CN111177930B - Aeroengine system identification modeling method based on mixed sequence - Google Patents
Aeroengine system identification modeling method based on mixed sequence Download PDFInfo
- Publication number
- CN111177930B CN111177930B CN201911416076.1A CN201911416076A CN111177930B CN 111177930 B CN111177930 B CN 111177930B CN 201911416076 A CN201911416076 A CN 201911416076A CN 111177930 B CN111177930 B CN 111177930B
- Authority
- CN
- China
- Prior art keywords
- pressure rotor
- speed
- polynomial
- identified
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 57
- 239000000446 fuel Substances 0.000 claims description 10
- 238000000611 regression analysis Methods 0.000 claims description 5
- 230000005284 excitation Effects 0.000 claims description 3
- 239000000295 fuel oil Substances 0.000 claims description 3
- 241000208340 Araliaceae Species 0.000 claims 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 claims 1
- 235000003140 Panax quinquefolius Nutrition 0.000 claims 1
- 235000008434 ginseng Nutrition 0.000 claims 1
- 238000012417 linear regression Methods 0.000 abstract description 5
- 239000011159 matrix material Substances 0.000 description 10
- 230000004044 response Effects 0.000 description 9
- 238000013461 design Methods 0.000 description 6
- 238000012795 verification Methods 0.000 description 6
- 238000013016 damping Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 238000002156 mixing Methods 0.000 description 5
- 238000002790 cross-validation Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000002945 steepest descent method Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 229910002056 binary alloy Inorganic materials 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000002485 combustion reaction Methods 0.000 description 1
- 238000000695 excitation spectrum Methods 0.000 description 1
- 239000012530 fluid Substances 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Landscapes
- Feedback Control In General (AREA)
Abstract
The invention discloses an aeroengine system identification modeling method based on a mixed sequence, which comprises the steps of firstly establishing a component-level model of an aeroengine, and converting the linear time-invariant modeling problem of a single working state into a system identification problem for solving least square according to the input/output and state quantity data of each working state point; secondly, selecting a proper model class, an estimation method and a signal source to solve a least square problem, and establishing a linear time-invariant model of a single working state; and finally, establishing a polynomial nonlinear system under the ground condition in a linear regression mode. The method can effectively solve the problem of state space modeling of the aircraft engine in the field of multivariable control.
Description
Technical Field
The invention relates to the technical field of aerospace, in particular to an aeroengine system identification modeling method based on a mixed sequence.
Background
Because the aero-engine is a complex fluid machine integrating multiple disciplines such as aerodynamics, combustion science, structural strength, automatic control and materials, and a dynamic system of the aero-engine has the characteristics of strong nonlinearity and complexity and variability, modeling and control of the aero-engine system have certain difficulty, and obtaining a state variable model beneficial to design of a multivariable control system is challenging and very important work.
At present, domestic research on modeling of the aero-engine basically stays in an aero-thermodynamic nonlinear model of the aero-engine based on a component method. The basic principle of the model is that according to the Brayton cycle principle followed by the aero-engine and the working characteristics of each component, an iterative algorithm is adopted to solve three nonlinear balance equations of a flow continuous equation, a pressure balance equation and a power balance equation to establish a common working model of the aero-engine.
The state variable model is of great importance in solving the problems of multivariable control rule design and state estimation of the aircraft engine, but because the domestic limit is the experimental condition, and the research on state variable modeling by adopting test data is little, the research on the state variable model mainly aims at the model established by a component method, and the linearization is realized on the basis. Some scholars obtain the step response of the nonlinear engine model by using the step disturbance signal, and then calculate the state space model by using a partial derivative method and a fitting method. However, the modeling method obtains a small-disturbance state variable model, and the model has certain limitation on full-envelope multivariable control.
In view of strong nonlinearity of an aero-engine model, but the existing research results are limited to a linear parametric system, a nonlinear model conforming to characteristics of the aero-engine needs to be established.
Disclosure of Invention
The invention provides an aeroengine system identification modeling method based on a mixed sequence aiming at the problem of aeroengine modeling of multivariable control, and the method establishes an aeroengine polynomial nonlinear state space model by using the mixed sequence and a system identification method of a least square theory.
