LF-STOVL engine multi-nozzle thrust matching control method and system

Through the six-degree-of-freedom STOVL aircraft model and multivariable controller, the problem of multi-nozzle thrust matching of the STOVL engine was solved, rapid thrust matching and balance control under different flight conditions were achieved, and engine performance was improved.

CN118838160BActive Publication Date: 2025-09-05NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202410809970.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-06-21
Publication Date
2025-09-05
Estimated Expiration
2044-06-21

AI Technical Summary

Technical Problem

Existing technologies cannot effectively achieve matching control of the multi-nozzle thrust of STOVL engines when the flight attitude and flight conditions change, especially the multi-variable control instructions of lift fan type STOVL engines are insufficiently scheduled, resulting in poor thrust control.

Method used

A six-degree-of-freedom STOVL aircraft model is used to generate multi-nozzle thrust commands, and closed-loop control is performed through three-variable block TVB and single-variable block SVB controllers. A direct thrust controller is designed by combining the linear matrix inequality method with the optimization algorithm, and the actual thrust is estimated using a dynamic deep neural network to achieve multi-nozzle thrust matching.

Benefits of technology

It achieves rapid matching of multi-nozzle thrust under different flight attitudes and conditions, maintains aircraft balance, suppresses control loop coupling, and improves the overall performance of the engine.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118838160B_ABST
    Figure CN118838160B_ABST
Patent Text Reader

Abstract

This invention discloses a multi-nozzle thrust matching control method for an LF-STOVL (lift fan short takeoff / vertical landing) engine. First, based on flight attitude and flight conditions, and by solving a six-degree-of-freedom STOVL aircraft model, a core thrust command for the three-bearing vector tail nozzle, a lift fan thrust command for the lift fan nozzle, and roll thrust commands for the left and right roll nozzles are generated. Closed-loop control is then performed on the main fuel flow rate, tail nozzle throat area, lift fan nozzle outlet area, and roll nozzle outlet area based on these thrust commands and actual thrust. The invention also discloses a multi-nozzle thrust matching control system for an LF-STOVL engine. Compared to existing technologies, this invention effectively achieves multi-nozzle thrust matching control in STOVL mode as flight attitude and flight conditions change.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to an aero-engine control method, and in particular to a multi-nozzle thrust matching control method and system for an LF-STOVL engine. Background Art

[0002] Short Take Off / Vertical Landing (STOVL) aircraft combine the short takeoff and hover capabilities of rotary-wing aircraft with the long range, high speed, and maneuverability of fixed-wing aircraft. They hold significant military value, and my country has been committed to the development of STOVL engines.

[0003] STOVL engines must not only provide thrust for conventional missions but also sufficient lift for short takeoff and vertical landing (STOVL). This significantly increases the demand for multi-nozzle thrust-matching control, particularly for lift-fan STOVL (LF-STOVL) engines, which utilize a three-bearing vectoring tail nozzle, a lift fan, and left and right roll nozzles to provide both thrust and lift. Due to their complex structure and numerous adjustable actuators, single-variable control is no longer sufficient to fully leverage the various components and achieve optimal integration between aerodynamics and propulsion systems. Advanced multivariable control is required to improve overall engine performance.

[0004] However, previous research on STOVL engine control has not addressed the scheduling of control commands for multivariable control systems. Given that the primary function of an aircraft engine is to provide the required thrust for the aircraft, and that STOVL engine thrust is generated by multiple components, matching thrust control commands in the STOVL control system is crucial. Existing technologies can provide appropriate thrust commands for traditional engines with a single nozzle, but for STOVL engines, they are unable to provide commands that match the coordination requirements of multiple nozzles. Summary of the Invention

[0005] The purpose of the present invention is to overcome the shortcomings of the existing technology and provide a multi-nozzle thrust matching control method for an LF-STOVL engine to effectively realize multi-nozzle thrust matching control when the flight attitude and flight conditions change in STOVL mode.

[0006] The present invention specifically adopts the following technical solutions to solve the above technical problems:

[0007] The LF-STOVL engine multi-nozzle thrust matching control method first generates the core thrust command at the three-bearing vector tail nozzle, the lift fan thrust command at the lift fan nozzle, and the roll thrust command of the left and right roll nozzles according to the flight attitude and flight conditions by solving the six-degree-of-freedom STOVL aircraft model; then the main fuel flow rate W is adjusted according to the thrust command and the actual thrust. fm , tail nozzle throat area A8, lift fan nozzle outlet area A lf and the tumble nozzle exit area A gz Perform closed-loop control; the six-degree-of-freedom STOVL aircraft model is specifically shown in the following formula:

[0008]

