CN115758870A - Multi-source energy optimal configuration method of hypersonic aircraft - Google Patents
Multi-source energy optimal configuration method of hypersonic aircraft Download PDFInfo
- Publication number
- CN115758870A CN115758870A CN202211390198.XA CN202211390198A CN115758870A CN 115758870 A CN115758870 A CN 115758870A CN 202211390198 A CN202211390198 A CN 202211390198A CN 115758870 A CN115758870 A CN 115758870A
- Authority
- CN
- China
- Prior art keywords
- population
- covariance matrix
- energy
- mode
- optimal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Control Of Turbines (AREA)
Abstract
The invention discloses a multisource energy optimal configuration method of a hypersonic aircraft, which is characterized in that energy configuration with the lowest fuel consumption and the largest thrust is obtained by adopting a multi-target self-adaptive covariance matrix and a chaotic search cluster algorithm MOACCGA, the strong nonlinear change of a system is adapted by an improved method of the self-adaptive covariance matrix, local optimization is avoided by the improved method of chaotic search, multisource electric energy is accurately distributed by the hypersonic aircraft in a dynamic environment, and the technical problem of optimal configuration of the hypersonic aircraft energy is solved.
Description
Technical Field
The invention relates to an aircraft engine, in particular to a multi-source energy optimal configuration method of a hypersonic aircraft, and belongs to the technical field of aerospace.
Background
The hypersonic aerocraft has the characteristics of good stealth, high flying speed, wide striking range, large effective load and the like, and is called as a third revolutionary result of aviation history after a propeller and a jet propeller by military experts. Under the support of a novel combined cycle power technology, the hypersonic aircraft technology is developed in a crossing manner in recent years, at present, domestic research mostly focuses on the key technology of a main power core machine, and research on an effective electric energy generation method, efficient energy configuration and the like of the hypersonic aircraft is weak.
With the development of aircrafts in the direction of multi-electric power/full-electric power, the hypersonic aircraft has higher and higher power consumption requirements, which puts higher requirements on electric energy generation systems on the aircrafts, and multi-source energy extraction under turbine-based combined power is a core technology to be solved urgently. Turbine base combined power applied to hypersonic aircraft mainly improves through main power air inlet/oil gas total enthalpyExtracting shaft power, heat energy and cooling fuel oil
The method is used for realizing multi-source electric energy generation, the multi-source electric energy generation can affect the unit fuel consumption rate of the main power, the stability margin of the turbine power compressor, the intake stability margin of the stamping power, the combustion stability and other performances, and the influence degrees and the influence mechanisms of different energy extraction modes and parameters thereof on the performance of the main power are different. Therefore, the influence mechanism of multi-source energy extraction on the main power performance is strengthened, the multi-source energy optimization configuration method is researched, the balanced development of the hypersonic aircraft system technology is facilitated, and the requirements of a comprehensive optimization design method for higher thrust-weight ratio and lower fuel consumption rate are met.
Disclosure of Invention
The invention aims to provide a multi-source energy optimal configuration method of a hypersonic aircraft, aiming at the defects of the prior art. The method adopts an energy configuration scheme with the lowest oil consumption rate and the largest thrust based on a multi-objective adaptive covariance matrix and a chaotic search cluster algorithm MOACCGA, adapts to strong nonlinear change of a system by an improved method of the adaptive covariance matrix, avoids local optimization by the improved method of chaotic search, realizes accurate distribution of multi-source electric energy of the hypersonic aircraft in a dynamic environment, and solves the technical problem of optimal energy configuration of the hypersonic aircraft.
In order to achieve the purpose, the invention adopts the following technical scheme:
a multisource energy optimal configuration method of a hypersonic aircraft is based on a parallel turbine-based combined cycle TBCC engine and a multisource energy extraction system, and comprises two working modes of a turbofan engine and a ramjet engine;
the method is characterized in that: the method comprises the steps of taking the generated energy of each energy extraction mode as an optimization design variable, taking the oil consumption rate and the thrust as optimization objective functions, taking the rotating speed of a mechanical shaft, the temperature in front of a turbine, the stability margin and the like as constraint conditions, adopting a MOACCGA based on a multi-target self-adaptive covariance matrix and a chaotic search cluster algorithm, adapting to the strong nonlinear change of a system by an improved method of the self-adaptive covariance matrix, avoiding the local optimization by the improved method of chaotic search, realizing the accurate distribution of multi-source electric energy of the hypersonic aerocraft in a dynamic environment, obtaining the energy user configuration with the lowest oil consumption rate and the highest thrust, and finally obtaining the optimal multi-source electric energy distribution of the oil consumption rate and the Pareto thrust in the whole flight envelope; the method comprises the following steps:
s1, determining a working mode in a current flight state and selecting a corresponding multi-source energy extraction mode
Confirming the working mode of a current engine according to the current flight altitude H and the flight Mach number Ma of the airplane and selecting a corresponding multi-source energy extraction mode; there are three energy extraction modes in turbofan engine mode: (1) Shaft work is extracted for power generation, and the shaft work is extracted from a high-pressure shaft and a low-pressure shaft; (2) Air entraining and power generation are carried out, and the air entraining and power generation are extracted from three parts of a high-pressure stage, a middle stage and an outer duct; (3) Gas power generation, which is extracted from a high-pressure shaft and a low-pressure shaft; the shaft work is extracted to generate power, namely, the transmission ratio is adjusted by a gear box from the high/low pressure of a turbofan engine to drive a motor to generate power; bleed air power generation, namely, gas is led out from a high-pressure stage/middle stage/outer duct of a turbofan engine to drive a motor to generate power; gas power generation, namely, after gas is led out, the gas is combusted in a combustion chamber and then drives a motor to generate power; four energy extraction modes are involved in the ramjet mode: (1) carrying out air-entraining power generation; (2) generating power by semiconductor temperature difference; (3) generating power by an oil-gas turbine; (4) a storage battery; bleed air power generation, namely leading out a gas driving motor from an isolation section of a ramjet; the semiconductor temperature difference power generation is to draw heat from the wall surface of the combustion chamber, and the formed temperature difference enables the P-type thermocouple and the N-type thermocouple to generate potential difference, so that power is supplied to a load; the oil-gas turbine generates electricity, namely, hydrocarbon fuel absorbs heat through a cooling channel of the combustion chamber and is cracked into mixed oil gas, and the mixed oil gas is introduced into the oil-gas turbine to generate electricity; the storage battery does not need to extract energy from the engine, and only the weight causes fuel compensation;