In order to solve the problems, the invention is realized by the following technical scheme:
the aeroengine system identification modeling method based on the mixed sequence specifically comprises the following steps:
and 6, establishing a polynomial nonlinear model under the ground condition of the aircraft engine by using the linear time invariant model established at the plurality of working state points in the step 5, taking the rotating speed of a high-pressure rotor of the aircraft engine as a scheduling parameter, and adopting a regression analysis mode based on the idea of gain scheduling.
The common working equation of the aircraft engine in the step 1 with respect to flow continuity and power balance comprises: a balance equation of the flow of the fan outlet and the sum of the flow of the compressor and the inlet of the bypass; a balance equation of the inlet flow of the low-pressure turbine and the outlet flow of the high-pressure turbine; a balance equation of the sum of the inlet flow of the high-pressure turbine, the outlet flow of the compressor and the fuel quantity; a balance equation of the outlet flow of the tail nozzle and the sum of the outlet flow of the outer culvert, the outlet flow of the low-pressure turbine and the boosted fuel oil amount; a high pressure rotor power balance equation; low pressure rotor power balance equations.
In step 2, the model types required for system identification are:
in the formula, nLFor the low-pressure rotor speed, the speed of the rotor,
is the differential of the low-pressure rotor speed, nHThe rotation speed of the high-pressure rotor is set,
is the differential of the high-pressure rotor speed, WfbA main amount of fuel, A8Is the throat area, pi is the pressure increase ratio, a11,a12,a21,a22,b11,b12,b21,b22,c31,c32,d31,d32Are parameters to be identified.
In the step 4, the established polynomial nonlinear model under the ground condition of the aircraft engine is as follows:
in the formula, nLFor the low-pressure rotor speed, the speed of the rotor,
is the differential of the low-pressure rotor speed, nHThe rotation speed of the high-pressure rotor is set,
is the differential of the high-pressure rotor speed, WfbA main amount of fuel, A8Is the area of the throat, f11(nH) For the parameter a to be identified11Polynomial f fitted to the rotation speed of the high-pressure rotor12(nH) For the parameter a to be identified12Polynomial f fitted to the rotation speed of the high-pressure rotor21(nH) For the parameter a to be identified21Polynomial f fitted to the rotation speed of the high-pressure rotor22(nH) For the parameter a to be identified22Polynomial, g, fit-to-high-pressure rotor speed11(nH) Parameter b to be identified11Polynomial, g, fit-to-high-pressure rotor speed12(nH) Parameter b to be identified12Polynomial, g, fit-to-high-pressure rotor speed21(nH) Parameter b to be identified21Polynomial, g, fit-to-high-pressure rotor speed22(nH) Parameter b to be identified22Polynomial h fitted to the rotation speed of the high-pressure rotor21(nH) Parameter c to be identified21Polynomial h fitted to the rotation speed of the high-pressure rotor22(nH) Parameter c to be identified22Polynomial fit-up to high-pressure rotor speed21(nH) Parameter d to be identified21Polynomial fit-up to high-pressure rotor speed22(nH) Parameter d to be identified22Polynomial fitting at high pressure rotor speed.
In the step 4, the pseudo-random signal of the first segment of the mixed sequence is an M sequence.
The principle of the invention is as follows: firstly, establishing a component-level model of the aircraft engine, and converting the linear time-invariant modeling problem of a single working state into a system identification problem for solving least squares according to the input/output and state quantity data of each working state point; secondly, selecting a proper model class, an estimation method and a signal source to solve a least square problem, and establishing a linear time-invariant model of a single working state; and finally, establishing a polynomial nonlinear system under the ground condition in a linear regression mode. The method can effectively solve the problem of state space modeling of the aircraft engine in the field of multivariable control.
Compared with the prior art, the invention has the following characteristics:
1. and the state space models of all working state points are scheduled according to the rotating speed of the high-pressure rotor of the aircraft engine system, so that the conservatism of the models is effectively reduced.
2. The actual model of the aircraft engine can be more closely reflected by taking the actual high-pressure rotor rotation speed of the aircraft engine instead of the increment of the high-pressure rotor rotation speed as the state quantity.
3. The hybrid sequence is used as the input of the component model, so that the dynamic characteristics of the system can be fully excited, the model identified by the system has excellent generalization performance, and the comprehensive performance of the model is superior to that of other input sequences.
