A design method for component-level airborne dynamic model of a turbofan engine with adjustable guide vanes
By introducing compressor guide vane characteristic correction and LPV model in the aero engine model, combined with the calculation strategy of optimized state parameter inertia links, the problem of insufficient real-time performance of component-level models is solved, and a high-precision and high-reality airborne dynamic model is realized , significantly reduces calculation time and controls errors at lower levels.
Patent Information
- Application Number
- CN202210424287.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-04-21
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2042-04-21
AI Technical Summary
The nonlinear component-level models of existing aircraft engines are insufficient in real-time, and the linearized model has large errors, making it difficult to meet the real-time simulation verification requirements of the control algorithm.
Based on the original component-level model, the iterative solution algorithm for nonlinear joint working equations in the form of LPV is designed, and the working conditions containing the force width of the compression ratio state parameters after the force is turned on are established. Range LPV model, and an optimization state parameter inertia link calculation strategy is designed to obtain a wide operating range dynamic model with smooth switching between different flight states within the envelope.
Without loss of accuracy, the real-timeness of the engine model is significantly improved, the error caused by coefficient matrix replacement is reduced, and the calculation time-consuming reduction of more than 80% is achieved. The maximum steady-state error and dynamic error are not exceeding 5%.
Smart Images

Figure CN114912191B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of aeroengine modeling and simulation, and in particular to the design of an onboard dynamic model of an afterburner turbofan engine with adjustable guide vanes. Background Art
[0002] Modern aircraft engines are extremely complex nonlinear aerodynamic thermodynamic systems, which require a lot of manpower and capital investment in the research and development process. The mathematical model of aircraft engines is the basis of control system design research, which includes three types of aircraft engine mathematical models: component level model, state variable model and LPV model.
[0003] Component-level model: It has high accuracy and can fully reflect the working status of each section of the engine. However, since it requires repeated flow iteration calculations during calculation, its real-time performance is poor. When verifying the control algorithm in real-time simulation, the mathematical model is required to have both high accuracy and good real-time performance.
[0004] State variable model: A model established by using the local linearization of a nonlinear component-level model. It is currently widely used in the design of variable engine controllers. The methods for establishing aero-engine state variable models include partial derivative method, fitting method and least squares method.
[0005] LPV model: Since the state variable model is a small-range linear model, it is difficult to meet the wide-range accuracy requirements of the engine, so the LPV modeling method needs to be applied. The LPV model describes the dynamic characteristics of the system through measurable external real-time parameters as scheduling parameters. The LPV controller is directly designed using linear robust control theory to ensure that the designed controller has robust stability. The LPV system has high potential value in both engineering applications and theoretical research. Therefore, in recent years, LPV modeling technology has received great attention from the control community.
[0006] Afterburner combustion is generally completed in the afterburner chamber after the low-pressure turbine of the engine and before the nozzle. The afterburner chamber can greatly increase the thrust of the engine under certain circumstances, so that the aircraft can gain a temporary advantage, and its role cannot be ignored. The present invention combines the models of various components of the turbofan engine and the established LPV model of the speed and pressure ratio, and proposes a method for constructing an onboard real-time model of a turbofan engine with afterburner and adjustable guide vanes, so as to improve the real-time performance of the engine model with less loss of accuracy. Summary of the invention
[0007] The technical problem to be solved by the present invention is to provide an airborne dynamic real-time model with high-precision adjustable guide vanes in response to the defects of the background technology, so as to solve the problems of insufficient real-time performance of the original nonlinear component-level model and large linear model error.
[0008] The present invention adopts the following technical solutions to solve the above technical problems:
[0009] Step A), based on the original component-level model, introduce the correction of the compressor guide vane characteristics; design a non-iterative solution algorithm for the nonlinear joint working equations in the LPV form, and according to the data of the component-level model when the engine is turned on, take the afterburner fuel quantity as one of the control quantities, and establish an LPV model with a wide operating range including afterburner for the pressure ratio state parameters after the afterburner is turned on;
[0010] Step B), design, introduce and optimize the state parameter inertia link calculation strategy, thereby obtaining a wide range of airborne dynamic models for smooth switching of different flight states within the envelope.
