Dynamics modeling and control method of a distributed thrust tilt-wing aircraft

Abstract

The present invention discloses a dynamic modeling and control method for a distributed thrust tilt-wing aircraft. The modeling method comprises the steps of establishing a multi-body dynamic tensor model and designing a smooth trajectory for synchronous tilting of a wing and a canard in a modal conversion stage. The control method adopts a hybrid control framework in which an inner and outer loop are connected in series, comprising an outer loop trajectory tracking control module, an inner loop attitude control module, a logic switching module and a distributed thrust control allocation module. Finally, the control method is simulated and verified by designing a reference trajectory. The structural weight layout design in which the center of mass of the wing and the canard, the center of mass of the fuselage and the total center of mass of the whole aircraft coincide with each other reduces the difficulty of modeling and control. A fine dynamic model of a distributed tilt-wing aircraft in a tensor form is established by using a multi-body dynamics method. Compared with a traditional single-body dynamics modeling method, the model accuracy is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle control, and in particular, relates to a dynamic modeling and control method for a distributed thrust tilt-wing aircraft. Background Art

[0002] Conventional aircraft, especially fixed-wing aircraft, generally do not undergo major changes in their airframe structure, so single rigid body dynamics methods are often used in modeling. The research object is considered as a rigid body as a whole, and the relative motion between multiple internal components is not considered. However, during the mode conversion stage, due to the rotation of active parts such as wings, tilt-wing aircraft will produce additional dynamic characteristics compared to fixed-wing aircraft, such as the anti-torque generated by wing rotation, the gyroscopic torque generated by the random wing tilt of high-speed rotating rotors, etc., which will continuously reconstruct the dynamic model of the aircraft and seriously affect flight safety. Therefore, the single-body dynamics modeling method based on the fixed center of mass and single rigid body assumption is prone to modeling errors in the face of parameter changes and inertial forces caused by tilting, and is not suitable for the modeling of distributed thrust tilt-wing aircraft;

[0003] The multi-body dynamics method models each active part of the aircraft separately and projects it to the center of mass of the body, fully considering the impact of the tilting motion of the active parts on the aircraft dynamics model. Therefore, the modeling accuracy is higher and it is more suitable for the modeling of distributed thrust tilt-wing aircraft. At the same time, the introduction of tensor symbols has clear physical meaning, simplifies the equation expression and derivation process, and makes the model have the characteristic of tensor form independent of the coordinate system, which is convenient for computer programming and solving using matrix operations.

[0004] The distributed thrust tilt-wing aircraft is a highly nonlinear, strongly coupled, under-driven controlled system. The rotation of the moving parts causes the system to be time-varying. At the same time, the aerodynamic changes in the modal conversion stage are also complex. These characteristics ultimately put forward strict design requirements for the controller. Therefore, a controller that can not only solve the uncertainty problems such as system time variation, unmodeled dynamics and external interference, but also has the characteristics of fast response speed and high control accuracy is crucial for the stable control and trajectory tracking of the aircraft.

[0005] Commonly used methods for traditional aircraft include PID control, backstepping control, and dynamic inversion control. PID control is a linear control method commonly used in engineering. It has the simplest structure and low requirements for model accuracy. However, it also has the disadvantages of poor anti-interference ability and low control accuracy, especially when the model parameters change or there is external interference, the control effect is often poor; backstepping is a nonlinear controller that can meet the control requirements of nonlinear systems with uncertainties such as external interference, but due to the online update of control parameters, its real-time performance is reduced and it is not suitable for the attitude subsystem of the fast loop, but is mostly used for the position subsystem of the slow loop. Dynamic inversion offsets the nonlinear characteristics of the original system to obtain a new linearized system, which can be controlled by a simple linear method. However, this method has extremely high requirements for model accuracy. Uncertainties such as unmodeled dynamics and parameter perturbations will seriously affect the control effect, and it is not suitable for the attitude control of tilt-rotating aircraft.

[0006] Therefore, in order to improve the robustness of nonlinear control systems, speed up system response speed, and eliminate the impact of uncertainties such as unmodeled dynamics, inertial parameter perturbations and external interference, the present invention adopts a hybrid control strategy in the form of inner and outer loops. The outer loop is designed with an L1 adaptive trajectory tracking controller, and the inner loop is designed with an attitude controller based on robust error integration. These two controllers have relatively simple structures, fast convergence, and can effectively solve uncertainty problems. Therefore, they are more suitable for attitude control of variable configuration aircraft with higher requirements on response speed and control accuracy than the above-mentioned PID, backstepping and dynamic inversion. Summary of the invention

[0007] In view of the above defects, the present invention provides a dynamic modeling method for a distributed thrust tilt-wing aircraft, comprising the following steps:

[0008] S1. Establish a time-varying multi-body dynamic tensor model of a distributed thrust tilt-wing aircraft;

[0009] S1-1. Establish the general translational dynamics equation of the whole machine in the inertial system in tensor form;

