DESIGN METHOD FOR OPTIMIZATION OF TRANSIENT CONTROL LAW OF AERO-ENGINE

Abstract

A design method for optimization of a transient control law of the aero-engine is disclosed, and performs the transient schedule optimization for the aero-engine by adopting an SQP algorithm, to realize the design of the transient control law along a constrained boundary condition. The fuel flow rate value is adjusted, other constraints remain unchanged, and the transient control law is designed under different limits. The transient time under each transient control law is calculated by constructing the transient time evaluation function. A lookuptable interpolation table is established by using the calculated transient time and corresponding fuel flow, to realize the fuel flow scheduling under different transient time. The fuel flow obtained by scheduling in the expected time is taken as an acceleration and deceleration control schedule of the closed-loop control of the aero-engine, and the output thereof is taken as the reference instruction of an acceleration process.

Claims

1. A design method for optimization of a transient control law of an aero-engine, wherein the steps are as follows: S1 performing the transient schedule optimization for the aero-engine based on an SQP algorithm, to realize the design of the transient control law of the aero-engine running along the constrained boundary; the steps for the design of the transient control law of the aero-engine running along the constrained boundary are as follows: S1.1 determining the number of optimal nodes and the length of the entire optimization process time; S1.2 taking the fuel flow obtained by the aero-engine under the action of a closed-loop controller as a reference, and selecting the initial value of the fuel flow, so that each output of the aero-engine does not exceed the limit value after the finally determined initial value of the fuel flow is loaded into an aero-engine model; S1.3 setting the parameters of the number of calculations of the maximum function MaxFunEvals, the maximum number of iterations Maxlter and the accuracy of function TolFun related to optimization options of an SQP algorithm; S1.4 determining boundary conditions for the optimization of the transient control law of the aero-engine, comprising high-pressure rotor speed N.sub.2 limit, low-pressure rotor speed N.sub.1 limit, compressor outlet total pressure P.sub.3 limit, low-pressure turbine outlet temperature T.sub.5 limit, surge margin of fan surge SMF, surge margin of compressor surge SMC, fuel flow W.sub.f limit and transient fuel flow rate W.sub.f limit, and establishing corresponding constraint function according to the constraint condition, wherein the form thereof is shown as follows:
N.sub.2,minN.sub.2N.sub.2,max
N.sub.1N.sub.1,max
P.sub.3P.sub.3,max
T.sub.5T.sub.5,max
SMFSMF.sub.min
SMCSMC.sub.min
W.sub.f,minW.sub.fW.sub.f,max
W.sub.fW.sub.f,max where, N.sub.2,min is minimum high-pressure rotor speed, N.sub.2,max is maximum high-pressure rotor speed, N.sub.1,max is minimum low-pressure rotor speed, P.sub.3,max is maximum value of compressor outlet total pressure, T.sub.5,max is maximum value of low-pressure turbine outlet temperature, SMC.sub.min is minimum value of compressor surge margin, SMF.sub.min is minimum value of fan surge margin, W.sub.f,min and W.sub.f,max are respectively minimum value and maximum value, and W.sub.f,max is maximum value of fuel flow rate; S1.5 according to the time requirement for the transient state of the aero-engine, establishing the objective function for optimization of the transient control law of the aero-engine, wherein the form thereof is shown as follows:
J=100*norm(N.sub.2N.sub.2,cmd, 2)+100*norm(N.sub.1N.sub.1,cmd, 2) where, N.sub.1,cmd and N.sub.2,cmd are expected values of the low-pressure rotor speed and the high-pressure rotor speed; S2. on the premise of not exceeding the limit boundary of the aero-engine, adjusting the limit value of the fuel flow rate of the aero-engine while other constraint conditions remain unchanged, thereby establishing the transient control law of the aero-engine under different fuel rate limits; the steps for the design of the established transient control law of the aero-engine under different fuel rate limits are as follows: S2.1 keeping the number of optimal nodes, the whole optimization process time, the initial value of the fuel flow and the setting of each optimization option consistent with the transient schedule optimization