s2: multiple optimization targets for determining multi-source energy optimization configuration of hypersonic aircraft
The fuel consumption rate Wf and the thrust FN of the engine are used as optimization targets, so that the fuel economy and the thrust of the system are considered, the cruising ability of the supersonic aircraft is improved, and the flight quality is improved;
s3: determining constraint conditions and working requirements of multi-source energy optimization configuration of hypersonic aircraft
The constraints of the turbofan engine mode are: turbine front temperature T, high pressure shaft speed N LPS Low pressure shaft speed N HPS Stability margin SM of high pressure compressor HPC And stability margin SM of low pressure compressor LPS (ii) a The operating requirements for turbofan engine modes are: electric energy demand P w ；
The constraints for the ramjet mode are: combustion stability SM of a combustion chamber COV (ii) a The operating requirements for the ramjet mode are: electric energy demand P c ；
S4: optimizing energy distribution schemes for multi-source energy extraction
Within a flight envelope, aiming at the flight altitude H, the flight Mach number Ma and the electric energy demand P which dynamically changes along with the working mode of an engine w And P c And distributing the output power of the multi-source energy extraction system by using a multi-target self-adaptive covariance matrix and a chaotic search cluster algorithm MOACCGA.
Further, the fuel penalty and thrust loss due to the engine are both caused by extracting energy from the engine. In the turbofan engine mode, the multi-source energy optimization configuration for a hypersonic aircraft with optimization targets of fuel consumption Wf and thrust FN can be described as:
in the formula (1), P z,1 、P z,2 Extracting the output power of power generation for the shaft work of the high-pressure shaft and the low-pressure shaft in the turbine engine mode; p y,1 、P y,2 、P y,3 The output power of the bleed air power generation of the high-pressure stage, the middle stage and the outer duct in the turbine engine mode; p r,1 、P r,2 、P r,1 The output power of the turbine engine mode high-pressure stage, intermediate stage and outer duct gas power generation;
in ramjet mode, the multi-source energy optimized configuration for a hypersonic aircraft can be described as:
in formula (2), P yq 、P wc 、P wl 、P dc The method comprises the following steps of respectively carrying out air-entraining power generation, semiconductor temperature difference power generation, oil-gas turbine power generation and output power of a storage battery in a ramjet engine mode;
the method comprises the steps that the output power of a multi-source energy extraction system is reasonably distributed to carry out multi-source energy optimal configuration, namely, in an environment with a change of a flight state, an optimal energy distribution mode is obtained through optimization calculation based on a multi-target self-adaptive covariance matrix and a chaotic search cluster algorithm MOACCGA, and a Pareto optimal solution of the fuel consumption rate and the thrust is obtained under the constraint that an engine does not overtemperature, overturn and surge, so that the electric energy requirement of a hypersonic aircraft is met, and the lowest fuel consumption rate and the highest thrust of the system are realized; the population in MOACCGA is divided into 3 types: discoverers, followers and patrols, which manifest in the search behavior of the animal. The discoverer takes an optimized object as a characteristic and seeks an optimal fitness value through rapid non-dominated sorting; the follower adopts an evolutionary strategy of a self-adaptive covariance matrix to obtain a reliable estimator of the path; in addition, patrolmen explore huge search space by using chaotic search; in an MOACCGA algorithm, obtaining a Pareto optimal front edge through rapid non-dominated sorting; the evolutionary strategy of the adaptive covariance matrix improves the response speed of the algorithm so as to adapt to strong nonlinear change in a high dynamic environment; the high-ergodic chaotic search enhances the local search capability of the algorithm and overcomes the problem that the algorithm is partially optimal in a multi-peak solution space; the method specifically comprises the following steps:
s2-1, setting initialization parameters: setting a population size m =100; setting the maximum iteration number T under the same environment max =100; setting an objective function dimension d 1 =2; in turbofan engine mode, the population dimension d is set 2 =7, set variable upper bound X max ＝P w (ii) a In the turbofan engine mode, the engine is,setting a population dimension d 2 =4, set variable upper bound X max ＝P c (ii) a Setting a variable lower bound X min =0; the chaotic variable u =4 is set. In the actual working condition acquisition process, each individual in the population is an energy distribution mode of multi-source energy optimal configuration;
s2-2, initializing a population: initializing an initial generation population and initializing a covariance matrix, selecting 70% of individuals in the initial generation population as patrols to perform chaotic search in order to enable the position distribution of the initial generation population to be more uniform, performing rapid non-dominated sorting on the whole population after the chaotic search, and obtaining the mean value of fitness values of all the individuals in the population
S2-3, generating subgroups: after rapid non-dominated sorting, selecting the optimal individuals in the population as discoverers, selecting the first 50% individuals in the population as followers, and performing adaptive covariance matrix evolution in a tracking manner;
s2-4, forming a new population: combining the parent population with the sub population after the adaptive covariance evolution to form a new population, and selecting 70% of individuals in the new population as patrols to carry out chaotic search. Performing fast non-dominated sorting on the population after the chaotic search, and selecting a new finder and a follower;
s2-5, iteration: the steps S2-1 to S2-4 are repeatedly executed in an iteration mode until the iteration times in the iteration formula reach the maximum iteration times T preset in the current environment max Obtaining an optimal power generation distribution mode in the current environment;
s2-6, recording: recording the optimal multi-source electric energy distribution scheme of each environment, and forming dynamic multi-source electric energy distribution data with optimal oil consumption rate and thrust Pareto in the whole flight envelope according to the obtained optimal multi-source electric energy distribution scheme.