4. The state space model of the component model at each working state point can be accurately and quickly identified by adopting an L-M least square estimation method and a mixed sequence, and the model on the cross validation display can reflect the reality of the aircraft engine with high precision.
Drawings
FIG. 1 is a flow chart of a hybrid sequence based aircraft engine system identification modeling method.
FIG. 2 is a state quantity tracking diagram of a component level model at design point: (a) a response image representing the high pressure rotor speed, and (b) a response image representing the low pressure rotor speed.
FIG. 3 is a graph comparing the recognition results using the G-N algorithm: y is1Representing the result of the identification of the rotational speed of the low-pressure rotor, y2Representing the result of the identification of the speed of rotation of the high-pressure rotor, y3Representing the result of the pressure ratio identification.
FIG. 4 is a comparison graph of identification results using a step signal and L-M least squares estimation method: y is1Representing the result of the identification of the rotational speed of the low-pressure rotor, y2Representing the result of the identification of the speed of rotation of the high-pressure rotor, y3Representing the result of the pressure ratio identification.
FIG. 5 is a verification plot of the state space model and component model identified using the step signal and L-M least squares estimation method: (a) a response image representing the high pressure rotor speed, (b) a response image representing the boost ratio, (c) an error image representing the high pressure rotor speed, and (d) an error image representing the boost ratio.
Fig. 6 is a signal diagram of a pseudo-random sequence.
FIG. 7 is a comparison graph of identification results using a pseudo-random sequence and an L-M least squares estimation method: y is1Representing the result of the identification of the rotational speed of the low-pressure rotor, y2Representing the result of the identification of the speed of rotation of the high-pressure rotor, y3Representing the result of the pressure ratio identification.
FIG. 8 is a verification plot of the state space model and part model identified using the pseudo-random sequence and L-M least squares estimation method: (a) a response image representing the high-pressure rotor rotation speed, and (b) a response image representing the pressure increase ratio.
Fig. 9 is a signal diagram of a mixing sequence.
FIG. 10 is a comparison graph of the recognition results using the mixed sequence and L-M least squares estimation method: y is1Representing the result of the identification of the rotational speed of the low-pressure rotor, y2Representing the result of the identification of the speed of rotation of the high-pressure rotor, y3Representing the result of the pressure ratio identification.
FIG. 11 is a verification plot of the state space model and part model identified using the hybrid sequence and L-M least squares estimation method: (a) a response image representing the high pressure rotor speed, (b) a response image representing the boost ratio, (c) an error image representing the high pressure rotor speed, and (d) an error image representing the boost ratio.
FIG. 12 is a plot of a fit of a linear regression of the identified state space model b matrix: (a) a graph of a fitted curve of the parameter b11 representing the b matrix, (b) a graph of a fitted curve of the parameter b12 representing the b matrix, (c) a graph of a fitted curve of the parameter b21 representing the b matrix, and (d) a graph of a fitted curve of the parameter b22 representing the b matrix.
FIG. 13 is a verification image of a polynomial nonlinear state space model at a certain operating point: (a) a tracking curve representing the rotation speed of the high-pressure rotor, (b) a tracking curve representing the pressure increase ratio, (c) an error curve representing the rotation speed of the high-pressure rotor and the component model, and (d) an error curve representing the pressure increase ratio and the component model.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to specific examples.
The component model of the aircraft engine is modeled into a polynomial nonlinear model by taking a component-level model of the aircraft engine as an identification object and adopting an L-M least square estimation method and taking a mixed sequence as input at each working state point. The modeling problem of the aircraft engine is converted into a system identification problem based on an L-M least square estimation method, and the calculation complexity can be effectively reduced while the system identification precision is improved by improving a signal source mode.
Therefore, the invention designs an aeroengine system identification modeling method based on a mixed sequence, which establishes an aeroengine polynomial nonlinear model based on the mixed sequence and a system identification method of L-M least square estimation, as shown in FIG. 1, and specifically comprises the following steps:
step 1: a component-level model of the aircraft engine is established.
Solving a common working equation of the aero-engine about flow continuity and power balance based on a component method to establish a component-level model of the aero-engine, namely solving the common working equation of the aero-engine about flow continuity and power balance by a Newton-Raphson (Newton-Raphson) iteration method according to the Brayton cycle principle followed by the aero-engine and the working characteristics of each component to establish the common working model of the aero-engine. Where FIG. 2 is a state quantity tracking diagram of the component level model at the design point.
The common operating equation of an aircraft engine with respect to flow continuity and power balance includes the following 6: a balance equation of the flow of the fan outlet and the sum of the flow of the compressor and the inlet of the bypass; a balance equation of the inlet flow of the low-pressure turbine and the outlet flow of the high-pressure turbine; a balance equation of the sum of the inlet flow of the high-pressure turbine, the outlet flow of the compressor and the fuel quantity; a balance equation of the outlet flow of the tail nozzle and the sum of the outlet flow of the outer culvert, the outlet flow of the low-pressure turbine and the boosted fuel oil amount; a high pressure rotor power balance equation; low pressure rotor power balance equations.
Step 2: and determining the model class required by system identification.
The input and output data of the component-level model are identified by a system identification method, and the model class required by the system identification is determined according to the existing experience, as follows:
wherein n isLFor the low-pressure rotor speed, the speed of the rotor,
is the differential of the low-pressure rotor speed, nHThe rotation speed of the high-pressure rotor is set,
is the differential of the high-pressure rotor speed, WfbA main amount of fuel, A8Is the throat area, and pi is the pressure increase ratio (connotative outlet pressure P)6And fan inlet pressure P2A ratio of) a11,a12,a21,a22,b11,b12,b21,b22,c31,c32,d31,d32Are parameters to be identified.
The output quantity of the model class in the identification process is the rotating speed of the low-pressure rotor, the rotating speed of the high-pressure rotor and the pressure increasing ratio. The output quantity of the finally established model is the rotating speed and the pressure increasing ratio of the high-pressure rotor, and the output quantity of the rotating speed of the low-pressure rotor is added in the identification process, so that the identification result is closer to an optimal value.
And step 3: and determining an iterative estimation algorithm and a criterion adopted by system identification.
And converting the modeling problem of the aircraft engine into a least square problem identified by a system, and identifying an actual rotation speed linear time invariant model of each steady-state point by an L-M least square estimation method.
The L-M least squares method is a method in which the steepest descent method and Gauss-Newton (G-N) are combined. The steepest descent method is suitable for an algorithm in which the parameter estimation value is far from the optimal value at the initial stage of iteration, and the gauss-newton algorithm is suitable for a situation in which the parameter estimation value is in the neighborhood of the optimal value at the later stage of iteration. The L-M least square method overcomes the limitation that the G-N algorithm requires the full rank of matrix columns when processing the nonlinear problem.
L-M is calculated as follows, and the general least squares problem is given by:
wherein, f (x) is the residual between the output of the component-level model and the output of the recognition result.
Solving the least squares problem to find an x*So that x is*＝argminxF(x)
The G-N algorithm solves the least square problem by choosing h such that F (x) is at xkA near second order approximation, then
The iterative formula of gauss newtons is:
wherein h is the change of each iteration step after a given initial state,
then is an approximation of the Hessian matrix of the reaction gradient.
In the iteration of the G-N algorithm, when
When singular, the iteration cannot continue. L-M algorithm with appropriate addition
The measure of diagonal bins allows the iteration to continue.
L-M is also called damping least square method, and a damping term mu is introduced on the basis of the G-N method. The iterative formula is as follows:
if the error is reduced by the h obtained currently, the algorithm enters the next iteration and the damping term mu is reduced. Conversely, if the current increment increases the function value, the damping value is increased.
Each iteration of the L-M algorithm adjusts the damping term to ensure that the error is reduced. When the optimal solution is far away, mu is gradually increased, the algorithm is close to the steepest descent method, global search can be carried out, and the step length is reduced
On the contrary, the G-N algorithm is approached, and the optimal solution can be converged quickly by using the information similar to the second derivative. As can be seen in fig. 3, when the G-N algorithm is employed, the final solution is far from the optimal solution. L-M is an adaptive algorithm that converges quickly in the neighborhood of the optimal solution if the current value drops slowly away from the optimal solution. The invention can solve the least square problem in the System identification modeling of the aeroengine by using a Tool kit identification Tool in Matlab.
And 4, step 4: determining that the system recognizes the input and output data source.
Under the condition that the high-pressure rotor rotating speeds of all working state points of an aircraft engine system are different, the amplitude of the random signal can influence the precision of least square identification and the precision of subsequent cross validation.
The problem of low fitting degree by using a least square method occurs when a nonlinear input and output data source obtained by adopting a step input signal is used for system identification. As can be seen from fig. 4 and 5, a better recognition effect and a better verification result can be achieved by only using the step signal as an input, but a multi-step system recognition iterative process is required to obtain a higher-precision recognition effect, so that a larger amount of calculation is generated, and the dynamic process effect is general.
The Pseudo Random Sequence (PRBS) is a signal simulating white noise, and has the characteristics of reproducibility, easy generation, large excitation spectrum coverage width and uniform power spectral density distribution. Such as the signal diagram of the pseudo-random sequence shown in fig. 6. The aircraft engine model is a system with strong nonlinearity and strong dynamic performance, and a PRBS (pseudo random binary system) can be adopted to fully excite a component system, so that a better result can be obtained when a nonlinear least square problem is solved. The state space model can be obtained by the output of the component model obtained by the PRBS input through system identification, but the model has poor generalization performance, and the state variables can be diverged when cross validation is carried out by adopting step signals commonly used in engineering. As can be seen from fig. 7 and 8, the system identification result using only the pseudorandom sequence as input has high accuracy, but has poor generalization performance, and in the cross validation, the state quantity of the system diverges.
Based on the problem, the invention adopts the mixed sequence as the excitation of the component-level model to obtain the data source required by system identification. That is, for any one operating state, a corresponding blending sequence is generated, and a set of blending sequences suitable for the operating state point is used as the component-level model input to obtain the data source recognized by the system. The mixed sequence adopted by the invention is divided into three sections: the first section is a pseudo-random signal with strong dynamic performance under a certain amplitude, the pseudo-random sequence adopted by the invention is an M sequence, and the M sequence is generated by Matlab. For different operating points, M sequences with different amplitudes are selected to improve the precision of the least square method. The second segment is a step signal. The third segment is a sinusoidal signal at a certain amplitude and frequency. The mixing sequence generated by the present invention is shown in fig. 9. As can be seen from fig. 10 and 11, a hybrid sequence is generated at each operating state point, so that an optimal solution can be generated when system identification is performed, that is, the error between the state quantity and the output quantity and the component method model is less than 2%. And adjusting the frequency and amplitude of the mixing sequence according to the identification result to obtain an optimal solution. The invention adopts the upper and lower bound amplitude of the mixed signal to be about +/-2% of the steady-state fuel amplitude, and the linear time-invariant model obtained by identifying the working state points of different high-pressure rotor rotating speeds can better approximate the component model in the step 1 by adopting the signal.
And 5: and (3) establishing a linear time-invariant model with the actual high-pressure rotor rotating speed and the low-pressure rotor rotating speed of the aircraft engine as state variables at each working state point under the ground condition according to the model type, the identification criterion and the input and output data source determined in the step 2-4.
The least square problem is solved by utilizing a Tool box identification Tool provided by Matlab, and the component-level System is quickly and linearly identified by combining a mixed sequence. FIG. 12 is a plot of a fit of a linear regression of the identified state space model b matrix.
Step 6: a linear time-invariant system established at a plurality of working state points is utilized, the rotating speed of a high-pressure rotor is taken as a scheduling parameter, and a polynomial nonlinear model under the ground condition of the aircraft engine is established in a regression analysis mode based on the idea of gain scheduling.
And (3) establishing a quantitative relation of the mutual dependence between the high-pressure rotor rotating speed and the state space matrix parameters by using unitary regression analysis, wherein the linear regression or the nonlinear regression is selected according to actual data because the quantitative relation between the high-pressure rotor rotating speed and the state space matrix parameters is unknown. And selecting a cost function, and finding an optimal regression model to minimize the corresponding cost function.
The method is characterized in that a linear state space system established according to a plurality of working states takes the rotating speed of a high-pressure rotor of an aeroengine as a scheduling parameter, a polynomial nonlinear system under the ground condition of the aeroengine is established in a regression analysis mode based on the idea of gain scheduling, and the model is shown as follows
In the formula, nLFor the low-pressure rotor speed, the speed of the rotor,
is the differential of the low-pressure rotor speed, nHThe rotation speed of the high-pressure rotor is set,
is the differential of the high-pressure rotor speed, WfbA main amount of fuel, A8Is the area of the throat, f11(nH) For the parameter a to be identified11Polynomial f fitted to the rotation speed of the high-pressure rotor12(nH) For the parameter a to be identified12Polynomial f fitted to the rotation speed of the high-pressure rotor21(nH) For the parameter a to be identified21Polynomial f fitted to the rotation speed of the high-pressure rotor22(nH) For the parameter a to be identified22Polynomial, g, fit-to-high-pressure rotor speed11(nH) Parameter b to be identified11Polynomial, g, fit-to-high-pressure rotor speed12(nH) Parameter b to be identified12Polynomial, g, fit-to-high-pressure rotor speed21(nH) Parameter b to be identified21Polynomial, g, fit-to-high-pressure rotor speed22(nH) Parameter b to be identified22Polynomial h fitted to the rotation speed of the high-pressure rotor21(nH) Parameter c to be identified21Polynomial h fitted to the rotation speed of the high-pressure rotor22(nH) Parameter c to be identified22Polynomial fit-up to high-pressure rotor speed21(nH) Parameter d to be identified21Polynomial fit-up to high-pressure rotor speed22(nH) Parameter d to be identified22Polynomial fitting at high pressure rotor speed.
A polynomial nonlinear model of the aircraft engine system under ground conditions is established. FIG. 13 is a verification image of a polynomial nonlinear state space model at a certain operating point.
And 7: the established polynomial nonlinear model can embody strong nonlinearity and time-varying characteristics of the aero-engine, so that the method can be used for engine multivariable control rule design, state estimation problems and health management problems.
It should be noted that, although the above-mentioned embodiments of the present invention are illustrative, the present invention is not limited thereto, and thus the present invention is not limited to the above-mentioned embodiments. Other embodiments, which can be made by those skilled in the art in light of the teachings of the present invention, are considered to be within the scope of the present invention without departing from its principles.
Claims (5)
1. The aeroengine system identification modeling method based on the mixed sequence is characterized by comprising the following steps:
step 1, solving a common working equation of the aero-engine about flow continuity and power balance based on a component method to establish a component-level model of the aero-engine;
step 2, adopting a system identification method for input and output data of the component-level model, and determining a model type required by system identification according to the existing experience;
step 3, converting the modeling problem of the aircraft engine into a least square problem identified by a system, and identifying an actual rotation speed linear time invariant model of each steady-state point by an L-M least square estimation method;
step 4, respectively generating a corresponding mixed sequence aiming at each working state point of the aircraft engine, wherein the mixed sequence is divided into three sections, the first section is a pseudo-random signal, the second section is a step signal, and the third section is a sinusoidal signal; then, the generated mixed sequence is used as the excitation of a component-level model to obtain a data source required by system identification;
step 5, based on the model type required by system identification determined in the step 2-4, the system identification criterion based on L-M least square estimation and the data source required by system identification, establishing a linear time invariant model of each working state point under the ground condition, wherein the linear time invariant model takes the actual high-pressure rotor rotating speed and low-pressure rotor rotating speed of the aircraft engine as state variables;
and 6, establishing a polynomial nonlinear model under the ground condition of the aircraft engine by using the linear time invariant model established at the plurality of working state points in the step 5, taking the rotating speed of a high-pressure rotor of the aircraft engine as a scheduling parameter, and adopting a regression analysis mode based on the idea of gain scheduling.
2. The hybrid sequence-based aircraft engine system identification modeling method of claim 1, wherein in step 1, the common working equations of the aircraft engine with respect to flow continuity and power balance comprise: a balance equation of the flow of the fan outlet and the sum of the flow of the compressor and the inlet of the bypass; a balance equation of the inlet flow of the low-pressure turbine and the outlet flow of the high-pressure turbine; a balance equation of the sum of the inlet flow of the high-pressure turbine, the outlet flow of the compressor and the fuel quantity; a balance equation of the outlet flow of the tail nozzle and the sum of the outlet flow of the outer culvert, the outlet flow of the low-pressure turbine and the boosted fuel oil amount; a high pressure rotor power balance equation; low pressure rotor power balance equations.
