Multivariable performance optimization method based on stabilization-optimization double loops

Abstract

The invention provides a multivariable performance optimization method based on stabilization-optimization double loops, and relates to the technical field of aviation aircraft power system control. The method comprises the steps that an inner layer control loop of an aircraft engine is designed according to control requirements, and the control requirements comprise the steps that ideal thrust closed-loop bandwidth is achieved, A8 and A163 have the centering characteristic, disturbance suppression is carried out, and servo tracking is carried out; an outer layer performance optimization loop of the aircraft engine is designed, and the fuel consumption rate is globally optimized; determining parameters of a feedforward-feedback combined controller according to the designed inner-layer control loop; according to the method, parameters are adopted to control thrust, so that the thrust is maintained at an expected value, a designed outer layer performance optimization loop is adopted, and delta A8ref and delta A163ref are dynamically adjusted, so that the engine reaches the lowest fuel consumption rate under the specific thrust. According to the optimization method provided by the invention, the highest fuel efficiency output of the engine under a specific flight condition is realized through online correction of the steady state control plan.

Description

Technical Field

[0001] The present invention relates to the technical field of aviation vehicle power system control, and in particular to a multivariable performance optimization method based on a stabilization-optimization dual loop. Background Art

[0002] Performance Seeking Control (PSC) technology was proposed in the Highly Integrated Digital Electronic Control (HIDEC) project to achieve online adaptive optimization of aircraft power system performance. By using onboard models and control algorithms, PSC can dynamically adjust the engine operating point to meet the performance requirements of different flight phases, thereby achieving multi-objective optimization such as low specific fuel consumption, high thrust or low low-pressure turbine inlet temperature.

[0003] In the existing technology, there are many studies on performance optimization control. For example, NASA's related research shows that the use of PSC technology in F-15 flight tests can significantly improve engine performance, such as reducing fuel consumption under specific conditions (2%-10%) or increasing thrust under the same fuel conditions (10%-15%).

[0004] However, traditional performance optimization methods rely on multiple online linearizations, which results in large computational complexity and high reliance on high-precision airborne models. In addition, current research mainly focuses on optimization under steady-state conditions, while how to optimize the comprehensive performance of aircraft power systems under installed and non-installed conditions remains an urgent problem to be solved. Summary of the invention

[0005] Based on the theoretical basis and shortcomings of the prior art, the present invention achieves optimal engine performance while ensuring safe engine operation. On this basis, it breaks through the limitations of the prior art, reduces dependence on precise airborne models, and improves system robustness and adaptability through a two-layer architecture design. The multivariable performance optimization method based on dual loops proposed in the present invention achieves the highest fuel efficiency output of the engine under specific flight conditions by online correction of the steady-state control plan. Simulation results show that this method significantly improves engine performance, has anti-disturbance capabilities and fuel efficiency optimization effects, and provides a solution and practical path for optimizing engine performance in the wing state.

[0006] Specifically, the present invention provides a multivariable performance optimization method based on a stabilization-optimization dual loop, comprising:

[0007] Design the inner control loop of the aircraft engine according to the control requirements;

[0008] The inner control loop is a feedforward-feedback combined control loop, wherein the feedforward control loop is used to improve the rapidity of thrust adjustment and respond to input changes in advance, and includes an A8 loop based on the tail nozzle throat area and an A loop based on the rear duct ejector area. 163 Loop; The feedback control loop is used to correct the error based on thrust feedback, which is W based on the fuel flow f loop; loop

[0009] The control requirements include the following four aspects: achieving the ideal thrust closed-loop bandwidth; after the thrust setting value changes, the tail nozzle throat area A8 and the rear duct ejector area A 163 It can always return to the set value; avoid changing A8 and A during the optimization process of the outer performance optimization loop 163 Disturbance thrust; A8 loop can be reference value of A8 ΔA 8_ref Respond quickly, A 163 The circuit can 163 Reference value ΔA 163_ref Respond quickly;

[0010] Design the outer performance optimization loop of the aircraft engine to globally optimize the fuel consumption rate;

[0011] Determine the parameters of the feedforward-feedback combination controller based on the designed inner control loop;

[0012] The thrust is controlled by the parameters to maintain it at the desired value, and the outer performance optimization loop is designed to dynamically adjust ΔA 8_ref and ΔA 163_ref , achieving a balance between fuel efficiency and thrust, so that the engine can achieve the lowest fuel consumption rate under a specific thrust.