[0009] Where, F cnr 、F lfr 、F gz1r 、F gz2r Respectively represent the core thrust command, lift fan thrust command and the roll thrust command of the left and right roll nozzles; F gztr The vector sum of the roll thrust commands of the left and right roll nozzles; δ cn and δ lf Core thrust F cn and lift fan thrust F lf The angle between the projection on the plane symmetry plane Oxz and the x-axis, when the projection is below the x-axis, δ cn and δ lf is positive; cny Core thrust F cn The angle between the projection on the Oxy plane and the symmetry plane Oxz, when the projection is on the left side of the symmetry plane Oxz, δ cny is positive; x, y, z are the vector force arms of the corresponding thrust; subscripts cn, lf, gz1, gz2 represent the three-bearing vector tail nozzle, lift fan nozzle and left and right roll nozzle respectively; subscripts x, y, z represent the three directional components in the body coordinate system respectively; subscript I represents the engine inlet; F I 、F p Respectively represent the engine inlet resistance and the total thrust generated by the four nozzles of the engine; L p 、M p 、N p It represents the three directions of thrust moments caused by the engine thrust in the body coordinate system, namely rolling moment, pitching moment and yaw moment.

[0010] Furthermore, the controller for executing the closed-loop control is composed of a three-variable block TVB controller and a single-variable block SVB controller; in the TVB controller, r m =[F cnr ,A8r ,F lfr ] T For input command, y m =[F cn ,A8,F lf ] T is the actual output of the engine, and the error e is obtained by making the difference between the two m =r m -y m Obtained through the integration phase

[0011] ∫e m dt,∫e m dt and the state quantity x of the TVB controller output by the engine m =[W fm ,A8,A lf ,N1,N2] T After the controller gain is applied, we can get Finally, u mc =[W fmr ,A 8r ,A lfr ] T The main fuel flow command, tail nozzle throat area command, and lift fan nozzle outlet area command act on the engine, where N1 and N2 are the low pressure shaft speed and high pressure shaft speed respectively; in the SVB controller, r s =F gzr For input command, y s =F gz is the actual output of the engine, and the error e is obtained by making the difference between the two s =r s -y s After the integration step, we get ∫e s dt, the state quantity and error integral of the TVB controller are added to the state quantity of the SVB controller to obtain the state quantity of the SVB controller ∫e s dt and x s After the controller gain is applied, we can get Finally, u sc =A gzr The left and right roll nozzle area instructions act on the engine.

[0012] Preferably, the actual thrust is estimated by a dynamic deep neural network thrust estimator based on similarity transformation.

[0013] Based on the same inventive concept, the following technical solutions can also be obtained:

[0014] The LF-STOVL engine multi-nozzle thrust matching control system includes a thrust instruction solver module and a controller; the thrust instruction solver module is used to generate the core thrust instruction at the three-bearing vector tail nozzle, the lift fan thrust instruction at the lift fan nozzle, and the roll thrust instruction of the left and right roll nozzles according to the flight attitude and flight conditions by solving the six-degree-of-freedom STOVL aircraft model; the controller is used to adjust the main fuel flow W according to the thrust instruction and the actual thrust. fm , tail nozzle throat area A8, lift fan nozzle outlet area A lf and the tumble nozzle exit area A gz Perform closed-loop control; the six-degree-of-freedom STOVL aircraft model is specifically shown in the following formula:

[0015]

[0016] Where, F cnr 、F lfr 、F gz1r 、F gz2r Respectively represent the core thrust command, lift fan thrust command and the roll thrust command of the left and right roll nozzles; F gztr The vector sum of the roll thrust commands of the left and right roll nozzles; δ cn and δ lf Core thrust F cn and lift fan thrust F lf The angle between the projection on the plane symmetry plane Oxz and the x-axis, when the projection is below the x-axis, δ cn and δ lf is positive; cny Core thrust F cn The angle between the projection on the Oxy plane and the symmetry plane Oxz, when the projection is on the left side of the symmetry plane Oxz, δ cny is positive; x, y, z are the vector force arms of the corresponding thrust; subscripts cn, lf, gz1, gz2 represent the three-bearing vector tail nozzle, lift fan nozzle and left and right roll nozzle respectively; subscripts x, y, z represent the three directional components in the body coordinate system respectively; subscript I represents the engine inlet; F I 、F p Respectively represent the engine inlet resistance and the total thrust generated by the four nozzles of the engine; L p 、M p 、N p It represents the three directions of thrust moments caused by the engine thrust in the body coordinate system, namely rolling moment, pitching moment and yaw moment.

[0017] Furthermore, the controller is composed of a three-variable block TVB controller and a single-variable block SVB controller; in the TVB controller, rm =[F cnr ,A 8r ,F lfr ] T For input command, y m =[F cn ,A8,F lf ] T is the actual output of the engine, and the error e is obtained by making the difference between the two m =r m -y m After the integration step, we get ∫e m dt,∫e m dt and the state quantity x of the TVB controller output by the engine m =[W fm ,A8,A lf ,N1,N2] T After the controller gain is applied, we can get Finally, u mc =[W fmr ,A 8r ,A lfr ] T The main fuel flow command, tail nozzle throat area command, and lift fan nozzle outlet area command act on the engine, where N1 and N2 are the low pressure shaft speed and high pressure shaft speed respectively; in the SVB controller, r s =F gzr For input command, y s =F gz is the actual output of the engine, and the error e is obtained by making the difference between the two s =r s -y s After the integration step, we get ∫e s dt, the state quantity and error integral of the TVB controller are added to the state quantity of the SVB controller to obtain the state quantity of the SVB controller ∫e s dt and x s After the controller gain is applied, we can get Finally, u sc =A gzr The left and right roll nozzle area instructions act on the engine.

[0018] Preferably, the actual thrust is estimated by a dynamic deep neural network thrust estimator based on similarity transformation.

[0019] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0020] Aiming at the special structure of multiple nozzles in the LF-STOVL propulsion system, the present invention establishes a six-degree-of-freedom LF-STOVL aircraft model to generate thrust control instructions for each nozzle according to the aircraft's flight attitude, thus achieving thrust matching of multiple nozzles to maintain the balance of the aircraft body.

[0021] The present invention further combines the linear matrix inequality (LMI) method with an optimization algorithm to design a direct thrust controller, which can achieve coupling suppression between control loops and fast tracking of thrust commands. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a schematic diagram of the LF-STOVL propulsion system structure distribution;

[0023] Figure 2 This is the thrust command model structure diagram;

[0024] Figure 3 It is the structure diagram of the decoupling control system;

[0025] Figure 4 This is the structural diagram of the multi-nozzle thrust matching control system for the STOVL engine;

[0026] Figure 5 The control system response results when the roll angle command changes at the working point H = 0 km, Ma = 0, (a) is the response curve of the controlled quantity thrust; (b) is the response curve of the controlled quantity;

[0027] Figure 6 The control system responds to changes in the pitch angle command at the operating point H = 0 km and Ma = 0. (a) is the response curve of the controlled quantity thrust; (b) is the response curve of the controlled quantity.

[0028] Figure 7 The control system response results when the roll angle command changes at the working point H = 1.6 km and Ma = 0.24. (a) is the response curve of the controlled quantity thrust; (b) is the response curve of the controlled quantity;

[0029] Figure 8 The control system responds to changes in the pitch angle command at the operating point H = 1.6 km and Ma = 0.24. (a) is the response curve of the controlled quantity thrust; (b) is the response curve of the controlled quantity. DETAILED DESCRIPTION

[0030] In response to the technical problem that the existing STOVL engine multi-nozzle thrust control cannot match the flight attitude, the solution of the present invention is to establish an LF-STOVL six-degree-of-freedom aircraft model to obtain thrust instructions that match the flight conditions and flight attitude, and further combine the linear matrix inequality (LMI) method with an optimization algorithm to design a direct thrust controller to achieve rapid tracking of the thrust instructions.

[0031] To facilitate public understanding, the technical solution of the present invention is described in detail below through a preferred embodiment with reference to the accompanying drawings:

[0032] like Figure 1 As shown in the figure, in STOVL mode, the propulsion system generates thrust and lift through four nozzles: core thrust from the three-bearing vectoring tail nozzle, lift fan thrust from the lift fan nozzle, and roll thrust from the left and right roll nozzles. These four thrusts directly affect the aircraft's pitch, yaw, and roll angles. The three-bearing vectoring tail nozzle can deflect downward by 0° to 95° within the aircraft's plane of symmetry and can also deflect laterally by ±12°. The lift fan, driven by the engine's low-pressure shaft, can deflect downward by 34° to 95° within the aircraft's plane of symmetry. The two roll nozzles, extending from the engine duct, are used for roll control. Therefore, adjusting the thrust and deflection angles of the tail nozzle and lift fan enables pitch control, while adjusting the thrust of the two roll nozzles enables roll control.

[0033] The LF-STOVL six-degree-of-freedom aircraft model consists of four modules, the structure is as follows Figure 2 As shown, based on the flight conditions (altitude H, Mach number Ma) and flight attitude (roll angle φ, pitch angle θ, yaw angle ψ), the thrust instructions for each nozzle are obtained. Among them, the aerodynamic module is used to calculate the aerodynamic force and aerodynamic torque acting on the aircraft; the particle motion module and the angular motion module are used to calculate the external force and external torque required by the aircraft; the thrust instruction solution module calculates the thrust instructions of the engine tail nozzle, lift fan and roll nozzle according to the force and torque balance of the aircraft. The flight attitude angle includes the Euler angle θ and ψ. The Euler angle is between the body coordinate system Oxyz and the ground coordinate system O g x g y g z g between.

[0034] The aerodynamic module determines the aerodynamic force and aerodynamic torque acting on the aircraft based on the aircraft's flight state and flight attitude. Σ Along the airflow coordinate system Ox a y a za It can be decomposed into drag D, side force Y and lift L. Total aerodynamic moment M Σ Decomposed into the rolling moment L around the x-axis along the body coordinate system a , pitch moment M around the y-axis a and the yaw moment N about the z axis a , the polarity of the torque in the three directions is determined by the right-hand rule. Σ Decomposing along the airflow coordinate system yields

[0035]

[0036] Among them, c D 、c Y and c L They are respectively the drag coefficient, side force coefficient and lift coefficient, dynamic pressure Q = ρV 2 / 2, ρ and V are air density and airspeed respectively, S is the wing reference area. Rolling moment L a , pitching moment M a and yaw moment N a for

[0037]

[0038] Among them, c l 、c m 、c n are the rolling moment coefficient, pitching moment coefficient and yaw moment coefficient, c A and c b are the average aerodynamic chord and span of the wing, respectively.

[0039] Considering that the aerodynamic force is in the airflow coordinate system, in order to simplify the thrust calculation, it is converted to the body coordinate system. The conversion matrix S of the airflow coordinate system to the body coordinate system is ba for

[0040]

[0041] Among them, α is the aircraft's angle of attack, and β is the aircraft's sideslip angle.

[0042] Therefore, the aerodynamic force converted to the body coordinate system is

[0043]

[0044] The particle motion module treats the aircraft as a particle and uses the flight attitude to obtain the angular velocity [p, q, r] of the aircraft in the body coordinate system. T , and then solve for the external force on the aircraft. Therefore, the angular velocity of the aircraft is

[0045]

[0046] The external force F in three directions of the aircraft is obtained by linear motion and Newton's second law of motion x ,F y ,F z The expression is

[0047]

[0048] Where m is the mass of the aircraft.

[0049] In addition, due to the uniqueness of the STOVL aircraft structure, the engine also has an inlet resistance of

[0050]

[0051] in, is the air flow rate change rate of the intake duct, x I and z I is the moment arm of the inlet resistance.

[0052] According to the above analysis, the forces acting on the aircraft include aerodynamic force, gravity, inlet resistance and engine thrust. According to the force balance equation, we can get:

[0053]

[0054] Among them, F px , F py , F pz is the total thrust generated by each nozzle. S bg is the transformation matrix from the ground coordinate system to the body coordinate system, which can be expressed as

[0055]

[0056] According to formula (8), the total thrust of the engine can be expressed as

[0057]

[0058] The angular motion module calculates the total torque of the aircraft according to the change of the aircraft's flight attitude. According to the dynamic equation of rotation around the center of mass, the expression of the external torque can be obtained as follows:

[0059]

[0060] Among them I x , I y , I z is the moment of inertia of the aircraft, I xz is the product of the vehicle's moment of inertia.

[0061] The torque that the engine should provide [L p ,M p,N p ] T It can be obtained in the body coordinate system, and its expression is

[0062]

[0063] Assuming that the thrust commands of the rolling nozzles on both sides change in differential form and the sum remains unchanged, combining equations (10) and (12), we obtain the following equation:

[0064]

[0065] Where, F cnr 、F lfr 、F gz1r 、F gz2r Respectively represent the core thrust command, lift fan thrust command and the roll thrust command of the left and right roll nozzles; F gztr The vector sum of the roll thrust commands of the left and right roll nozzles; δ cn and δ lf Core thrust F cn and lift fan thrust F lf The angle between the projection on the plane symmetry plane Oxz and the x-axis, when the projection is below the x-axis, δ cn and δ lf is positive; cny Core thrust F cn The angle between the projection on the Oxy plane and the symmetry plane Oxz, when the projection is on the left side of the symmetry plane Oxz, δ cny is positive; x, y, z are the vector force arms of the corresponding thrust; subscripts cn, lf, gz1, gz2 represent the three-bearing vector tail nozzle, lift fan nozzle and left and right roll nozzle respectively; subscripts x, y, z represent the three directional components in the body coordinate system respectively; subscript I represents the engine inlet; F I 、F p Respectively represent the engine inlet resistance and the total thrust generated by the four nozzles of the engine; L p 、M p 、N p It represents the three directions of thrust moments caused by the engine thrust in the body coordinate system, namely rolling moment, pitching moment and yaw moment.

[0066] The thrust command calculation module solves Equation (13) to obtain the thrust commands of each nozzle (nozzle) with thrust matching.

[0067] The above thrust command model was simulated and verified when the aircraft attitude angle was 0°, H = 0 km, and Ma = 0. The thrust command was obtained and compared with the literature data. The results are shown in Table 1.

[0068] Table 1 Thrust command model verification

[0069]

[0070] It can be seen from Table 1 that the generated thrust command is relatively consistent with the thrust output of the STOVL engine in the literature, with a maximum error of 0.23%, indicating that the thrust command model designed in the present invention has high accuracy.

[0071] In multivariable control, the design of the control structure is key to ensuring the performance of the control system, especially when there are more than two controlled variables. In STOVL mode, the left and right roll nozzles operate in a differential form, which means that the change amplitude of the roll nozzle outlet area on both sides is equal and the direction is opposite. Therefore, only one roll nozzle outlet area is considered in the controller design. The closed-loop control adopts the main fuel flow W fm , tail nozzle throat area A8, lift fan nozzle outlet area A lf and the tumble nozzle exit area A gz As the control variable. Take the tail nozzle thrust F cn , lift fan nozzle thrust F lf and the unilateral rolling nozzle thrust F gz is the control variable. The state space model of the STOVL engine in vertical take-off and landing mode is as follows:

[0072]

[0073] where a ij ,b ip ,c qj ,d qp (i, j = 1, 2, p = 1, 2, 3, 4, q = 1, 2, 3) are the state variable model parameters, N1 and N2 are the low-pressure shaft speed and high-pressure shaft speed respectively.

[0074] Since there are many variables involved in closed-loop control, the coupling between loops will affect the performance of the control system. In order to ensure system performance and simplify the system structure, the RGA method is used to perform coupling analysis on the control system.

[0075] The RGA method is mainly used to analyze the coupling between multiple loops in a multivariable control system, and uses a relative gain matrix to describe the coupling information. ij It represents the relative gain from the jth input to the ith output, and its calculation formula is

[0076]

[0077] Where g ij is the element in the i-th row and j-th column of the matrix G(0), where G(0) is the static gain of the engine transfer matrix.

[0078] Convert the state space model into a transfer matrix and then calculate the RGA using the following formula:

[0079]

[0080] in, represents the Hadamard product.

[0081] A 0 km, 0 Ma state space model was established for RGA analysis, and the results are shown in Table 2.

[0082] Table 2 RGA calculation results

[0083]

[0084] From Table 2, we can see that A gz Only for F gz has a greater impact, while F gz receive A gz The impact is far greater than W fm , A8 and A lf In addition, W fm , A8 and A lf F lf The effects of F lf Therefore, the system can be divided into two control blocks, where W fm , A8, A lf Control F cn 、F lf , A gz Control F gz .

[0085] In the design fm ,A8,A lf ; F cn ,F lf}Triple-variable Block (TVB) controller, set {A gz ; F gz Input A of the Single-variable Block (SVB) controller gz As a disturbance to be suppressed, design {A gz ; F gz When using a single-variable block (SVB) controller, the reference to the three-variable block is also considered as a disturbance of the single-variable block.

[0086] In the TVB controller, W fm , A8, A lf is the input, F cn 、F lfFor output. When designing, it is necessary to suppress A gz Interference, so the H2 / H with strong disturbance suppression ability is used ∞ control method, the TVB model of the actuator can be expressed as

[0087]

[0088] where x m =[W fm ,A8,A lf ,N1,N2] T is the state vector of the TVB controller, u mc =[W fmr ,A 8r ,A lfr ] T is the input vector, where W fmr 、A 8r 、A lfr is the instruction for the main fuel flow, tail nozzle throat area, and lift fan nozzle outlet area. Let u s =A gz is the disturbance of TVB controller, y m =[F cn ,F lf ] T is the output vector.

[0089] In order to achieve robust tracking characteristics, it is necessary to augment the output tracking error into the state variable. First, Equation (17) is transformed into

[0090]

[0091] Where r m Represents the control instruction of the TVB controller, which can be regarded as a step signal, and its derivative is zero.

[0092] make Formula (18) can be expressed as

[0093]

[0094] in

[0095] In order to achieve H2 / H ∞ Control method, enhance the robustness of the system, and construct the output equation:

[0096]

[0097] in Where Q and R are the weighted matrices of the following quadratic performance indicators

[0098]

[0099] Where Q∈R 7×7 is a positive definite matrix, R∈R 3×3 is a positive semidefinite matrix.

[0100] The standard H2 / H can be obtained from equations (19) and (20): ∞ Question format

[0101]

[0102] where w m is the system noise.

[0103] According to Parseval's equation, the energy of the signal in the time domain and the frequency domain is equal, so

[0104] ||z m (t)||2=||z m (jω)||2 (23)

[0105] According to the superposition principle and the norm inequality property, we can get

[0106]

[0107] in is the system noise w m to z m The closed-loop transfer function.

[0108] The state adjustment controller designed according to system formula (22) In addition to meeting the quadratic index (24), the following interference suppression requirements must also be met

[0109]

[0110] The sensitivity function for arrive The closed-loop transfer function.

[0111] For the system (22), if and only if there exists a positive definite matrix X, a positive definite matrix Z and a matrix W, there exists a state feedback H2 / H ∞ Controller, so that the following linear matrix inequality holds true

[0112]

[0113] If Equation (26) has the optimal solution X, Z and W, then the H2 / H of system (22) ∞ The control problem can be solved, and the H2 / H of the system (22) ∞ The state feedback control law is

[0114]

[0115] The resulting state feedback controller And satisfy the performance index (25) and make the quadratic index (21) small enough.

[0116] In the SVB controller, A gz is the input, F gz is the output. Its state equation is the same as that in equation (17), and the output equation is

[0117] y s =C s x m +D sm u mc +D ss u sc (28)

[0118] Among them C s 、D sm 、D ss is the corresponding coefficient matrix. Let y s =F gz ,u sc =A gzr , A gzr It is the control instruction of the tumbling nozzle exit area.

[0119] According to Equations (19) and (27), the state equation for SVB controller design is:

[0120]

[0121] in

[0122] Referring to the TVB controller design, the tracking error of the SVB controller is extended to the state vector of the system (29), and A is considered. gz The actuator, so the state vector of the SVB controller design model is The state equation used for SVB controller design is

[0123]

[0124] in And u sc =A gzr A gz The input command of the actuator. Let T gz A gz The time constant of the actuator, whose transfer function is A gz / A gzr=1 / (T gz s+1).

[0125] According to the TVB controller design steps, during the TVB design process, the SVB controller can be obtained by repeating the process from equation (20) to equation (26). The SVB controller is solved using the linear matrix inequality method to obtain

[0126]

[0127] in

[0128] In the TVB controller design process, the parameters of the state variable model (14) are determined and the actuator is simplified to a first-order inertia link with a known time constant. Therefore, for the standard H2 / H ∞ To solve the control problem, we only need to set the weight matrices Q and R corresponding to Equation (20). Different controller parameters can be obtained by using different Q and R. This paper uses differential evolution algorithm to optimize Q and R to achieve the purpose of fast tracking. The optimization goal is to minimize the tracking error, which can be expressed as

[0129]

[0130] Where, F cnr (t), F lfr (t) are F cn and F lf instruction, t1 is the total simulation time.

[0131] In order to simplify the optimization process, the matrices Q and R are set as diagonal matrices, and then there are 10 parameters to be optimized, which correspond to the diagonal elements of matrices Q and R. Similarly, the weight matrices Q and R of the SVB controller are set as diagonal matrices and optimized to achieve the F gzr Fast tracking.

[0132] The final controller designed is as follows Figure 3 As shown, it consists of a three-variable block TVB controller and a single-variable block SVB controller; in the TVB controller, r m =[F cnr ,A 8r ,F lfr ] T For input command, y m =[F cn ,A8,F lf ] T is the actual output of the engine, and the error e is obtained by making the difference between the two m =r m -y m After the integration step, we get ∫em dt,∫e m dt and the state quantity x of the TVB controller output by the engine m =[W fm ,A8,A lf ,N1,N2] T After the controller gain is applied, we can get Finally, u mc =[W fmr ,A 8r ,A lfr ] T The main fuel flow command, tail nozzle throat area command, and lift fan nozzle outlet area command act on the engine, where N1 and N2 are the low pressure shaft speed and high pressure shaft speed respectively; in the SVB controller, r s =F gzr For input command, y s =F gz is the actual output of the engine, and the error e is obtained by making the difference between the two s =r s -y s After the integration step, we get ∫e s dt, the state quantity and error integral of the TVB controller are added to the state quantity of the SVB controller to obtain the state quantity of the SVB controller ∫e s dt and x s After the controller gain is applied, we can get Finally, u sc =A gzr The left and right roll nozzle area instructions act on the engine.

[0133] The structure of the LF-STOVL engine multi-nozzle thrust matching control system established in this embodiment is as follows: Figure 4 As shown, based on a six-degree-of-freedom STOVL aircraft model, matching thrust commands are generated according to the flight state and attitude. Since the thrust required by the engine control system feedback loop cannot be measured during flight, this embodiment estimates it by constructing a dynamic deep neural network thrust estimator based on similarity transformations. The STOVL engine is represented by a component-level model, which is constructed by solving eight equilibrium equations using the Newton-Raphson and Euler methods.

[0134] According to public media reports, the F135 can open the lift fan at an altitude of 5000 feet (about 1524 meters) and a flight speed of 288 kilometers per hour (about 0.24 Mach number). Therefore, two working points (H = 0km Ma = 0 and H = 1.6km Ma = 0.24) are selected to verify the effect of the technical solution of the present invention. First, the input parameter is set to W fmis 2kg / s, A8 is 0.374339m 2 , A lf 0.43223m 2 , A gz is 95%.

[0135] First, the STOVL engine multi-nozzle thrust matching closed-loop control system is verified at the operating point H = 0km, Ma = 0. The command changes linearly from 0° to 10° within 2 to 12 seconds and remains at 10° thereafter. The command for the pitch angle θ and the yaw angle ψ remains unchanged at 0°. The system response curve is shown in Figure 1. Figure 5 As shown, the subscript "r" represents the command value and "e" represents the estimated value.

[0136] Roll angle Increase means that the left side of the fuselage is raised and the right side is lowered, where the left and right sides are distinguished based on the tail view of the aircraft. lf and F gz When always on the Oxz surface of the body, y g F in the positive direction of the axis lfr and F gzr The weight increases, z g The negative component of the axis decreases. In order to maintain the balance of the aircraft, the tail nozzle deflects to the right, F cnr Increase, so that y g axis and z g The component in the negative direction of the axis increases. At the same time, F lfr and F gzr It shows an opposite downward trend. cn For instruction F cnr The maximum dynamic tracking error is 0.220%; F lf For instruction F lfr The maximum dynamic tracking error is 0.085%; F gz For instruction F gzr The maximum dynamic tracking error is 0.027%, and the steady-state error is almost 0. The simulation results show that the controlled quantity can quickly track the control command, suppress disturbances, and the controlled quantity changes smoothly without overshoot.

[0137] Then, the pitch angle θ is linearly changed from 0° to 10° within 2 to 12 seconds, and then maintained at 10°. The command of the yaw angle ψ remains unchanged at 0°. The system response curve is as follows Figure 6 shown.

[0138] An increase in pitch angle θ means the nose is raised and the tail is lowered. cn and F gzAt the same time, the F on the other side of the center of gravity and close to the nose lf Increase, change in the opposite direction. cn For instruction F cnr The maximum dynamic tracking error is 0.641%; F lf For instruction F lfr The maximum dynamic tracking error is 0.502%; F gz For instruction F gzr The maximum dynamic tracking error is 0.396%, and the steady-state error is almost 0. Although the control effect is slightly worse than that when the roll angle command changes, it can still achieve fast tracking of the command and respond without overshoot.

[0139] Then, a simulation verification is performed at the working point H = 1.6 km, Ma = 0.24. The command adopts the same change, keeping the command of θ and ψ at 0°. The response result is as follows Figure 7 As shown. cn For instruction F cnr The maximum dynamic tracking error is 0.202%; F lf For instruction F lfr The maximum dynamic tracking error is 0.079%; F gz For instruction F gzr The maximum dynamic tracking error is 0.118%, the steady-state error is almost 0, and the controlled quantity can quickly track the instruction. Figure 5 and Figure 7 , F cn and F lf The maximum tracking errors of F gz The maximum tracking error of F increases significantly by 0.091%. gz The control effect is greatly affected.

[0140] Finally, at the working point H = 1.6km, Ma = 0.24, verify the control effect when the pitch angle θ changes. The pitch angle command adopts the same change, and the roll angle and yaw angle commands are kept at 0°. The response result curve is as follows Figure 8 As shown. cn For instruction F cnr The maximum dynamic tracking error is 0.612%; F lf For instruction F lfr The maximum dynamic tracking error is 0.90%; F gz For instruction F gzr The maximum dynamic tracking error is 0.166%, and the steady-state error is almost 0. Figure 6 and Figure 8 , F cn 、F lf and Fgz The tracking errors of the 10 ... Figure 5 、 Figure 6 、 Figure 7 、 Figure 8 It can be seen that F gz The control effect is better than F cn and F lf , and the control error is much smaller.

[0141] In the STOVL state, when the flight attitude angle changes, the thrust command model solves the reasonable thrust command for each nozzle. fm ,A8,A lf ; F cn ,F lf} and {A gz ; F gz The block controller can quickly track instructions and suppress the coupling effects between loops.

Claims

1. The LF-STOVL engine multi-nozzle thrust matching control method is characterized by: First, according to the flight attitude and flight conditions, the core thrust instruction at the three-bearing vector tail nozzle, the lift fan thrust instruction at the lift fan nozzle, and the roll thrust instruction at the left and right roll nozzles are generated by solving the six-degree-of-freedom STOVL aircraft model; then, the main fuel flow rate W is adjusted according to the thrust instructions and the actual thrust. fm , tail nozzle throat area A8, lift fan nozzle outlet area A lf and the tumble nozzle exit area A gz Perform closed-loop control; the six-degree-of-freedom STOVL aircraft model is specifically shown in the following formula: Where, F cnr 、F lfr 、F gz1r 、F gz2r Respectively represent the core thrust command, lift fan thrust command and the roll thrust command of the left and right roll nozzles; F gztr The vector sum of the roll thrust commands of the left and right roll nozzles; δ cn and δ lf Core thrust F cn and lift fan thrust F lf The angle between the projection on the plane symmetry plane Oxz and the x-axis, when the projection is below the x-axis, δ cn and δ lf is positive; cny Core thrust F cn The angle between the projection on the Oxy plane and the symmetry plane Oxz, when the projection is on the left side of the symmetry plane Oxz, δ cny is positive; x, y, z are the vector force arms of the corresponding thrust; subscripts cn, lf, gz1, gz2 represent the three-bearing vector tail nozzle, lift fan nozzle and left and right roll nozzle respectively; subscripts x, y, z represent the three directional components in the body coordinate system respectively; subscript I represents the engine inlet; F I 、F p Respectively represent the engine inlet resistance and the total thrust generated by the four nozzles of the engine; L p 、M p 、N p It represents the three directions of thrust moments caused by the engine thrust in the body coordinate system, namely rolling moment, pitching moment and yaw moment.

2. The LF-STOVL engine multi-nozzle thrust matching control method according to claim 1, characterized in that: The controller for executing the closed-loop control is composed of a three-variable block TVB controller and a single-variable block SVB controller; in the TVB controller, r m =[F cnr ,A 8r ,F lfr ] T For input command, y m =[F cn ,A8,F lf ] T is the actual output of the engine, and the error e is obtained by making the difference between the two m =r m -y m After the integration step, we get ∫e m dt,∫e m dt and the state quantity x of the TVB controller output by the engine m =[W fm ,A8,A lf ,N1,N2] T After the controller gain is applied, we can get Finally, u mc =[W fmr ,A 8r ,A lfr ] T The main fuel flow command, tail nozzle throat area command, and lift fan nozzle outlet area command act on the engine, where N1 and N2 are the low pressure shaft speed and high pressure shaft speed respectively; in the SVB controller, r s =F gzr For input command, y s =F gz is the actual output of the engine, and the error e is obtained by making the difference between the two s =r s -y s After the integration step, we get ∫e s dt, the state quantity and error integral of the TVB controller are added to the state quantity of the SVB controller to obtain the state quantity of the SVB controller ∫e s dt and x s After the controller gain is applied, we can get Finally, u sc =A gzr The left and right roll nozzle area instructions act on the engine.

3. The LF-STOVL engine multi-nozzle thrust matching control method according to claim 1 or 2, characterized in that: The actual thrust is estimated by a dynamic deep neural network thrust estimator based on similarity transformation.

4. LF-STOVL engine multi-nozzle thrust matching control system, characterized by: The system comprises a thrust instruction calculation module and a controller; the thrust instruction calculation module is used to generate the core thrust instruction at the three-bearing vector tail nozzle, the lift fan thrust instruction at the lift fan nozzle, and the roll thrust instruction of the left and right roll nozzles according to the flight attitude and flight conditions and by solving the six-degree-of-freedom STOVL aircraft model; the controller is used to adjust the main fuel flow W according to the thrust instruction and the actual thrust fm , tail nozzle throat area A8, lift fan nozzle outlet area A lf and the tumble nozzle exit area A gz Perform closed-loop control; the six-degree-of-freedom STOVL aircraft model is specifically shown in the following formula: Where, F cnr 、F lfr 、F gz1r 、F gz2r Respectively represent the core thrust command, lift fan thrust command and the roll thrust command of the left and right roll nozzles; F gztr The vector sum of the roll thrust commands of the left and right roll nozzles; δ cn and δ lf Core thrust F cn and lift fan thrust F lf The angle between the projection on the plane symmetry plane Oxz and the x-axis, when the projection is below the x-axis, δ cn and δ lf is positive; cny Core thrust F cn The angle between the projection on the Oxy plane and the symmetry plane Oxz, when the projection is on the left side of the symmetry plane Oxz, δ cny is positive; x, y, z are the vector force arms of the corresponding thrust; subscripts cn, lf, gz1, gz2 represent the three-bearing vector tail nozzle, lift fan nozzle and left and right roll nozzle respectively; subscripts x, y, z represent the three directional components in the body coordinate system respectively; subscript I represents the engine inlet; F I 、F p Respectively represent the engine inlet resistance and the total thrust generated by the four nozzles of the engine; L p 、M p 、N p It represents the three directions of thrust moments caused by the engine thrust in the body coordinate system, namely rolling moment, pitching moment and yaw moment.

5. The LF-STOVL engine multi-nozzle thrust matching control system according to claim 4, characterized in that: The controller is composed of a three-variable block TVB controller and a single-variable block SVB controller; in the TVB controller, r m =[F cnr ,A 8r ,F lfr ] T For input command, y m =[F cn ,A8,F lf ] T is the actual output of the engine, and the error e is obtained by making the difference between the two m =r m -y m After the integration step, we get ∫e m dt,∫e m dt and the state quantity x of the TVB controller output by the engine m =[W fm ,A8,A lf ,N1,N2] T After the controller gain is applied, we can get Finally, u mc =[W fmr ,A 8r ,A lfr ] T The main fuel flow command, tail nozzle throat area command, and lift fan nozzle outlet area command act on the engine, where N1 and N2 are the low pressure shaft speed and high pressure shaft speed respectively; in the SVB controller, r s =F gzr For input command, y s =F gz is the actual output of the engine, and the error e is obtained by making the difference between the two s =r s -y s After the integration step, we get ∫e s dt, the state quantity and error integral of the TVB controller are added to the state quantity of the SVB controller to obtain the state quantity of the SVB controller ∫e s dt and x s After the controller gain is applied, we can get Finally, u sc =A gzr The left and right roll nozzle area instructions act on the engine.

6. The LF-STOVL engine multi-nozzle thrust matching control system according to claim 4 or 5, characterized in that: The actual thrust is estimated by a dynamic deep neural network thrust estimator based on similarity transformation.

Citation Information

Patent Citations

  • Power system suitable for small vertical take-off and landing fixed-wing aircraft

    CN110641695A

  • Distributed propulsion system test bed for short-distance / vertical take-off and landing aircraft

    CN111619823A