Further, in S2-2, the specific steps of initializing the population are as follows:
generating the location of the first generation population by equation (3):
X＝X min +r·(X min +X max ) (3)
in formula (3), r represents a random number within 0 to 1;
initializing a covariance matrix C, and setting an orthonormal base B = [ B ] of the matrix C with eigenvectors 1 ,b 2 ,...,b n ]And has a corresponding characteristic value lambda 1 2 ,λ 2 2 ,...,λ n 2 Feature root matrix D = diag (λ) 1 ,λ 2 ,...,λ n ). The covariance matrix is expressed as follows:
in order to make the position distribution of the initial generation population more uniform, 70% of individuals in the initial generation population can be selected as patrollers to carry out chaotic search, and the chaotic search process comprises the following steps:
step1 generating chaotic variables
In the nth search, D chaotic variables cx with different tracks are randomly generated n d (D =1, 2.. Multidot., D), 5 motionless points that do not contain chaotic iterative equations
Step2 Linear mapping
C x is n d Linear mapping to an optimized variable interval [ X ] min ,X max ]To obtain the position rx n d . The process of linear mapping is represented by the following equation:
rx n d ＝X min +cx n d (X max -X min ) (5)
step3 chaotic search
For target object x n d Chaotic search is carried out to obtain a new value x n+1 d . The process of chaotic search is represented by the following equation:
x n+1 d ＝x n d +β·rx n d (6)
step4: update
Updated cx n d The value, update procedure is represented by:
cx n d ＝u·cx n d (1-cx n d ) (7)
in the formula (7), u is a chaotic variable;
after chaotic search, the whole population is subjected to rapid non-dominated sorting, and the mean value of the fitness values of all individuals in the population is obtained through a formula (8)
Further, in S2-3, the specific steps of generating the subgroups are as follows:
after fast non-dominated sorting, selecting the optimal individuals in the population as discoverers, selecting the first 50% individuals in the population as followers, and performing adaptive covariance matrix evolution by following, wherein the evolution process of the adaptive covariance matrix is as follows:
step1: updating mean values
Adaptive covariance matrix evolution strategy utilizing learning rate alpha μ The update speed of the mean value is controlled. The updating process is represented as follows:
in the formula, x (t) Is the sample of the t generation.
Step2: step size control
During the evolution process, the update speed of the standard deviation σ depends largely on the step value. Adaptive covariance matrix evolution strategy construction evolution path P σ By comparing the evolution paths P σ The length of the standard deviation sigma can be adjusted correspondingly to the expected length under random selection;
the update process of the evolutionary path is represented by the following formula:
in the formula, alpha σ Is an updated learning rate;
under random distribution, the expected length of the evolutionary path is
According to the proportion
And adjusting the step size. The update procedure for the standard deviation σ is thus represented by:
in the formula (d) σ Is an updated damping factor;
step3: adaptive covariance matrix
By average step size y i Re-evaluating the covariance matrix C:
the sample capacity of the population determines the accuracy and the iteration speed of the algorithm, and the accuracy and the iteration speed of the algorithm are two contradictory indexes. To solve this problem, two independent paths are used to adaptively update the covariance matrix C:
3) Rank-mu update
In order to make the algorithm converge faster, the population information of previous generations is used for making up the defect of small population scale, and the mean value of the covariance matrix C of each generation is used for updating:
C (t+1) ≈avg(C λ (i) ；i＝1,2,...t) (13)
will learn the rate alpha cμ Introducing population information, and the updating process of C can be expressed as:
in equation (14), n is the dimension of the search space.
4) Rank-1 updates
In order to obtain a better probability of the optimal sample in the sampling process, an exponential smoothing method is used for constructing an evolution path P c ，P c The update process of (a) can be expressed as:
in the formula (15), α cp Is P c The learning rate of (c);
according to an evolution path P c Updating the covariance matrix C:
in summary, combining Rank-mu and Rank-1, the update process of the covariance matrix C is represented as:
the invention has the advantages and remarkable effects that:
(1) Aiming at multi-source energy extraction of the hypersonic aircraft, the method considers multiple evaluation targets such as the fuel consumption rate and the thrust, determines the optimization constraint condition, and reduces the fuel consumption rate and improves the thrust while meeting the power requirement of the hypersonic aircraft. The method is characterized in that the generated energy of each energy extraction mode is used as an optimization design variable, the variable is solved by adopting a multi-objective adaptive covariance matrix and a chaotic search cluster algorithm MOACCGA (mean solution computer aided algorithm), and finally, a multi-source electric energy distribution scheme with optimal oil consumption rate and thrust Pareto in the whole flight envelope is obtained, so that necessary technical support is provided for energy configuration of the hypersonic aircraft, the oil consumption rate is reduced, and the thrust is improved on the basis of ensuring that an engine does not overrun, does not overheat and does not surge.
(2) In the aspect of algorithm, according to the characteristics of the hypersonic flight vehicle, on the premise that the two characteristics of strong nonlinear change of fuel consumption and thrust of the hypersonic flight vehicle and multi-peak distribution in a solution space are considered under the condition that optimization design variables are changed, the energy distribution trend of the hypersonic flight vehicle is analyzed, the thrust and the fuel consumption rate are taken as targets, safety is taken as constraint, an improved method of adaptive covariance matrix evolution is adopted to adapt to the strong nonlinear change of a system, the problem of falling into local optimum is solved by adopting an improved method of chaotic search, and the problem of energy configuration of the hypersonic flight vehicle is solved by adopting a multi-target adaptive covariance matrix and a chaotic search cluster algorithm MOACCGA.