[0013] As a further explanation of the present invention, the feedforward-feedback combined controller includes an A8 controller, a W f Controller and A 163 The A8 circuit includes an A8 controller, an A8 actuator and an engine. 163 The circuit contains A 163 Controller, A 163 Actuator and motor, the W f The circuit contains W f Controller, W f Actuators and motors;

[0014] The feedforward-feedback combined controller receives a thrust reference value ΔF n,ref As input, the input signal passes through W f The transfer function of the controller is g c2 After processing, send to W factuator, and at the same time, the A8 controller transfers its function g c11 and g c12 Process the input signal and output it to the A8 actuator, A 163 The controller transfers the function g c31 and g c32 Process the input signal and output it to A 163 Actuator, A8 actuator through its transfer function g a1 The tail nozzle throat area A8 changes, W f The actuator passes through its transfer function g a2 Driving fuel flow W f Change, A 163 The actuator passes through its transfer function g a3 Driving rear duct ejector area A 163 Change, the engine is based on the A8 actuator, W f Actuator and A 163 The change in actuator output generates thrust deviation ΔF through transfer functions g1, g2 and g3 n , where g1 is the deviation from A8 to ΔF n The transfer function of g2 is from W f Deviation ΔW f to ΔF n The transfer function of g3 is from A 163 Deviation ΔA 163 to ΔF n The transfer function, finally, converts the thrust deviation ΔF n Feedback to the input of the feedforward-feedback combined controller and the thrust reference value ΔF n,ref Form a closed-loop control.

[0015] As a further explanation of the present invention, the ideal thrust closed-loop bandwidth is specifically:

[0016] Using the loop shaping method, the ideal open-loop transfer function L of the inner control loop is specified as:

[0017] (1);

[0018] s is the complex variable of Laplace transform, ω b1 is the closed-loop transfer function bandwidth of the inner control loop, which is specifically set to 10 rad / s. The corresponding closed-loop transfer function is the inertia link, and the corresponding closed-loop transfer function G y1 (s) is the thrust reference value ΔF n,ref To thrust deviation ΔF n The transfer function is defined as:

[0019] (2);

[0020] Design the performance index of the ideal thrust bandwidth T1, and the control target is:

[0021] (3);

[0022] Among them, γ1 is the error tolerance and K is the controller.

[0023] As a further illustration of the present invention, the open-loop transfer function .

[0024] As a further illustration of the present invention, after the thrust setting value changes, the tail nozzle throat area A8 and the rear duct ejector area A 163 It is always possible to return to the setpoint, which is met by pre-specifying the controller structure:

[0025] W f The controller is a PI controller, A 163 The controller and A8 controller are composed of a PD controller and a lead-lag compensator without an integrator, where g c2 is the transfer function of the PI controller; g c11 and g c12 are the transfer function of the lead-lag compensator without integrator and the transfer function of the PD controller in the A8 controller, respectively; g c31 and g c32 A 163 The transfer function of the lead-lag compensator and the transfer function of the PD controller without the integrator in the controller.

[0026] As a further illustration of the present invention, W f Controller, A8 Controller, A 163 The transfer function of the controller is specifically:

[0027] (4);

[0028] (5);

[0029] (6);

[0030] Among them, K p is the proportionality coefficient, K i is the integration coefficient, K d is the differential coefficient, T is the time constant, s is the complex variable of Laplace transform, and θ1 and θ1 are the coefficients to be determined.

[0031] As a further illustration of the present invention, the method of avoiding changing the outer performance optimization loop during the optimization process A 8 andA 163 Disturbance thrust, specifically: definition ΔA 8 and ΔA 163 To thrust deviation Δ F n The transfer function g 1 and g 3 are and , Design the following disturbance suppression T 2 and T 3 performance indicators, control objectives are:

[0032] (7);

[0033] Among them, s is the complex variable of Laplace transform, γ2, γ3 are error tolerances, and K is the controller.

[0034] As a further illustration of the present invention, the A 8 circuits can A Reference value of 8 ΔA 8_ref Respond quickly, A 163 The circuit can A 163 Reference value ΔA 163_ref Respond quickly, specifically: remember G y2 For reference input ΔA 8,ref To the actual output ΔA The transfer function of 8 is G y3 For reference input Δ A 163,ref To the actual output ΔA 163 The transfer function is:

[0035] (8);

[0036] Among them, ω b2 and ω b3 G y2 and G y3 The bandwidth of , s is the complex variable of Laplace transform; Designing servo tracking T 4 and T 5 performance indicators, the control objectives are:

[0037] (9); Where s is the complex variable of Laplace transform, γ 4, γ 5 is the error tolerance, K For the controller.