[0011] As a further optimization scheme of the component-level airborne dynamic model design method of a turbofan engine with adjustable guide vanes according to the present invention, the specific steps of step A) are as follows:
[0012] Step A1), based on the original component-level model, the first three stages and the last six stages of the compressor are modeled separately, and the compressor guide vane characteristic correction is introduced.
[0013] Step A2), based on the data of the component-level model when the engine is powered on, the fuel quantity of the power-on is used as one of the control quantities, and a power-on state variable model containing the pressure ratio state parameters after the power-on is turned on is established.
[0014] Step A3), at different inlet temperatures, at different throat areas A8 and different guide vane angles Afa, the state variable models at different speeds at the throat areas and guide vane angles are combined into an LPV model.
[0015] As a method for constructing an onboard dynamic real-time model of a turbofan engine with afterburner, an optimization scheme for reducing errors is proposed. The specific steps of step B) are as follows:
[0016] Step B1), establish a flight envelope division scheme based on isotherms. According to the flight envelope division scheme, optimize and select several envelope points under T2 to establish an LPV model with a wide range of operating conditions including afterburner and pressure ratio state parameters after afterburner is turned on;
[0017] Step B2), for the flight points within the full envelope, the design introduces and optimizes the state parameter inertia link calculation strategy. The LPV model is used to solve the converted speed and pressure ratio, which are provided to the aerodynamic thermodynamic model of each component of the gas path to complete the calculation, thereby obtaining an airborne dynamic model with a wide range of operating conditions and smooth state switching.
[0018] Furthermore, the flight envelope division scheme in step B1) is as follows: within the envelope, an isotherm is established at an inlet temperature T2 every 20K, thereby obtaining an LPV model with a wide range of afterburner conditions within the envelope.
[0019] Furthermore, the specific steps in step B2) are as follows:
[0020] Step B2.1), according to the inlet total temperature T2 of the turbofan engine, calculate the difference between T2 and T2 of the established LPV model. When the difference between T2 and T2 of the established LPV model does not exceed 20K, the LPV model with the smallest difference can be selected. Select the polynomial coefficients under the corresponding isotherm, calculate the elements in the coefficient matrix, and then obtain the current high-pressure speed n H The coefficient matrices A, B, C, and D under the current throat cross-sectional area A8 are interpolated to calculate the speed-pressure ratio.
[0021] Step B2.2), based on the original model, the state parameter inertia link calculation strategy is designed, introduced and optimized. The state parameter inertia link calculation strategy is shown in Figure 1(b).
[0022] Step B2.3), the pressure ratio and speed are sent back to the nonlinear model and provided to the aerodynamic thermodynamic model of each gas path component to complete the calculation, thereby obtaining an airborne dynamic model with a wide operating range and smooth state switching.
[0023] Furthermore, the specific steps of the optimization strategy in step B2.2) are as follows: in the process of switching the coefficient matrix, determine whether the change of the input parameters requires changing the matrix coefficients. If not, continue to use the matrix coefficients of the previous moment. If the matrix coefficients need to be changed, repeat the calculation several times to achieve a steady-state effect; after obtaining the speed and pressure ratio, introduce the state parameter inertia link of the speed and pressure ratio to reduce the output error of the model in the process of switching the matrix. The transfer function of the first-order inertia link is expressed as follows:
[0024]
[0025] Where T is the time constant and s represents the complex frequency domain.