[0010] S1-2, adjusting the structural weight layout so that the mass centers of the wing and canard, the body mass center and the total mass center of the whole aircraft coincide, and the eccentric forces of the wing / canard / rotor relative to the total mass center of the whole aircraft are offset against each other, and the simplified translational dynamic equation under the body axis system is obtained according to step S1-1;

[0011] S1-3, transform the above equation into an inertial coordinate system to obtain the final translational dynamics equation after transformation;

[0012] S1-4. Establish the full machine rotation dynamics equation in tensor form relative to the origin of the inertial coordinate system;

[0013] S1-5, based on the structural weight layout adjustment result of S1-2, the additional eccentric moment caused by the moving parts such as the wing, canard and rotor and coupled with the translational dynamics is offset, and the simplified rotational dynamics equation is obtained according to step S1-4;

[0014] S1-6, transforming the above equation according to the conversion relationship between angular velocity and attitude angle to obtain the final rotational dynamics equation after transformation;

[0015] S2. Design the smooth trajectory of the synchronous tilting of the wing and canard during the mode conversion stage, including the following designs:

[0016] S2-1. Design of the tilting angular acceleration from vertical take-off to level flight and from level flight to vertical take-off;

[0017] S2-2. Design of the tilt angular velocity from vertical take-off to level flight and from level flight to vertical take-off;

[0018] S2-3. Design of the tilt angle from vertical takeoff to level flight and from level flight to vertical takeoff.

[0019] Furthermore, in step S1-1, the specific process of establishing the general translational dynamics equation of the whole machine in the inertial system in tensor form is:

[0020] Based on the idea of ​​multi-body dynamics, each moving part on the aircraft is regarded as a separate rigid body. According to Newton's second law, the force analysis is performed on each rigid body. Finally, the general translational dynamic equation of the whole aircraft in the inertial system is obtained by superposition. Its tensor form is expressed as follows:

[0021] ;

[0022] In the above formula, It represents the total external force on the whole aircraft. Respectively, body, Wing and canard and The rotor mass, Respectively represent the body, Wing and canard and The position of the center of mass of each rotor in the inertial coordinate system, They represent the body, Wing and canard and The linear velocity of the rotor relative to the inertial coordinate system is as follows: It represents the first-order rotational time derivative relative to the coordinate system “*”.

[0023] Furthermore, in step S1-2, the structural weight layout is adjusted so that the wing and canard mass centers, the body mass center and the total mass center of the whole aircraft coincide with each other, and the general translational dynamics equation is simplified to obtain the translational dynamics equation under the body axis system:

[0024] In the above formula, is the total mass of the machine, They represent the projections of the body's center of mass velocity and acceleration vector on the body axis system, is the attitude angular velocity of the aircraft, is the sum of all rotor thrusts, is the synchronous tilt angle of the wing and canard, is the aerodynamic force vector acting on the aircraft, is the acceleration due to gravity, is the transformation matrix from the ground coordinate system to the body axis system.

[0025] Furthermore, in step S1-3, the translational dynamics equation obtained in step S1-2 is converted into an inertial coordinate system:

[0026] ;

[0027] In the above formula, is the acceleration vector of the aircraft in the ground coordinate system, is the transformation matrix from the body axis system to the ground coordinate system.

[0028] Furthermore, the tensor-form dynamic equation of the whole machine rotation relative to the origin of the inertial coordinate system in step S1-4 is:

[0029] ;

[0030] In the above formula, is the total external torque on the entire aircraft, They are fuselage, Wing and canard and The angular velocity of the rotor relative to the ground coordinate system is They are fuselage, Wing and canard and The moment of inertia of the rotor relative to its own coordinate system.

[0031] Furthermore, the rotational dynamics equation in step S1-5 is simplified to:

[0032] .

[0033] Furthermore, the conversion step in step S1-6 is:

[0034] According to the angular velocity and attitude angle The conversion relationship is: ,

[0035] The simplified rotational dynamics equation obtained in S1-5 is transformed into:

[0036] ;

[0037] In the above formula, To express the total external torque on the entire aircraft, and are the direction cosine transformation matrix and its first-order derivative, respectively. and are the first and second order derivatives of the attitude angle respectively.

[0038] Furthermore, the present invention also discloses a control method for a distributed thrust tilt-wing aircraft, including the above-mentioned dynamic modeling method for the distributed thrust tilt-wing aircraft, and further including the following steps:

[0039] P1, the preset speed expectation value and position expectation value are compared with the speed vector calculated in real time by the time-varying multi-body dynamics tensor model established in step S1 P and position vector V After subtracting them, we get the error vector e , the resulting error vector e The L1 adaptive controller inputs to the outer loop trajectory tracking control module for control solution and outputs a virtual control signal U ;

[0040] P2, according to the virtual control signal U The definition of tension extraction is performed to obtain the expected total thrust T ;

[0041] P3, the virtual control signal obtained in step P1 U Input to the logic switching module, select different expected attitude acquisition methods according to different modes (VTOL vertical take-off and landing mode or Tilt tilt mode conversion stage), and finally output the expected attitude ;