[0038] As a further illustration of the present invention, the outer performance optimization loop uses a Newton-Raphson iterative algorithm to calculate the first-order derivative of the objective function and quickly approach the global optimal solution.

[0039] As a further illustration of the present invention, the control objectives of the outer performance optimization loop are:

[0040] (10);

[0041] That is, by adjusting ΔA 8,ref and ΔA 163,ref Make J minimum while keeping the thrust at the desired value, where SFC is the fuel consumption rate.

[0042] Compared with the prior art, the present invention has the following beneficial technical effects:

[0043] The present invention proposes a performance optimization control architecture for calculating gradients as an improvement method: 1) based on direct thrust estimation, it reduces the dependence on high-precision airborne models; 2) by introducing an inner and outer loop integrated architecture, it solves the problem of not explicitly considering the integration of performance optimization and steady-state point control in traditional methods, thereby improving the system robustness and adaptability.

[0044] Other features and advantages of the technical solution will be described in the subsequent description, and partly become apparent from the description, or understood by implementing the technical solution. The purpose and other advantages of the technical solution can be achieved and obtained through the structures specifically pointed out in the written description and the drawings.

[0045] The technical solution of the present technical solution is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The accompanying drawings are used to provide a further understanding of the technical solution and constitute a part of the specification. Together with the embodiments of the technical solution, they are used to explain the technical solution and do not constitute a limitation of the technical solution. In the accompanying drawings:

[0047] Figure 1 The overall framework diagram of multivariable performance optimization based on the stabilization-optimization dual loop provided by the present invention.

[0048] Figure 2 It is a block diagram of the feedforward-feedback combined control system in the present invention.

[0049] Figure 3 Find the optimal isopotential diagram for any point of fixed thrust (75%) in the present invention.

[0050] Figure 4(a)-Figure 4(c) The physical parameters of the engine in the present invention vary with time, wherein FIG4 (a) is SFC, α8 and α 163 Figure 4(b) shows the time-varying graph of F n , N l With N h Figure 4(c) shows the time-varying graph of T 41 , S.M. HPC With SM LPC Graph of changes over time.

[0051] Figure 5 It is a schematic diagram of the variable cycle engine structure of the present invention.

[0052] Figure 6(a)-Figure 6(b) It is a schematic diagram of the ground steady-state characteristics of the present invention, wherein FIG6(a) is a SFC surface diagram of multiple equal thrusts (thrust from 55% to 95% of the design point thrust), and FIG6(b) is a cross-sectional diagram of the SFC minimum point under each thrust.

[0053] Figure 7(a)-Figure 7(e) Figure 7(a) is the closed-loop system characteristic diagram, where Figure 7(b) is the A8 gain tuning target, and Figure 7(c) is the A 163 Gain tuning target, Figure 7(d) is the A8 step tracking tuning target, Figure 7(e) is the A 163 Step tracking tuning target. DETAILED DESCRIPTION

[0054] The preferred embodiments of the present technical solution are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present technical solution, and are not used to limit the present technical solution.

[0055] The present invention provides a multivariable performance optimization method based on a stabilization-optimization dual loop. The method performs control at the inner layer to accurately adjust the engine thrust in real time to reduce thrust fluctuations; and performs optimization at the outer layer to perform global performance optimization under safety conditions to improve fuel efficiency and mission adaptability.

[0056] Specifically, Figure 1 As shown, the method comprises the following steps:

[0057] Step 1: Design the inner control loop of the aircraft engine according to the control requirements.

[0058] The inner control loop is a feedforward-feedback combined control loop to improve the regulation efficiency of the variable cycle engine under constant thrust conditions. Among them, the feedforward control loop is used to improve the rapidity of thrust regulation, respond to input changes in advance, and reduce system delays. It includes the A8 loop based on the tail nozzle throat area and the A based on the rear duct ejector area. 163 The feedback control loop is used to correct the error based on thrust feedback to ensure the stability and accuracy of thrust. It is based on the fuel flow W f loop.

[0059] It should be noted that Figure 1 The subscript "ref" in the symbol is used to indicate the reference value, while "cmd" refers to the instruction sent by the controller to the actuator, Δ represents the difference of the physical quantity relative to its steady-state value, and W f,ss Represents the steady-state value of fuel flow, A 8,ss represents the steady-state value of the tail nozzle throat area, A 163,ss Represents the steady-state value of the rear duct ejector area.

