A method for quickly identifying same-numbered individuals in a grid

Abstract

The application discloses a method for quickly identifying same-numbered individuals in a grid, and belongs to the technical field of aero-engine optimization; the method steps comprise the following: in a model for multi-target optimization of the overall performance of an aero-engine, each target value of each individual in population design parameters is normalized; the grid number of each individual in each target dimension is calculated; multiple grid numbers of each individual in all targets are converted into a comprehensive number; all individuals in the population are sorted according to the comprehensive number of the individuals; and the individuals in the same grid are identified according to the same comprehensive number by traversing the sorted population. The application solves the problems of long time and large complexity in identifying same-numbered individuals in a grid during the optimization of the overall performance of an aero-engine.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of aero-engine optimization, and particularly relates to a method for quickly identifying same-numbered individuals in a grid. BACKGROUND

[0002] The design parameters of each component of an aero-engine have a very great influence on the overall thrust, fuel consumption and stability of the engine, and many parameters need to be optimized. The traditional optimization method mainly uses engine performance simulation analysis software to analyze the performance of the engine, and based on existing experience and knowledge, the optimization is realized by repeatedly modifying the design parameters and performing numerical simulation. The manual iterative method is low in efficiency, and cannot fully consider the coupling relationship between disciplines, so it is difficult to seek the optimal solution of a multi-objective problem. The multi-objective optimization algorithm is used to select appropriate design parameters to make multiple objectives optimal at the same time under given conditions, so as to perform multi-objective optimization on the comprehensive performance of the engine under different working conditions, and thus obtain the optimal engine design scheme.

[0003] However, for the multi-objective optimization algorithm, the uniformity of the population needs to be maintained through a certain strategy multiple times. The grid method is an effective method for maintaining the uniformity of the population, which needs to traverse and identify the same-numbered individuals in the grid through a double-loop, so that the individuals with smaller aggregation density in each grid can be retained into the new population. For the optimization problem of the overall performance of a complex engine, a larger population size is usually needed to ensure that the optimization algorithm can obtain the global optimal solution. However, with the increase in the number of individuals in the population, the calculation amount of the double-loop for identifying the same-numbered individuals in the grid method increases sharply. The conventional grid method identifies the same-numbered individuals in the grid through a double-loop, and for the case where the number of individuals in the population is N, the time complexity is O(N 2 ). This makes the multi-objective optimization algorithm based on the grid method low in efficiency when solving complex optimization problems, thereby limiting the applicability of the grid method.

[0004] Therefore, in order to improve the applicability of the grid method to complex optimization problems, the application proposes a method for quickly identifying same-numbered individuals in a grid applied to the overall performance optimization of an aero-engine, which improves the process of identifying the same-numbered individuals in the grid in the grid method, thereby effectively improving the calculation efficiency of the grid method. SUMMARY

[0005] The technical problem to be solved is:

[0006] In order to avoid the defects of the prior art, the application provides a method for quickly identifying same-numbered individuals in a grid, when identifying the same-numbered individuals in the grid, the numbers of the individuals in each dimension of the grid are arranged in different positions of the same comprehensive number, the numbers of different bits of the comprehensive number are used to represent the numbers of the individuals in different dimensions, the process of identifying the same-numbered individuals is converted into a sorting problem of the comprehensive number, and the comprehensive number is sorted by using a quick sorting method, so that the individuals with the same comprehensive number can be quickly gathered in the same comprehensive number area. The application solves the problems of long time and large complexity in identifying the same-numbered individuals in the grid during the overall performance optimization of the aero-engine.

[0007] The technical scheme of the application is a method for quickly identifying same-numbered individuals in a grid, and the specific steps are as follows:

[0008] In the model of the multi-objective optimization of the overall performance of the aero-engine, each target value of each individual in the population design parameter is normalized;

[0009] The grid numbers of each individual in each target dimension are calculated;

[0010] The multiple grid numbers of each individual in all targets are converted into a comprehensive number;

[0011] According to the comprehensive number of the individual, all the individuals in the population are sorted;

[0012] The population after sorting is traversed, and the individuals in the same grid are identified according to the same comprehensive number.

[0013] The further technical scheme of the application is that the population design parameter includes N individuals, each individual is a vector including D optimization variables, and the expression is:

[0014] X m =(x m,1 ,x m,2 ,…,x m,n ,…,x m,D ) T

[0015] Wherein, m = 1, 2, … N, x m,n represents the nth optimization variable of the mth individual.

[0016] The further technical scheme of the application is that the individual X m The target value of the target i is f i (X m ), and the target value after normalization is The expression is:

[0017]

[0018] wherein f i,max and f i,min are the maximum and minimum values of the target i in the population, respectively, and i represents the serial number of the target.

[0019] A further technical solution of the present application is that the method for calculating the grid number of each individual on each target dimension is that the grid number on the target i is denoted as G i , f i * (X m )×G i is rounded to obtain the grid number O m of the individual X i on the i-th target dimension. m

[0020] A further technical solution of the present application is that the comprehensive number is O C (X m ), and the expression is as follows:

[0021]

[0022] wherein O1(X m ) represents the grid number of the individual X m on the first target dimension; O i (X m ) represents the grid number of the individual X m on the i-th target dimension; W j represents the bit number of the grid number G j of the target j; and p represents the total number of targets.

[0023] A further technical solution of the present application is that the method for sorting all individuals in the population is the quick sort method.

[0024] A system for quickly identifying individuals with the same number in a grid comprises a normalization processing module, a grid number module, a comprehensive number module, a sorting module and an identification module; the system realizes the quick identification of individuals with the same number in a multi-target model grid of engine overall performance optimization through the cooperation of the modules.

[0025] The normalization processing module is used for performing normalization processing on each target value of each individual in the population design parameters.

[0026] The grid number module is used for calculating the grid number of each individual on each target dimension.

[0027] The comprehensive number module is used for converting the multiple grid numbers of each individual on all targets into a comprehensive number.

[0028] ​The sorting module is used for sorting all individuals in the population.

[0029] The identification module identifies the individuals in the same grid according to the same comprehensive number.

[0030] A multi-objective optimization method of an aero-engine overall performance adopts the method for quickly identifying the individuals with the same number in a grid.

[0031] An electronic device comprises at least one processor and a memory in communication connection with the at least one processor, wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the method for quickly identifying the individuals with the same number in a grid.

[0032] A computer readable storage medium stores computer instructions for enabling a processor to implement the method for quickly identifying the individuals with the same number in a grid when the processor executes the computer instructions.

[0033] Advantages

[0034] The present application has the advantages that in the process of identifying the individuals with the same number in a grid in the multi-objective optimization model of an aero-engine overall performance, the individuals with the same number in each dimension of the grid are no longer directly found by a double-loop traversal, but the numbers of the individuals in each dimension of the grid are placed in different positions of a comprehensive number, the numbers in different positions of the comprehensive number represent the numbers of the individuals in different dimensions, and the individuals in the same grid have the same comprehensive number. The comprehensive numbers of all the individuals are sorted by using a quick sorting method, so that the individuals with the same comprehensive number can be quickly gathered in the same comprehensive number area. The method for quickly identifying the individuals with the same number in a grid can convert the process of identifying the individuals with the same number into a sorting problem, so that a high-efficiency quick sorting method can be introduced to overcome the problem of slow speed in identifying the individuals with the same number in a grid.

[0035] The method for quickly identifying the individuals with the same number in a grid can convert the double-loop process of identifying the individuals with the same number in a grid into a sorting problem based on a comprehensive number, and can introduce a high-efficiency sorting method such as a quick sorting method to improve the sorting efficiency, thereby overcoming the problem of large calculation amount of the traditional grid method and solving the problem of poor applicability of the traditional grid method to the complex multi-objective optimization problem of an aero-engine. Based on the standard multi-objective optimization algorithm NSGA-||, when optimizing the performance of an aero-engine in three typical working conditions, compared with the conventional grid method, the improved grid method of the present application can reduce the optimization time by 14.44% to 17.77%. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 is a flow chart of a method for quickly identifying same-numbered individuals in a grid according to an embodiment of the present application;

[0037] Figure 2 is a schematic diagram of individual distribution in a 2-dimensional grid method.

[0038] Legend: f1: target function 1; f2: target function 2; A: code of an area in the grid. DETAILED DESCRIPTION

[0039] The embodiments described below with reference to the drawings are exemplary and are intended to explain the present application, and cannot be understood as limiting the present application.

