Optimization Method for Aircraft Engine Fuel Consumption Control Based on TMLSAA

Abstract

The present invention relates to an optimization method for controlling the fuel consumption of an aeroengine based on TMLSAA, which comprises the following steps: Step 1: Establish a non-linear model of the target fuel consumption control process of the aeroengine; Step 2: Establish a target fuel consumption control model of the aeroengine; Step 3: Calculate the optimal solution of the target fuel consumption control model of the aeroengine by using a dual-mode linear spiral attraction optimization algorithm; Step 4: Complete the optimal control of the target fuel consumption control model of the variable cycle aeroengine. The present invention compensates for the defect of being easily trapped in local optimal solutions caused by single-mode optimization in traditional intelligent optimization algorithms by designing a dual-mode optimization structure of an exploration mode and a tracking mode, and improves the optimization speed and optimization accuracy of the algorithm by designing two tracking optimization modes of linear and spiral based on speed update. On the premise of ensuring the safe operation of the engine, the present invention realizes the optimization of the fuel consumption control mode, outputs the optimal control parameters to the aeroengine, and improves the flight distance of the aircraft.

Description

Technical Field

[0001] The present application relates to the field of aircraft engine technology, and in particular to an aircraft engine fuel consumption control optimization method based on TMLSAA. Background Art

[0002] As the heart of an aircraft, the reliability and safety of an aircraft engine are extremely important. However, an aircraft engine is a very complex aerodynamic and thermal rotating machine with many parts. Many parts work under high temperature, high pressure, high speed rotation, strong vibration and complex and changeable environmental conditions. They are often subjected to high loads and thermal shocks. The working environment is extremely harsh, so it is easy to fail. Since the working process of an aircraft engine is complex and changeable, and it has strong nonlinearity, multiple control variables, time-varying, and complex structural characteristics, the study of engine control problems is more difficult than general control systems.

[0003] At the same time, with the rapid development of science and technology, higher requirements are put forward for aircraft engines, which are mainly reflected in the wider flight envelope, improved maneuverability and flexibility, increased thrust-to-weight ratio, reduced fuel consumption, short-distance starting, improved reliability and operability, etc. The purpose of the engine fuel consumption control mode is to ensure that the engine thrust remains unchanged, reduce the engine fuel consumption rate, and increase the working radius of the aircraft under the premise of ensuring the safe operation of the engine.

[0004] Although some achievements have been made in the research on the optimization control of the minimum fuel consumption of engines at home and abroad, there are still many unresolved technical problems or areas for improvement. The difficulty lies in finding an optimization algorithm that has both strong global convergence ability and fast convergence. Therefore, the present invention proposes an optimization method for the fuel consumption control of aircraft engines based on TMLSAA. Summary of the invention

[0005] In order to overcome the shortcomings of the prior art, the present invention realizes the optimization of aircraft engine fuel consumption control by designing a two-mode linear spiral attraction algorithm TMLSAA (Two-Mode Linear Spiral Attraction Algorithm), and makes up for the defect of single-mode optimization in traditional intelligent optimization algorithms that is easily trapped in local optimal solutions by using a dual-mode optimization structure of exploration mode and tracking mode. In addition, by designing two tracking optimization modes, linear and spiral, based on speed update, the optimization speed and optimization accuracy of the algorithm are improved. Under the premise of ensuring the safe operation of the engine, the fuel consumption control mode is optimized, the optimal control parameters are output to the aircraft engine, and the aircraft's range is improved.

[0006] To achieve the above object, the solution adopted by the present invention is: to provide an aircraft engine fuel consumption control optimization method based on TMLSAA, which comprises the following steps:

[0007] Step 1: Establish a non - linear model of the target fuel consumption control process of an aero - engine. The target fuel consumption usually refers to the minimum fuel consumption. The non - linear model of the target fuel consumption control process of the aero - engine is as follows:

[0008] y = H(u);

[0009] Where: y represents the output vector of the target fuel consumption control model of the aero - engine; u represents the input vector of the target fuel consumption control model of the aero - engine; H represents the non - linear vector function that generates the system output;

[0010] Step 2: Establish a target fuel consumption control model for the aero - engine;

[0011] The described target fuel consumption control model for the aero - engine is, on the premise of ensuring the safe operation of the engine, to ensure that the engine thrust remains unchanged and reduce the specific fuel consumption of the engine by changing the control parameters. Its mathematical model description is as follows:

[0012]

[0013] Where: const represents the expected thrust value, which is a constant; g i (u) represents the constraint condition of the i - th control parameter; g imin and g imax represent the lower limit value and the upper limit value of the constraint condition of the i - th control parameter respectively; F represents the aero - engine thrust; sfc represents the aero - engine fuel consumption; min represents taking the minimum value; n represents the number of constraint conditions; f represents the non - linear vector function that generates the system output; x represents the control parameter variable; W f represents adjusting the main fuel flow; A9 represents the area of the tail nozzle; dvgl represents the fan guide vane angle; dvgh represents the compressor guide vane angle;

[0014] The constraint conditions of the described control parameters are as follows:

[0015]

[0016] Where: T t4 represents the aero - engine temperature; T t4max represents the upper limit value of the aero - engine temperature; n F represents the fan speed; n Fmax represents the upper limit of the fan speed; n H represents the high - pressure compressor speed; n Hmax represents the upper limit of the high - pressure compressor speed; SMF represents the fan surge margin; SMF min represents the minimum allowable value of the fan surge margin; SMC represents the high - pressure compressor surge margin; SMC min represents the minimum allowable value of the high - pressure compressor surge margin; ROGmin Represents the lower limit of the fuel-air ratio in the combustion chamber; R OG Represents the fuel-air ratio in the combustion chamber; R OGmax Represents the upper limit of the fuel-air ratio in the combustion chamber; W f Represents the fuel supply of the main combustion chamber; W fmax Represents the upper limit of the fuel supply of the main combustion chamber;

[0017] Step 3: Calculate the optimal solution of the aero-engine target fuel consumption control model using the dual-mode linear spiral attraction optimization algorithm;

[0018] Step 31: Initialize the initial value of the solution of the aero-engine target fuel consumption control model and determine the control space of the determined fuel consumption control model;

[0019] Step 32: Determine the exploration mode and tracking mode of the aero-engine fuel consumption control model;

[0020] Step 33: Determine the particle velocity and particle position of the aero-engine fuel consumption control model;

[0021] The method for determining the particle velocity of the fuel consumption control model is as follows;

[0022] V i (t + 1) = ωV i (t) + μ[p g (t) - U i (t)];

[0023] In the formula: μ represents the learning factor; V i (t + 1) and V i (t) respectively represent the next iteration and the current particle velocity; p g (t) represents the optimal position searched by the current entire control point group so far; U i (t) represents the value of the i-th particle of the current entire particle swarm; ω represents the weight adjustment coefficient matrix of the constraint function;

[0024] The method for determining the particle position of the fuel consumption control model is as follows;

[0025]

[0026] In the formula: U i (t + 1) represents the particle position at the updated time t + 1; U best represents the original particle position; η represents a random number between [0.45, 0.95]; β represents a random number within [-1, 1]; U r1 (t) represents the position of the r1-th randomly selected particle; if represents satisfying the condition; rand represents a random number between [0, 1]; else represents other cases where the judgment condition is not satisfied.

[0027] Step 4: Complete the optimal control of the target fuel consumption control model for the variable cycle aeroengine;

[0028] Obtain the optimal solution of the fuel consumption control model determined in Step 3, and output the optimal solution to the aeroengine to achieve fuel consumption control under the condition of constant thrust of the variable cycle aeroengine.

[0029] y best = H(u best );

[0030] Where: u best represents the optimal solution of the fuel consumption control model determined in Step 3, which is the optimal control quantity output to the aeroengine, and y best represents the optimal output of the aeroengine under the optimal control quantity.

[0031] Preferably, the output vector of the fuel consumption control model in Step 1 is specifically:

[0032] The output vector of the target fuel consumption control model of the aeroengine is as follows:

[0033] y = [sfc F] T ∈ R 2 ;

[0034] Where: sfc represents the specific fuel consumption; F represents the engine thrust value; R 2 represents the two-dimensional real number field of the output vector.

[0035] Preferably, the input vector of the fuel consumption control model in Step 1 is specifically:

[0036] The input vector of the target fuel consumption control model of the aeroengine is as follows:

[0037] u = [W f , A8, dvgl, dvgh] T ∈ R 4 ;

[0038] Where: W f represents the adjustment of the main fuel flow; A8 represents the area of the tail nozzle; dvgl represents the fan guide vane angle; dvgh represents the compressor guide vane angle; R 4 represents the four-dimensional real number field of the input vector.