Drawings
FIG. 1 is a schematic diagram of a TBCC engine and multi-source energy extraction system;
FIG. 2 is a flow chart of the overall optimization of the multi-source energy configuration of the hypersonic aircraft;
FIG. 3 is a graph of thrust distribution in solution space;
FIG. 4 is a graph of fuel consumption in a solution space;
FIG. 5 is a flow chart of a multi-target adaptive covariance matrix and a chaotic search clustering algorithm MOACCGA.
Detailed Description
The technical scheme of the invention is explained in detail in the following with reference to the attached drawings.
The multi-source energy optimal configuration method of the hypersonic aircraft is realized based on a parallel turbine-based combined cycle (TBCC) engine and a multi-source energy extraction system. As shown in fig. 1, the parallel TBCC engine includes two operating modes: turbofan engine mode, ramjet mode, the multiple energy extraction mode that the multisource energy extraction system contains respectively is: shaft work extraction power generation, air-entraining power generation, gas power generation, semiconductor temperature difference power generation, oil-gas turbine power generation and storage batteries. There are three energy extraction modes involved in turbofan engine mode: (1) Shaft work extraction power generation (extraction from a high-pressure shaft and a low-pressure shaft); (2) Bleed air power generation (extracted from three parts of a high-pressure stage, a middle stage and an outer duct); (3) Gas power generation (extracted from two parts of a high-pressure shaft and a low-pressure shaft). The shaft work is extracted to generate power, namely, the transmission ratio is adjusted by a gear box from the high/low pressure of a turbofan engine to drive a motor to generate power; bleed air power generation, namely, gas is led out from a high-pressure stage/a middle stage/an outer duct of a turbofan engine to drive a motor to generate power; the gas power generation is that after the gas is led out, the gas is burnt by the combustion chamber and then drives the motor to generate power. Four energy extraction modes are involved in the ramjet mode: (1) carrying out air-entraining power generation; (2) generating power by semiconductor temperature difference; (3) generating power by an oil-gas turbine; and (4) a storage battery. Bleed air power generation, namely, a gas driving motor is led out from an isolation section of a ramjet; the semiconductor temperature difference power generation is to draw heat from the wall surface of the combustion chamber, and the formed temperature difference enables the P-type thermocouple and the N-type thermocouple to generate potential difference, so that power is supplied to a load; the oil-gas turbine generates electricity, namely, hydrocarbon fuel absorbs heat through a cooling channel of the combustion chamber and is cracked into mixed oil gas, and the mixed oil gas is introduced into the oil-gas turbine to generate electricity; the storage battery does not need to extract energy from the engine, and only the weight causes fuel compensation.
As shown in FIG. 2, the overall optimization process of the multi-source energy configuration of the hypersonic aircraft comprises four steps S1-S4.
S1: determining working mode and multi-source energy extraction mode in current flight state
And confirming the current working mode of the engine according to the current flight altitude H and the flight Mach number Ma of the airplane and selecting a corresponding multi-source energy extraction mode.
S2: multiple optimization targets for determining multi-source energy optimization configuration of hypersonic aircraft
The multi-source energy extraction can affect the fuel consumption rate Wf and the thrust FN of the engine, and the fuel consumption rate and the thrust can directly reflect the economy and the performance of the engine as main indexes. Therefore, the invention takes the primary fuel consumption rate and the thrust as optimization targets, so that the system has good fuel economy and sufficient thrust, the cruising ability of the supersonic aircraft can be improved, and the supersonic aircraft has good flight quality.
S3: constraint condition and working requirement for determining multi-source energy optimal configuration of hypersonic aircraft
The constraints of the turbofan engine mode are: turbine front temperature T, high pressure shaft speed N LPS Low pressure shaft speed N HPS Stability margin SM of high pressure compressor HPC And stability margin SM of low pressure compressor LPS (ii) a The operating requirements for the turbofan engine mode are: electric energy demand P w 。
The constraints for the ramjet mode are: combustion stability SM of a combustion chamber COV (ii) a The operating requirements for the ramjet mode are: electric energy demand P c 。
S4: optimizing energy distribution schemes for multi-source energy extraction
In a flight envelope, aiming at the dynamic environment of flight altitude, flight Mach number and working mode change, the output power of the multi-source energy extraction system is distributed by using a MOACCGA based on a multi-target self-adaptive covariance matrix and a chaotic search cluster algorithm.
Specifically, both the fuel penalty loss and the thrust loss from the engine are caused by the extraction of energy from the engine. In the turbofan engine mode, the multi-source energy optimization configuration for the hypersonic aircraft with optimization targets of fuel consumption Wf and thrust FN can be described as:
in the formula (1), P z,1 、P z,2 Extracting the output power of the power generation for the shaft work of the high-pressure shaft and the low-pressure shaft in the turbine engine mode; p is y,1 、P y,2 、P y,3 The output power of the bleed air power generation of the high-pressure stage, the middle stage and the outer duct in the turbine engine mode; p r,1 、P r,2 、P r,1 The power is the output power of the turbine engine mode high-pressure stage, intermediate stage and outer duct gas power generation.
In the ramjet mode, the multi-source energy optimization configuration for a hypersonic aircraft can be described as:
in the formula (2), P yq 、P wc 、P wl 、P dc The power generation system is respectively the output power of a ramjet engine mode bleed air power generation, a semiconductor temperature difference power generation, an oil-gas turbine power generation and a storage battery.