[0026] Compared with the existing solutions, the present invention adopts the above technical solutions and has the following technical effects:
[0027] (1) The component-level model used in the present invention is based on the original component-level model, and the first six stages and the last three stages of the compressor are modeled separately, and the compressor guide vane characteristic correction is introduced;
[0028] (2) The onboard dynamic real-time model used in the present invention takes the afterburner fuel quantity as one of the control quantities and establishes an onboard afterburner combustion chamber model;
[0029] (3) After solving the LPV model, the present invention introduces and optimizes the state parameter inertia link calculation strategy, which can effectively reduce the error caused by the coefficient matrix replacement. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] In Figure 1, (a) is a schematic diagram of the principle of the onboard dynamic real-time model of a turbofan engine with afterburner, and (b) is a detailed description of the optimization calculation strategy;
[0031] Figure 2 It is an envelope partitioning scheme based on isotherms;
[0032] Figure 3 H = 0 km, Ma = 0, T t2 = Normalized fuel flow rate W at 288.15K f , afterburner fuel flow W fa and the tail nozzle throat interface area A8 changes;
[0033] Figure 4 H = 0 km, Ma = 0, T t2 = Output parameter comparison when 288.15K;
[0034] Figure 5 H = 11km, Ma = 0.8, T t2 = Normalized fuel flow rate W at 244.38K f , afterburner fuel flow W fa and the tail nozzle throat interface area A8 changes;
[0035] Figure 6 H = 11km, Ma = 0.8, T t2 = Output parameter comparison when 244.38K;
[0036] Figure 7 H = 11km, Ma = 1.5, T t2 = Normalized fuel flow rate W at 314.3K f , afterburner fuel flow W fa and the tail nozzle throat interface area A8 changes;
[0037] Figure 8 H = 11km, Ma = 1.5, T t2 = Output parameter comparison when 314.3K. DETAILED DESCRIPTION
[0038] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings.
[0039] The idea of the present invention is to introduce the correction of the compressor guide vane characteristics on the basis of the original component-level model; according to the data of the component-level model when the engine is turned on, the afterburner fuel amount is used as one of the control quantities, and a wide operating range LPV model of the pressure ratio state parameters after the afterburner is turned on is established; the corresponding LPV model is determined according to the flight point in the envelope, and the calculation strategy of the inertia link of the state parameter is designed, introduced and optimized, thereby obtaining a wide operating range airborne dynamic model with smooth state switching.
[0040] The specific implementation of the present invention takes the design of a component-level airborne dynamic real-time model of a turbofan engine with afterburner as an example. FIG. 1( a ) is a schematic diagram of the design of an airborne dynamic model of a turbofan engine. The present invention describes a method for designing a component-level airborne dynamic model of a turbofan engine with adjustable guide vanes, which specifically includes the following steps:
[0041] Step A), based on the original component-level model, introduce the correction of the compressor guide vane characteristics; design a non-iterative solution algorithm for the nonlinear joint working equations in the LPV form, and according to the data of the component-level model when the engine is turned on, take the afterburner fuel quantity as one of the control quantities, and establish an LPV model with a wide operating range including afterburner for the pressure ratio state parameters after the afterburner is turned on;
[0042] Step A1), based on the original component-level model, the first three stages and the last six stages of the compressor are modeled separately, and the compressor guide vane characteristic correction is introduced.
[0043] When the guide vane angle is less than 10°, the guide vane angle can be used for interpolation calculation:
[0044] W=W1-(W1-W2)·Afa / 10
[0045] eff=eff1-(eff1-eff2)·Afa / 10
[0046] W and eff are the flow rate and efficiency under this guide vane angle, W1 and eff1 are the flow rate and efficiency obtained by interpolation using the first characteristic diagram, and W2 and eff2 are the flow rate and efficiency obtained by interpolation using the second characteristic diagram.
[0047] Step A2), based on the data of the component-level model when the engine is powered on, the fuel quantity of the power-on is used as one of the control quantities, and a power-on state variable model containing the pressure ratio state parameters after the power-on is turned on is established.
[0048] The state variable model of the engine at the steady-state point (x0, u0, y0) is:
[0049]
[0050] Δy=CΔx+DΔu
[0051] In the formula, the state quantity x is the high and low pressure rotor speed N H and N L , the control quantity u is the fuel supply to the afterburner chamber W fa and tail nozzle throat area A8, output y is fan pressure ratio π fan 、Low pressure compressor pressure ratio π lcomp 、High pressure compressor pressure ratio π hcomp 、High-pressure turbine pressure ratio π hturb and low-pressure turbine pressure ratio π lturb , Δx=x-x0, Δy=y-y0.