[0042] P4, the desired posture obtained in step P3 The attitude angle calculated in real time by the time-varying multi-body dynamics tensor model established in step S1 Subtract to get the attitude error , attitude error Input to the robust error integral controller of the inner loop attitude control module to obtain the desired torque ;

[0043] P5: The expected total thrust of the 8 rotors obtained in step P2T and the desired torque obtained in step P4 Input to the distributed thrust control allocation module to obtain the rotor speed of each ducted fan ;

[0044] P6: The rotor speed of each ducted fan obtained in step P5 is Then input it into the time-varying multi-body dynamics tensor model established in step S1 to obtain the position vector P , velocity vector V and attitude angle The feedback is then fed back to the outer-loop trajectory tracking control module and the inner-loop attitude control module to iterate the cycle.

[0045] Furthermore, the outer loop trajectory tracking control process is a self-circulating process, and the specific processing method is:

[0046] The error vector obtained in real time is solved together with the adaptive estimated values ​​of the time-varying unknown coefficient matrix and the time-varying disturbance vector through the control law with low-pass filtering to obtain the outer loop control input vector. The outer loop control input vector, the time-varying unknown coefficient matrix and the adaptive estimated values ​​of the time-varying disturbance vector are then simultaneously input into the state observer to obtain the estimated value of the error vector. The estimated value of the error vector is subtracted from the error vector to obtain the estimated error, which is then input into the adaptive law to update the time-varying unknown coefficient matrix and the adaptive estimated values ​​of the time-varying disturbance vector, and the operation is repeated in an iterative cycle.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] 1. The structural weight layout design with the wing mass center-canard mass center, fuselage mass center and total mass center of the whole aircraft coincides reduces the difficulty of modeling and control, and uses the multi-body dynamics method to establish a tensor-form distributed tilt-wing aircraft fine dynamics model. Compared with the traditional single-body dynamics modeling method, the model accuracy is improved;

[0049] 2. The synchronous tilting trajectory of the wing and canard is designed to ensure a continuous and smooth transition during the tilting process;

[0050] 3. A hybrid control method is adopted, in the form of inner and outer loops, to connect the trajectory tracking controller and the attitude controller in series through the desired attitude switching logic, and to distribute the virtual control force and torque to each rotor in the form of rotation speed through the control distribution scheme, so as to adjust the attitude and position of the aircraft and achieve the purpose of trajectory tracking;

[0051] 4. Stable tracking control of attitude can be achieved through the robust error integration attitude controller. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1It is a control system framework diagram of the distributed thrust tilt-wing aircraft in the present invention;

[0053] Figure 2 It is a schematic diagram of the three-dimensional trajectory tracking effect of the circular reference trajectory in the present invention;

[0054] Figure 3 It is a schematic diagram of position tracking comparison of a circular reference trajectory in the present invention;

[0055] Figure 4 A schematic diagram of posture tracking comparison of a circular reference trajectory in the present invention;

[0056] Figure 5 It is a schematic diagram of three-dimensional trajectory tracking of the full-stage take-off, landing and tilting compound flight in the present invention;

[0057] Figure 6 It is a schematic diagram of position tracking comparison of the full-stage take-off, landing and tilting compound flight in the present invention;

[0058] Figure 7 It is a schematic diagram of attitude tracking comparison of full-stage take-off, landing and tilting compound flight in the present invention;

[0059] Figure 8 It is a schematic diagram of the left rotor speed distribution of the full-stage take-off, landing and tilting compound flight in the present invention;

[0060] Fig. 9 It is a schematic diagram of the right rotor speed distribution of the full-stage take-off, landing and tilt-rotation compound flight in the present invention. DETAILED DESCRIPTION

[0061] In order to facilitate the understanding of the present invention, the device of the present invention will be described more fully below with reference to the relevant drawings. Embodiments of the device are given in the drawings. However, the device can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0062] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "set" should be understood in a broad sense, for example, they can be fixedly connected or set, or detachably connected or set, or integrally connected or set. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0063] Example

[0064] This embodiment provides a dynamic modeling method for a distributed thrust tilt-wing aircraft, including:

[0065] S1. Establish a time-varying multi-body dynamic tensor model of a distributed thrust tilt-wing aircraft;

[0066] S1-1. Establish the general translational dynamics equation of the whole machine in the inertial system in tensor form. The specific process is as follows:

[0067] Based on the idea of ​​multi-body dynamics, each moving part on the aircraft is regarded as a separate rigid body. According to Newton's second law, the force analysis is performed on each rigid body. Finally, the general translational dynamic equation of the whole aircraft in the inertial system is obtained by superposition. Its tensor form is expressed as follows:

[0068] ;

[0069] In the formula, It represents the total external force on the whole aircraft. Respectively, body, Wing and canard and The rotor mass, Respectively represent the body, Wing and canard and The position of the center of mass of each rotor in the inertial coordinate system, They represent the body, Wing and canard and The linear velocity of the rotor relative to the inertial coordinate system is as follows: Then it represents the first-order rotation time derivative relative to the coordinate system “*”;