Output power of the multi-source energy extraction system is reasonably distributed to carry out multi-source energy optimal configuration, namely, in an environment with a change of a flight state, an optimal energy distribution mode is obtained through optimization calculation based on a multi-target self-adaptive covariance matrix and a chaotic search cluster algorithm MOACCGA, and a Pareto optimal solution of the fuel consumption rate and the thrust is obtained under the constraint that an engine does not overtemperature, overtravel and surge, so that the electric energy requirement of a hypersonic aircraft is met, and the lowest fuel consumption rate and the highest thrust of the system are realized. The populations in MOACCGA can be divided into 3 types: discoverers, followers and patrols, which manifest as animal search behavior. The discoverer is characterized by an optimized object and seeks an optimal fitness value through a fast non-dominated sorting. And the follower adopts an evolutionary strategy of the self-adaptive covariance matrix to obtain the reliable estimator of the path. In addition, patrolmen explore a huge search space using chaotic search. In an MOACCGA algorithm, obtaining a Pareto optimal front edge through rapid non-dominated sorting; the evolution strategy of the adaptive covariance matrix can improve the response speed of the algorithm so as to adapt to strong nonlinear change in a high dynamic environment; the high-ergodic chaotic search enhances the local search capability of the algorithm and overcomes the problem that the algorithm is trapped into local optimization in a multi-peak solution space.
As shown in fig. 3, the thrust varies strongly non-linearly in the solution space and has a plurality of maximum points; as shown in fig. 4, the fuel consumption varies strongly non-linearly in the solution space and has a plurality of minimum points. Therefore, the following characteristics can be obtained along with the change of the thrust and the oil consumption along with the change of the generated energy:
1) The changes in thrust and fuel consumption are highly non-linear and unpredictable. Multiple source energy extraction can cause sudden changes in compressor efficiency, changing the thermal cycle inside the engine, further affecting the performance of the engine. When the power generation capacity reaches a certain value, and may suddenly rise or fall.
2) The generated energy of multi-source pumping energy is changed, so that the oil consumption rate and the thrust are suddenly changed, multi-peak distribution is presented on a solution space, and a plurality of extreme values are presented. The range of the dashed line is the maximum and minimum values in the solution space.
As shown in fig. 5, the MOACCGA optimization process based on the multi-objective adaptive covariance matrix and the chaotic search cluster algorithm includes the following steps:
s4-1: setting initialization parameters
Setting a population size m =100; setting the maximum iteration number T under the same environment max =100; setting an objective function dimension d 1 =2; in turbofan engine mode, the population dimension d is set 2 =7, set variable upper bound X max ＝P w (ii) a In turbofan engine mode, the population dimension d is set 2 =4, set variable upper bound X max ＝P c (ii) a Setting a variable lower bound X min =0; the chaotic variable u =4 is set. In the actual working condition acquisition process, each individual in the population is an energy distribution mode of multi-source energy optimal configuration.
S4-2: initializing a population
Generating the location of the first generation population by equation (3):
X＝X min +r·(X min +X max ) (3)
in formula (3), r represents a random number within the range of 0 to 1.
Initializing a covariance matrix C, and setting an orthonormal base B = [ B ] of the matrix C with eigenvectors 1 ,b 2 ,...,b n ]And has a corresponding characteristic value lambda 1 2 ,λ 2 2 ,...,λ n 2 Characteristic root matrix D = diag (λ) 1 ,λ 2 ,...,λ n ). The covariance matrix is expressed as follows:
in order to make the position distribution of the initial generation population more uniform, 70% of individuals in the initial generation population are selected as patrollers to carry out chaotic search, and the chaotic search process comprises the following steps:
step1 generating chaotic variables
In the nth search, D chaotic variables cx with different tracks are randomly generated n d (D =1, 2.. Multidot., D), 5 motionless points that do not contain chaotic iterative equations
Step2 Linear mapping
Will cx n d Linear mapping to an optimized variable Interval [ X ] min ,X max ]To obtain the position rx n d . The process of linear mapping is represented by the following equation:
rx n d ＝X min +cx n d (X max -X min ) (5)
step3 chaotic search
For the target object x n d Chaotic search is carried out to obtain a new value x n+1 d . The process of chaotic search is represented by the following equation:
x n+1 d ＝x n d +β·rx n d (6)
step4: update
Updated cx n d The value, update procedure is represented by:
cx n d ＝u·cx n d (1-cx n d ) (7)
in the formula (7), u is a chaotic variable.
After chaotic search, the whole population is subjected to rapid non-dominated sorting, and the mean value of the fitness values of all individuals in the population is obtained through a formula (8)
S4-3: generating subgroups
After fast non-dominated sorting, selecting the optimal individuals in the population as discoverers, selecting the first 50% individuals in the population as followers, and performing adaptive covariance matrix evolution by following, wherein the evolution process of the adaptive covariance matrix is as follows:
step1: updating mean values
Adaptive covariance matrix evolution strategy utilizing learning rate alpha μ The update speed of the mean value is controlled. The updating process is represented as follows:
in the formula, x (t) Is the sample of the t generation.
Step2: and controlling the step size.
The update speed of the standard deviation σ during evolution depends strongly on the step value. Self-adaptive covariance matrix evolution strategy construction evolution path P σ By comparing the evolution paths P σ Can be adjusted to the standard deviation sigma correspondingly to the expected length under random selection.
The updating process of the evolutionary path is represented by the following formula:
in the formula, alpha σ Is an updated learning rate.
Under random distribution, the expected length of the evolutionary path is
According to the proportion
And adjusting the step size. The update procedure for the standard deviation σ is thus represented by:
in the formula (d) σ Is an updated damping factor.
Step3: adaptive covariance matrix
By average step size y i Re-evaluating the covariance matrix C:
when the samples are large enough, the evaluation of the covariance matrix is reliable. However, in each generation population, a fast iteration is performed with a smaller sample population. To solve this problem, the covariance matrix C is adaptively updated using two independent paths.
5) Rank-mu update
In order to make the algorithm converge more quickly, the population information of previous generations is utilized to make up the defect of small population scale, and the mean value of the covariance matrix C of each generation is used for updating
C (t+1) ≈avg(C λ (i) ；i＝1,2,...t) (13)
Will learn the rate alpha cμ Introducing population information, and the updating process of C can be expressed as:
in equation (14), n is the dimension of the search space.
