

kron() represents matrix cross product.

It can be seen that the virtual control command U ₁ '(t)=U _d1 (t)+U _a1 (t) obtained by the dynamic inverse neural network of the speed loop, and the virtual control command U ₂ '(t) obtained by the dynamic inverse neural network of the attitude loop =U _d2 (t)+U _a2 (t).

Step 3.3. In order to enable the flight system of the composite rotorcraft to complete the system command tracking within a limited time and have good robustness, design the adaptive terminal sliding mode compensation value, including the speed loop adaptive terminal sliding mode compensation value U _t1 and attitude loop adaptive terminal sliding mode compensation value U _t2 to improve the command tracking speed and stability of the flight system;

The steps for implementing the adaptive terminal sliding mode are as follows:

Step 3.3.1. Set the compound rotorcraft to track the control command within a limited time T _d , and design the terminal sliding surface of the flight system;

Define the speed command tracking error as:

E ₁ (t)=x _r1 -x _c1 =[e ₁₁ , e ₂₁ , e ₃₁ ] ^T (7)

Similarly, the attitude command error can be expressed as:

E ₂ (t)=x _r2 -x _c2 =[e ₁₂ , e ₂₂ , e ₃₂ ] ^T (8)

为飞行系统姿态角跟踪指令，

θ_r和ψ_r分别表示俯仰角跟踪指令、偏航角跟踪指令和滚转速度跟踪指令，

为期望姿态控制指令，e₁₁、e₂₁和e₃₁分别表示前飞速度、升降速度和偏航速度指令跟踪误差，e₁₂、e₂₂和e₃₂分别表示俯仰角、偏航角和滚转角指令跟踪误差；In the formula, x _r1 =[ur , v _r , w _r ] ^T is the flight system speed tracking command, _{ur , v r and wr represent} the forward flight speed tracking command, the ascending and descending speed tracking command and the yaw speed tracking command respectively , x _c1 =[u _c , v _c , w _c ] ^T is the desired speed tracking command,

is the flight system attitude angle tracking command,

θ _r and ψ _r represent pitch angle tracking command, yaw angle tracking command and roll speed tracking command, respectively,

is the desired attitude control command, e ₁₁ , e ₂₁ and e ₃₁ represent the forward flight speed, lift speed and yaw speed command tracking error respectively, e ₁₂ , e ₂₂ and e ₃₂ represent the pitch angle, yaw angle and roll angle commands respectively tracking error;

The terminal sliding surface of the flight system is designed as:

S(x)=CE _j (t)-CP _j (t),j=1,2 (9)

The present invention selects C as a third-order unit matrix, and P _j (t)=[p _1j (t), p _2j (t), p _3j (t)] ^T represents a time-varying compensation function.

(10)

In the formula, i=1, 2, 3, j=1, 2, e _ij (0) represents the initial instruction tracking error when t=0;

设计速度回路自适应终端滑膜补偿值U_t1和姿态回路自适应终端滑膜补偿值U_t2，补偿系统期望控制指令x_r1和x_r2；Step 3.3.2, by constructing the Lyapunov function

Design the speed loop adaptive terminal synovial compensation value U _t1 and the attitude loop adaptive terminal synovial compensation value U _t2 , and the compensation system expects control commands x _r1 and x _r2 ;

关于时间的导数为：Lyapunov function

The derivative with respect to time is:

u_v为复合式旋翼飞行器虚拟控制指令；In the formula, j=1, 2, K is a constant, Δ _j =E _j (t)-P _j (t), A _j and B _j represent the state matrix and the approximate linearization model of the compound rotorcraft respectively. control matrix,

u _v is the virtual control command of the compound rotorcraft;

The design speed loop adaptive terminal synovial compensation value is:

In the formula, A ₁ and B ₁ respectively represent the state matrix and control matrix corresponding to the approximate linearized model velocity state quantity of the compound rotorcraft.