[0060] Specifically, Figure 1 and Figure 2 As shown, the feedforward-feedback combined controller includes the A8 controller, W f Controller and A 163 Controller, A8 circuit includes A8 controller, A8 actuator and engine, A 163 The circuit contains A 163 Controller, A 163 Actuator and motor, W f The circuit contains W f Controller, W f Actuators and motors;

[0061] The feedforward-feedback combined controller receives the thrust reference ΔF n,ref As input, the input signal passes through W f The transfer function of the controller is g c2 After processing, send to W f actuator, and at the same time, the A8 controller transfers its function g c11 and g c12 Processes input signals and outputs them to the A8 actuator. 163 The controller transfers the function g c31 and g c32 Processes the input signal and outputs it to A 163 Actuator, A8 actuator through its transfer function g a1 The tail nozzle throat area A8 changes, W f The actuator passes through its transfer function ga2 Driving fuel flow W f Change, A 163 The actuator passes through its transfer function g a3 Driving rear duct ejector area A 163 Change, the engine is based on the A8 actuator, W f Actuator and A 163 The change in actuator output generates thrust deviation ΔF through transfer functions g1, g2 and g3 n , where g1 is the deviation from A8 to ΔF n The transfer function of g2 is from W f Deviation ΔW f to ΔF n The transfer function of g3 is from A 163 Deviation ΔA 163 to ΔF n The transfer function, finally, is the thrust deviation ΔF n Feedback to the input of the feedforward-feedback combined controller and the thrust reference value ΔF n,ref Form a closed-loop control.

[0062] The design of the above controller needs to meet the following control requirements:

[0063] 1) Achieve the ideal thrust closed-loop bandwidth. 2) A8 and A 163 It has a return-to-center characteristic, that is, after the thrust setting value changes, the tail nozzle throat area A8 and the rear duct ejector area A 163 It can always return to the set value to ensure continuous modulation capability. 3) Disturbance suppression: Avoid changing A8 and A during the optimization process of the outer performance optimization loop. 163 4) Servo tracking: The A8 loop can track the reference value of A8 ΔA 8_ref Respond quickly, A 163 The circuit can 163 Reference value ΔA 163_ref Respond quickly.

[0064] In view of the above control requirements, the following design method is adopted:

[0065] 1) To achieve the ideal thrust closed-loop bandwidth:

[0066] Using the loop shaping method, the ideal open-loop transfer function L of the inner control loop is specified as:

[0067] ;

[0068] in, s is the complex variable of Laplace transform, ω b1is the closed-loop transfer function bandwidth of the inner control loop, which is specifically set to 10 rad / s. The corresponding closed-loop transfer function is the inertia link, and the corresponding closed-loop transfer function G y1 (s) is the thrust reference value ΔF n,ref To thrust deviation ΔF n The transfer function is defined as:

[0069] .

[0070] In order to ensure that the system can quickly respond to the thrust command in a short time, the present invention designs the following performance indicators related to the ideal thrust bandwidth T1, so that the closed-loop transfer function G y1 It can be approximated to a first-order inertia link, and the specific control objectives are:

[0071] (3);

[0072] Among them, γ1 is the error tolerance, K is the controller, and this design indicator requires that the closed-loop response of the system can be close to the ideal first-order response, so as to achieve fast and stable thrust adjustment in actual operation.

[0073] 2) A8 and A 163 The return feature: This requirement requires W f Circuit types are 1, A8 and A 163 The loop type is 0.

[0074] For A8 and A 163 The requirement of the return-to-center characteristic can be met by pre-specifying the controller structure: f The controller is a PI controller, A 163 The controller and A8 controller are composed of a PD controller and a lead-lag compensator without an integrator, where g c2 is the transfer function of the PI controller; g c11 and g c12 are the transfer function of the lead-lag compensator without integrator and the transfer function of the PD controller in the A8 controller, respectively; g c31 and g c32 A 163 The transfer function of the lead-lag compensator and the transfer function of the PD controller without the integrator in the controller.

[0075] The above W f Controller, A8 Controller, A 163 The transfer function of the controller is specifically:

[0076] (4);

[0077] (5);

[0078] (6);

[0079] Among them, K p is the proportionality coefficient, K i is the integration coefficient, K d is the differential coefficient, T is the time constant, s is the complex variable of Laplace transform, and θ1 and θ1 are the coefficients to be determined.

[0080] 3) Disturbance suppression: To ensure A8, A 163 The impact on thrust is minimal, and the controller needs to be designed to have a suppressive effect on disturbances. Define ΔA8 and ΔA 163 To thrust deviation ΔF n The transfer functions g1 and g3 are and .