[0039] Preferably, the control points of the optimization model of the initial aeroengine in Step 31 are specifically:

[0040] Initialize to find S D-dimensional particles U1, U2,..., U s, where \(i = 1, 2, \cdots, S\), and generate \(S\) initial velocities \(V_1, V_2, \cdots, V_S\). s , where \(i = 1, 2, \cdots, S\), to form the initial solution of the control variables, as shown below:

[0041] \(P_0=\{U_1, U_2, \cdots, U_S\}\), where \(i = 1, 2, \cdots, S\); s

[0042] In the formula: \(P_0\) represents the set of model control points; \(U_1, U_2, \cdots, U_S\) s represent the 1st, 2nd, \(\cdots\), \(S\)th model control points respectively; \(i\) represents the number of the control point; \(S\) represents the number of control points.

[0043] Preferably, the control space of the determined fuel consumption control model in step 31 is specifically:

[0044] The control space of the fuel consumption control model includes the optimal control point \(U_{opt}\) best and the worst control point \(U_{worst}\), worst The method for obtaining the optimal control point \(U_{opt}\) is as follows: best

[0045]

[0046] In the formula: \(U_{opt}\) best represents the optimal control point of the aeroengine; \(min\) represents taking the minimum value;

[0047] The method for obtaining the worst control point \(U_{worst}\) is as follows: worst

[0048]

[0049] In the formula: \(U_{worst}\) worst represents the worst control point of the aeroengine; \(max\) represents taking the maximum value;

[0050] The convergence condition for judging the optimization process is: \(\|H(U_{k + 1})-H(U_k)\|\leq\varepsilon\), where \(\varepsilon\) represents the convergence accuracy; or the number of iterations satisfies \(t < T\) best ; worst If the above convergence condition is satisfied, then output \(U_{opt}\), \(H(U_{opt})\) as the optimal control point of the aeroengine, and end the optimization process, otherwise continue to the next step. max

[0051] Preferably, the exploration mode and tracking mode of the determined optimization model of the aeroengine in step 32 are specifically: best worst

[0052]

[0053] The exploration mode of the optimization model is as follows:

[0054] Copy the S-dimensional control point P t ={U1, U2, …, U s} to obtain P t '={U1, U2, …, U s}. The control point P t ={U1, U2, …, U s} is used for the exploration mode, and the backup control point P t '={U1, U2, …, U s} is used for the tracking mode; Copy the current positions of the S-dimensional particle swarm P t ={U1, U2, …, U s} j copies respectively and place them in S storage pools, where j = G, that is, the size of the storage pool is j; The method for obtaining the number of copies of the storage pool position is as follows:

[0055] j=(G - 1);

[0056] In the formula: j represents the number of copies of the position; G represents the storage pool dimension;

[0057] Keep the current position as the candidate value;

[0058] For each individual copy in each storage pool, update it to replace the original value, as follows:

[0059] P Si ={U' 1,i , U' 2,i , …, U' s,i}, i = 1, 2, …, S;

[0060] In the formula: P Si represents; U' 1,i , U' 2,i , …, U' s,i respectively represent the updated 1st, 2nd, …, Sth model control points;

[0061] Calculate the fitness values H(U' b,i ) of all candidate values in the storage pool respectively, i = 1, 2, …, S, b = 1, 2, …, j; Select the candidate value with the highest fitness value from each storage pool to replace the current position of the particle, and complete the position update of the particle;

[0062] The tracking mode of the optimization model is as follows:

[0063] Attraction operator; Use the particle swarm P t '={U1, U2, …, U sThe control point in {} calculates the attraction operator λ as follows:

[0064]

[0065] Where: λ represents the attraction operator; H max and H min respectively represent the worst and best fitness function values of the control point P t ' = {U1, U2, …, U s}; H(i) represents the i-th fitness function value;

[0066] When the attraction operator is equal to or less than 0.4, this control point will not be selected and will be replaced by the second control point, as follows:

[0067] U i (t) = U best + aU r1 +(1 - a)U r2 (t)

[0068] Where: U r1 (t) and U r2 (t) represent the positions of the r1-th and r2-th randomly selected particles, r1 ≠ r2; U i (t) represents the position of the current particle; a represents a random number between [0, 1].