6) Rank-1 updates
In order to obtain a better probability of the optimal sample in the sampling process, an exponential smoothing method is used for constructing an evolution path P c ，P c Can be expressed as
In the formula (15), α cp Is P c The learning rate of (2).
According to an evolution path P c Updating the covariance matrix C:
in summary, combining Rank-mu and Rank-1, the update process of the covariance matrix C can be expressed as:
s4-4: forming a new population
Combining the parent population with the sub population after the adaptive covariance evolution to form a new population, and selecting 70% of individuals in the new population as patrols to carry out chaotic search. And (4) performing fast non-dominated sorting on the chaotically searched population, and selecting new discoverers and followers.
S4-5: iteration
And repeating the steps S4-3 to S4-4 until the iteration number in the iteration formula reaches the maximum iteration number T preset by the current environment max And obtaining the optimal power generation distribution mode under the current environment.
S4-6: recording
Recording the optimal multi-source electric energy distribution scheme of each environment, and forming dynamic multi-source electric energy distribution data with optimal oil consumption rate and thrust Pareto in the whole flight envelope according to the obtained optimal multi-source electric energy distribution scheme.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
The above examples are only intended to illustrate the technical solution of the present invention, and not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications and substitutions are intended to be included within the scope of the present invention.
Claims (4)
1. A multisource energy optimal configuration method of a hypersonic aircraft is based on a parallel turbine-based combined cycle TBCC engine and a multisource energy extraction system, and comprises two working modes of a turbofan engine and a ramjet engine;
the method is characterized in that: the method comprises the steps of taking the generated energy of each energy extraction mode as an optimization design variable, taking the oil consumption rate and the thrust as optimization objective functions, taking the rotating speed of a mechanical shaft, the temperature in front of a turbine, the stability margin and the like as constraint conditions, adopting a MOACCGA based on a multi-target self-adaptive covariance matrix and a chaotic search cluster algorithm, adapting to the strong nonlinear change of a system by an improved method of the self-adaptive covariance matrix, avoiding the local optimization by the improved method of chaotic search, realizing the accurate distribution of multi-source electric energy of the hypersonic aerocraft in a dynamic environment, obtaining the energy user configuration with the lowest oil consumption rate and the highest thrust, and finally obtaining the optimal multi-source electric energy distribution of the oil consumption rate and the Pareto thrust in the whole flight envelope;
the method comprises the following steps:
s1, determining a working mode in a current flight state and selecting a corresponding multi-source energy extraction mode
Confirming the working mode of a current engine according to the current flight altitude H and the flight Mach number Ma of the airplane and selecting a corresponding multi-source energy extraction mode; turbofan engine modes include three energy extraction modes: (1) Shaft work is extracted for power generation, and the shaft work is extracted from a high-pressure shaft and a low-pressure shaft; (2) Air is introduced to generate electricity, and the electricity is extracted from a high-pressure stage, a middle stage and an outer duct; (3) Gas power generation, which is extracted from a high-pressure shaft and a low-pressure shaft; the shaft work is extracted to generate power, namely, the transmission ratio is adjusted by a gear box from the high/low pressure of a turbofan engine to drive a motor to generate power; bleed air power generation, namely, gas is led out from a high-pressure stage/a middle stage/an outer duct of a turbofan engine to drive a motor to generate power; gas power generation, namely, after gas is led out, the gas is combusted in a combustion chamber and then drives a motor to generate power; the ramjet mode includes four energy extraction modes: carrying out air-entraining power generation; (2) generating electricity by semiconductor temperature difference; (3) generating power by an oil-gas turbine; (4) a storage battery; bleed air power generation, namely leading out a gas driving motor from an isolation section of a ramjet; the semiconductor temperature difference power generation is to draw heat from the wall surface of the combustion chamber, and the formed temperature difference enables the P-type thermocouple and the N-type thermocouple to generate potential difference, so that power is supplied to a load; the oil-gas turbine generates electricity, namely, hydrocarbon fuel absorbs heat through a cooling channel of the combustion chamber and is cracked into mixed oil gas, and the mixed oil gas is introduced into the oil-gas turbine to generate electricity; the storage battery does not need to extract energy from the engine, and only the weight causes fuel compensation;
s2: multiple optimization targets for determining multi-source energy optimization configuration of hypersonic aircraft
The fuel consumption rate Wf and the thrust FN of the engine are used as optimization targets, so that the fuel economy and the thrust of the system are considered, the cruising ability of the supersonic aircraft is improved, and the flight quality is improved;
s3: constraint condition and working requirement for determining multi-source energy optimal configuration of hypersonic aircraft
The constraints of the turbofan engine mode are: turbine front temperature T, high pressure shaft speed N LPS Low pressure shaft speed N HPS Stability margin SM of high pressure compressor HPC And stability margin SM of low pressure compressor LPS (ii) a The operating requirements for the turbofan engine mode are: electric energy demand P w ；
The constraints for the ramjet mode are: combustion stability SM of a combustion chamber COV (ii) a The operating requirements for the ramjet mode are: electric energy demand P c ；
S4: optimizing an energy distribution scheme for multi-source energy extraction
Aiming at the flight altitude H, the flight Mach number Ma and the electric energy demand P which dynamically changes along with the working mode of an engine in a flight envelope w And P c The multi-source energy extraction system is subjected to MOACCGA based on a multi-target self-adaptive covariance matrix and a chaotic search cluster algorithmThe output power is distributed.