along the constrained boundary; S2.2 keeping the high-pressure rotor speed N.sub.2 limit, the low-pressure rotor speed N.sub.1 limit, the compressor outlet total pressure P.sub.3 limit, the low-pressure turbine outlet temperature T.sub.5 limit, the surge margin of the fan surge SMF, the surge margin of the compressor surge SMC and the fuel flow W.sub.f limit in the transient state of the aero-engine unchanged, on the premise of not exceeding the maximum limit value W.sub.f,max of the transient fuel flow rate, changing the fuel flow rate limit, respectively set as W.sub.f,1, W.sub.f,2, W.sub.f,3 . . . W.sub.f,N, and performing the transient schedule optimization under different fuel flow rate limits again, to obtain the control law under different conditions, that is, fuel flow of W.sub.f,1, W.sub.f,2, W.sub.f,3, . . . W.sub.f,N; S2.3 saving the fuel flow of W.sub.f,1, W.sub.f,2, W.sub.f,3 . . . W.sub.f,N of the aero-engine obtained through optimization under different fuel flow rate limits, to prepare for later fuel flow scheduling; S3. after completing the design of the transient control law under different fuel rate limits, constructing the transient time evaluation function of the aero-engine, thereby determining the transient time under different transient control laws; the steps for constructing transient time evaluation function of the aero-engine are as follows: S3.1 producing aero-engine model outputs: loading the fuel flow W.sub.f of the aero-engine into the aero-engine model, to obtain the aero-engine outputs: low-pressure rotor speed N.sub.1, low-pressure rotor conversion speed N.sub.1cor, high-pressure rotor speed N.sub.2 and high-pressure rotor conversion speed of imported aero-engine N.sub.2cor; S3.2 taking high-pressure rotor speed N.sub.2 in the aero-engine output as the evaluation variable of the transient time of the aero-engine, and when the speed fluctuates 0.2% above and below the steady state value, entering a steady state by default, wherein the entered time is the transient time of the aero-engine; S3.3 loading the fuel flow of W.sub.f,1, W.sub.f,2, W.sub.f,3 . . . W.sub.f,N in the transient control law of the aero-engine under different limits into the aero-engine model, to obtain the aero-engine model outputs of N.sub.2,1, N.sub.2,2, N.sub.2,3 . . . N.sub.2,N under different limits, and calculating the transient time of T.sub.s,1, T.sub.s,2, T.sub.s,3 . . . T.sub.s,N under different limits and saving; S4. sorting the fuel flow under different transient control laws and the corresponding transient time from smallest to biggest, and establishing a lookuptable interpolation table; and according to the established interpolation table, implementing online transient-state control scheduling, that is, inputting the expected transient time of aero-engine, and then obtaining the corresponding fuel flow under the transient time by scheduling; the steps for implementing the online transient-state control scheduling are as follows: S4.1 sorting the fuel flow under limits and the corresponding transient time of T.sub.s,1, T.sub.s,2, T.sub.s,3 . . . T.sub.s,N from smallest to biggest; S4.2 taking each transient time of T.sub.s,1, T.sub.,2, T.sub.s,3 . . . T.sub.s,N as an interpolation node 1, taking the corresponding fuel flow of W.sub.f,1, W.sub.f,2, W.sub.f,3 . . . W.sub.f,N as Tabledata, since the fuel flow rate loaded into the aero-engine model is in time series format, taking the time series of the fuel flow as the interpolation node 2, and establishing the lookuptable interpolation table; S4.3 selecting the expected transient time T*.sub.s, and interpolating through the established interpolation table, to obtain a group of fuel flow W*.sub.f corresponding to the transient time, thereby implementing the online transient time scheduling; S4.4 taking the fuel flow W*.sub.f under the expected transient time as the aero-engine input, and saving the relevant aero-engine output, to take the saved output as reference instruction of the closed-loop controller; S5. taking the fuel flow under the expected transient time as the aero-engine input to obtain the aero-engine output, and taking the relevant output parameters as the reference instruction of the closed-loop control of the aero-engine, to achieve the full closed-loop control of the aero-engine.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0033] FIG. 1 is a structural diagram of a design method for optimization of a transient control law of an aero-engine;