[0081] The present invention designs the following performance indicators of disturbance suppression T2 and T3, and the control objectives are:

[0082] (7);

[0083] Where s is the complex variable of Laplace transform, γ2, γ3 are error tolerances, and K is the controller. By properly designing the controller parameters, and Gain suppression reduces the impact of disturbances on thrust.

[0084] 4) Servo tracking: record G y2 is the input ΔA from the reference 8,ref The transfer function to the actual output ΔA8 is denoted by G y3 is the input ΔA from the reference 163,ref To the actual output ΔA 163 The transfer function is:

[0085] (8);

[0086] Among them, ω b2 and ω b3 G y2 and G y3 bandwidth to ensure the fast tracking capability of the system, and s is the complex variable of the Laplace transform.

[0087] In order to ensure the tracking capability of the system to the reference input, the present invention designs the performance indicators of servo tracking T4 and T5, so that the tracking error of the system to the reference input is kept within a small range. The specific control objectives are:

[0088] (9);

[0089] Among them, s is the complex variable of Laplace transform, γ4, γ5 are error tolerances, which ensure that the system can quickly and accurately track the command input to achieve precise thrust control, and K is the controller.

[0090] Step 2: Design the outer performance optimization loop of the aircraft engine to globally optimize the fuel consumption rate.

[0091] The aforementioned inner controller maintains the thrust at the desired value through feedforward-feedback combined control. Based on the inner loop, the present invention proposes to add an outer loop to globally optimize the specific fuel consumption (SFC).

[0092] Step 3: Determine the parameters of the feedforward-feedback combination controller based on the designed inner control loop.

[0093] For A8 and A 163 The requirement of the return-to-center characteristic can be met by pre-specifying the controller structure. For the other requirements mentioned above, the controller parameters need to be tuned. This problem is a parameter tuning problem under the constraints of a given controller structure, which can be expressed as:

[0094] (10)

[0095] Where T i It is the control requirement proposed in the previous article, and K is the controller.

[0096] This is a controller parameter optimization problem with structural constraints, which is solved by the present invention using a non-smooth optimization algorithm. The so-called "non-smooth" means that the objective function or constraint condition does not have continuous derivatives at certain points, that is, there are discontinuous gradient changes, so it is difficult to apply traditional gradient optimization methods.

[0097] Step 4: Use the parameters of the feedforward-feedback combined controller obtained above to control the thrust to maintain it at the desired value, and use the designed outer performance optimization loop to dynamically adjust ΔA 8_ref and ΔA 163_ref , achieving a balance between fuel efficiency and thrust, so that the engine can achieve the lowest fuel consumption rate under a specific thrust.

[0098] In order to achieve optimal control of engine performance, the outer performance optimization loop uses the Newton-Raphson iterative algorithm, which can quickly approach the global optimal solution by calculating the first-order derivative (gradient) of the objective function. The advantage of the Newton-Raphson method is that it can effectively handle convex optimization problems in multidimensional space, ensuring that the engine can maintain optimal performance under various flight conditions.

[0099] The control objectives of the outer performance optimization loop are:

[0100] (11);

[0101] That is, by adjusting ΔA 8,ref and ΔA 163,ref Make J minimum while keeping the thrust at the desired value, where SFC is the fuel consumption rate.

[0102] In summary, the outer performance optimization loop in the present invention can estimate the gradient of the objective function relative to each variable by applying a small disturbance to the control variable. This online disturbance method can avoid the influence of inaccurate engine model and is applicable to actual flight conditions. Due to the stabilizing effect of the inner loop on the thrust, the influence of this disturbance on the thrust can be effectively reduced, so that the outer layer can be optimized without significantly affecting the thrust. The outer optimization loop continuously adjusts the geometric variable area (A8 and A 163 ), achieve a balance between fuel efficiency and thrust, and ensure that the engine converges to the minimum fuel consumption point under a specific thrust.

[0103] The method provided by the present invention is verified and explained below in conjunction with a specific simulation example:

[0104] The engine used in this example is a variable cycle engine, and its main components include an air inlet, a low-pressure compressor, a high-pressure compressor, a combustion chamber, a high-pressure turbine, a low-pressure turbine, a mixing chamber, and a tail nozzle. Figure 5 shown.

[0105] The engine performance parameters used in the simulation are: maximum thrust of about 84.51 kN, high-pressure rotor design point speed of 15000 r / min, low-pressure rotor design point speed of 9000 r / min, and the nozzle adopts a convergent form.