2. The multi-source energy optimal configuration method for hypersonic aircraft according to claim 1, characterized in that: since both the fuel penalty loss and the thrust loss of the engine are caused by extracting energy from the engine, in the turbofan engine mode, the multi-source energy optimization configuration for the hypersonic aircraft with the fuel consumption Wf and the thrust FN as optimization targets is described as follows:
in the formula (1), P z,1 、P z,2 Extracting the output power of power generation for the shaft work of the high-pressure shaft and the low-pressure shaft in the turbine engine mode; p y,1 、P y,2 、P y,3 The output power of the turbine engine mode high-pressure stage, the middle stage and the outer duct is used for air entraining and power generation; p r,1 、P r,2 、P r,1 The output power of the turbine engine mode high-pressure stage, intermediate stage and outer duct gas power generation;
in the ramjet mode, the multi-source energy optimization configuration of a hypersonic aircraft is described as:
in formula (2), P yq 、P wc 、P wl 、P dc The method comprises the following steps of respectively carrying out air-entraining power generation, semiconductor temperature difference power generation, oil-gas turbine power generation and output power of a storage battery in a ramjet engine mode;
output power of the multi-source energy extraction system is reasonably distributed to carry out multi-source energy optimal configuration, namely, in an environment with changing flight states, an optimal energy distribution mode is obtained through optimization calculation based on a multi-target self-adaptive covariance matrix and a chaotic search cluster algorithm MOACCGA, and a Pareto optimal solution of oil consumption and thrust is obtained under the constraint that an engine does not exceed the temperature, does not exceed the rotation and does not surge, so that the electric energy requirement of a hypersonic aircraft is met, and the lowest oil consumption and the largest thrust of the system are realized; the populations in MOACCGA are divided into 3 types: discoverers, followers and patrollers, which are represented by the search behavior of animals, and seek an optimal fitness value by fast non-dominated sorting by taking an optimized object as a characteristic; the follower adopts an evolutionary strategy of a self-adaptive covariance matrix to obtain a reliable estimator of the path; in addition, patrolmen explore huge search space by using chaotic search; in an MOACCGA algorithm, obtaining a Pareto optimal front edge through rapid non-dominated sorting; the evolutionary strategy of the adaptive covariance matrix improves the response speed of the algorithm so as to adapt to strong nonlinear change in a high dynamic environment; the high-ergodic chaotic search enhances the local search capability of the algorithm and overcomes the problem that the algorithm is partially optimal in a multi-peak solution space; the method specifically comprises the following steps:
s2-1, setting initialization parameters: setting a population size m =100; setting the maximum iteration number T under the same environment max =100; setting the dimension d of the objective function 1 =2; in turbofan engine mode, the population dimension d is set 2 =7, set variable upper bound X max ＝P w (ii) a In turbofan engine mode, the population dimension d is set 2 =4, set variable upper bound X max ＝P c (ii) a Setting a variable lower bound X min =0; setting a chaotic variable u =4, wherein each individual in the population is an energy distribution mode of multi-source energy optimal configuration in the actual working condition acquisition process;
s2-2, initializing a population: initializing an initial generation population and initializing a covariance matrix, selecting 70% of individuals in the initial generation population as patrols for chaotic search in order to enable the position distribution of the initial generation population to be more uniform, and after the chaotic search, performing rapid non-dominated sorting on the whole population to obtain the mean value of the fitness values of all the individuals in the population
S2-3, generating subgroups: after rapid non-dominated sorting, selecting the optimal individuals in the population as discoverers, selecting the first 50% individuals in the population as followers, and performing adaptive covariance matrix evolution in a tracking manner;
s2-4, forming a new population: combining the parent population with the sub-population after the adaptive covariance evolution to form a new population, selecting 70% of individuals from the new population as patrols to perform chaotic search, performing rapid non-dominated sorting on the population after the chaotic search, and selecting new discoverers and followers;
s2-5, iteration: the steps S2-1 to S2-4 are repeatedly executed in an iteration mode until the iteration times in the iteration formula reach the maximum iteration times T preset in the current environment max Obtaining an optimal power generation distribution mode in the current environment;
s2-6, recording: recording the optimal multi-source electric energy distribution scheme of each environment, and forming dynamic multi-source electric energy distribution data with optimal oil consumption rate and thrust Pareto in the whole flight envelope according to the obtained optimal multi-source electric energy distribution scheme.
3. The multi-source energy optimal configuration method for hypersonic aircraft according to claim 2, characterized in that: in the S2-2, the specific steps for initializing the population are as follows:
generating the location of the first generation population by equation (3):
X＝X min +r·(X min +X max ) (3)
in formula (3), r represents a random number within 0 to 1;
initializing a covariance matrix C, and setting an orthonormal base B = [ B ] of the matrix C with eigenvectors 1 ,b 2 ,...,b n ]And has a corresponding characteristic value lambda 1 2 ,λ 2 2 ,...,λ n 2 Feature root matrix D = diag (λ) 1 ,λ 2 ,...,λ n ) The covariance matrix is expressed as follows:
in order to make the position distribution of the initial generation population more uniform, 70% of individuals in the initial generation population are selected as patrols to carry out chaotic search, and the chaotic search process is as follows:
step1 generating chaotic variables
In the nth search, D chaotic variables cx with different tracks are randomly generated n d (D =1, 2.. Multidot., D), 5 motionless points that do not contain chaotic iterative equations
Step2 Linear mapping
C x is n d Linear mapping to an optimized variable interval [ X ] min ,X max ]To obtain the position rx n d The process of linear mapping is represented by the following equation:
rx n d ＝X min +cx n d (X max -X min ) (5)
step3 chaotic search
For target object x n d Chaotic search is carried out to obtain a new value x n+1 d The chaotic search process is represented by the following formula:
x n+1 d ＝x n d +β·rx n d (6)
step4: update
Updated cx n d The value, update procedure is represented by:
cx n d ＝u·cx n d (1-cx n d ) (7)
in the formula (7), u is a chaotic variable;
after chaotic search, the whole population is subjected to rapid non-dominated sorting, and the mean value of the fitness values of all individuals in the population is obtained through a formula (8)