[0070] S1-2. Adjust the weight layout of the structure so that the center of mass of the wing and canard (i.e. the wing and canard are regarded as one, both as "wing", and the total center of mass of the two), the center of mass of the fuselage and the total center of mass of the whole aircraft coincide (i.e. the three centers are in one, and "three centers in one" is used below to represent the same coincidence description) to offset the eccentric forces of the wing, canard and rotor relative to the center of mass of the fuselage, and simplify the general translational dynamics equation to obtain the translational dynamics equation under the body axis system:

[0071] ;

[0072] In the above formula, is the total mass of the machine, They represent the projections of the body's center of mass velocity and acceleration vector on the body axis system, is the attitude angular velocity of the aircraft, is the sum of all rotor thrusts, is the synchronous tilt angle of the wing and canard, is the aerodynamic force vector acting on the aircraft, is the acceleration due to gravity, is the transformation matrix from the ground coordinate system to the body axis system;

[0073] S1-3. To facilitate control law design and numerical simulation, the translational dynamics equation obtained in S1-2 is converted to the inertial coordinate system:

[0074] ;

[0075] In the above formula, is the acceleration vector of the aircraft in the ground coordinate system, is the transformation matrix from the body axis system to the ground coordinate system;

[0076] S1-4. Based on the idea of ​​multi-body dynamics and the momentum theorem, superimpose the coordinates of each moving part relative to the origin of the inertial coordinate system. The momentum of the whole machine relative to the origin of the inertial coordinate system is obtained The angular momentum of , the rotational dynamics equation is established:

[0077] ;

[0078] In the above formula, is the total external torque on the entire aircraft, They are fuselage, Wing and canard and The angular velocity of the rotor relative to the ground coordinate system is They are fuselage, Wing and canard and The moment of inertia of each rotor relative to its own coordinate system;

[0079] S1-5. After expanding, deriving and simplifying the rotational dynamics equation obtained in S1-4, we obtain the rotational dynamics equation in tensor form:

[0080] ;

[0081] In the above formula, Represents the center of mass of the whole machine relative to the body The momentum of is the angular velocity of the aircraft The first derivative of ;

[0082] The moment of inertia of the whole machine is: ;

[0083] In the above formula, the subscript * indicates the active part*, Represents the inertia offset matrix of component "*";

[0084] The gyroscopic torque generated by the relative motion of the moving parts includes and ,Right now:

[0085] ;

[0086] is the gyroscopic torque caused by the high-speed rotating rotor and tilting wing canard as the aircraft attitude changes:

[0087] + ;

[0088] In the above formula, represents the wing and canard tilt angular rate vector, is the wing and canard tilt angle The first derivative of ;

[0089] is the gyroscopic torque caused by the high-speed rotation of the rotor, the random wing and the canard wing:

[0090] ;

[0091] In the above formula, For the The rotor angular velocity, For the The rotation speed of each rotor;

[0092] It represents the dynamic characteristics caused by the rotation of active parts such as wings, canards and rotors around their own axes:

[0093] ;

[0094] is the additional eccentric moment caused by the moving parts such as wings, canards and rotors and coupled with the translational dynamics:

[0095] ;

[0096] In the above formula, is the relative position of the center of mass of the wing, canard and rotor moving parts;

[0097] Because of the "three centers in one", that is, the center of mass of the wing and canard, the center of mass of the fuselage and the total center of mass of the whole aircraft coincide, The terms in are symmetrically cancelled, that is, , the rotational dynamics equation can be simplified to:

[0098] .

[0099] S1-6, according to the angular rate and attitude angle The conversion relationship , the rotational dynamics equation is transformed into:

[0100] ;

[0101] In the above formula, To express the total external torque on the entire aircraft, and are the direction cosine transformation matrix and its first-order derivative, respectively. and are the first and second order derivatives of the attitude angle respectively;

[0102] Compared with the single rigid body dynamics equation of general fixed-wing aircraft, the transformed rotational dynamics equation has the additional anti-torque caused by the rotation of the moving parts. , and taking into account the gyroscopic torque Therefore, the accuracy of the model is improved.

[0103] It should be noted that in this modeling step, under the premise of ensuring that the real dynamic characteristics are reflected as much as possible, unnecessary modeling details should be reasonably ignored and the following assumptions should be made:

[0104] Assume that the fuselage, wings, canards, rotors and other components are rigid bodies, and ignore their elastic deformation; assume that the earth's surface has no curvature, that is, the flat earth assumption, and ignore its rotation; assume that the extension lines of the wing and canard rotation axes pass through the center of mass of each rotor; assume that there is no lag in the rotation response of the wing, canard and rotor.