[0106] The SFC steady-state characteristics of the engine under near-ground (Alt=0.4km, Mach number Ma=0) flight conditions are as follows: Figure 4(a)-4(c) As shown in the figure, α8 is the opening area of ​​the tail nozzle throat:

[0107] (12);

[0108] Set it to vary between 105% and 145%;

[0109] α 163 is the opening of the rear duct ejector:

[0110] (13);

[0111] Set it to vary between 40% and 130%.

[0112] Figure 6(a) shows several SFC surfaces with equal thrust (thrust from 55% to 95% of the design point thrust). Each surface represents the SFC curves of A8, A 163 The correspondence between the SFC and the convex surface. It can be seen that each surface is a convex surface, that is, there is a minimum point (shown as a red dot in the figure) representing the optimal fuel consumption point under the thrust, and this point is the optimal working point under the current thrust.

[0113] In order to more clearly show the distribution of the optimal working point under different thrusts, Figure 6(b) shows the cross-sectional view of the SFC minimum point under each thrust. n,dp represents the thrust setting value) This figure shows the A8, A 163 How does the change of affect the minimum position of fuel consumption? It can be seen that the optimal working point under different thrusts is A8, A 163 There are significant differences in the configurations of the two algorithms, but the overall trend shows consistent convex characteristics, which provides a theoretical basis and reference for the subsequent online performance optimization using gradient-based methods.

[0114] The present invention selects the 75% design point thrust in Figure 6(a) as the simulation condition to test the performance optimization effect. The inner loop controller will maintain the thrust at 75%, so the outer loop optimization will find the SFC optimal point on the 75% equal thrust surface.

[0115] In order to apply the engine control method proposed in the present invention, the engine system is first identified by the least square method, and the system matrix is ​​estimated using a small step disturbance to obtain the 75% thrust ( =100%, =100%) to support controller design and performance optimization. At this operating point, the system identification obtains g1, g2 and g3 respectively:

[0116] (14);

[0117] (15);

[0118] (16);

[0119] Take A8, W f and A 163 The actuator is a first-order inertia link:

[0120] (17);

[0121] The actuator bandwidth ω a1 =ω a2 =ω a3 =20rad / s

[0122] In view of the above-mentioned composite control requirements, the design parameters of the non-smooth optimization algorithm are taken as:

[0123] 1) Ideal thrust bandwidth: Ensure that the loop gain is .

[0124] 2) Disturbance suppression: Ensure A 8,ref and A 163,ref The maximum gain γ2,γ3≤0.1.

[0125] 3) Servo tracking: Ensure that the 8,ref and A 163,ref To A8 and A respectively 163 The target response bandwidth is ω b2 =ω b3 =5rad / s.

[0126] The non-smooth optimization algorithm is used to adjust the parameters and calculate the performance of the feedforward-feedback combination controller. The closed-loop system characteristics are as follows: Figure 7(a)-Figure 7(e) As shown. Figure 7(a) shows the closed-loop response gain diagram of the entire loop. It can be seen from the figure that the loop meets the requirements ; Figure 7(b) and Figure 7(c) are A 8,ref and A 163,ref The maximum gain of the disturbance suppression amplitude of A8 and A1 are 1% and 1.37% respectively, both of which are less than -20dB. 163 The step response tuning target effect diagram shows that this tuning is close to The expected goal is to meet the first-order inertial link requirement of a target response bandwidth of 5 rad / s.

[0127] According to the above target adjustment, g is finally determined c11 , g c12 , g c2 , g c31 and g c32 for:

[0128] (18);

[0129] (19);

[0130] (20);

[0131] During the simulation, the step size of the Newton-Raphson solver was uniformly set to 0.1, and the perturbation amplitude of the online gradient was 5% (for A8) and -3% (for A 163 According to the simulation conditions, set any α8 and α under 75% thrust.163 The position is used as the initial point for optimization. Figure 3 The optimization results of 10 groups of different initial points are given. It can be seen from the figure that the outer loop optimization loop can start from any initial point and make the engine converge to the minimum fuel consumption point under the thrust, showing a good optimization effect.

[0132] Figure 4(a)-Figure 4(c) It is given in Figure 3 When the initial point is ①, the time evolution trajectory of each parameter. and ; Figure 4(b) is F n , low pressure rotor speed (N l ) and high pressure rotor speed (N h Figure 4(c) shows the temperature before the turbine (T 41 )、High pressure compressor surge margin(SM HPC ) and low pressure compressor surge margin (SM LPC ).