[0040] Based on the conventional grid method, same-numbered individuals in a grid are identified through a double-loop cycle, and for a population of N individuals, the time complexity is O(N 2 ). This makes the multi-objective optimization algorithm based on the grid method inefficient in solving complex optimization problems, thereby limiting the scope of application of the grid method, etc. The present application provides a method for quickly identifying same-numbered individuals in a grid, and the specific steps are as follows:

[0041] Step 1: In the model of the overall performance multi-objective optimization of an aero-engine, each target value of each individual in the population design parameters is normalized.

[0042] Step 2: Calculate the grid number of each individual in each target dimension.

[0043] Step 3: Convert the multiple grid numbers of each individual in all targets into a comprehensive number.

[0044] Step 4: Sort all individuals in the population according to the comprehensive number of the individuals.

[0045] Step 5: Traverse the sorted population, and identify the individuals in the same grid according to the same comprehensive number.

[0046] The application is a system for quickly identifying same-numbered individuals in a grid, comprising a normalization processing module, a grid numbering module, a comprehensive numbering module, a sorting module and an identification module; the system realizes the quick identification of same-numbered individuals in a grid of an aero-engine overall performance multi-objective optimization model through the cooperation of the modules; the normalization processing module is used for normalizing each target value of each individual in a population design parameter; the grid numbering module is used for calculating the grid number of each individual in each target dimension; the comprehensive numbering module is used for converting the multiple grid numbers of each individual in all targets into one comprehensive number; the sorting module is used for sorting all individuals in the population; and the identification module identifies the individuals in the same grid according to the same comprehensive number.

[0047] The application is a multi-objective optimization method for the overall performance of an aero-engine, which adopts the method for quickly identifying same-numbered individuals in a grid.

[0048] The application is an electronic device, comprising at least one processor and a memory in communication connection with the at least one processor; the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the method for quickly identifying same-numbered individuals in a grid.

[0049] The application is a computer-readable storage medium, which stores computer instructions for enabling a processor to execute the method for quickly identifying same-numbered individuals in a grid when the processor executes the computer instructions.

[0050] The above technical solutions are further described below in combination with the drawings and specific experiments:

[0051] Referring to Figure 1 , the specific steps of the method for quickly identifying same-numbered individuals in a grid according to the embodiment are as follows:

[0052] The embodiment is described by taking the 2-target optimization of the thrust and specific fuel consumption of an aero-engine as an example, and the optimization target is the maximum thrust and the lowest specific fuel consumption. Figure 2 For the individual distribution in the 2-dimensional grid in formula (1), the target function f1 and the target function f2 correspond to the target functions of the thrust and the specific fuel consumption respectively.

[0053] Step 1: normalizing each target value of each individual in the population.

[0054] For a population comprising N individuals, each individual can be expressed as a vector X m =(x m,1 ,x m,2 ,…,x m,n,..., x m,D ) T where m = 1, 2,... N, x m,n represents the nth optimization variable of the mth individual. Let the individual X m the objective value on the target i be f i (X m ), then the normalized objective value f i * (X m ) is:

[0055]

[0056] where f i,max and f i,min are the maximum and minimum values of the target i in the population respectively. After normalization, the maximum and minimum values of the objective function 1 and the objective function 2 are 1 and 0 respectively.

[0057] Step two, calculate the grid number of each individual in each target dimension.

[0058] Let the number of grids on the target i be G i , then f i * (X m ) × G i is rounded to obtain the grid number O i (X m ) of the individual X A in the i-th target dimension. Take the individual X A in the region A in Figure 2 for example, the belonging interval of f1(X A ) is [0, 0.2), the belonging interval of f2(X A ) is [0.6, 0.8), and G1 and G2 are both 5, then the belonging interval of f1(X A ) × G1 is [0, 1), and the belonging interval of f2(X A ) × G2 is [3, 4), the rounding of f1(X A ) × G1 can obtain O1(X A ) as 0, and the rounding of f2(X A ) × G2 can obtain O2(X C ) as 3.

[0059] Step three, let the total number of targets be p, and convert the multiple grid numbers of each individual on the p targets into a comprehensive number O m (X m ) according to the following formula:

[0060]

[0061] Among them, O1(X) m ) represents individual X m The grid number in the first target dimension; O i (X m ) represents individual X m The grid number in the i-th target dimension; W j G represents the number of grid cells on target j. j The number of digits.