[0069] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0070] (1) The present invention compensates for the defect of being easily trapped in local optimal solutions caused by single-mode optimization in traditional intelligent optimization algorithms by designing a dual-mode optimization structure of exploration mode and tracking mode, and improves the optimization speed and accuracy of the algorithm by designing two tracking optimization modes of linear and spiral based on speed update;

[0071] (2) On the premise of ensuring the safe operation of the engine, the present invention realizes the optimization of the fuel consumption control mode, outputs the optimal control parameters to the aeroengine, and improves the flight distance of the aircraft. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 is the control block diagram of the aeroengine fuel consumption control optimization method based on TMLSAA according to the embodiment of the present invention;

[0073] Figure 2 is the schematic diagram of the variable cycle aeroengine according to the embodiment of the present invention;

[0074] Figure 3 is the schematic diagram of the fuel consumption control mode principle according to the embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0075] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

[0076] Based on the fuel consumption model of an aeroengine, the embodiment of the present invention makes up for the defect of being easily trapped in local optimal solutions caused by single-mode optimization in traditional intelligent optimization algorithms through dual-mode optimization structures of a design exploration mode and a tracking mode. Moreover, by designing two tracking optimization modes of linear and spiral based on speed update, the optimization speed and optimization accuracy of the algorithm are improved. On the premise of ensuring the safe operation of the engine, the mode optimization of fuel consumption control is realized, and the optimal control parameters are obtained for the aeroengine, improving the flight distance and endurance time of the aircraft. As Figure 1 shown is the control block diagram of the aeroengine fuel consumption control optimization method based on TMLSAA according to the embodiment of the present invention.

[0077] The embodiment of the present invention provides an aeroengine fuel consumption control optimization method based on TMLSAA. To prove the applicability of the present invention, it is applied to an example, which specifically includes the following steps:

[0078] S1: Establish a nonlinear model of the target fuel consumption control process of the aeroengine;

[0079] As Figure 2 shown is a schematic diagram of a variable cycle aeroengine according to the embodiment of the present invention. In the figure, 1 represents the secondary bypass duct, 2 represents the primary bypass duct, 3 represents the total bypass duct, 4 represents the tail-end duct, 5 represents the inlet duct, 6 represents the fan, 7 represents the CDFS, 8 represents the high-pressure compressor, 9 represents the high-pressure turbine, 10 represents the low-pressure turbine, 11 represents the mixing chamber, 12 represents the afterburner, and 13 represents the tail nozzle; the nonlinear model of the target fuel consumption control process of the aeroengine is as follows:

[0080] y = H(u);

[0081] In the formula: y represents the output vector of the target fuel consumption control model of the aeroengine; u represents the input vector of the target fuel consumption control model of the aeroengine; H represents the nonlinear vector function generating the system output;

[0082] The output vector of the target fuel consumption control model of the aeroengine is as follows:

[0083] y = [sfc F] T ∈R 2 ;

[0084] In the formula: sfc represents the specific fuel consumption; F represents the engine thrust value; R 2 represents the two-dimensional real number field of the output vector.

[0085] The input vector of the target fuel consumption control model of the aeroengine is as follows:

[0086] u = [Wf , A8, dvgl, dvgh] T ∈R 4 ;

[0087] Where: W f represents the adjustment of the main fuel flow; A8 represents the area of the tail nozzle; dvgl represents the fan guide vane angle; dvgh represents the compressor guide vane angle; R 4 represents the four-dimensional real number field of the input vector.

[0088] S2: Establish an aero-engine target fuel consumption control model;

[0089] The aero-engine target fuel consumption control model is to ensure that the engine thrust remains unchanged while ensuring the safe operation of the engine, and reduce the specific fuel consumption of the engine by changing the control parameters. Its mathematical model is described as follows:

[0090]

[0091] Where: const represents the expected thrust value, which is a constant; g i (u) represents the constraint condition of the i-th control parameter; g imin and g imax represent the lower limit value and the upper limit value of the constraint condition of the i-th control parameter respectively; F represents the aero-engine thrust; sfc represents the aero-engine fuel consumption; min represents taking the minimum value; n represents the number of constraint conditions; f represents the non-linear vector function that generates the system output; x represents the control parameter variable; W f represents the adjustment of the main fuel flow; A9 represents the area of the tail nozzle; dvgl represents the fan guide vane angle; dvgh represents the compressor guide vane angle;