Taking u _v =U _t1 , and bringing it into equation (11), the derivative of the Lyapunov function in the velocity loop with respect to time can be obtained as:

In the same way, the design attitude loop adaptive terminal synovial compensation value is:

In the formula, A ₂ and B ₂ respectively represent the state matrix and control matrix corresponding to the attitude angle of the approximate linearized model of the compound rotorcraft.

Taking u _v =U _t2 , and bringing it into equation (11), the derivative of the Lyapunov function in the attitude loop with respect to time can be obtained as:

负定，故可以确定运动状态可在有限时间内沿着滑模面运动到平衡点，确保了飞行系统的鲁棒性。Through analysis, it can be seen that the Lyapunov function V(x) is positive definite, and its derivative with respect to time

Negative definite, so it can be determined that the motion state can move to the equilibrium point along the sliding surface within a limited time, which ensures the robustness of the flight system.

偏航角θ、滚转角ψ、俯仰角速度p、偏航角速度q和滚转角速度r，并重复步骤一至四。Step 4: Detect the flight status of the compound rotorcraft in real time, the flight status mainly includes: forward flight speed u, lateral speed v, vertical speed w, pitch angle

Yaw angle θ, roll angle ψ, pitch angular velocity p, yaw angular velocity q and roll angular velocity r, and repeat steps 1 to 4.

The invention carries out the speed command tracking simulation for the full-mode flight of the compound rotorcraft, and the control simulation command is set as: firstly take the low-speed flight of 30m/s as the initial state, accelerate to 40m/s after 10s, enter the transition mode flight, and then After 20s of acceleration, the composite rotorcraft can fly at a speed of 70m/s and then switch to the fixed-wing flight mode of high-speed flight. Finally, through the acceleration command, the composite rotorcraft can fly smoothly at a cruising speed of 80m/s after 10s. The simulation structure is shown in Figure 4 and Figure 5, and the shaded part in the figure is the transition flight mode.

Figure 4 shows that after 40s of acceleration, the composite rotorcraft completes the speed transition from a low speed of 40m/s to a high speed of 80m/s, and after 40s, the yaw speed and pitch speed approach zero, while the forward flight speed At 80m/s, the flight speed remains unchanged, and the compound rotorcraft starts to fly in a steady straight line at 40m/s.

Figure 5 shows that after the compound rotorcraft completes the speed tracking command, except for the slight disturbance of the attitude angle of the roll channel, the other attitude angles will remain approximately unchanged, the attitude angular velocity will approach zero, and the compound rotorcraft will fly stably .

The above are only the preferred embodiments of the present invention. It should be pointed out that for those skilled in the art, without departing from the principles of the present invention, several improvements and modifications can be made. It should be regarded as the protection scope of the present invention.

Claims (2)

1. a composite rotorcraft all-mode flight control method, is characterized in that, comprises the steps: Step 1: Select the composite _{rotorcraft dynamics} model, and obtain the desired speed tracking command of the flight system according to the composite rotorcraft flight mission. speed w _c , and take the desired speed tracking command as the input of the controller; Step 2: Design the speed loop and the attitude loop control structure respectively. The speed loop is used as the outer loop to provide the desired attitude control command for the attitude loop. The desired attitude control command includes: the desired pitch angle θ _c , the desired yaw angle

Desired roll angle ψ _c ; the attitude loop acts as an inner loop, which can decouple the channels of the compound rotorcraft and enhance the stability of the system; Step 3: Design the dynamic inverse adaptive terminal sliding mode control method of the speed loop, and obtain the manipulated variables of the compound rotorcraft through the dynamic inverse controller

At the same time, an adaptive terminal sliding mode controller module combined with a neural network is designed to obtain the speed increment U _a1 +U _t1 , where U _a1 is the compensation value of the speed loop neural network, and U _t1 is the speed loop adaptive terminal sliding mode compensation value , and obtain the desired attitude control command signal through the control distribution calculation

The dynamic inverse adaptive terminal sliding mode control method of the attitude loop is designed. First, the composite rotor manipulation variable δ ₂ = [A _1s B _1s θ _wr θ _wl θ ₁ θ ₂ ] is obtained through the dynamic inverse controller, and combined with the control obtained by the speed loop The variable δ ₁ , as the desired rudder control signal of the compound rotorcraft

At the same time, an adaptive terminal sliding mode controller module combined with a neural network is designed to obtain the attitude angle increment U _a2 +U _t2 , where U _a2 is the compensation value of the attitude loop neural network, and U _t2 is the attitude loop adaptive terminal sliding mode compensation value, and calculate the rudder surface control signal increment Δδ through the control distribution, and take U=Δδ+δ _d as the actual rudder surface control signal of the compound rotorcraft; Among them, T is the thrust vector of the ducted fan,

is the rotor collective pitch, A _1s is the lateral cyclic pitch, B _1s is the longitudinal cyclic pitch, θ _wr is the right flaperon declination angle, θ _wl is the left flaperon declination angle, and θ ₁ is the thrust vector and the body coordinate system The angle between the lower XOY plane, θ ₂ is the angle between the projection of the thrust vector on the horizontal plane and the X axis; Step 4: Detect the flight status of the compound rotorcraft in real time, the flight status mainly includes: forward flight speed u, lateral speed v, vertical speed w, pitch angle

Yaw angle θ, roll angle ψ, pitch angular velocity p, yaw angular velocity q and roll angular velocity r, and repeat steps 1 to 4; Step 3 The specific design steps include: Step 3.1. Desired speed tracking command signal

or desired attitude control command signal

Through the command filtering module, the amplitude and frequency of the input quantity are limited, and the first-order derivative of the desired speed tracking command and the second-order derivative of the desired attitude control command are obtained; Among them, the second derivative obtained by the attitude loop is

The first derivative of the velocity loop is

The instruction filtering module adopts a second-order filter, and the expression is as follows:

In the formula, select the natural frequency _wn = 3, the damping ratio ξ = 0.7; x _c1 = [u _c , _{vc , w c ]} ^T is the desired speed tracking command,

is the desired attitude control command; Step 3.2, design a dynamic inverse controller according to the first derivative of the desired speed tracking command, the second derivative of the desired attitude control command and the flight state of the compound rotorcraft, and at the same time, design a neural network control method to compensate the system model error; Step 3.3. In order to enable the flight system of the composite rotorcraft to complete the system command tracking within a limited time and have good robustness, design the adaptive terminal sliding mode compensation value, including the speed loop adaptive terminal sliding mode compensation value U _t1 and attitude loop adaptive terminal sliding mode compensation value U _t2 to improve the command tracking speed and stability of the flight system; The design steps of the dynamic inverse controller and neural network control method described in step 3.2 are: Step 3.2.1. Establish a nonlinear dynamic model of the compound rotor, select the equilibrium point, and use the quasi-Newton iteration method to perform dynamic trimming and small disturbance linearization analysis to obtain an approximate linearization model of the compound rotor; Since the compound rotorcraft has three flight modes, including helicopter flight mode, transition flight mode and fixed-wing flight mode, and the values of the manipulated variables in each flight mode are quite different, the forward flight speed in each mode is selected for dynamics Trim and small disturbance linearization analysis; It is assumed that the nonlinear dynamic model of the composite rotor is expressed as:

In the formula,

represents the first derivative of x, δ represents the control signal of the rudder surface of the compound rotorcraft, The approximate linearized model obtained by trimming can be expressed as:

in,

Indicates the approximate rudder surface control signal of the composite rotorcraft; Step 3.2.2, in order to meet the dynamic inverse control conditions, according to the approximate linearization model of the compound rotorcraft, inversely solve the virtual control command

And based on this, the dynamic inverse controller is designed, that is, the dynamic inverse controller of the speed loop approximates the control signal of the rudder surface.

Attitude Loop Dynamic Inverse Controller Approximate Rudder Surface Control Signal

Step 3.2.3. Design a single-layer sigma-pi neural network to compensate for the model error, The compensation value of the speed loop neural network is: U _a1 =W ₁ ^T β ₁ (4) The compensation value of the attitude loop neural network is: U _a2 =W ₂ ^T β ₂ (5) Among them, β ₁ and β ₂ are basis function vectors, W ₁ and W ₂ are weight coefficient vectors, and the values of β ₁ and β ₂ are as follows:

In the formula, C ₁ =C ₁ '=[0.1 VV ² ],

C ₃ =[u vw],

kron() represents matrix cross product. 2. The all-mode flight control method for a composite rotorcraft according to claim 1, characterized in that, The adaptive terminal sliding mode design steps described in step 3.3 include: Step 3.3.1. Set the compound rotorcraft to track the control command within a limited time T _d , and design the terminal sliding surface of the flight system; Define the speed command tracking error as: E ₁ (t)=x _r1 -x _c1 =[e ₁₁ , e ₂₁ , e ₃₁ ] ^T (7) Similarly, the attitude command error can be expressed as: E ₂ (t)=x _r2 -x _c2 =[e ₁₂ , e ₂₂ , e ₃₂ ] ^T (8) In the formula, x _r1 =[ur , v _r , w _r ] ^T is the flight system speed tracking command, _{ur , v r and wr represent} the forward flight speed tracking command, the ascending and descending speed tracking command and the yaw speed tracking command respectively ,

is the flight system attitude angle tracking command,

θ _r and ψ _r respectively represent pitch angle tracking command, yaw angle tracking command and roll speed tracking command, e ₁₁ , e ₂₁ and e ₃₁ respectively represent the forward flight speed, lift speed and yaw speed command tracking error, e ₁₂ , e ₂₂ and e ₃₂ represent pitch angle, yaw angle and roll angle command tracking error respectively; The terminal sliding surface of the flight system is designed as: S(x)=CE _j (t)-CP _j (t),j=1,2 (9) The present invention selects C as a third-order unit matrix, P _j (t)=[p _1j (t), p _2j (t), p _3j (t)] ^T represents a time-varying compensation function,

In the formula, i=1, 2, 3, j=1, 2, e _ij (0) represents the initial instruction tracking error when t=0; Step 3.3.2, by constructing the Lyapunov function

Design the speed loop adaptive terminal synovial compensation value U _t1 and the attitude loop adaptive terminal synovial compensation value U _t2 , and the compensation system expects control commands x _r1 and x _r2 ; Lyapunov function

The derivative with respect to time is:

In the formula, j=1, 2, K is a constant, Δ _j =E _j (t)-P _j (t), A _j and B _j represent the state matrix and the approximate linearization model of the compound rotorcraft respectively. control matrix,

u _v is the virtual control command of the compound rotorcraft; The design speed loop adaptive terminal synovial compensation value is:

In the formula, A ₁ and B ₁ respectively represent the state matrix and control matrix corresponding to the approximate linearized model velocity state quantity of the compound rotorcraft, Taking u _v =U _t1 , and bringing it into equation (11), the derivative of the Lyapunov function in the velocity loop with respect to time can be obtained as:

In the same way, the design attitude loop adaptive terminal synovial compensation value is:

In the formula, A ₂ and B ₂ respectively represent the state matrix and control matrix corresponding to the attitude angle of the approximate linearized model of the composite rotorcraft, Taking u _v =U _t2 , and bringing it into equation (11), the derivative of the Lyapunov function in the attitude loop with respect to time can be obtained as:

Through analysis, it can be seen that the Lyapunov function V(x) is positive definite, and its derivative with respect to time

Negative definite, so it can be determined that the motion state can move to the equilibrium point along the sliding surface within a limited time, which ensures the robustness of the flight system.

Images