[0062] by Figure 2 Individual X in region A A For example, if G1 and G2 are both 5, then W1 and W2 are both 1. Since the target number p is 2, then:

[0063]

[0064] Among them, O C (X A The units digit represents the grid number O1(X) of the first target. A The tens digit represents the grid number O2(X) of the second target. A ).

[0065] Step 4: Sort all individuals in the population using quicksort based on their composite ID.

[0066] Step 5: Traverse the sorted population and identify individuals in the same grid based on the same composite ID. For example... Figure 2 The composite ID of all individuals in region A is 30. The occurrence of three consecutive individuals with the composite ID 30 indicates that all three individuals are located in region A.

[0067] To verify the effectiveness of this invention, the performance of the conventional mesh method and the improved mesh method of this invention in optimizing the performance of aero-engines under different operating conditions was compared based on the standard multi-objective optimization algorithm NSGA-II. The aero-engine type selected was a mixed-exhaust turbofan engine; the operating conditions were ground takeoff, supersonic cruise, and subsonic cruise; the optimization objectives were maximum thrust and minimum fuel consumption. The optimization times for the conventional mesh method and the improved mesh method of this invention under different operating conditions are shown in the table below. It can be seen that compared to the conventional mesh method, the improved mesh method of this invention can reduce the optimization time by 14.44% to 17.77%.

[0068] Table 1. Time (s) for different grid methods in optimizing aero-engine performance.

[0069]

[0070] Although the embodiments of the present application have been shown and described above, it is understood that the above-described embodiments are exemplary and are not to be construed as limiting the present application, and that changes, modifications, substitutions and variations can be made by those skilled in the art without departing from the spirit and scope of the present application.

Claims

1. A method for quickly identifying the same number of individuals in a grid, characterized in that The specific steps are as follows: In a model for multi-objective optimization of overall performance of an aero-engine, each objective value of each individual in a population of design parameters is normalized; the population of design parameters includes N individuals, each of which is a vector including D optimization variables, expressed as: wherein , represents the jth optimization variable of the ith individual; The individual In the target i The target value is The target value after normalization is The expression is: in, and The target in the population i The maximum and minimum values, i Indicates the target's sequence number; Calculate the grid number for each individual in each target dimension; the method for calculating the grid number for each individual in each target dimension is as follows: Let Gi be the number of grids on target i, and for... Rounding down yields the individual values. Grid number in the i-th target dimension ; The multiple grid numbers of each individual across all targets are converted into a single comprehensive number; the comprehensive number is... The expression is as follows: in, Represents an individual The grid number in the first target dimension; Represents an individual In the i Grid numbering in each target dimension; Indicate target j Number of grids on The number of digits; p Indicates the total number of targets; Based on the individual's composite ID, all individuals in the population are sorted. Traverse the sorted population and identify individuals in the same grid based on the same comprehensive number.

2. The method for quickly identifying individuals with the same ID number in a grid according to claim 1, characterized in that: The method used to sort all individuals in the population is quicksort.

3. A system for rapidly identifying individuals with the same ID number in a grid, characterized in that: The method for rapidly identifying individuals with the same number in a grid as described in claim 1 or 2 includes a normalization processing module, a grid numbering module, a comprehensive numbering module, a sorting module, and an identification module; the modules work together to achieve rapid identification of individuals with the same number in the grid of a multi-objective optimization model for overall performance of an aero-engine. The normalization module is used to normalize each target value of each individual in the population design parameters; The grid numbering module is used to calculate the grid number of each individual in each target dimension; The integrated numbering module is used to convert multiple grid numbers of each individual across all targets into a single integrated number. The sorting module is used to sort all individuals in the population; The identification module identifies individuals within the same grid based on the same comprehensive number.

4. A multi-objective optimization method for the overall performance of an aero-engine, characterized in that: The multi-objective optimization method employs the method for rapidly identifying individuals with the same number in a grid as described in claim 1 or 2.

5. An electronic device, characterized in that: The device includes at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the method for rapidly identifying individuals with the same number in a grid as described in claim 1 or 2.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions that enable a processor to execute the method for rapidly identifying individuals with the same number in a grid as described in claim 1 or 2.

Images