[0092] The constraint conditions of the control parameters are as follows:

[0093]

[0094] Where: T t4 represents the aero-engine temperature; T t4max represents the upper limit value of the aero-engine temperature; n F represents the fan speed; n Fmax represents the upper limit of the fan speed; n H represents the high-pressure compressor speed; n Hmax represents the upper limit of the high-pressure compressor speed; SMF represents the fan surge margin; SMF min represents the minimum allowable value of the fan surge margin; SMC represents the high-pressure compressor surge margin; SMC min represents the minimum allowable value of the high-pressure compressor surge margin; R OGmin represents the lower limit of the fuel-air ratio in the combustion chamber; R OGRepresents the air-fuel ratio of the combustion chamber; R OGmax Represents the upper limit of the air-fuel ratio of the combustion chamber; W f Represents the fuel supply of the main combustion chamber; W fmax Represents the upper limit of the fuel supply of the main combustion chamber;

[0095] S3: Calculate the optimal solution of the aero-engine target fuel consumption control model using the dual-mode linear spiral attraction optimization algorithm;

[0096] S31: Initialize the initial value of the solution of the aero-engine target fuel consumption control model and determine the control space of the fuel consumption control model; Initialize to find S D-dimensional particles U1, U2, …, U s , i = 1, 2, ..., S, and generate S initial velocities V1, V2, …, V s , i = 1, 2, ..., S, to form the initial solution of the control variables, as shown below:

[0097] P0 = {U1, U2, …, U s}, i = 1, 2, ..., S;

[0098] In the formula: P0 represents the set of model control points; U1, U2, …, U s respectively represent the 1st, 2nd … the Sth model control points; i represents the number of the control point; S represents the number of control points.

[0099] The control space of the fuel consumption control model includes the optimal control point U best and the worst control point U worst , and the method for obtaining the optimal control point U best is as shown below:

[0100]

[0101] In the formula: U best represents the optimal control point of the aero-engine; min represents taking the minimum value;

[0102] The method for obtaining the worst control point U worst is as shown below:

[0103]

[0104] In the formula: U worst represents the worst control point of the aero-engine; max represents taking the maximum value;

[0105] The convergence condition for judging the optimization process is: ||H(U best ) - H(U worst )|| ≤ ε, where ε represents the convergence accuracy, and in this embodiment, it is taken as 10 -3 ; or the number of iterations satisfies t < T max;

[0106] If the above convergence conditions are met, output U best , H(U worst ) as the optimal control point of the aeroengine and end the optimization process; otherwise, continue to the next step.

[0107] S32: Determine the exploration mode and tracking mode of the fuel consumption control model of the aeroengine;

[0108] The exploration mode of the optimization model is as follows:

[0109] Copy the S-dimensional control point P t = {U1, U2, …, U s} to obtain P t ' = {U1, U2, …, U s}; the control point P t = {U1, U2, …, U s} is used for the exploration mode, and the backup control point P t ' = {U1, U2, …, U s} is used for the tracking mode; copy the current position of the S-dimensional particle swarm P t = {U1, U2, …, U s} j copies and place them in S storage pools respectively, where j = G, that is, the size of the storage pool is j; the method for obtaining the number of copies of the storage pool position is as follows:

[0110] j = (G - 1);

[0111] In the formula: j represents the number of copies of the position; G represents the dimension of the storage pool;

[0112] Keep the current position as the candidate value;

[0113] Update each individual copy in each storage pool to replace the original value, as follows:

[0114] P Si = {U' 1,i , U' 2,i ,..., U' s,i}, i = 1, 2,..., S;

[0115] In the formula: P Si represents; U' 1,i , U' 2,i ,..., U' s,i respectively represent the updated first, second... the Sth model control points;

[0116] Calculate the fitness values H(U' b,i), i = 1, 2, ..., S, b = 1, 2, ..., j; Select the candidate value with the highest fitness value from each storage pool to replace the current particle's position, and complete the update of the particle's position;

[0117] The tracking mode of the optimization model is as follows:

[0118] Attraction operator; Use the control points in the particle swarm P t ' = {U1, U2, …, U s} to calculate the attraction operator λ, as follows:

[0119]

[0120] In the formula: λ represents the attraction operator; H max and H min respectively represent the worst and best fitness function values of the control points P t ' = {U1, U2, …, U s}; H(i) represents the i-th fitness function value;

[0121] When the attraction operator is equal to or less than 0.4, this control point will not be selected and will be replaced by the second control point, as follows:

[0122] U i (t) = U best + aU r1 +(1 - a)U r2 (t)

[0123] In the formula: U r1 (t) and U r2 (t) represent the positions of the r1-th and r2-th randomly selected particles, r1 ≠ r2; U i (t) represents the position of the current particle; a represents a random number between [0, 1], which is randomly generated by the computer.

[0124] S33: Determine the particle velocity and particle position of the fuel consumption control model of the aeroengine;

[0125] The method for determining the particle velocity of the fuel consumption control model is as follows;

[0126] V i (t + 1) = ωV i (t)+μ[p g (t)-U i (t)];

[0127] In the formula: μ represents the learning factor; V i (t + 1) and V i (t) respectively represent the next iteration and the current particle velocity; p g(t) represents the optimal position found so far for the entire current control point group; U i (t) represents the value of the i-th particle in the current entire particle swarm; ω represents the weight adjustment coefficient matrix of the constraint function;

[0128] The method for determining the particle position of the fuel consumption control model is as follows;

[0129]

[0130] In the formula: U i (t + 1) represents the particle position at the updated time t + 1; U best represents the original particle position; η represents a random number between [0.45, 0.95]; β represents a random number within [-1, 1]; U r1 (t) represents the position of the r1-th randomly selected particle; if represents satisfying the condition; rand represents a random number between [0, 1], randomly generated by the computer; else represents other situations where the judgment condition is not satisfied.

[0131] S4: Complete the optimal control of the variable cycle aeroengine target fuel consumption control model;

[0132] Obtain the optimal solution of the fuel consumption control model determined in step 3, output the optimal solution to the aeroengine, and achieve fuel consumption control under the condition of constant thrust of the variable cycle aeroengine.

[0133] y best = H(u best );

[0134] In the formula: u best represents the optimal solution of the fuel consumption control model determined in step 3, which is the optimal control quantity output to the aeroengine, and y best represents the optimal output of the aeroengine under the optimal control quantity.

[0135] Compare the optimization algorithm proposed in the present invention with the traditional genetic algorithm and particle swarm optimization algorithm, and optimize the minimum fuel consumption respectively. Taking the supersonic cruise point, 9 km, 1.5 Ma, and throttle lever angle of 70° as an example, the simulation results are shown in Table 1 for the comparison of the simulation results of this method and other methods. Therefore, the intelligent optimization algorithm proposed in this application has a lower fuel consumption rate and a lower thrust error accuracy under the premise of satisfying the constraint conditions. As Figure 3 shown is the schematic diagram of the fuel consumption control mode determined according to the embodiment of the present invention.

[0136] Table 1 Comparison table of simulation results of the optimization method of this application and other methods

[0137] Specific fuel consumption Thrust error accuracy Surge margin lowest boundary Maximum temperature before turbine Optimization method of this application 0.84 0 0.31 1530K Genetic algorithm 0.88 0.05％ 0.33 1519K Particle swarm optimization algorithm 0.87 0.04％ 0.32 1525K

[0138] In summary, the prediction results of the optimization method for aircraft engine fuel consumption control based on TMLSAA in this case prove to have good effects.

[0139] (1) The embodiment of the present invention compensates for the defect of being easily trapped in local optimal solutions caused by single-mode optimization in traditional intelligent optimization algorithms through the dual-mode optimization structure of the design exploration mode and the tracking mode. Moreover, by designing two tracking optimization modes, namely linear and spiral, based on speed update, the optimization speed and accuracy of the algorithm are improved. The comparison of the calculation results of different methods proves that this method has good performance.

[0140] (2) The embodiment of the present invention realizes the mode optimization of fuel consumption control while ensuring the safe operation of the engine through the designed TMLSAA method, obtains the optimal control parameters for the aircraft engine, and improves the flight distance and endurance time of the aircraft. The data comparison in the embodiment proves that this method can meet the actual use requirements and has good application effects.

[0141] The above-described embodiments are only descriptions of the preferred embodiments of the present invention and do not limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of the present invention shall fall within the protection scope determined by the claims of the present invention.

Claims

1. An optimization method for controlling the fuel consumption of an aero-engine based on TMLSAA, characterized in that It includes the following steps: Step 1: Establish a non - linear model of the target fuel consumption control process of an aero - engine; The non - linear model of the target fuel consumption control process of the aero - engine is as follows: y = H(u); Where: y represents the output vector of the target fuel consumption control model of the aero - engine; u represents the input vector of the target fuel consumption control model of the aero - engine; H represents the non - linear vector function that generates the system output; Step 2: Establish a target fuel consumption control model of the aero - engine; The described target fuel consumption control model of the aero - engine is to ensure that the engine thrust remains unchanged while ensuring the safe operation of the engine, and reduce the fuel consumption rate of the engine by changing the control parameters. Its target fuel consumption control model is described as follows: Where: const represents the expected value of thrust, which is a constant; g k (u) represents the constraint condition of the k-th control parameter; g kmin and g kmax respectively represent the lower limit value and the upper limit value of the constraint condition of the k-th control parameter; F represents the thrust of the aeroengine; sfc represents the fuel consumption of the aeroengine; min represents taking the minimum value; n represents the number of constraint conditions; The constraint conditions of the control parameters are as follows: Where: T t4 represents the temperature of the aero-engine; T t4max represents the upper limit value of the aero-engine temperature; n F represents the fan speed; n Fmax represents the upper limit of the fan speed; n H represents the high-pressure compressor speed; n Hmax represents the upper limit of the high-pressure compressor speed; SMF represents the fan surge margin; SMF min represents the minimum allowable value of the fan surge margin; SMC represents the high-pressure compressor surge margin; SMC min represents the minimum allowable value of the high-pressure compressor surge margin; R OGmin represents the lower limit of the fuel-air ratio in the combustor; R OG represents the fuel-air ratio in the combustor; R OGmax represents the upper limit of the fuel-air ratio in the combustor; W f represents the fuel supply amount of the main combustion chamber; W fmax represents the upper limit of the fuel supply amount to the main combustion chamber; Step 3: Calculate the optimal solution of the target fuel consumption control model of the aero - engine with a dual - mode linear spiral attraction optimization algorithm; Step 31: Initialize the initial value of the target fuel consumption control model of the aero - engine and determine the control space of the target fuel consumption control model; Step 32: Determine the exploration mode and tracking mode of the fuel consumption control model of the aero - engine; Step 33: Determine the particle velocity and particle position of the fuel consumption control model of the aero - engine; The method for determining the particle velocity of the fuel consumption control model is as follows; V i (t + 1)= ωV i (t)+ μ[p g (t)- U i (t)]; Where: μ represents the learning factor; V i (t + 1) and V i (t) represent the next iteration and the current particle velocity respectively; p g (t) represents the optimal position searched so far by the current entire control point group; U i (t) represents the position of the i-th particle of the current entire particle swarm; ω represents the weight adjustment coefficient matrix of the constraint function; The method for determining the particle position of the fuel consumption control model is as follows; Where: U i (t + 1) represents the particle position at the updated time t + 1; U best represents the original particle position; η represents a random number between [0.45, 0.95]; β represents a random number within [-1, 1]; U r1 (t) represents the position of the r1-th randomly selected particle; if represents satisfying the condition; rand represents a random number between [0, 1]; else represents other cases where the judgment condition is not satisfied; Step 4: Complete the optimal control of the target fuel consumption control model of the variable - cycle aero - engine; Obtain the optimal solution of the fuel consumption control model determined in Step 3, and output the optimal solution to the aero - engine to achieve fuel consumption control under the condition that the thrust of the variable - cycle aero - engine remains unchanged; y best = H(u best ); Where: u best represents the optimal solution of the fuel consumption control model determined in step 3, and is the optimal control quantity output to the aero-engine, y best represents the optimal output of the aero-engine under the optimal control quantity.

2. The optimization method for controlling the fuel consumption of an aeroengine based on TMLSAA according to claim 1, wherein The output vector of the fuel consumption control model in Step 1 is specifically: The output vector of the target fuel consumption control model of the aero - engine is as follows: y = [sfc F] T ∈R 2 ; Where: sfc represents the fuel consumption of an aeroengine; F represents the engine thrust value; R 2 represents the two-dimensional real number field of the output vector.

