A monotonic smoothing design method for aeroengine throttling control law

Abstract

The invention discloses a monotonic smoothing design method for a throttling control law of an aero-engine, and relates to the technical field of overall performance simulation of an aero-engine. The method comprises the following steps: determining the upper and lower bounds of engine control law variables and their values ​​according to the characteristics of an engine configuration; performing a preliminary throttling control law analysis based on an existing control law variable combination, determining the direction of change of each control law in a throttling process, recording the control law variable values ​​at an initial operating point, and then obtaining an initial adjustment range; dividing the initial adjustment range of each control law variable into N equal parts, and calculating the equal division length of the coding space of the control law variable; determining the upper and lower bounds of thrust according to the throttling process, and obtaining a discretized throttling thrust point set to be optimized; starting to solve the control law variable combination corresponding to the i-th throttling optimized thrust point based on a genetic optimization algorithm; integrating the optimization results of each thrust throttling point, and obtaining a final throttling process control law scheme.

Description

Technical Field

[0001] The invention relates to the technical field of overall performance simulation of aircraft engines, and in particular to a monotonic smoothing design method for a throttling control law of an aircraft engine. Background Art

[0002] During the long cruising phase of the aircraft, as the fuel is continuously consumed, the weight of the aircraft continues to decrease. When the lift-to-drag ratio remains unchanged, the drag encountered by the aircraft will continue to decrease. Therefore, the demand for engine thrust is also gradually decreasing. At this time, the engine needs to achieve a match with the aircraft's drag through thrust throttling to maintain the aircraft's stable and uniform flight speed.

[0003] There are generally two ways to throttle engine thrust. The more direct way is to actively control the fuel flow to reduce the speed of the compression component, reduce the inlet flow and thus reduce the thrust. If the engine intake duct is not adjustable, during the down-throttling process, the air flow in the intake duct is greater than the air flow required by the engine, so a large overflow resistance will be generated. For geometrically adjustable engines, in the early stage of throttling, the engine working state can also be affected by adjusting adjustable components such as the nozzle throat area and the fan guide vane angle. While maintaining the same inlet flow, the unit thrust of the engine is reduced to reduce the thrust. This method is called engine equal flow throttling, which can largely avoid intake duct overflow, thereby improving the engine installation performance.

[0004] Whether it is down-throttling or equal-flow throttling, it is necessary to design the engine control law during the throttling process. At present, some research and analysis have been carried out on the design of engine throttling control law. In some studies, the control law is obtained by researchers manually debugging. It is necessary to understand the different effects of various control law variables on engine thrust and then make comprehensive trade-offs and adjustments. At the same time, it is necessary to ensure that the engine does not have risks such as overheating, overspeed, and surge during the adjustment process. The debugging process is cumbersome, and it puts forward high professional knowledge requirements for researchers of control laws. At the same time, at the beginning of the engine design, it is necessary to repeatedly adjust the engine design parameters and component characteristics. The cumbersomeness of manual debugging of control laws will seriously restrict the efficient development of the design process. Therefore, some studies use optimization algorithms to realize the automatic optimization design of control laws. However, although the control law combination obtained by optimization can meet various throttling thrust conditions, there is a situation in which a single control law variable in a discrete thrust throttling sequence is not smoothly and monotonously adjusted during the entire throttling process. This is determined by the global search characteristics of the optimization algorithm itself. Although the control law combination sequence obtained in this way makes the engine throttling performance numerically better, it is not suitable for engineering implementation.

[0005] Therefore, it is necessary to study a new design method for monotonic smoothing of engine throttling control law, which will facilitate the design and implementation of engine controllers. Summary of the invention

[0006] The purpose of the present invention is to provide a monotonic smoothing design method for an aircraft engine throttling control law. An improved genetic optimization algorithm based on monotonic smoothing can realize automatic optimization design of the engine throttling control law, including preprocessing and monotonicity restriction of the control law variable encoding space, smoothness restriction of the genetic algorithm objective function and fitness function value, and the program algorithm flow of the entire control law design method.

[0007] The present invention provides a monotonic smoothing design method for an aircraft engine throttling control law, comprising the following steps:

[0008] Step 1: according to the characteristics of the engine configuration, determine the engine control law variables involved in the optimization design of the throttling control law, and determine the upper and lower bounds of the corresponding engine control law variable values ​​for each control law variable;

[0009] Step 2: Based on the existing control law variable combination, a preliminary throttling control law analysis is performed by means of sensitivity analysis to determine the change direction of each control law in the throttling process, and the control law variable values ​​at the initial operating point are recorded. The initial adjustment range is obtained by combining the initial value, change direction and upper and lower limits of each control law variable;

[0010] Step 3: Divide the initial adjustment range of each control law variable into N equal parts, and calculate the equal division length of the coding space of the control law variable;

[0011] Step 4: Determine the upper and lower limits of thrust according to the throttling process, and discretize the throttling thrust point set to be optimized;

[0012] Step 5: Based on the genetic optimization algorithm, start solving the control law variable combination corresponding to the i-th throttling optimization thrust point, where i changes from 1 to end, until the control law combination of all throttling thrust point sets is optimized;

[0013] Step 6: Integrate the optimization results of each thrust throttling point to obtain the final throttling process control law solution.

[0014] Preferably, the control law variables in step 1 include guide vane angle, duct ejector area, and nozzle throat area.

[0015] Preferably, in step 2, the guide vane angle α and the duct ejector area A VABI and nozzle throat area A Noz The initial adjustment range is:

[0016] D0:{αAFS ∈[α AFS,min ,α AFS,0 ],θ RVABI ∈[θ RVABI,min ,θ RVABI,0 ],A Noz ∈[A Noz,0 ,A Noz,max ]}.

[0017] Preferably, in step 3, the guide vane angle α and the duct ejector area A VABI and nozzle throat area A Noz The lengths of the equal parts of the coding space are:

[0018] R(α)=(α0-α min ) / N;

[0019] R(A VABI )=(A VABI,0 -A VABI,min ) / N;

[0020] R(A Noz )=(A Noz,max -A VABI,0 ) / N.

[0021] Preferably, the throttling thrust point set to be optimized in step 4 is F=F1, F2, ..., F end-1 ,F end}.

[0022] Preferably, in step 5, based on the genetic optimization algorithm, the control law variable combination corresponding to the i-th throttling optimization thrust point is solved, where it changes from 1 to end until the specific content of the optimization of the control law combination of all throttling thrust point sets is:

[0023] Step 501: for the i-th thrust point, determine the coding space of the control law variables under the round optimization;

[0024] Step 502: According to the coding space D of the control law variable i , generate the initial population P of genetic algorithm i,0 , including NIND individuals, each of which is different in having different genes, and different genes in the scheme of the present invention are different combinations of control law variables whose values ​​are in the control law variable encoding space;

[0025] Step 503: Calculate the objective function value of each individual in the population, calculate the steady-state performance of the target engine based on the control law variable combination corresponding to the genes of different individuals, and import the engine performance calculation result into the objective function calculation to obtain the objective function value corresponding to the individual;

[0026] Step 504: determine whether the optimal individual objective function value of the population meets the convergence condition. If so, directly proceed to step 501 to optimize the next thrust point i+1. If not, proceed to step 505.

[0027] Step 505: Based on the objective function value of each individual in the population, the fitness function set FIT of the population is calculated by the fitness calculation function. i,j , and sum the gene deviation based on the fitness function set, that is, FIT' i,j =FIT i,j +X delta ;

[0028] Step 506: Based on the fitness function set FIT' i,j , perform selection, crossover and mutation to obtain the next generation population P i,j+1 , and proceed to step 503 at the same time.

[0029] Preferably, the specific content of step 501 is: if the optimization result of the control law variable of the i-1th throttling thrust point is {α=α i-1 ,A VABI =A VAB1,i-1 ,A N0z =A NOZ,i-1}, then combined with the throttling change direction of each control law variable obtained by previous analysis, the encoding space of the control law variable of the i-th thrust point can be directionally adjusted as:

[0030]

[0031] Preferably, the objective function of step 503 is constructed as follows:

[0032] ObjV=abs|FF i \+SFC+X delta ;

[0033] Where X delta =(XX initial ) 2 is the genetic deviation, X is the genetic value of the individual in the population, and X initial is the optimization result of the control law variables of the i-1th throttling thrust point, that is, {α=α ,i-1 ,A VABI =A VABI,i-1 ,A NOZ =A NOz,i-1}.

[0034] Therefore, the present invention adopts a monotonic smoothing design method for the throttling control law of an aircraft engine in the above steps, which is applicable to aircraft engines of various configurations. Compared with the traditional control law design method, the scheme of the present invention has the following two advantages:

[0035] 1. The method of the present invention can effectively ensure the monotonicity and smoothness of the control law in the thrust throttling process by adaptively improving the genetic algorithm, thereby making it easier to design and execute the engine controller.

[0036] 2. The method of the present invention can realize the automatic optimization design of the engine throttling control law, avoiding the tediousness and inefficiency of manually adjusting the design control law. Compared with the traditional optimization algorithm, it also has a higher optimization convergence speed and improves the design efficiency of the control law.

[0037] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a program algorithm flow chart of a monotonic smoothing design method for aero-engine throttling control law of the present invention;

[0039] Figure 2 It is an adaptability improvement diagram of the genetic algorithm optimization control law of the present invention;

[0040] Figure 3 This is a schematic diagram of the structure of the adaptive intercooling cycle engine of the present invention;

[0041] Figure 4 This is a comparison chart of the optimization results of the throttling control law of the improved algorithm of the present invention and the traditional algorithm.

[0042] Reference numerals

[0043] 1. Fan 1; 2. Fan 2; 3. Intercooler heat exchanger; 4. Fan 3; 5. Compressor; 6. Combustion chamber 1; 7. Turbine 1; 8. Turbine 2; 9. Nozzle 1; 10. Nozzle 2; 11. Combustion chamber 2; 10-1. Mode selection valve; 12. Turbine 2; 13. Ejector 2; S3, first duct; S2, second duct; S3, third duct. DETAILED DESCRIPTION

[0044] The technical solution of the present invention is further described below by means of the accompanying drawings and embodiments. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and numerical values ​​described in these embodiments do not limit the scope of the present application.

[0045] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present application, its application, or uses.

[0046] Technologies, systems, and devices known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, systems, and devices should be considered part of the specification.

[0047] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0048] Unless otherwise defined, technical or scientific terms used in the present invention shall have the common meanings understood by one having ordinary skills in the field to which the present invention belongs.

[0049] The present invention provides a monotonic smoothing design method for an aircraft engine throttling control law, comprising the following steps:

[0050] Step 1: according to the characteristics of the engine configuration, determine the engine control law variables involved in the optimization design of the throttling control law, and determine the upper and lower bounds of the corresponding engine control law variable values ​​for each control law variable;

[0051] Preferably, the control law variables in step 1 include guide vane angle, duct ejector area, and nozzle throat area.

[0052] In fact, the control law variables may also include any controllable variables such as engine speed, and the guide vane angle, throat area, and duct ejector area are just general summaries. In fact, there are many variables. For example, the guide vane angle includes fan guide vanes, CDFS guide vanes, compressor guide vane angle, etc.

[0053] Step 2: Based on the existing control law variable combination, a preliminary throttling control law analysis is performed by means of sensitivity analysis to determine the change direction of each control law in the throttling process, and the control law variable values ​​at the initial operating point are recorded. The initial adjustment range is obtained by combining the initial value, change direction and upper and lower limits of each control law variable;

[0054] Preferably, in step 2, the guide vane angle α and the duct ejector area A VABI and nozzle throat area A Noz The initial adjustment range is:

[0055] D0:{α AFS ∈[α AFS,min ,α AFS,0 ],θ RVABI ∈[θ RVABI,min ,θ RVABI,0 ],A Noz ∈[A Noz,0 ,A Noz,max ]}.

[0056] Step 3: Divide the initial adjustment range of each control law variable into N equal parts, and calculate the equal division length of the coding space of the control law variable;

[0057] Preferably, in step 3, the guide vane angle α and the duct ejector area A VABI and nozzle throat area A Noz The lengths of the equal parts of the coding space are:

[0058] R(α)=(α0-α min ) / N;

[0059] R(A VABI )=(A VABI,0 -A VABI,min ) / N;

[0060] R(A Noz )=(A Noz,max -A VABI,0 ) / N.

[0061] Step 4: Determine the upper and lower limits of thrust according to the throttling process, and discretize the throttling thrust point set to be optimized;

[0062] Preferably, the throttling thrust point set to be optimized in step 4 is F=F1, F2, ..., F end-1 ,F end}.

[0063] Step 5: Based on the genetic optimization algorithm, start solving the control law variable combination corresponding to the i-th throttling optimization thrust point, where i changes from 1 to end, until the control law combination of all throttling thrust point sets is optimized;

[0064] The present invention can effectively ensure the monotonic smoothing design of the control law by adding two improvements, namely, monotonicity restriction and smoothness restriction, to the adaptability of the genetic algorithm. Figure 2 The principles of two adaptive improvements are specifically demonstrated. The monotonicity restriction is mainly achieved by dynamically adjusting the coding space according to the change direction of the control law variables analyzed in advance, so that the coding space of the next throttling thrust point is a subset of the coding space of the previous throttling thrust point, which can effectively ensure the monotonicity of the change of the control law variables. In addition, the scheme of the present invention also divides the coding space of the control law variables into appropriate equal parts, greatly reducing the adjustable range of each round of optimization while ensuring the existence of feasible solutions, further avoiding the problem of large jumps in the optimization results of the control law variables, and at the same time, a smaller search optimization range can also greatly improve the convergence speed.

[0065] The optimization goal of the genetic optimization algorithm is to find a control law variable combination so that the calculated objective function value is as small as possible and close to 0, that is, the thrust is equal to the current target throttling thrust, and the fuel consumption rate is as small as possible. The smoothness restriction is to add the gene deviation to the objective function value and fitness calculation, so that individuals that meet the optimization goal and have a small interval between optimization variables can be more easily selected and inherited, so that the control law has fewer mutations and is more stable and smooth.

[0066] Preferably, in step 5, based on the genetic optimization algorithm, the control law variable combination corresponding to the throttling optimization thrust point is solved, where it changes from 1 to end until the specific content of the control law combination optimization of all throttling thrust point sets is:

[0067] Step 501: for the i-th thrust point, determine the coding space of the control law variables under the round optimization;

[0068] Step 502: According to the coding space D of the control law variable i , generate the initial population P of genetic algorithm i,0 , including NIND individuals, each of which is different in having different genes, and different genes in the scheme of the present invention are different combinations of control law variables whose values ​​are in the control law variable encoding space;

[0069] Step 503: Calculate the objective function value of each individual in the population, calculate the steady-state performance of the target engine based on the control law variable combination corresponding to the genes of different individuals, and import the engine performance calculation result into the objective function calculation to obtain the objective function value corresponding to the individual;

[0070] Step 504: determine whether the optimal individual objective function value of the population meets the convergence condition. If so, directly proceed to step 501 to optimize the next thrust point i+1. If not, proceed to step 505.

[0071] Step 505: Based on the objective function value of each individual in the population, the fitness function set FIT of the population is calculated by the fitness calculation function. i,j , and sum the gene deviation based on the fitness function set, that is, FIT' i,j =FIT i,j +X delta ;

[0072] Step 506: Based on the fitness function set FIT' i,j , perform selection, crossover and mutation to obtain the next generation population P i,j+1 , and proceed to step 503 at the same time.

[0073] Preferably, the specific content of step 501 is: if the optimization result of the control law variable of the i-1th throttling thrust point is {α=α i-1 ,A VABI =A VAB1,i-1 ,A N0z =A NOZ,i-1}, then combined with the throttling change direction of each control law variable obtained by previous analysis, the encoding space of the control law variable of the i-th thrust point can be directionally adjusted as:

[0074] D i α AFS ∈[α AFS,i-1 -R(α AFS )α AFS,i-1 ]θ RVABI ∈[θ RVABI,i-1 -R(θ RVABI )θ RVABI,i-1 ]A Noz ∈[A Noz,i-1 ,A Noz,i-1 +R(A Noz )]}.

[0075] Preferably, the objective function of step 503 is constructed as follows:

[0076] ObjV=abs|FF i |+SFC+X delta ;

[0077] Where X delta =(XX initial ) 2 is the genetic deviation, X is the genetic value of the individual in the population, and X initial is the optimization result of the control law variables of the i-1th throttling thrust point, that is, {α=α ,i-1 ,A VABI =A VABI,i-1 ,A NOZ =A NOz,i-1}.

[0078] Step 6: Integrate the optimization results of each thrust throttling point to obtain the final throttling process control law solution.

[0079] like Figure 3 As shown in the figure, a schematic diagram of the structure of an adaptive intercooled cycle aircraft engine shows its main components and duct structure. As an engine with a novel configuration and many adjustable geometric components, one of the advantages of the adaptive cycle engine is that it can achieve the conversion between different working modes by adjusting the adjustable geometric components, so as to adapt to various combat environments and combat missions. At the same time, the increase in adjustable geometric components makes it more difficult to determine the control law. Its adjustable geometric components include: rear variable fan adjustable stator blade angle αRFan , CDFS adjustable stator blade angle α CDFS , High pressure compressor adjustable stator blade angle α HPC , low pressure turbine guide vane angle α LPT , main nozzle throat area A Noz , front duct ejector area A FVABI and the area of ​​the rear duct ejector A RVABI .

[0080] Taking the equal flow throttling at high altitude and high speed (2.5Ma 20km) as an example, after sensitivity analysis of the above control law variables, it can be seen that the adjustable stator angle of the high-pressure compressor, the angle of the low-pressure turbine guide vane and the area of ​​the front duct ejector have little effect on the equal flow throttling of the engine, so they are not involved in the optimization of the equal flow throttling control law. At the same time, due to the equal flow throttling, the engine speed is not included in the control law variables. Table 1 lists the upper and lower bounds, initial values, change directions, initial adjustment ranges and equal division lengths of the coding space of other control law variables, where the coding space is divided into four equal parts.

[0081] Table 1

[0082]

[0083] The initial thrust of the engine under this condition is 3349daN, and the cruise process requires an equal flow throttling depth of 70%. Therefore, the throttling thrust point set to be optimized can be discretized according to the throttling thrust discrete percentage of 5%:

[0084] F={3181.5,3014.1,2846.7,2679.2,2511.8,2344.3}.

[0085] Subsequently, based on the genetic optimization algorithm improved by monotonic smoothing, the control law combinations corresponding to each throttling thrust point are optimized in turn, and the optimization results of each thrust throttling point are integrated to obtain the throttling process control law.

[0086] Figure 4 The comparison between the throttling control law obtained by the traditional optimization algorithm and the throttling control law obtained by the monotonic smoothing design method proposed in the solution of the present invention is shown.

[0087] Depend on Figure 4 It can be seen that although the control law obtained by using the traditional optimization algorithm can achieve the throttling optimization goal, the control law variables jump repeatedly during the throttling process, which is very unfavorable for the design and implementation of the engine's adjustable actuator.

[0088] Compared with the traditional optimization algorithm, while achieving the same optimization goal, the equal flow throttling control law obtained by the design method proposed in the scheme of the present invention shows a trend of monotonous and stable change, meeting the design requirements of monotonicity and smoothness of the control law. At the same time, in terms of optimization time, the traditional optimization algorithm takes 611.3s, while the scheme of the present invention takes 316.7s, with a higher optimization convergence speed.

[0089] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A monotonic smoothing design method for aircraft engine throttling control law, characterized in that: Here are the steps: Step 1: according to the characteristics of the engine configuration, determine the engine control law variables involved in the optimization design of the throttling control law, and determine the upper and lower bounds of the corresponding engine control law variable values ​​for each control law variable; Step 2: Based on the existing control law variable combination, a preliminary throttling control law analysis is performed by means of sensitivity analysis to determine the change direction of each control law in the throttling process, and the control law variable values ​​at the initial operating point are recorded. The initial adjustment range is obtained by combining the initial value, change direction and upper and lower limits of each control law variable; Step 3: Divide the initial adjustment range of each control law variable into N equal parts, and calculate the equal division length of the coding space of the control law variable; Step 4: Determine the upper and lower limits of thrust according to the throttling process, and discretize the throttling thrust point set to be optimized; Step 5: Based on the genetic optimization algorithm, start solving the control law variable combination corresponding to the i-th throttling optimization thrust point, where i changes from 1 to end, until the control law combination of all throttling thrust point sets is optimized; Step 6: Integrate the optimization results of each thrust throttling point to obtain the final throttling process control law solution; In step 5, based on the genetic optimization algorithm, the control law variable combination corresponding to the i-th throttling optimization thrust point is solved, where it changes from 1 to end until the specific content of the control law combination optimization of all throttling thrust point sets is: Step 501: for the i-th thrust point, determine the coding space of the control law variables under the round optimization; Step 502: Based on the coding space of the control law variables , generate the initial population of genetic algorithm , which includes NIND individuals, each of which has different genes. Different genes in the scheme [1] are different combinations of control law variables in the control law variable encoding space; Step 503: Calculate the objective function value of each individual in the population, calculate the steady-state performance of the target engine based on the control law variable combination corresponding to the genes of different individuals, and import the engine performance calculation result into the objective function calculation to obtain the objective function value corresponding to the individual; Step 504: determine whether the optimal individual objective function value of the population meets the convergence condition. If so, directly proceed to step 501 to optimize the next thrust point i+1. If not, proceed to step 505. Step 505: Based on the objective function value of each individual in the population, the fitness function set of the population is calculated by the fitness calculation function. , and sum the gene deviation based on the fitness function set, that is, ; Step 506: Based on the fitness function set , perform selection, crossover and mutation to obtain the next generation population , and proceed to step 503; The objective function of step 503 is constructed as follows: ； in is the gene deviation, is the gene value of the individual in the population, is the optimization result of the control law variables of the i-1th throttling thrust point, that is, .

2. The monotonic smoothing design method for aircraft engine throttling control law according to claim 1 is characterized by: The control law variables in step 1 include guide vane angle, duct ejector area, and nozzle throat area.

3. The monotonic smoothing design method for aero-engine throttling control law according to claim 2 is characterized by: Guide vane angle in step 2 , duct ejector area and nozzle throat area The initial adjustment range is: 。 4. The monotonic smoothing design method for aircraft engine throttling control law according to claim 1, characterized in that: Guide vane angle in step 3 , duct ejector area and nozzle throat area The lengths of the equal parts of the coding space are: ； ； 。 5. The monotonic smoothing design method for aircraft engine throttling control law according to claim 1, characterized in that: The throttling thrust point set to be optimized in step 4 is .

6. The monotonic smoothing design method for aircraft engine throttling control law according to claim 1, characterized in that: The specific content of step 501 is: if The optimization result of the control law variables of the throttling thrust point is: , then combined with the throttling change direction of each control law variable obtained by previous analysis, the encoding space of the i-th thrust point control law variable can be adjusted directionally as follows: 。

Images