[0105] S2. Designing a synchronous tilting smooth trajectory of the wing and the canard during the mode conversion stage, specifically comprising the following steps:

[0106] The tilt angular acceleration in the transition from vertical takeoff to level flight is:

[0107] ;

[0108] The tilt angular velocity in the transition from vertical takeoff to level flight is:

[0109] ;

[0110] The tilt angle in the transition from vertical takeoff to level flight is:

[0111] ;

[0112] The tilt angular acceleration in the state of level flight to vertical take-off is:

[0113] ;

[0114] The tilt angular velocity in the state of level flight to vertical take-off is:

[0115] ;

[0116] The tilt angle in the transition from level flight to vertical take-off is:

[0117] ;

[0118] In the above formula, express The tilt angular acceleration between The time taken for this segment to tilt; express The tilt angular acceleration between

[0119] It should be noted that in this process, the wing and canard need to rotate synchronously along the desired tilt trajectory without lag. At the same time, in order to ensure a continuous and smooth transition during the tilt process, the tilt motion of the wing and canard is divided into an acceleration motion starting from rest and a deceleration motion stopping at the target angle. In addition, according to different states, the tilt motion is divided into two states: vertical take-off to level flight and level flight to vertical take-off. At the same time, in order to improve the lift coefficient during level flight, the tilt motion is retained. The wing and canard installation angles are set to Between (and including and ), tilt angle It is in the vertical state.

[0120] Figure 1 As shown, based on the dynamic modeling of the above-mentioned distributed thrust tilt-wing aircraft, a control method for the distributed thrust tilt-wing aircraft is also disclosed, and the control method adopts a hybrid control framework of inner and outer loops in series, including an outer loop trajectory tracking control module, an inner loop attitude control module, a logic switching module and a distributed thrust control allocation module, and finally the above control method is simulated and verified by designing a reference trajectory;

[0121] The outer loop trajectory tracking control is implemented by an L1 adaptive controller, which includes: a state observer, an adaptive law, and a control law with a low-pass filter;

[0122] Specifically, the design of the state observer is:

[0123] ;

[0124] In the above formula, and They are The estimated value of and its first-order derivative, , , , and is the position vector and velocity vector The expected value of is the outer loop control input vector, They are respectively the time-varying unknown coefficient matrices and the time-varying disturbance vector The corresponding adaptive estimate,

[0125] , is the third-order identity matrix, is a positive definite diagonal matrix;

[0126] Specifically, the design of the adaptive law is:

[0127] ;

[0128] ;

[0129] In the above formula, for The first derivative of Characterizing the projection operator, estimating the error , is the adaptive gain, is a positive symmetric matrix, according to the Lyapunov stability equation get, take is the 6th-order identity matrix;

[0130] Specifically, the control law with low-pass filtering is designed as:

[0131] ;

[0132] In the above formula, is the feedback gain diagonal matrix, for The Laplace transform of ;

[0133] According to the above controller design, the virtual control signal can be obtained for:

[0134] ,

[0135] At the same time, according to the virtual control signal Definition:

[0136] ,

[0137] Taking the square of both sides of the equation, we can see that the expected total thrust T of the 8 ducted fan rotors is:

[0138] ;

[0139] The inner loop attitude control is implemented by a robust error integral controller, and its design process is as follows:

[0140] First, according to the attitude angle , define the attitude error:

[0141] ,

[0142] Design the following control law:

[0143] ,

[0144] The auxiliary function is defined as follows:

[0145] ,

[0146] ,

[0147] ,

[0148] In the above formula, is a positive definite constant-valued diagonal matrix.

[0149] Finally, the desired torque input into the mathematical model is obtained :

[0150] ;

[0151] A logic switching module, including an attitude extraction rule for the vertical take-off and landing mode and a zero attitude fixing rule for the tilt mode transition phase;

[0152] The distributed thrust control allocation module is used to solve the control allocation problem under constraints by using the redistribution weighted pseudo-inverse method to solve the problem of severe coupling between control channels due to the fact that each rotor participates in the control of all channels. The specific design steps are as follows:

[0153] First, assuming the ducted thrust and reaction torque Rotor speed The square of is linearly proportional to:

[0154] ,

[0155] In the above formula, Respectively The thrust and reaction torque of each rotor, For the The rotation speed of the rotor, are the tension coefficient and torque coefficient respectively,

[0156] Since the rotors are installed symmetrically relative to the fuselage, the two rotors in symmetrical positions are regarded as a group, and the thrust between the ducted fans in the same group is Difference and mean As a new variable, defined as:

[0157] ;

[0158] ;

[0159] The control equations of the rotor can be expressed as four sets of equations:

[0160] ;

[0161] in, , , Indicates The Y displacement of the rotor center of mass relative to the fuselage center of mass, , ;

[0162] ;

[0163] in, , , , and They represent the X-direction displacement and Z-direction displacement of the canard mass center relative to the fuselage mass center, respectively. and They represent the X-direction displacement and Z-direction displacement of the wing mass center relative to the fuselage mass center, respectively. ;

[0164] Considering the saturation problem of the actuator (rotor speed), the distribution problem of distributed thrust can be described in the following mathematical form:

[0165] ,

[0166] In the above formula, is the unknown output quantity, is a positive definite symmetric weight matrix, is the coefficient (efficiency) matrix, is the known desired control quantity.