[0052] The parameters are similarly normalized as follows:
[0053]
[0054]
[0055] Where PN L PN H , PW fa , PA8 are respectively the normalized high pressure rotor speed, low pressure rotor speed, afterburner fuel supply, tail nozzle throat area, T2 is the engine inlet total temperature, subscript d s Represents the engine design point parameters, and the engine state variable model represented by similar normalized parameters is as follows:
[0056]
[0057]
[0058] In the formula, Denote matrices A, B, C, and D respectively, and the small perturbation method is used to obtain the initial solutions of the coefficient matrices A, B, C, and D of this working point.
[0059] Compared with the direct calculation of the state variable model without afterburner, one of the control variables is A8, but when calculating the state variable model with afterburner, the fuel quantity in the main combustion chamber W f At this time, the maximum value needs to be reached, so the fuel volume in the main combustion chamber W f The default is the maximum value. When using the afterburner, W is not used. f is the control quantity, and the other control quantity is the fuel supply to the afterburner W fa ;
[0060] Step A3), at different inlet temperatures, different throat areas A8 and different guide vane angles Afa, a large number of state variable models are combined into an LPV model:
[0061]
[0062] The state variable models of the engine under different throat areas A8, guide vane angles afa and different high-pressure speeds constitute the LPV model of speed and pressure ratio. Then, the similarity conversion theory is used to expand the model within the full envelope, and the LPV model of the engine is described as:
[0063]
[0064] In the formula, x is the state quantity, y is the output quantity, and the subscript cor represents the similarity conversion to the ground point parameters.
[0065] In order to reduce the space for data storage, a third-order polynomial is fitted for each element in the ABCD matrix with a number of k:
[0066]
[0067] Obtained by multiple trials using the fitting method Therefore, a dynamic real-time model of LPV regarding pressure ratio is established when boost is turned on over a wide range of operating conditions.
[0068] Step B), design, introduce and optimize the state parameter inertia link calculation strategy, thereby obtaining a wide range of airborne dynamic models for smooth switching of different flight states within the envelope.
[0069] Step B1), establish a flight envelope division scheme based on isotherms, the flight envelope division scheme is shown in Figure (2). According to the flight envelope division scheme, optimize and select several envelope points under T2 to establish a wide range of LPV models containing afterburner pressure ratio state parameters after afterburner is turned on;
[0070] In step B1), the flight envelope division scheme is as follows: within the envelope range, an isotherm is established at an inlet temperature T2 every 20K. The flight envelope division scheme is as follows: Figure 2 As shown, the LPV model with a wide operating range including boost is obtained.
[0071] Step B2), for the flight points within the full envelope, the design introduces and optimizes the state parameter inertia link calculation strategy. The LPV model is used to solve the converted speed and pressure ratio, which are provided to the aerodynamic thermodynamic model of each component of the gas path to complete the calculation, thereby obtaining an airborne dynamic model with a wide range of operating conditions and smooth state switching.
[0072] Step B2.1), using the LPV dynamic real-time model of the pressure ratio state parameter including force, record the five pressure ratios solved by the iterative convergence of the nonlinear model under different working conditions.
[0073] According to the total inlet temperature T2 of the turbofan engine, when the difference between T2 and the T2 of the established LPV model does not exceed 20K, the LPV model with a smaller adjacent temperature difference can be used. Select the polynomial coefficients under the corresponding isotherm, calculate the elements in the coefficient matrix, and then obtain the current high-pressure speed n H The coefficient matrices A, B, C, and D under the current throat area A8 are interpolated to calculate the speed pressure ratio under the current throat area A8. If the current throat area is A8, when A 8i ≤A8≤A 8i+1 When, according to A 8i+1 and A 8i Calculate the current Afa and W respectively fa The elements in the next two groups of coefficient matrices are used to obtain the current high-pressure speed n H The coefficient matrices A, B, C, and D under the above formula are solved by the LPV model to calculate the speed and the pressure ratio of each component as follows;
[0074]
[0075]
[0076] In the formula, n=[n L ,n H ] T ,π=[π Fan ,π Lcomp ,π Hcomp ,π HTurb ,π LTurb ] T is the pressure ratio of the five rotating parts, u=[W fa ,A8] T , subscript 0 is the steady-state point. According to A 8i and A 8i+1 Calculated here W fa Two sets of speed and pressure ratio below.
[0077] The speed-pressure ratio under the current throat cross-sectional area A8 is calculated by interpolation based on the current A8.
[0078]
[0079] Step B2.2), since the output parameters of the component-level dynamic real-time model have jitter errors before correction, the design introduces and optimizes the state parameter inertia link calculation strategy based on the original model.
[0080] The optimization strategy is shown in Figure 1(b). The specific optimization strategy is: in the process of switching the coefficient matrix, determine whether the change of the input parameters requires changing the matrix coefficient. If not, continue to use the matrix coefficient of the previous moment. If the matrix coefficient needs to be changed, repeat the calculation 100 times to achieve a steady-state effect. After obtaining the speed and pressure ratio, the state parameter inertia link of the speed and pressure ratio is introduced to reduce the output error of the model in the process of switching the matrix. The transfer function of the first-order inertia link is expressed as follows:
[0081]
[0082] Where T is the time constant;
[0083] In step B2.3), the obtained pressure ratio and speed are sent back to the nonlinear model to form a component-level airborne dynamic model together with the aerodynamic thermodynamic models of each component of the gas path, so that the nonlinear model with afterburner can obtain the key output parameters P25, T25, Ps3, P6, T6 of the engine at different working points without iteration. If the error does not exceed 2%, an airborne dynamic model with a wide operating range and smooth state switching can be obtained.
[0084] In order to verify the effectiveness of the design of the onboard dynamic model of the turbofan engine with adjustable guide vanes designed in the present invention, a digital simulation of the output of the onboard dynamic model with adjustable guide vanes within the envelope was carried out in the MATLAB environment.
[0085] First, the inlet temperature T t2 The six isotherms covering the entire envelope are [244.38K, 259K, 273K, 288.15K, 314.3K, 352.9K] respectively. The envelope division scheme is as follows: Figure 2 shown.
[0086] Under these six isotherms, the following data of turbofan engine with afterburner at different high pressure speeds are calculated respectively: A8=[0.191, 0.211, 0.218, 0.266, 0.314, 0.362, 0.41], afa=[0, 0.156, 0.279, 0.351, 0.369, 0.4, 1.4, 2.4, 3.4, 4.4, 5.4, 6.4, 7.4, 8, 9, 10]. The state variable model of speed and pressure ratio, i.e., coefficient matrix ABCD, is established. The corresponding elements of the coefficient matrix at different high pressure speeds are fitted with a third-order polynomial to obtain the polynomial fitting coefficients of the ABCD matrix elements at different A8, afa and different high pressure speeds in the afterburner mode.
[0087] The simulation test was carried out at three working points: (H = 0 km, Ma = 0), (H = 11 km, Ma = 0.8), and (H = 11 km, Ma = 1.5). The fuel flow rate W f, afterburner fuel flow W fa The interface area A8 of the nozzle throat changes as Figure 3 , Figure 5 , Figure 7 As shown in the figure, the guide vane angle Afa is kept constant at 0. The output comparison between the normalized airborne dynamic real-time model and the dynamic model is as follows: Figure 4 , Figure 6 , Figure 8 The maximum steady-state error and the maximum dynamic error are shown in Tables 1 and 2. In Tables 1 and 2, the maximum steady-state error does not exceed 1%, and the maximum steady-state error does not exceed 5%, and the results are highly accurate. The time consumption comparison is shown in Table 3. In Table 3, the time consumption of the airborne dynamic real-time model is reduced by more than 80% compared with the component-level model. Based on the above simulation results, this method achieves the purpose of obtaining higher real-time performance with less loss of accuracy.
[0088] Table 1 Maximum steady-state error (%)
[0089]
[0090] Table 2 Maximum dynamic error (%)
[0091]
[0092] Table 3 Time comparison
[0093]
[0094] It should be pointed out that the above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes and substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.