[0133] As can be seen from the figure, after ten gradient descents, Finally, it converged to 114%. Finally, it converges to 88.15%. At this time, SFC drops from the initial 0.596 (kg / (daN*h)) to 0.566 (kg / (daN*h)), a decrease of 5.03% compared to the initial value, which means that the aircraft's cruising range is extended and its combat effectiveness is improved. n Fluctuation shows that F n The maximum fluctuation of does not exceed 0.5%, indicating that the inner loop control can effectively suppress the disturbance of thrust. These fluctuations are mainly caused by online disturbances, which are used to estimate the gradient and will not adversely affect the stability and safety of the system.

[0134] Figure 4(a)-Figure 4(c) The figure shows the process of the Newton-Raphson solver reaching the minimum fuel consumption point after 300 seconds. When the thrust finally stabilizes at 63.38 kN, the low-pressure rotor speed (N l ) from 87.1% to 91.8%, high pressure rotor speed (N h ) from 92.6% to 92.2%, T 41 From 1727K to 1497K, SM HPC and SM LPC remained at 23.8% and 53.7% respectively. Figure 7(a)-Figure 7(e) It can be seen that there is no overspeed, overheating or surge in the whole process. Specifically, the low-pressure rotor speed (N l ) and high pressure rotor speed (N h ) is always kept within a safe range, and the turbine inlet temperature (T41 ) does not exceed the design limit, and the surge margin (SM HPC and SM LPC ) did not drop to dangerous levels, which shows that the optimization method effectively improves engine performance while ensuring safety.

[0135] The above simulation results show that this method can effectively reduce the fuel consumption rate under constant speed cruising conditions, thereby achieving long-term performance optimization. Therefore, this new double-layer architecture provides new ideas and means for the performance optimization of modern aviation aircraft power systems, and shows broad application prospects in improving aircraft cruise performance and fuel economy.

[0136] Obviously, those skilled in the art can make various changes and modifications to the technical solution without departing from the spirit and scope of the technical solution. Thus, if these modifications and variations of the technical solution fall within the scope of the claims of the technical solution and their equivalents, the technical solution is also intended to include these modifications and variations.

Claims

1. A multivariable performance optimization method based on a stabilization-optimization dual loop, characterized in that: include: Design the inner control loop of the aircraft engine according to the control requirements; The inner control loop is a feedforward-feedback combined control loop, wherein the feedforward control loop is used to improve the rapidity of thrust adjustment and respond to input changes in advance, and includes an A8 loop based on the tail nozzle throat area and an A loop based on the rear duct ejector area. 163 Loop; The feedback control loop is used to correct the error based on thrust feedback, which is W based on the fuel flow f loop; The control requirements include the following four aspects: achieving the ideal thrust closed-loop bandwidth; after the thrust setting value changes, the tail nozzle throat area A8 and the rear duct ejector area A 163 It can always return to the set value; avoid changing A8 and A during the optimization process of the outer performance optimization loop 163 Disturbance thrust; A8 circuit can be reference value of A8 ΔA 8_ref Respond quickly, A 163 The circuit can 163 Reference value ΔA 163_ref Respond quickly; Design the outer performance optimization loop of the aircraft engine to globally optimize the fuel consumption rate; Determine the parameters of the feedforward-feedback combination controller based on the designed inner control loop; The thrust is controlled by the parameters to maintain it at the desired value, and the outer performance optimization loop is designed to dynamically adjust ΔA 8_ref and ΔA 163_ref , achieving a balance between fuel efficiency and thrust, so that the engine can achieve the lowest fuel consumption rate under a specific thrust.

2. The multivariable performance optimization method based on the stabilization-optimization dual loop as claimed in claim 1, characterized in that: The feedforward-feedback combined controller includes an A8 controller, a W f Controller and A 163 The A8 circuit includes an A8 controller, an A8 actuator and an engine. 163 The circuit contains A 163 Controller, A 163 Actuator and engine, the W f The circuit contains W f Controller, W f Actuators and motors; The feedforward-feedback combined controller receives a thrust reference value ΔF n,ref As input, the input signal passes through W f The transfer function of the controller is g c2 After processing, send to W f actuator, and at the same time, the A8 controller transfers its function g c11 and g c12 Process the input signal and output it to the A8 actuator, A 163 The controller transfers the function g c31 and g c32 Process the input signal and output it to A 163 Actuator, A8 actuator through its transfer function g a1 The tail nozzle throat area A8 changes, W f The actuator passes through its transfer function g a2 Driving fuel flow W f Change, A 163 The actuator passes through its transfer function g a3 Driving rear duct ejector area A 163 Change, the engine is based on the A8 actuator, W f Actuator and A 163 The change in actuator output generates thrust deviation ΔF through transfer functions g1, g2 and g3 n , where g1 is the deviation from A8 to ΔF n The transfer function of g2 is from W f Deviation ΔW f to ΔF n The transfer function of g3 is from A 163 Deviation ΔA 163 to ΔF n The transfer function, finally, converts the thrust deviation ΔF n Feedback to the input of the feedforward-feedback combined controller and the thrust reference value ΔF n,ref Form a closed-loop control.