3. The hybrid sequence-based aircraft engine system identification modeling method of claim 1, wherein in step 2, the model types required for system identification are:
in the formula, nLFor the low-pressure rotor speed, the speed of the rotor,
is the differential of the low-pressure rotor speed, nHThe rotation speed of the high-pressure rotor is set,
is the differential of the high-pressure rotor speed, WfbA main amount of fuel, A8Is the throat area, pi is the pressure increase ratio, a11,a12,a21,a22,b11,b12,b21,b22,c31,c32,d31,d32Are parameters to be identified.
4. The hybrid sequence-based aircraft engine system identification modeling method according to claim 1, wherein in step 6, the established polynomial nonlinear model under the aircraft engine ground conditions is:
in the formula, nLFor the low-pressure rotor speed, the speed of the rotor,
is the differential of the low-pressure rotor speed, nHThe rotation speed of the high-pressure rotor is set,
is the differential of the high-pressure rotor speed, WfbA main amount of fuel, A8Is the area of the throat, f11(nH) For the parameter a to be identified11Polynomial f fitted to the rotation speed of the high-pressure rotor12(nH) For the parameter a to be identified12Polynomial f fitted to the rotation speed of the high-pressure rotor21(nH) For the parameter a to be identified21Polynomial f fitted to the rotation speed of the high-pressure rotor22(nH) For the parameter a to be identified22Polynomial, g, fit-to-high-pressure rotor speed11(nH) Parameter b to be identified11Polynomial, g, fit-to-high-pressure rotor speed12(nH) Parameter b to be identified12Polynomial, g, fit-to-high-pressure rotor speed21(nH) Parameter b to be identified21Polynomial, g, fit-to-high-pressure rotor speed22(nH) Parameter b to be identified22Polynomial h fitted to the rotation speed of the high-pressure rotor21(nH) Parameter c to be identified21Polynomial h fitted to the rotation speed of the high-pressure rotor22(nH) For the ginseng to be identifiedNumber c22Polynomial fit-up to high-pressure rotor speed21(nH) Parameter d to be identified21Polynomial fit-up to high-pressure rotor speed22(nH) Parameter d to be identified22And (3) fitting a polynomial with the rotating speed of the high-pressure rotor, wherein pi is a supercharging ratio.
5. The hybrid sequence-based aircraft engine system identification modeling method of claim 1, wherein in step 4, the pseudo-random signal of the first segment of the hybrid sequence is an M-sequence.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911416076.1A CN111177930B (en) | 2019-12-31 | 2019-12-31 | Aeroengine system identification modeling method based on mixed sequence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911416076.1A CN111177930B (en) | 2019-12-31 | 2019-12-31 | Aeroengine system identification modeling method based on mixed sequence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111177930A CN111177930A (en) | 2020-05-19 |
| CN111177930B true CN111177930B (en) | 2021-10-22 |
Family
ID=70650644
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911416076.1A Expired - Fee Related CN111177930B (en) | 2019-12-31 | 2019-12-31 | Aeroengine system identification modeling method based on mixed sequence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111177930B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111965981B (en) * | 2020-09-07 | 2022-02-22 | 厦门大学 | Aeroengine reinforcement learning control method and system |
| CN112363411A (en) * | 2020-11-14 | 2021-02-12 | 大连理工大学 | Design method of aeroengine dynamic matrix controller |
| CN112859606B (en) * | 2021-01-13 | 2022-03-11 | 厦门大学 | Turbofan engine transition state disturbance suppression method based on preset performance |
| CN114013665B (en) * | 2021-11-19 | 2024-02-27 | 中国航发沈阳发动机研究所 | State transition control method and device for aviation gas turbine engine |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107066673A (en) * | 2017-01-17 | 2017-08-18 | 大连理工大学 | The sampling anti-interference identification modeling method of industrial time lag response process |
| CN108416141A (en) * | 2017-08-31 | 2018-08-17 | 北京理工大学 | A kind of linear time-varying structural modal vibration shape discrimination method |
| CN108920868A (en) * | 2018-07-23 | 2018-11-30 | 国网江西省电力有限公司电力科学研究院 | Load based on intelligent electric meter data recognizes model parameter correction method |
| CN109829238A (en) * | 2019-02-02 | 2019-05-31 | 哈尔滨工业大学 | Aero-engine system identifying method based on equivalent Expansion Model Based on Equilibrium Manifold |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103701414B (en) * | 2012-09-28 | 2017-03-01 | 富士通株式会社 | The selecting device of nonlinear terms and method, identification system and compensation system |
-
2019
- 2019-12-31 CN CN201911416076.1A patent/CN111177930B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107066673A (en) * | 2017-01-17 | 2017-08-18 | 大连理工大学 | The sampling anti-interference identification modeling method of industrial time lag response process |
| CN108416141A (en) * | 2017-08-31 | 2018-08-17 | 北京理工大学 | A kind of linear time-varying structural modal vibration shape discrimination method |
| CN108920868A (en) * | 2018-07-23 | 2018-11-30 | 国网江西省电力有限公司电力科学研究院 | Load based on intelligent electric meter data recognizes model parameter correction method |
| CN109829238A (en) * | 2019-02-02 | 2019-05-31 | 哈尔滨工业大学 | Aero-engine system identifying method based on equivalent Expansion Model Based on Equilibrium Manifold |
Non-Patent Citations (2)
| Title |
|---|
| 基于PSO算法的递归网络航空发动机随动系统辨识;彭靖波 等;《系统仿真学报》;20081231;第6818-6820页 * |
| 基于系统辨识的冲压发动机动态模型分析;梁俊龙 等;《弹箭与制导学报》;20170630;第147-151页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111177930A (en) | 2020-05-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111177930B (en) | Aeroengine system identification modeling method based on mixed sequence | |
| Basso et al. | NARX models of an industrial power plant gas turbine | |
| CN106874569B (en) | Variable-geometry split-shaft type gas turbine real-time simulation modeling method | |
| US20100030547A1 (en) | System, program product, and related methods for global targeting of process utilities under varying conditions | |
| CN101364084A (en) | Systems and methods for model-based sensor fault detection and isolation | |
| CN111507637A (en) | Parameter modeling and testing method for water turbine and water diversion system for power system stability calculation | |
| Kidikian et al. | Off-design prediction of transonic axial compressors: part 1—mean-line code and tuning factors | |
| CN115392133A (en) | Wind power plant optimal clustering equivalence method adopting Gaussian mixture model | |
| CN106840685B (en) | A kind of gas turbine dynamic operation condition database sample point choosing method | |
| Rainer et al. | Iterative DoE-improved emission models and better optimisation results within a shortened measurement time | |
| Abbasi et al. | Performance optimization of an axial turbine with a casing injection based on response surface methodology | |
| CN110098610B (en) | Real-time identification method and system for oscillation leading mode of power system under fault disturbance | |
| CN110826182B (en) | Aeroelastic design method of aircraft structure based on vertex method and sequential optimization strategy | |
| Fang et al. | Optimal-behavior-based dynamic calibration of the automotive diesel engine | |
| CN109344522B (en) | Method and system for calculating grinding quantity of stationary blade of axial flow compressor | |
| Zito et al. | Narmax model identification of a variable geometry turbocharged diesel engine | |
| CN106529041B (en) | Turbine refined modeling and identification method | |
| Monk | A computational analysis of the aerodynamic and aeromechanical behavior of the purdue multistage compressor | |
| Bertram et al. | Augmentation of an Artificial Neural Network (ANN) model with expert knowledge of critical combustion features for optimizing a compression ignition engine using multiple injections | |
| Dogga et al. | Reverse engineering EEE transonic compressor fan blade | |
| Cerri et al. | Gas path analysis and gas turbine re-mapping | |
| Rivera et al. | Systematic CFD Parameter Approach to Improve Torque Converter Simulation | |
| KR20150054245A (en) | Linear approximation of the model within the design space optimal design method | |
| CN114997073B (en) | Impulse turbine nozzle structure parameter optimization method and system | |
| Kazemi Bakhshmand et al. | Experimental Energy and Exergy Analysis of an Automotive Turbocharger Using a Novel Power-Based Approach. Energies 2021, 14, 6572 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20211022 |