[0034] FIG. 2 is an overall flow chart of a design method for optimization of a transient control law of an aero-engine;

[0035] FIG. 3 is a flow chart of a design algorithm of a transient control law of an aero-engine running along a constrained boundary;

[0036] FIG. 4 is a flow chart of online transient-state control scheduling;

[0037] FIG. 5 is a three-dimensional curve of an interpolation table established based on the fuel flow obtained through optimization and the calculated transient time;

[0038] FIG. 6 is a transient control law diagram of a micro gas turbine obtained through optimization based on this method.

DETAILED DESCRIPTION

[0039] The present invention will be further described below in combination with the drawings, and a structural diagram of a system is shown in FIG. 1.

[0040] As shown in FIG. 2, a design method for optimization of a transient control law of an aero-engine comprises the following steps:

[0041] S1. performing the transient schedule optimization for the aero-engine by adopting an SQP algorithm, to realize the design of the transient control law of the aero-engine running along the constrained boundary;

[0042] S2. on the premise of not exceeding the limit boundary of the aero-engine, adjusting the limit value of the fuel flow rate of the aero-engine while other constraint conditions remain unchanged, thereby establishing the transient control law under different limits through optimization;

[0043] S3. after completing the design of the transient control law under different fuel rate limits, constructing the transient time evaluation function of the aero-engine, thereby determining the transient time under different transient control laws;

[0044] S4. sorting the fuel flow under different transient control laws and the corresponding transient time from smallest to biggest, and establishing a lookuptable interpolation table; and according to the established interpolation table, implementing online transient-state control scheduling, that is, inputting the expected transient time of aero-engine, and then obtaining the corresponding fuel flow under the transient time by scheduling;

[0045] S5. taking the fuel flow under the expected transient time as the aero-engine input to obtain the aero-engine output, and taking the relevant output parameters as the reference instruction of the closed-loop control of the aero-engine, to achieve the full closed-loop control of the aero-engine;

[0046] as shown in FIG. 3, the steps of the design for a transient control law of an aero-engine running along the constraint boundary are as follows:

[0047] S1. determining the optimized transient time to be T=8 seconds, and because the model thereof calculates the step length to be 0.025 seconds, determining the number of the optimal nodes to be 0:0.025:8, with a total of 321, by adopting a principle that the number of model calculation points is equal to the number of optimal nodes;

[0048] S2. taking the fuel flow of the aero-engine obtained under the action of a closed-loop controller as a reference, and taking the optimized initial value of the fuel flow as the numerical value of 59.1 Kg/h and a constant sequence with the length of 321, wherein after verification, under the action of the fuel, each output of the aero-engine does not exceed the limit value, which is a feasible solution;

[0049] S3. setting the parameters of the number of calculations of the maximum function MaxFunEvals, the maximum number of iterations Maxlter, the accuracy of function TolFun, etc., related to optimization options of an SQP method, and taking MaxFunEvals as INF, Maxlter as INF, and TolFun as 1 e-8;

[0050] S4 determining boundary conditions for the optimization of the transient control law of the aero-engine, mainly comprising high-pressure rotor speed N.sub.2 limit, low-pressure rotor speed N.sub.1 limit, compressor outlet total pressure P.sub.3 limit, low-pressure turbine outlet temperature T.sub.5 limit, surge margin of fan surge SMF, surge margin of compressor surge SMC, fuel flow W.sub.f limit, transient fuel flow rate W.sub.f limit, etc., and establishing corresponding constraint function according to the constraint condition, wherein the form thereof is shown as follows:

N.sub.2,minN.sub.2N.sub.2,max

N.sub.1N.sub.1,max

P.sub.3P.sub.3,max

T.sub.5T.sub.5,max

SMFSMF.sub.min

SMCSMC.sub.min

W.sub.f,minW.sub.fW.sub.f,max

W.sub.fW.sub.f,max

where, N.sub.2,min is minimum high-pressure rotor speed, N.sub.2,max is maximum high-pressure rotor speed, N.sub.1,max is minimum low-pressure rotor speed, P.sub.3,max is maximum value of compressor outlet total pressure, T.sub.5,max is maximum value of low-pressure turbine outlet temperature, SMC.sub.min is minimum value of compressor surge margin, SMF.sub.min is minimum value of fan surge margin, W.sub.f,min and W.sub.f,max are respectively minimum value and maximum value of fuel flow, and W.sub.f,max is maximum value of fuel flow rate, N.sub.2,min is taken as 68.9%, N.sub.2,max is taken as 102%, N.sub.1,max is taken as 102%, P.sub.3,max is taken as 1310 kPa, T.sub.5,max is taken as 873K, W.sub.f,min is taken as 39.6 kg/h, W.sub.f,max is taken as 465 kg/h, W.sub.f,max is taken as 2.5 kg/h/25 ms, SMF.sub.min is taken as 3%, SMC.sub.min is taken as 3%, and a constraint function is expressed as;

68.9%N.sub.2102%

N.sub.1102%

P.sub.31310

T.sub.5873

SMF3%

SMC3%

39.6W.sub.f465

W.sub.f2.5

[0051] S5. according to the time requirement for the aero-engine transient state, establishing an objective function for optimization of the transient control law of the aero-engine, wherein the form thereof is shown as follows:

J=100*norm(N.sub.2N.sub.2,cmd, 2)+100*norm(N.sub.1N.sub.1,cmd, 2)

[0052] where, N.sub.1,cmd and N.sub.2,cmd are expected values of the low-pressure rotor speed and the high-pressure rotor speed, N.sub.2,cmd is taken as 100 and N.sub.1,cmd is taken as 100, and the objective function is expressed as;

[0053] J=100*norm(N.sub.2100,2)+100*norm(N.sub.1100,2)

[0054] the steps for establishing the transient control law under different limits are as follows:

[0055] S1. keeping the setting of optimal time and the number of optimal nodes, selection of initial value of the fuel flow and the setting of each optimization option consistent with the transient schedule optimization along the constrained boundary;

[0056] S2. keeping the high-pressure rotor speed N.sub.2 limit, the low-pressure rotor speed N.sub.1 limit, the compressor outlet total pressure P.sub.3 limit, the low-pressure turbine outlet temperature T.sub.5 limit, the surge margin of the fan surge SMF, the surge margin of the compressor surge SMC and the fuel flow W.sub.f in the aero-engine transient state limit unchanged, on the premise of not exceeding the maximum limit value W.sub.f,max=2.5 of the transient fuel flow rate, changing the fuel flow rate limit, respectively set as W.sub.f,1=2.2, W.sub.f,2=2.0,W.sub.f,3=1.8, W.sub.f,3=1.6,W.sub.f,3=1.4 and W.sub.f3,=1.2, and performing the transient schedule optimization under different fuel flow rate limits again, to obtain the control law under different conditions, that is, fuel flow of W.sub.f,1, W.sub.f,2, W.sub.f,3 , W.sub.f,4 and W.sub.f,5;

[0057] S3. saving the fuel flow of W.sub.f,1, W.sub.f,2, W.sub.f,3, W.sub.f,4 and W.sub.f,5 of the aero-engine obtained through optimization under different fuel flow rate limits;

[0058] the steps for constructing transient time evaluation function of the aero-engine are as follows:

[0059] S1 producing aero-engine model outputs: loading the optimized fuel flow W.sub.f of the aero-engine into the aero-engine model, to obtain the aero-engine outputs: low-pressure rotor speed N.sub.1, low-pressure rotor conversion speed N.sub.1cor, high-pressure rotor speed N.sub.2, high-pressure rotor conversion speed of imported aero-engine N.sub.2cor, etc.;

[0060] S2 taking high-pressure rotor speed N.sub.2 in the aero-engine output as the evaluation variable of the aero-engine transient time, and when the speed fluctuates 0.2% above and below the steady state value, entering a steady state by default, wherein the entered time is the aero-engine transient time;

[0061] S3 loading the fuel flow W.sub.f,1, W.sub.f,2, W.sub.f,3, W.sub.f,4, and W.sub.f,5 in the transient control law of the aero-engine under different limits into the aero-engine model, to obtain the aero-engine model outputs of N.sub.2,1, N.sub.2,2, N.sub.2,3, N.sub.2,4, and N.sub.2,5 under different limits, and calculating the transient time of T.sub.s,1, T.sub.s,2, T.sub.s,3, T.sub.s,4 and T.sub.s,5 under different limits and saving;

[0062] as shown in FIG. 4, the steps for implementing the online transient-state control scheduling are as follows:

[0063] S1. sorting the fuel flow under limits and the corresponding transient time T.sub.s,1, T.sub.s,2, T.sub.s,3 . . . T.sub.s,N from smallest to biggest;

[0064] S2. taking each transient time of T.sub.s,1, T.sub.s,2, T.sub.s,3 . . . T.sub.s,N as an interpolation node 1, taking the corresponding fuel flow of W.sub.f,1, W.sub.f,2, W.sub.f,3 . . . W.sub.f,N as Tabledata, since the fuel flow loaded into the model is in a time series format, taking the time series of the fuel flow as the interpolation node 2, and establishing the lookuptable interpolation table, as shown in FIG. 5, taking the transient time as the amount of scheduling, and scheduling the corresponding fuel flow according to the expected transition time;

[0065] S3. selecting the expected transient time T*.sub.s, and interpolating through the established interpolation table, to obtain a group of fuel flow W*.sub.f corresponding to the transient time, thereby implementing the online transient time scheduling;

[0066] S4. taking the fuel flow W*.sub.f under the expected transient time as the aero-engine input, and saving the relevant aero-engine output, to take the saved output as reference instruction of the closed-loop controller;

[0067] FIG. 5 is a three-dimensional curve of an interpolation table established based on fuel flow obtained through optimization and the calculated transient time;

[0068] FIG. 6 is a transient control law diagram of a micro gas turbine obtained through optimization based on this method;

[0069] In summary, it can be seen that an optimization method for the transient control law of the aero-engine proposed in the present invention is effective and feasible, and has universal applicability, which can be applied to the optimization of the transient control law of other types of aero-engines.