Claims
1. A method for designing a component-level airborne dynamic model of a turbofan engine with adjustable guide vanes, characterized in that: The following steps are involved: Step A), based on the original component-level model, introduce the correction of the compressor guide vane characteristics; design a non-iterative solution algorithm for the nonlinear joint working equations in the LPV form, and according to the data of the component-level model when the engine is turned on, take the afterburner fuel quantity as one of the control quantities, and establish an LPV model with a wide operating range including afterburner for the pressure ratio state parameters after the afterburner is turned on; Step B), designing, introducing and optimizing the state parameter inertia link calculation strategy, thereby obtaining a wide range of airborne dynamic models for smooth switching of different flight states within the envelope; The specific steps of step A) are as follows: Step A1), based on the original component-level model, the first three stages and the last six stages of the compressor are modeled separately, and the compressor guide vane characteristic correction is introduced; Step A2), based on the data of the component-level model when the engine is powered on, the fuel quantity of the power-on is used as one of the control quantities, and a power-on state variable model containing the pressure ratio state parameters after the power-on is turned on is established; Step A3), at different inlet temperatures, at different throat areas A8 and different guide vane angles Afa, the state variable models at different speeds at the throat areas and guide vane angles are combined into an LPV model; The specific steps of step B) are as follows: Step B1), establishing a flight envelope division scheme based on isotherms, and according to the flight envelope division scheme, optimizing and selecting a number of envelope points at the inlet total temperature T2 to establish an afterburner-containing wide operating range LPV model for pressure ratio state parameters after afterburner is turned on; Step B2), for the flight points within the full envelope, the design introduces and optimizes the state parameter inertia link calculation strategy; the converted speed and pressure ratio are obtained by solving the LPV model, and provided to the aerodynamic thermodynamic model of each component of the gas path to complete the calculation, thereby obtaining an airborne dynamic model with a wide operating range and smooth state switching.
2. The method for designing a component-level airborne dynamic model of a turbofan engine with adjustable guide vanes according to claim 1, characterized in that: The flight envelope division scheme in step B1) is as follows: within the envelope, an isotherm is established at an inlet temperature T2 every 20K, and a wide-range LPV model containing afterburner is established within the envelope based on the isotherms.
3. The method for designing a component-level airborne dynamic model of a turbofan engine with adjustable guide vanes according to claim 1, characterized in that: The specific steps of step B2) are as follows: Step B2.1), according to the inlet total temperature T2 of the turbofan engine, calculate the difference between T2 and T2 of the established LPV model. When the difference between T2 and T2 of the established LPV model does not exceed 20K, the LPV model with the smallest difference can be selected; select the polynomial coefficient under the corresponding isotherm, calculate the elements in the coefficient matrix, and then obtain the current high-pressure speed n H The coefficient matrices A, B, C, and D under the current throat cross-sectional area A8 are interpolated to calculate the speed-pressure ratio; Step B2.2), based on the original model, design, introduce and optimize the state parameter inertia link calculation strategy; Step B2.3), the pressure ratio and speed are sent back to the nonlinear model and provided to the aerodynamic thermodynamic model of each gas path component to complete the calculation, thereby obtaining an airborne dynamic model with a wide operating range and smooth state switching.
4. The method for designing a component-level airborne dynamic model of a turbofan engine with adjustable guide vanes as claimed in claim 3, characterized in that: The specific steps of the optimization strategy in step B2.2) are as follows: in the process of switching the coefficient matrix, determine whether the change of the input parameters requires changing the matrix coefficient. If not, continue to use the matrix coefficient of the previous moment. If the matrix coefficient needs to be changed, repeat the calculation several times to achieve a steady-state effect; after obtaining the speed and pressure ratio, introduce the state parameter inertia link of the speed and pressure ratio to reduce the output error of the model in the process of switching the matrix. The transfer function of the first-order inertia link is expressed as follows: Where T is the time constant and s represents the complex frequency domain.
Citation Information
Patent Citations
Afterburner overall excess air coefficient design method
CN107122537A
Aero-engine system identification method based on multi-cell folded equilibrium manifold expansion model
CN110362960A
Physical and data-driven aero-engine digital twin modeling method
CN111475889A
Fuzzy reliability optimization method for power assembly suspension system
CN112733266A
Turbofan engine airborne dynamic model full envelope precision expansion method
CN114297786A