4. The multi-source energy optimal configuration method for the hypersonic aircraft according to claim 2, characterized in that: in S2-3, the specific steps of generating the subgroups are as follows:
after the rapid non-dominated sorting, selecting the optimal individuals in the population as discoverers, selecting the first 50% individuals in the population as followers, and performing adaptive covariance matrix evolution by following, wherein the evolution process of the adaptive covariance matrix is as follows:
step1: updating mean values
Adaptive covariance matrix evolution strategy utilizing learning rate alpha μ The updating speed of the mean value is controlled, and the updating process is represented as the following formula:
in the formula, x (t) Is the sample of the t generation;
step2: step size control
In the evolution process, the updating speed of the standard deviation sigma greatly depends on the step value, and the evolution path P is constructed by the adaptive covariance matrix evolution strategy σ By comparing the evolution paths P σ The length of the standard deviation sigma can be adjusted correspondingly to the expected length under random selection;
the update process of the evolutionary path is represented by the following formula:
in the formula, alpha σ Is an updated learning rate;
under random distribution, the expected length of the evolutionary path is
According to the proportion
The step size is adjusted so that the update procedure of the standard deviation σ is represented by:
in the formula (d) σ Is an updated damping factor;
step3: adaptive covariance matrix
By average step size y i Re-evaluating the covariance matrix C:
the sample capacity of the population determines the accuracy and the iteration speed of the algorithm, the accuracy and the iteration speed of the algorithm are two contradictory indexes, and in order to solve the problem, two independent paths are adopted to update the covariance matrix C in a self-adaptive manner;
1) Rank-mu update
In order to make the algorithm converge faster, the population information of previous generations is used for making up the defect of small population scale, and the mean value of the covariance matrix C of each generation is used for updating:
C (t+1) ≈avg(C λ (i) ；i＝1,2,...t) (13)
will learn the rate alpha cμ Introducing population information, and the updating process of C can be expressed as:
in equation (14), n is the dimension of the search space;
2) Rank-1 updates
In order to obtain a better probability of the optimal sample in the sampling process, an exponential smoothing method is used for constructing an evolution pathP c ，P c The update process of (a) can be expressed as:
in the formula (15), α cp Is P c The learning rate of (c);
according to an evolution path P c Updating the covariance matrix C:
C (t+1) ＝(1-α c1 )C (t) +α c1 P c (t+1) P c (t+1) (16)
in summary, combining Rank-mu and Rank-1, the update process of the covariance matrix C is represented as:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211390198.XA CN115758870A (en) | 2022-11-08 | 2022-11-08 | Multi-source energy optimal configuration method of hypersonic aircraft |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211390198.XA CN115758870A (en) | 2022-11-08 | 2022-11-08 | Multi-source energy optimal configuration method of hypersonic aircraft |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115758870A true CN115758870A (en) | 2023-03-07 |
Family
ID=85357437
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211390198.XA Pending CN115758870A (en) | 2022-11-08 | 2022-11-08 | Multi-source energy optimal configuration method of hypersonic aircraft |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115758870A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116859756A (en) * | 2023-09-05 | 2023-10-10 | 中国航空工业集团公司金城南京机电液压工程研究中心 | Aviation comprehensive electromechanical system optimization model construction method and device |
-
2022
- 2022-11-08 CN CN202211390198.XA patent/CN115758870A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116859756A (en) * | 2023-09-05 | 2023-10-10 | 中国航空工业集团公司金城南京机电液压工程研究中心 | Aviation comprehensive electromechanical system optimization model construction method and device |
| CN116859756B (en) * | 2023-09-05 | 2023-11-21 | 中国航空工业集团公司金城南京机电液压工程研究中心 | Aviation comprehensive electromechanical system optimization model construction method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106970523B (en) | Energy management strategy of airplane self-adaptive power and heat management system | |
| CN107885891B (en) | Method for optimizing overall performance of aircraft engine | |
| CN115758870A (en) | Multi-source energy optimal configuration method of hypersonic aircraft | |
| CN108416421B (en) | Dynamic fire power distribution method based on DDE improved bat algorithm | |
| CN111679574A (en) | Variable-cycle engine transition state optimization method based on large-scale global optimization technology | |
| CN114675535A (en) | Aero-engine transition state optimization control method based on reinforcement learning | |
| CN112182789B (en) | Multi-objective optimization design method for self-adaptive variable cycle engine | |
| CN113032899B (en) | Turbofan engine design parameter optimization method | |
| CN113255138A (en) | Load distribution optimization method for power system | |
| Magrini et al. | Ultra-high bypass nacelle geometry design space exploration | |
| CN113901573A (en) | Game strategy-based airplane comprehensive integrated heat/energy multi-target optimization method | |
| Lee et al. | Performance design of a turbofan engine using multi-objective particle swarm optimization (MOPSO) | |
| CN117738799A (en) | Aeroengine dynamic performance maintaining control method and device | |
| Hao et al. | A new design method for mode transition control law of variable cycle engine | |
| CN115859831A (en) | Multi-objective optimization method for environment control mode of thermal management type combined power device | |
| Yan et al. | A Newly Improved Collaborative Optimization Strategy: Application to Conceptual Multidisciplinary Design Optimization of a Civil Aero-Engine | |
| CN113671829B (en) | Instantaneous energy efficiency optimization method for aircraft comprehensive integrated heat/energy system | |
| CN112949159A (en) | IPSO-based optimal control method for acceleration process of variable cycle engine | |
| Zhang et al. | Optimization of adaptive cycle engine performance based on improved particle swarm optimization | |
| CN113569319A (en) | PSO _ DE intelligent algorithm-based design method for adaptive corrector of aircraft engine model | |
| CN107566180B (en) | Turbine base combined cycle engine dcs communication network topology structural optimization method | |
| Atashkari | Aero-Thermodynamic optimization of turboprop engines using Multi-Objective genetic algorithms | |
| CN115470575A (en) | Control parameter optimization method for variable-cycle aircraft engine acceleration process | |
| CN113536592B (en) | Method for optimizing thrust performance of ramjet based on chaotic particle swarm algorithm | |
| CN110058635A (en) | The MPPT method combined based on improvement particle swarm algorithm with fuzzy algorithmic approach |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |