Full-mode flight control method of combined type rotor craft
Technical Field
The invention belongs to the technical field of flight mechanics and flight simulation, and particularly relates to a full-mode flight control technology of a combined type rotor craft.
Background
The combined rotor aircraft has the thrust device, the fixed wings and the rotor wings, so that the combined rotor aircraft can not only realize vertical take-off and landing and hovering flight in a helicopter flight mode, but also has the high-speed and long-range cruising flight capability of the fixed wing aircraft, and is attracted by helicopter research institutions and researchers in all countries in the world. At present, the research on the carrier-based unmanned aerial vehicle of the composite rotor aircraft at home and abroad is still in the starting stage, and due to the reasons of technical secrecy and the like, the problems of the multi-mode flight control of the high-speed helicopter at home and abroad are less related at present.
The flight control system of the composite rotor craft is the key and guarantee for safe and stable flight of the craft and completion of various flight tasks. Designing a flight control system of a composite rotary wing aircraft is not easy, and both the problem of helicopter flight control and the problem of fixed wing flight control are considered, the problem of transitional flight control is also considered, and meanwhile, the problem of control switching during switching of different flight modes is also considered. When the flight mode of the aircraft changes, the aerodynamic characteristics change quite complexly, a complete and accurate mathematical model is difficult to establish for design verification of a control system, and the designed flight control system has to have good adaptivity and robustness.
The design method of the flight control law of the specific flight state based on the small-disturbance linearized model is a common method at present, and for the problem of multi-mode flight control of the composite rotor aircraft, the design method faces challenges and is difficult to achieve an expected control target by adopting a traditional linear control method and using a single parameter controller. In the flight process, mode conversion, variation of the inertia moment of an aircraft body, fluctuation of the rotating speed of a rotor wing, external interference and other factors bring uncertainty to the system, so that the aircraft system has the characteristics of nonlinearity, strong coupling and uncertainty. The change in flight mode determines the change in flight dynamics, which means that the flight control system needs to be adjusted to ensure that the aircraft flies smoothly. The adjustment of the control system is not only an adjustment of the control parameters but may also require an adjustment of the control structure. This creates significant difficulties in the design of the control system. The flight dynamics characteristic change caused by the flight mode change is abstracted into the uncertainty of the dynamic characteristic of the controlled object, a set of flight control system which is suitable for the full flight mode and has self-adaptive capacity and robustness is designed by applying the nonlinear control theory, and the method is an ideal method for solving the flight control of the combined rotor craft.
The invention takes the typical flight state as a base point, designs the nonlinear intelligent flight control system capable of compensating the inverse error of the model on line, and ensures that the composite rotor aircraft has good control performance under different flight modes, so that the system has real-time performance, stability and robustness. A nonlinear dynamic inverse model solving method is explored, and the rapidity, stability and accuracy of an inverse model controller algorithm are solved.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a full-mode flight control technology of a composite rotary wing aircraft, so that the system has self-adaptive capacity and robustness, the problem of controller switching in the multi-mode flight process is solved, and the full-mode stable flight of the composite rotary wing aircraft in the full-envelope range is realized.
In order to solve the problems, the technical scheme adopted by the invention is as follows:
a full-mode flight control method of a combined rotor craft comprises the following specific steps:
step one, selecting a certain combined rotor aircraft dynamic model as a research object, and obtaining an expected speed tracking instruction of a flight system according to a flight task of the combined rotor aircraft, wherein the expected speed tracking instruction comprises the following steps: desired forward flight velocity ucDesired lifting speed vcAnd an estimated cruising speed wcAnd taking the expected speed tracking command as the input quantity of the controller;
step two, respectively designing a speed loop and an attitude loop control structure, wherein the speed loop is used as an outer loop and provides an expected attitude control instruction for the attitude loop, and the expected attitude control instruction comprises the following steps: desired pitch angle θcDesired yaw angleDesired roll angle psic(ii) a The attitude loop is used as an inner loop, so that each channel of the composite rotor craft can be decoupled, and the stability of the system is enhanced;
step three, designing a speed loop dynamic inverse self-adaptive terminal sliding mode control method, and obtaining a composite rotor aircraft manipulated variable through a dynamic inverse controllerAt the same time, design integrationObtaining a speed increment U by a self-adaptive terminal sliding mode controller module of a neural networka1+Ut1And obtaining the expected attitude control command signal through control distribution calculationA sliding mode control method for designing a dynamic inverse self-adaptive terminal of an attitude loop comprises the steps of firstly obtaining a composite rotor wing control variable delta through a dynamic inverse controller2=[A1sB1sθwrθwlθ1θ2]And combined with a manipulated variable delta derived from the speed loop1As a combined rotorcraft desired control surface steering signalMeanwhile, an adaptive terminal sliding mode controller module combined with a neural network is designed to obtain an attitude angle increment Ua2+Ut2And calculating control surface control signal increment delta through control distribution, and changing U into delta + deltadAs an actual control plane steering signal for the compound rotorcraft;
wherein T is the thrust vector of the ducted fan,is the total pitch of the rotor, A1sIs transversely periodically pitch-changed, B1sIs longitudinally and periodically pitch-changed, thetawrIs the right flap aileron deflection angle thetawlIs the offset angle theta of the left flaperon1Is the included angle theta between the thrust vector and the XOY surface under the coordinate system of the machine body2The included angle between the projection of the thrust vector on the horizontal plane and the X axis is shown.
Step four, the flight state of real-time detection combined type rotor craft, flight state mainly includes: front flying speed u, lateral speed v, vertical speed w and pitch angleYaw angle theta, roll angle psi, pitch angular velocity p, yaw angular velocity q and roll angular velocity r, and repeating steps one through four.
Further, the attitude loop control and the speed loop control in the second step both include an instruction filtering module, a dynamic inverse neural network module and a self-adaptive terminal sliding mode module, and specifically include the following steps:
step 3.1, expect speed to track command signalOr desired attitude control command signalLimiting the amplitude and frequency of the input quantity through an instruction filtering module to obtain a first derivative of an expected speed tracking instruction and a second derivative of an expected attitude control instruction;
wherein the second derivative obtained by the attitude loop isThe first derivative obtained by the speed loop isThe instruction filtering module adopts a second-order filter, and the expression is as follows:
in the formula, the natural frequency w is selectednThe damping ratio ξ is 0.7, 3.
3.2, designing a dynamic inverse controller according to a first derivative of the expected speed tracking instruction, a second derivative of the expected attitude control instruction and the flight state of the composite rotor aircraft, and simultaneously designing a neural network control method to compensate the system model error;
step 3.3, in order to enable the flight system of the composite rotor aircraft to complete system instruction tracking in limited time and have better robustness, a self-adaptive terminal is designedSliding mode compensation value comprising speed loop adaptive terminal sliding mode compensation value Ut1(t) and attitude loop adaptive terminal sliding mode compensation value Ut2The command tracking speed and the stability of the flight system are improved;
further, the step of designing the dynamic inverse controller and neural network control method in step 3.2 includes:
3.2.1, establishing a nonlinear dynamic model of the composite rotor wing, selecting a balance point, and performing dynamic balancing and small-disturbance linear analysis by adopting a quasi-Newton iteration method to obtain an approximate linearization model of the composite rotor wing aircraft;
because the composite rotor aircraft has three flight modes, including a helicopter flight mode, a transition flight mode and a fixed wing flight mode, and the value difference of the control variable in each flight mode is large, the dynamic balancing and small-disturbance linear analysis are respectively carried out by selecting the forward flight speed of 20m/s in the helicopter flight mode, the forward flight speed of 70m/s in the transition flight mode and the forward flight speed of 100m/s in the fixed wing flight mode;
assuming the composite rotor nonlinear dynamics model is expressed as:
in the formula, the first derivative of x is represented and δ represents the compound rotorcraft control plane steering signal.
The trim-derived approximate linearization model can be expressed as:
in the formula,representing the approximate control plane steering signal of the compound rotorcraft.
Step 3.2.2, in order to meet the dynamic inverse control condition, according to the approximate linear model of the composite rotor craft, the virtual control instruction is solved reverselyDesigning a dynamic inverse controller according to the dynamic inverse controller;
i.e. approximate control surface steering signal for speed loop dynamic inverse controllerApproximate control surface control signal of attitude loop dynamic inverse controller
And 3.2.3, because the approximate linearization model and the actual model of the combined rotor aircraft have deviation, the flight system is possibly unstable due to the dynamic inverse controller designed by the approximate linearization model, and the single-layer sigma-pi neural network is designed to compensate the model error.
The speed loop neural network compensation value is as follows:
Ua1=W1 Tβ1(4)
the attitude loop neural network compensation value is as follows:
Ua2=W2 Tβ2(5)
wherein, β1、β2Are vector of basis functions, W1、W2As a vector of weight coefficients, β1、β2The values are as follows:
in the formula, C1=C1'=[0.1V V2],C3=[u vw],kron () represents a matrix cross product.
Further, the step of designing the adaptive terminal sliding mode module according to step 3.3 includes:
step 3.3.1, setting the composite rotor craft in the limited time TdInternally tracking a control instruction, and designing a terminal sliding mode surface of the flight system;
defining the speed command tracking error as:
E1(t)=xr1-xc1=[e11,e21,e31]T(7)
similarly, the attitude command error can be expressed as:
E2(t)=xr2-xc2=[e12,e22,e32]T(8)
in the formula, xr1=[ur,vr,wr]TFor flight system velocity tracking commands, ur、vrAnd wrRespectively representing a forward flying speed tracking command, a lifting speed tracking command and a yawing speed tracking command, xc1=[uc,vc,wc]TIn order for the speed to be expected to track the command,as a flying systemA tracking command of the attitude angle of the system,θrand psirRespectively representing a pitch angle tracking command, a yaw angle tracking command and a roll speed tracking command,to expect attitude control commands, e11、e21And e31Representing the tracking errors of the forward flight speed, the lifting speed and the yaw speed commands, respectively, e12、e22And e32Respectively representing the command tracking errors of a pitch angle, a yaw angle and a roll angle;
the terminal sliding mode surface of the flight system is designed as follows:
S(x)=CEj(t)-CPj(t),j=1,2 (9)
the invention selects C as a third-order identity matrix, Pj(t)=[p1j(t),p2j(t),p3j(t)]TRepresenting a time-varying compensation function.
Wherein i is 1, 2, 3, j is 1, 2, eij(0) The initial instruction tracking error is represented when t is 0.
Step 3.3.2, by constructing the Lyapunov functionDesign speed loop control law Ut1And attitude loop control law Ut2Compensating for system expected control command xr1And xr2;
Wherein, Lyapunov functionThe derivative with respect to time is:
wherein j is 1 or 2, K is a normal number, and Δj=Ej(t)-Pj(t),Aj、BjRespectively represents a state matrix and a control matrix corresponding to the approximate linearization model of the composite rotor aircraft,uva virtual control command for the combined rotor craft;
the speed loop control law is designed as follows:
in the formula, A1、B1And respectively representing a state matrix and a control matrix corresponding to the approximate linearization model speed state quantity of the composite rotor aircraft.
Get uv=Ut1The derivative of the Lyapunov function in the derived velocity loop with equation (11) with respect to time is:
similarly, the attitude loop control law is designed as follows:
in the formula,A2、B2and respectively representing a state matrix and a control matrix corresponding to the attitude angle of the approximate linear model of the composite rotor aircraft.
Get uv=Ut2The derivative of the Lyapunov function in the attitude loop obtained by the equation (11) with respect to time is as follows:
analytically, the Lyapunov function V (x) is positive, its derivative with respect to timeNegative determination is carried out, so that the motion state can be determined to move to a balance point along the sliding mode surface in a limited time, and the robustness of the flight system is ensured.
The invention has the beneficial effects that:
the speed loop control and the attitude loop control are respectively designed, and the attitude loop can decouple all channels of the composite rotor craft, so that the stability of a flight system is enhanced; the speed loop takes the speed quantity as a design target, and the command tracking capability and the anti-interference capability of the flight system are improved.
According to the invention, dynamic inverse self-adaptive terminal sliding mode control methods are respectively designed in the attitude loop and the speed loop, so that the composite rotor aircraft can finish system instruction tracking within a set limited time, the convergence is better, and the flight system has better robustness.
In addition, the full-mode control method of the combined type rotor craft provided by the invention can realize full-envelope full-mode flight of the craft, and a controller does not need to be switched in the flight process, so that the complexity of a flight system is reduced, and the safety of mode switching of the craft is improved.
Drawings
FIG. 1 is a block diagram of the full mode flight control architecture of the composite rotary wing aircraft of the present invention;
FIG. 2 is a block diagram of an instruction filtering architecture according to the present invention;
FIG. 3 is a block diagram of a dynamic inverse neural network control architecture of the present invention;
FIG. 4 is a graph of velocity variation versus time for a hybrid rotorcraft simulating full mode flight, in accordance with an embodiment of the present invention;
FIG. 4 is a graph of forward flight speed versus time for a hybrid rotorcraft simulating full mode flight in accordance with an embodiment of the present invention; (b) the method is a graph of the relationship between the yaw speed and the time of the combined rotor craft under the simulated full-mode flight in the embodiment of the invention; (c) the relationship graph of the lifting speed and the time of the combined rotor craft under the simulated full-mode flight in the embodiment of the invention;
FIG. 5 is a graph of attitude angle and attitude angular velocity versus time for a hybrid rotorcraft simulating full mode flight, in accordance with an embodiment of the present invention;
FIG. 5 (a) is a plot of pitch angle versus time for a hybrid rotorcraft simulated full mode flight in accordance with an embodiment of the present invention; (b) in the embodiment of the invention, a relation graph of the yaw angle and the time of the combined rotor aircraft under the simulated full-mode flight is shown; (c) in the embodiment of the invention, a relation graph of the roll angle and the time of the composite rotor aircraft under the simulation full-mode flight is shown; (d) in the embodiment of the invention, a relationship graph of the pitch angle angular velocity and the time of the combined rotor craft under the simulation full-mode flight is provided; (e) in the embodiment of the invention, a relationship graph of the yaw angular velocity and the time of the combined rotor aircraft in the simulated full-mode flight is shown; (f) in the embodiment of the invention, a relation graph of the roll angular velocity and the time of the composite rotor aircraft in the full-mode flight is simulated;
the labels in the figure are: deg-degree (angle units), t-time, s-second (time units); m-meters (length unit),m·s-1-meters per second (speed unit), deg.s-1-degree per second (angular velocity unit)
Detailed Description
In order to facilitate understanding of those skilled in the art, the present invention will be further described with reference to the accompanying drawings.
The structural block diagram of the full-mode flight control of the compound rotorcraft shown in fig. 1 includes the following steps:
step one, selecting a dynamic model of a composite rotor aircraft as a research object, and obtaining an expected speed tracking instruction of a flight system according to a flight task of the composite rotor aircraft, wherein the expected speed tracking instruction comprises the following steps: desired forward flight velocity ucDesired lifting speed vcAnd desired yaw velocity wcAnd taking the expected speed tracking command as the input quantity of the controller;
step two, respectively designing a speed loop and an attitude loop control structure, wherein the speed loop is used as an outer loop and provides an expected attitude control instruction for the attitude loop, and the expected control instruction comprises the following steps: desired pitch angle θcDesired yaw angleDesired roll angle psic(ii) a The attitude loop is used as an inner loop, so that each channel of the composite rotor craft can be decoupled, and the stability of the system is enhanced;
step three, designing a speed loop dynamic inverse self-adaptive terminal sliding mode control method, and obtaining a composite rotor aircraft manipulated variable through a dynamic inverse controllerMeanwhile, a self-adaptive terminal sliding mode controller module combined with a neural network is designed to obtain a speed increment Ua1+Ut1And obtaining the expected attitude control command signal through control distribution calculationA sliding mode control method for designing a dynamic inverse self-adaptive terminal of an attitude loop comprises the steps of firstly obtaining a composite rotor wing control variable delta through a dynamic inverse controller2=[A1sB1sθwrθwlθ1θ2]And combined with a manipulated variable delta derived from the speed loop1As a combined rotorcraft desired control surface steering signalMeanwhile, an adaptive terminal sliding mode controller module combined with a neural network is designed to obtain an attitude angle increment Ua2+Ut2And calculating control surface control signal increment delta through control distribution, and changing U into delta + deltadAs an actual control plane steering signal for the compound rotorcraft;
wherein T is the thrust vector of the ducted fan,is the total pitch of the rotor, A1sIs transversely periodically pitch-changed, B1sIs longitudinally and periodically pitch-changed, thetawrIs the right flap aileron deflection angle thetawlIs the offset angle theta of the left flaperon1Is the included angle theta between the thrust vector and the XOY surface under the coordinate system of the machine body2The included angle between the projection of the thrust vector on the horizontal plane and the X axis is shown.
The specific implementation mode of the sliding mode control of the dynamic inverse self-adaptive terminal comprises the following steps:
step 3.1, the instruction filtering architecture block diagram shown in figure 2,
desired velocity tracking command signalOr desired attitude control command signalThrough instruction filtering moduleLimiting the amplitude and frequency of the input quantity to obtain a first derivative of an expected speed tracking command and a second derivative of an expected attitude control command;
wherein the second derivative obtained by the attitude loop isThe first derivative obtained by the speed loop isThe instruction filtering module adopts a second-order filter, and the expression is as follows:
in the formula, the natural frequency w is selectednThe damping ratio ξ is 0.7, 3.
3.2, designing a dynamic inverse controller according to a first derivative of the expected speed tracking instruction, a second derivative of the expected attitude control instruction and the flight state of the composite rotor aircraft, and simultaneously designing a neural network control method to compensate the system model error;
the structure block diagram of the dynamic inverse neural network shown in fig. 3 is specifically implemented as follows:
assuming the combined rotor dynamics model is expressed as:
in the formula, the first derivative of x is represented and δ represents the compound rotorcraft control plane steering signal.
Because the combined rotor aircraft model is complex and delta cannot be directly solved, the model approximation function is obtained by linearizing the combined rotor aircraft model:
in the formula,representing a composite rotorcraft approximate virtual control command.
Solving from the inverse of the approximation functionAnd thus a dynamic inverse controller is designed. Wherein, the speed loop dynamic inverse controller approximates the control surface control signalApproximate control surface control signal of attitude loop dynamic inverse controller
And because the approximation function has deviation with the actual model of the combined type rotor aircraft, a single-layer sigma-pi neural network is designed to compensate model errors.
The speed loop neural network compensation value is as follows:
Ua1=W1 Tβ1(4)
the attitude loop neural network compensation value is as follows:
Ua2=W2 Tβ2(5)
wherein, β1、β2Are vector of basis functions, W1、W2Is a weightCoefficient vector, β1、β2The values are as follows:
in the formula, C1=C1'=[0.1V V2], C3'=[u v w],kron () represents a matrix cross product.