3. The optimization method for controlling the fuel consumption of an aero-engine based on TMLSAA according to claim 2, wherein, The input vector of the fuel consumption control model in Step 1 is specifically: The input vector of the target fuel consumption control model of the aero - engine is as follows: u = [W f , A8, dvgl, dvgh] T ∈R 4 ; Where: W f represents the adjustment of the main fuel flow; A8 represents the area of the tail nozzle; dvgl represents the fan guide vane angle; dvgh represents the compressor guide vane angle; R 4 represents the four-dimensional real number field of the input vector.

4. The optimization method for controlling the fuel consumption of an aeroengine based on TMLSAA according to claim 3, characterized in that, The control point of the optimal model of the initial aero - engine in Step 31 is specifically: Initialize and find S D-dimensional particles U1, U2, …, US s , where i = 1, 2, ..., S, and generate S initial velocities V1, V2, …, VS s , where i = 1, 2, ..., S, to form the initial solution of the control variables as follows: P0 = {U1, U2, …, U s}, i = 1, 2, …, S; Where: P0 represents the set of model control points; U1, U2, …, U s respectively represent the 1st, 2nd, …, Sth model control points; i represents the number of the control point; S represents the number of control points.

5. The optimization method for controlling the fuel consumption of an aero-engine based on TMLSAA according to claim 4, characterized in that, The control space of the determined fuel consumption control model in Step 31 is specifically: The control space of the fuel consumption control model includes an optimal control point U best and a worst control point U worst . The method for obtaining the optimal control point U best is as follows: where: U best represents the optimal control point of the aero-engine; min represents taking the minimum value; The worst control point U worst is obtained as follows: Where: U worst represents the worst control point of the aeroengine; max represents taking the maximum value; The convergence condition for the optimization process is: ||H(U best ) - H(U worst )|| ≤ ε, where ε represents the convergence accuracy; or the number of iterations satisfies t < T max ; If the above convergence condition is satisfied, output U best , H(U worst ) as the optimal control point of the aero-engine, and end the optimization process; otherwise, continue to the next step.

6. The optimized method for controlling the fuel consumption of an aero-engine based on TMLSAA according to claim 5, wherein, The exploration mode and tracking mode of the determined optimal model of the aero - engine in Step 32 are specifically: The exploration mode of the optimal model is as follows: Copy the S-dimensional control point P t ={U1, U2, …, U s} to obtain P t '={U1, U2, …, U s}. The control point P t ={U1, U2, …, U s} is used for the exploration mode, and the backup control point P t '={U1, U2, …, U s} is used for the tracking mode; Copy the current positions of the S-dimensional particle swarm P t ={U1, U2, …, U s} j copies respectively and place them in S storage pools. j = G, that is, the size of the storage pool is j; The method for obtaining the number of copies of the storage pool position is as follows: j=(G - 1); Where: j represents the number of copies of the position; G represents the dimension of the storage pool; Keep the current position as a candidate value; For each individual copy in each storage pool, update it to replace the original value, as follows: P Si = {U' 1,i , U' 2,i ,..., U' s,i}, i = 1, 2,..., S; Where: P Si represents; U' 1,i , U' 2,i ,..., U' s,i respectively represent the updated first, second,..., S-th model control points; Calculate the fitness value \(H(U')\) of all candidate values in the storage pool respectively, where \(i = 1, 2, \cdots, S\) and \(b = 1, 2, \cdots, j\); select the candidate value with the highest fitness value from each storage pool to replace the position of the current particle, and complete the position update of the particle; b,i ​ The tracking mode of the optimal model is as follows: Attraction operator; using the particle swarm P t ' = {U1, U2, …, U s} to calculate the attraction operator λ as follows: Where: λ represents the attraction operator; H max and H min respectively represent the worst and best fitness function values of the control point P t ' = {U1, U2, …, U s}; H(i) represents the i-th fitness function value; When the attraction operator is equal to or less than 0.4, this control point will not be selected and will be replaced by the second control point, as follows: U i U(t) = U best + aU r1 +(1 - a)U r2 U(t) Where: U r1 (t) and U r2 (t) represent the positions of the r1-th and r2-th randomly selected particles, where r1 ≠ r2; U i (t) represents the position of the i-th particle in the current particle swarm; a represents a random number between [0, 1].

Images