[0167] Without considering the actuator saturation, the optimal solution of the above formula is:

[0168] , where:

[0169] ;

[0170] ;

[0171] ;

[0172] Since the outer rotor is farther away from the attitude axis, the torque generated on the fuselage at the same speed is greater. Therefore, the outer rotor is allocated first during attitude control to increase its weight ratio coefficient during control allocation. According to this principle, the weight matrix is ​​taken as:

[0173] , where:

[0174] ;

[0175] ;

[0176] However, the above solution cannot solve the problem of actuator saturation. Therefore, the redistribution method is used. After subtracting the saturated amount, the remaining amount to be allocated is redistributed to the unsaturated mechanism for execution until the allocation is completed or the actuator is fully saturated. The specific implementation steps are as follows:

[0177] In the first step, the weighted pseudo-inverse method is used to find the optimal allocation solution;

[0178] The second step is to determine whether there is saturation of the actuator. If so, execute the third step. If not, obtain the optimal solution.

[0179] The third step is to record the variables corresponding to the actuator that has reached saturation. and coefficient , the corresponding variable of the unsaturated actuator is recorded as , the coefficient is recorded as , then the control quantity achieved at this time is , the remaining ;

[0180] The fourth step is to continue to use the weighted pseudo-inverse method to solve the new allocation problem Solve and return to step 2.

[0181] The above solution results can be converted to finally obtain the rotor speed of 8 ducted fans.

[0182] It should be noted that due to the signed function used in the inner loop attitude control law Therefore, the output control quantity may have high-frequency jitter. Therefore, a second-order low-pass filter can be used to filter the control signal before control distribution.

[0183] Based on the above-mentioned control method of the distributed thrust tilt-wing aircraft, the specific steps are as follows:

[0184] P1, the preset speed expectation value and position expectation value are compared with the speed vector calculated in real time by the time-varying multi-body dynamics tensor model established in step S1 P and position vector V After subtracting them, we get the error vector e , the resulting error vector e The L1 adaptive controller inputs to the outer loop trajectory tracking control module for control solution and outputs a virtual control signal U The specific processing method is: first, the error vector obtained in real time e With the time-varying unknown coefficient matrix and the time-varying disturbance vector The adaptive estimate of ) together with the control law with low-pass filtering to obtain the outer loop control input vector , continue to At the same time, the state observer is input to obtain the error vector e Estimated value of , the error vector e Estimated value of With error vector e After subtraction, we get the estimated error , and then input the adaptive law to update the time-varying unknown coefficient matrix and the time-varying disturbance vector The adaptive estimate of ), in order to iterate the cycle operation;

[0185] P2, according to the virtual control signal U The definition of tension extraction is performed to obtain the expected total thrust T ;

[0186] P3, the virtual control signal obtained in step P1 U Input to the logic switching module, select the posture extraction rule or zero posture fixing rule according to the different modes to obtain the desired posture The specific processing process is as follows:

[0187] 1) In vertical take-off and landing mode, the desired attitude input of the inner loop is transmitted through the virtual control signal U Extraction is obtained, and the posture extraction rules are as follows:

[0188] make , take the desired heading angle , then the expected roll angle and the desired pitch angle From the following we can get:

[0189] ,

[0190] ;

[0191] 2) During the mode conversion stage, the wings and canards are constantly tilting, and the dynamic equations of the aircraft are constantly changing. In order to reduce the risk of flight accidents and reduce the difficulty of control, it is set that the aircraft attitude does not change and remains at zero during the mode conversion stage, and the zero attitude fixed rule is maintained. The expected attitude input is:

[0192] ,

[0193] That is, the aircraft flight is not controlled by attitude adjustment, but only relies on the tension adjustment of the ducted fan to balance the force changes in this stage;

[0194] P4, the desired posture obtained in step P3 The attitude angle calculated in real time by the time-varying multi-body dynamics tensor model established in step S1 Subtract to get the attitude error , attitude error Input to the robust error integral controller of the inner loop attitude control module to obtain the desired torque ;

[0195] P5: The expected total thrust of the eight ducted fan rotors obtained in step P2 T and the desired torque obtained in step P4 Input to the distributed thrust control allocation module to obtain the Rotor speed (That is, the rotor speed of each ducted fan is obtained );

[0196] P6, the speed obtained in step P5 Then input it into the time-varying multi-body dynamics tensor model established in step S1 to obtain the position vector P , velocity vector V and attitude angle , and feed back to the outer loop trajectory tracking control module and the inner loop attitude control module, and repeat the cycle;

[0197] Design flight reference trajectories for simulation testing to verify the trajectory tracking and attitude control performance of the above control method under different modes: The simulation verification system is used to simulate and verify the designed reference trajectories. Four reference trajectories are designed, namely, circular reference trajectory, vertical take-off reference trajectory, tilt flight reference trajectory and vertical descent reference trajectory, specifically:

[0198] Circular reference trajectory (used to test trajectory tracking performance in vertical take-off and landing mode):

[0199] ,

[0200] Pick , ;

[0201] The vertical take-off reference trajectory (used to achieve vertical take-off in vertical take-off and landing mode) is designed to ensure that the initial and final velocities of the aircraft in the vertical take-off and landing mode are both zero during the vertical take-off phase. The vertical take-off velocity of the sine function is converted into acceleration and displacement through integration and differentiation as follows:

[0202] ,

[0203] In the above formula, Total time for vertical lifting, for The vertical take-off target altitude reached at the moment is also the reference altitude for fixed altitude flight in the mode conversion phase. The reference displacement, velocity and acceleration in other directions are kept at 0;

[0204] Tilt flight reference trajectory (used to achieve constant altitude and stable forward flight in the mode conversion phase), design forward flight acceleration for:

[0205] ,

[0206] In the above formula, is the total aerodynamic drag acting on the aircraft, is the total aerodynamic lift acting on the aircraft, both of which depend on the wing / canard tilt angle and forward speed The forward flight speed in the mode conversion phase is obtained by integration. and reference displacement The reference displacement, velocity and acceleration in other directions remain at 0;

[0207] Descent reference trajectory (used for landing at a low speed before the end of the simulation in vertical take-off and landing mode, until touching the ground), setting the vertical acceleration during landing , descent speed and reference height for:

[0208] ,

[0209] In the above formula, is the altitude of the aircraft when landing begins.

[0210] It should be noted that in vertical take-off and landing mode, the wing and canard tilt angles should be kept at After setting up the simulation environment, the circular reference trajectory is simulated. The tracking effect is as follows: Figure 2-4 As shown;

[0211] Then, the full-stage take-off, landing and tilt-rotation composite flight is simulated. The aircraft first takes off vertically from the ground to a specified height according to the vertical take-off reference trajectory, hovers for a period of time, and then enters the mode conversion stage. The wings and canards begin to tilt synchronously from 90° to 15°, maintain the attitude for a short glide, and then turn back to 90° from 15°. During the mode conversion, it flies according to the tilt-rotation flight reference trajectory, switches to the vertical take-off and landing mode after aerodynamic deceleration, and lands vertically on the ground according to the vertical descent reference trajectory. The tracking effect is as follows Figure 6-Figure 9 shown.

[0212] It should be noted that the structure described in the present invention can be implemented in a variety of different forms and is not limited to the described embodiments. Any equivalent transformations made by ordinary technicians in this field using the contents of the present invention description and drawings, or directly or indirectly applied to other related technical fields, such as the loading and unloading of other items, are included in the protection scope of the present invention.

Claims

1. A dynamic modeling method for a distributed thrust tilt-wing aircraft, characterized by: The following steps are involved: S1. Establish a time-varying multi-body dynamic tensor model for a distributed thrust tilt-wing aircraft; S1-1. Establish the general translational dynamic equation of the whole machine in the inertial system in tensor form; S1-2. Adjust the structural weight distribution so that the center of mass of the wing and canard, the center of mass of the fuselage, and the total center of mass of the entire aircraft coincide, offsetting the eccentric forces of the wing / canard / rotor relative to the total center of mass of the entire aircraft. Determine the simplified translational dynamic equation for the body axis system according to step S1-1. S1-3. Transform the above equation into an inertial coordinate system to obtain the final translational dynamic equation after transformation; S1-4. Establish the tensor-form dynamic equations of the entire machine's rotation relative to the origin of the inertial coordinate system; S1-5. Based on the structural weight layout adjustment results of S1-2, the additional eccentric moments caused by the moving parts of the wings, canards, and rotors and coupled with the translational dynamics are offset, and the simplified rotational dynamic equations are obtained according to step S1-4; S1-6. Transform the above equation according to the conversion relationship between angular rate and attitude angle to obtain the final rotational dynamics equation after transformation; S2. Design the smooth trajectory of the wing and canard synchronous tilting during the modal conversion phase. Includes the following designs: S2-1. Design of the tilt angular acceleration from vertical takeoff to level flight; S2-2. Design of the tilt angular velocity from vertical takeoff to level flight and from level flight to vertical takeoff; S2-3. Design of the tilt angle from vertical takeoff to level flight and from level flight to vertical takeoff.

2. The dynamic modeling method for a distributed thrust tilt-wing aircraft according to claim 1, characterized in that: In step S1-1, the specific process of establishing the general translational dynamic equation of the entire machine in the inertial system in tensor form is as follows: Based on the concept of multi-body dynamics, each moving part on the aircraft is treated as a separate rigid body. According to Newton's second law, the force analysis of each rigid body is performed. Finally, the general translational dynamic equation of the entire aircraft in the inertial system is obtained by superposition. Its tensor form is expressed as follows: ； In the above formula, It represents the total external force acting on the entire aircraft. The body, Wing and canard and The rotor mass, Respectively represent the body, Wing and canard and The position of the center of mass of each rotor in the inertial coordinate system, They represent the body, Wing and canard and The linear velocity of the rotor relative to the inertial coordinate system is as follows: It represents the first-order rotation time derivative relative to the coordinate system "*".

3. The dynamic modeling method of a distributed thrust tilt-wing aircraft according to claim 2, characterized in that: In step S1-2, the structural weight distribution is adjusted so that the center of mass of the wing and canard, the center of mass of the fuselage, and the total center of mass of the entire aircraft coincide. The general translational dynamics equation is simplified to obtain the translational dynamics equation under the body axis system: ； In the above formula, is the total mass of the machine, They represent the projection of the body's center of mass velocity and acceleration vector on the body axis system, is the attitude angular velocity of the aircraft, is the sum of all rotor thrusts, is the synchronous tilt angle of the wing and canard, is the aerodynamic force vector acting on the aircraft, is the acceleration due to gravity, is the transformation matrix from the ground coordinate system to the body axis system.

4. The dynamic modeling method for a distributed thrust tilt-wing aircraft according to claim 3, characterized in that: In step S1-3, the translational dynamics equation obtained in step S1-2 is converted into an inertial coordinate system: ； In the above formula, is the acceleration vector of the aircraft in the ground coordinate system, is the transformation matrix from the body axis system to the ground coordinate system.

5. The dynamic modeling method of a distributed thrust tilt-wing aircraft according to claim 4, characterized in that: The tensor-form dynamic equation of the whole machine rotation relative to the origin of the inertial coordinate system in step S1-4 is: ； In the above formula, is the total external torque acting on the entire aircraft, They are fuselage, Wing and canard and The angular velocity of each rotor relative to the ground coordinate system, They are fuselage, Wing and canard and The moment of inertia of the rotor relative to its own coordinate system.

6. The dynamic modeling method of a distributed thrust tilt-wing aircraft according to claim 5, characterized in that: The rotational dynamics equation in step S1-5 is simplified to: 。 7. The method for dynamic modeling of a distributed thrust tilt-wing aircraft according to claim 6, characterized in that: The conversion step in step S1-6 is: According to the angular rate and attitude angle The conversion relationship is: , The simplified rotational dynamics equation obtained in S1-5 is transformed into: ； In the above formula, To express the total external torque acting on the entire aircraft, and are the direction cosine transformation matrix and its first-order derivative, and are the first and second order derivatives of the attitude angle respectively.

8. A control method for a distributed thrust tilt-wing aircraft, comprising the dynamic modeling method for a distributed thrust tilt-wing aircraft according to any one of claims 1 to 7, further comprising the following steps: P1, the preset speed expectation value and position expectation value are compared with the speed vector calculated in real time by the time-varying multi-body dynamics tensor model established in step S1 P and position vector V After subtracting them respectively, we get the error vector e , the error vector obtained e The L1 adaptive controller inputs to the outer loop trajectory tracking control module for control solution and outputs a virtual control signal U ; P2, according to the virtual control signal U The definition of tension extraction is performed to obtain the expected total thrust T ; P3, the virtual control signal obtained in step P1 U Input to the logic switching module, according to the different modes, that is, vertical take-off and landing mode or tilt mode conversion stage, select different desired attitude acquisition methods, and finally output the desired attitude ; P4, the desired posture obtained in step P3 The attitude angle calculated in real time by the time-varying multi-body dynamics tensor model established in step S1 Subtract to get the attitude error , attitude error Input to the robust error integral controller of the inner loop attitude control module to obtain the desired torque ; P5, the expected total thrust of the 8 rotors obtained in step P2 T and the desired torque obtained in step P4 Input to the distributed thrust control allocation module to obtain the rotor speed of each ducted fan ; P6: The rotor speed of each ducted fan obtained in step P5 is Then input it into the time-varying multi-body dynamics tensor model established in step S1 to obtain the position vector P , velocity vector V and attitude angle The feedback is then fed back to the outer-loop trajectory tracking control module and the inner-loop attitude control module for an iterative cycle.

9. The control method of a distributed thrust tilt-wing aircraft according to claim 8, characterized in that: The outer loop trajectory tracking control process is a self-circulating process, and the specific processing method is as follows: The error vector obtained in real time is solved together with the adaptive estimated values ​​of the time-varying unknown coefficient matrix and the time-varying disturbance vector through the control law with low-pass filtering to obtain the outer loop control input vector. The outer loop control input vector, the time-varying unknown coefficient matrix and the adaptive estimated values ​​of the time-varying disturbance vector are then input into the state observer at the same time to obtain the estimated value of the error vector. The estimated value of the error vector is subtracted from the error vector to obtain the estimated error. The adaptive law is then input to update the adaptive estimated values ​​of the time-varying unknown coefficient matrix and the time-varying disturbance vector, and the operation is repeated in this iterative cycle.

Images