From this, it is understood that the virtual control command U 'is obtained from the speed circuit dynamic inverse neural network'1(t)=Ud1(t)+Ua1(t), virtual control instruction U 'obtained by attitude loop dynamic inverse neural network'2(t)=Ud2(t)+Ua2(t)。
Step 3.3, in order to enable the flight system of the composite rotorcraft to complete system instruction tracking in limited time and have better robustness, designing a self-adaptive terminal sliding mode compensation value comprising a speed loop self-adaptive terminal sliding mode compensation value Ut1(t) and attitude loop adaptive terminal sliding mode compensation value Ut2The command tracking speed and the stability of the flight system are improved;
the self-adaptive terminal sliding mode implementation steps are as follows:
step 3.3.1, setting the composite rotor craft in the limited time TdInternally tracking a control instruction, and designing a terminal sliding mode surface of the flight system;
defining the speed command tracking error as:
E1(t)=xr1-xc1=[e11,e21,e31]T(7)
similarly, the attitude command error can be expressed as:
E2(t)=xr2-xc2=[e12,e22,e32]T(8)
in the formula, xr1=[ur,vr,wr]TFor flight system velocity tracking commands, ur、vrAnd wrRespectively representing a forward flying speed tracking command, a lifting speed tracking command and a yawing speed tracking command, xc1=[uc,vc,wc]TIn order for the speed to be expected to track the command,in order to track the command of the attitude angle of the flight system,θrand psirRespectively representing a pitch angle tracking command, a yaw angle tracking command and a roll speed tracking command,to expect attitude control commands, e11、e21And e31Representing the tracking errors of the forward flight speed, the lifting speed and the yaw speed commands, respectively, e12、e22And e32Respectively representing the command tracking errors of a pitch angle, a yaw angle and a roll angle;
the terminal sliding mode surface of the flight system is designed as follows:
S(x)=CEj(t)-CPj(t),j=1,2 (9)
the invention selects C as a third-order identity matrix, Pj(t)=[p1j(t),p2j(t),p3j(t)]TRepresenting a time-varying compensation function.
Wherein i is 1, 2, 3, j is 1, 2, eij(0) Representing the initial instruction tracking error when t is 0;
step 3.3.2, by constructing the Lyapunov functionDesign speed loop control law Ut1And attitude loop control law Ut2Compensating for system expected control command xr1And xr2;
Lyapunov functionThe derivative with respect to time is:
wherein j is 1 or 2, K is a normal number, and Δj=Ej(t)-Pj(t),Aj、BjRespectively represents a state matrix and a control matrix corresponding to the approximate linearization model of the composite rotor aircraft,uva virtual control command for the combined rotor craft;
the speed loop control law is designed as follows:
in the formula, A1、B1And respectively representing a state matrix and a control matrix corresponding to the approximate linearization model speed state quantity of the composite rotor aircraft.
Get uv=Ut1The derivative of the Lyapunov function in the derived velocity loop with equation (11) with respect to time is:
similarly, the attitude loop control law is designed as follows:
in the formula, A2、B2And respectively representing a state matrix and a control matrix corresponding to the attitude angle of the approximate linear model of the composite rotor aircraft.
Get uv=Ut2The derivative of the Lyapunov function in the attitude loop obtained by the equation (11) with respect to time is as follows:
analytically, the Lyapunov function V (x) is positive, its derivative with respect to timeNegative determination is carried out, so that the motion state can be determined to move to a balance point along the sliding mode surface in a limited time, and the robustness of the flight system is ensured.
Step four, the flight state of real-time detection combined type rotor craft, flight state mainly includes: front flying speed u, lateral speed v, vertical speed w and pitch angleYaw angle theta, roll angle psi, pitch angular velocity p, yaw angular velocity q and roll angular velocity r, and repeating steps one through four.
The invention carries out speed instruction tracking simulation on the full-mode flight of the composite rotor craft, and sets a control simulation instruction as follows: the method comprises the steps of taking low-speed flight at 30m/s as an initial state, accelerating to 40m/s through 10s, entering transition mode flight, accelerating for 20s, switching to a fixed wing flight mode of high-speed flight after the flight speed of the composite rotor aircraft reaches 70m/s, and enabling the composite rotor aircraft to stably fly at the cruising speed of 80m/s through an acceleration instruction after 10 s. The simulation structure is shown in fig. 4 and 5, wherein the hatched part is the transition flight mode.
Fig. 4 shows that after 40s of acceleration, the compound rotary wing vehicle has completed a speed transition from a low speed of 40m/s to a high speed of 80m/s, and after 40s, the yaw and pitch speeds approach zero, while the forward flight speed is 80m/s, the flight speed remains unchanged, and the compound rotary wing vehicle 40m/s starts to fly in a steady straight line.
Fig. 5 shows that after the speed tracking command is completed, the attitude angles of the combined rotor craft are approximately kept unchanged except for slight disturbance of the attitude angle of the roll channel, the attitude angular speed approaches zero, and the flight of the combined rotor craft is stable.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.