3. The multivariable performance optimization method based on the stabilization-optimization dual loop as claimed in claim 2, characterized in that: The ideal thrust closed-loop bandwidth is specifically: Using the loop shaping method, the ideal open-loop transfer function L of the inner control loop is specified as: （1）； s is the complex variable of Laplace transform, ω b1 is the closed-loop transfer function bandwidth of the inner control loop, which is specifically set to 10 rad / s. The corresponding closed-loop transfer function is the inertia link, and the corresponding closed-loop transfer function G y1 (s) is the thrust reference value ΔF n,ref To thrust deviation ΔF n The transfer function is defined as: （2）； Design the performance index of the ideal thrust bandwidth T1, and the control target is: （3）； Among them, γ1 is the error tolerance and K is the controller.

4. The multivariable performance optimization method based on the stabilization-optimization dual loop as claimed in claim 3, characterized in that: The open loop transfer function .

5. The multivariable performance optimization method based on the stabilization-optimization dual loop as claimed in claim 2, characterized in that: After the thrust setting value changes, the tail nozzle throat area A8 and the rear duct ejector area A 163 It is always possible to return to the setpoint, which is met by pre-specifying the controller structure: W f The controller is taken as PI controller, A 163 The controller and A8 controller are composed of a PD controller and a lead-lag compensator without an integrator, where g c2 is the transfer function of the PI controller; g c11 and g c12 are the transfer function of the lead-lag compensator without integrator and the transfer function of the PD controller in the A8 controller, respectively; g c31 and g c32 A 163 The transfer function of the lead-lag compensator and the transfer function of the PD controller without the integrator in the controller.

6. The multivariable performance optimization method based on the stabilization-optimization dual loop as claimed in claim 5, characterized in that: W f Controller, A8 Controller, A 163 The transfer function of the controller is specifically: （4）； （5）； （6）； Among them, K p is the proportionality coefficient, K i is the integration coefficient, K d is the differential coefficient, T is the time constant, s is the complex variable of Laplace transform, and θ1 and θ1 are the coefficients to be determined.

7. The multivariable performance optimization method based on the stabilization-optimization dual loop as claimed in claim 2, characterized in that: Avoid changing A8 and A during the optimization process of the outer performance optimization loop 163 Disturbance thrust, specifically: Defining ΔA8 and ΔA 163 To thrust deviation ΔF n The transfer functions g1 and g3 are and , The following performance indicators of disturbance suppression T2 and T3 are designed, and the control objectives are: （7）； Among them, s is the complex variable of Laplace transform, γ2, γ3 are error tolerances, and K is the controller.

8. The multivariable performance optimization method based on the stabilization-optimization dual loop as claimed in claim 2, characterized in that: The A8 loop can be used to determine the reference value ΔA of A8. 8_ref Respond quickly, A 163 The circuit can 163 Reference value ΔA 163_ref Respond quickly, specifically: Remember G y2 is the input ΔA from the reference 8,ref The transfer function to the actual output ΔA8 is denoted by G y3 is the input ΔA from the reference 163,ref To the actual output ΔA 163 The transfer function is: （8）； Among them, ω b2 and ω b3 G y2 and G y3 The bandwidth of , s is the complex variable of Laplace transform; Design the servo tracking performance indicators of T4 and T5, and the control objectives are: （9）； Among them, s is the complex variable of Laplace transform, γ4, γ5 are error tolerances, and K is the controller.

9. The multivariable performance optimization method based on the stabilization-optimization dual loop as claimed in claim 1, characterized in that: The outer performance optimization loop adopts the Newton-Raphson iterative algorithm to calculate the first-order derivative of the objective function and quickly approach the global optimal solution.

10. The multivariable performance optimization method based on the stabilization-optimization dual loop as claimed in claim 9, characterized in that: The control objectives of the outer performance optimization loop are: （10）； That is, by adjusting ΔA 8,ref and ΔA 163,ref Make J minimum while keeping the thrust at the desired value, where SFC is the fuel consumption rate.

Images