CN103164555B - A kind of optimization method of artificial electromagnetic material design data - Google Patents

Abstract

The invention discloses a method for optimizing design data of artificial electromagnetic materials. The method comprises: extracting simulation data of artificial electromagnetic material design to form a population; performing curve fitting on individuals in the population to form an objective function; constructing the objective function as Augmented objective function; preset comparison parameters; generate random parameters; compare the size of the penalty function value of two adjacent individuals or the size of the random parameter and the comparison parameter to sort two adjacent individuals; repeatedly generate random parameters and compare the two adjacent individuals The step of sorting adjacent individuals until all individuals in the population are sorted; the optimal data is obtained according to the final sorting results. The optimization method of artificial electromagnetic material design data of the present invention can simplify the processing process of artificial electromagnetic material design data optimization, can obtain the optimal data satisfying the conditions in the artificial electromagnetic material design data quickly and accurately, and promote the artificial electromagnetic material design. The process of industrialization has a very positive effect.

Description

An Optimization Method for Design Data of Artificial Electromagnetic Materials

technical field

The invention relates to the field of artificial electromagnetic material design, in particular to an optimization method for artificial electromagnetic material design data.

Background technique

Artificial electromagnetic material technology is a cutting-edge cross-technology. Artificial electromagnetic material refers to some artificial composite structures or composite materials with extraordinary physical properties that natural materials do not have. Through the orderly design of the structure on the key physical scale of the material, the limitation of some apparent natural laws can be broken through, so as to obtain the supernormal material function beyond the ordinary nature inherent in nature.

In the design stage of artificial electromagnetic materials, the established unit structure model requires a lot of data simulation work to obtain some physical properties of the designed model. After the simulation is completed, a series of data of artificial electromagnetic materials will be obtained, and a large amount of data processing work will be performed on these data to obtain the characteristics of artificial electromagnetic materials. If it is done manually, it will be a very huge project. The optimization design of the relevant data of the simulation results of artificial electromagnetic materials is currently at the stage of manual adjustment, and the fitting of the response curve corresponding to the structure of artificial electromagnetic materials is a necessary way to realize the optimization of artificial electromagnetic materials. The current algorithm is to use a statistical model to fit the response curve. For relevant data, the data optimization process of the model is often cumbersome and time-consuming. Therefore, it is urgent to find an optimization method for data processing related to the design of artificial electromagnetic materials. The problem.

Therefore, it is necessary to provide an optimization method for artificial electromagnetic material design data to solve the problem of heavy, cumbersome and time-consuming optimization process of the prior art on artificial electromagnetic material design-related data processing. The industrialization process of materials has a very positive effect.

Contents of the invention

The technical problem mainly solved by the present invention is to provide an optimization method of artificial electromagnetic material design data, which can quickly and accurately obtain the optimal data satisfying conditions in the artificial electromagnetic material design data.

In order to solve the above technical problems, the present invention provides a method for optimizing the design data of artificial electromagnetic materials. The optimization method includes: extracting the simulation data of artificial electromagnetic material design to form a population; performing curve fitting on individuals in the population to form a target Function, and given constraints; add penalty factor and penalty function to the objective function to construct an augmented objective function; preset comparison parameters; generate random parameters; according to the comparison results of the penalty function values of two adjacent individuals or random parameters and According to the comparison result of comparing the size of the parameters, sort the two adjacent individuals; repeat the above steps of generating random parameters and sorting the two adjacent individuals until all the individuals in the population are sorted; according to the sorting of the above repeated steps The result is optimal data that satisfies the constraints of the objective function.

According to a preferred embodiment of the present invention, the formula of the objective function formed by performing curve fitting on individuals in the population is expressed as follows:

minf(x), where x _i ≤ x ≤ x _n

$\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; R</span>}}^{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&ForAll;</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \}</span>$

Among them, f(x) is the objective function, x=(x ₁ ,...,x _n )∈R ⁿ is the population, x ₁ ,...,x _n are the individuals in the population, n and m are A positive integer, given that x _min ≤ x ≤ x _max is the constraint condition of the objective function, x _min is the minimum value of x, x _max is the maximum value of x, g _j (x) is the constraint function of x, and F is Feasible region of the objective function.

According to a preferred embodiment of the present invention, the augmented objective function formula constructed by increasing the penalty factor and the penalty function to the objective function is expressed as follows:

Ψ(x)=f(x)+r _g φ(g _j (x); j=1,...,m)

$\mathrm{<span\; class="notranslate">\; \&phi;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ;</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; max</span>}{<span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \}</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}$

Among them, r _g is the penalty factor, and φ is the penalty function.

According to a preferred embodiment of the present invention, the comparison parameter and the random parameter are between 0 and 1.

According to a preferred embodiment of the present invention, the value of the comparison parameter is preset as 0.45.

According to a preferred embodiment of the present invention, the step of sorting two adjacent individuals according to the comparison result of the penalty function value of the two adjacent individuals or the comparison result between the random parameter and the size of the comparison parameter includes:

When the penalty function values of two adjacent individuals are equal and equal to zero, it is concluded that both adjacent individuals are feasible solutions of the augmented objective function, and comparing the objective function values of the two adjacent individuals, according to the size of the objective function value, the Two adjacent individuals exchange order.

According to a preferred embodiment of the present invention, the step of sorting two adjacent individuals according to the comparison result of the penalty function value of the two adjacent individuals or the comparison result between the random parameter and the size of the comparison parameter includes:

When the penalty function values of two adjacent individuals are not equal and equal to zero, compare the size of the random parameter and the comparison parameter. If the random parameter is smaller than the comparison parameter, then compare the penalty function values of the two adjacent individuals. According to the penalty function value Size Exchange the order of two adjacent individuals; if the random parameter is greater than the comparison parameter, repeat the above steps of generating random parameters and sorting two adjacent individuals.

According to a preferred embodiment of the present invention, the step of sorting the two adjacent individuals according to the comparison result of the penalty function value of the two adjacent individuals or the comparison result of the size of the random parameter and the comparison parameter further includes:

According to the size of the objective function value or penalty function value, the two adjacent individuals are sorted in ascending order, that is, the individual with the smaller objective function value or penalty function value in the two adjacent individuals is ranked in the objective function value of the two adjacent individuals Or before another individual with a large penalty function value, and after the sorting is completed, repeat the above steps of generating random parameters and sorting two adjacent individuals.

According to a preferred embodiment of the present invention, the step of sorting the two adjacent individuals according to the comparison result of the penalty function value of the two adjacent individuals or the comparison result of the size of the random parameter and the comparison parameter further includes:

According to the size of the objective function value or penalty function value, the two adjacent individuals are sorted in descending order, that is, the individual with the larger objective function value or penalty function value in the two adjacent individuals is ranked in the objective function value of the two adjacent individuals or before another individual with a small penalty function value, and repeat the above steps of generating random parameters and sorting two adjacent individuals after the sorting is completed.

According to a preferred embodiment of the present invention, the step of obtaining the optimal data satisfying the constraints of the objective function according to the sorting results of the above-mentioned repeated steps includes:

According to the results of ascending order or descending order, the optimal data satisfying the constraints of the objective function are obtained.

The beneficial effects of the present invention are: different from the situation of the prior art, the artificial electromagnetic material design data optimization method of the present invention is based on the application of the penalty function method to deal with the artificial electromagnetic material design data constraint optimization problem, and proposes a random sorting method for processing The data results can simplify the processing process of artificial electromagnetic material design data optimization, and can quickly and accurately obtain the optimal data that meets the conditions in the artificial electromagnetic material design data, which has a very positive effect on promoting the industrialization process of artificial electromagnetic materials .

Description of drawings

Fig. 1 is the three-dimensional structure schematic diagram of the artificial electromagnetic material structural unit of the embodiment of the present invention;

Fig. 2 is a flowchart of a method for optimizing design data of artificial electromagnetic materials according to an embodiment of the present invention.

detailed description

The present invention will be described in detail below in conjunction with the accompanying drawings and embodiments.

Fig. 1 is a schematic diagram of the three-dimensional structure of the artificial electromagnetic material structural unit of the embodiment of the present invention.

The information of the structural unit of artificial electromagnetic materials includes length ax, span ay, thickness hs, attenuation, frequency, dielectric constant, magnetic permeability, etc. In the design stage of artificial electromagnetic materials, a large amount of data is required for each type of information Simulation work, after the simulation is completed, a set of data will be obtained for each type of information, and these data are extremely large and complex.

Fig. 2 is a flowchart of an optimization method for artificial electromagnetic material design data according to an embodiment of the present invention. As shown in Fig. 2, a method 200 for optimizing artificial electromagnetic material design data includes the following steps:

Step 201: Extracting simulation data of artificial electromagnetic material design to form a population.

The simulation data in step 201 is obtained by simulating the structural unit data of the artificial electromagnetic material in Fig. 1, and these simulation data are the individuals of the formed population, denoted as x ₁ , x ₂ ..., x _n , where , n is a positive integer.

Step 202: Carry out curve fitting on the individuals in the population to form an objective function and give constraints.

The objective function in step 202 is formed by performing curve fitting on the individuals in the population, which is recorded as:

minf(x), where x _i ≤ x ≤ x _n

$\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; R</span>}}^{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&ForAll;</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \}</span>$

Among them, f(x) is the objective function, x=(x ₁ ,...,x _n )∈R ⁿ is the population, x ₁ ,...,x _n are the individuals in the population, n and m are A positive integer, given that x _min ≤ x ≤ x _max is the constraint condition of the objective function, x _min is the minimum value of x, x _max is the maximum value of x, g _j (x) is the constraint function of x, and F is Feasible region of the objective function.

Through step 202, the problem of finding the minimum or maximum value of the objective function is transformed into a problem of processing constraint optimization.

Step 203: Adding a penalty factor and a penalty function to the objective function to construct an augmented objective function.

Step 203 is the process of converting the constrained optimization problem into an unconstrained optimization problem through the penalty function method. The augmented objective function is constructed by adding a penalty factor and a penalty function to the objective function, which is recorded as:

Ψ(x)=f(x)+r _g φ(g _j (x); j=1,...,m)

$\mathrm{<span\; class="notranslate">\; \&phi;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ;</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; max</span>}{<span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \}</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}$

Among them, r _g is the penalty factor, φ is the penalty function, and the penalty function is a function that takes the constraint function as a variable, calculates the maximum value of 0 and the square value of the constraint function, and then sums them item by item.

The penalty function method is widely used because of its simple execution when dealing with constrained optimization problems. Its main idea is to construct an augmented objective function by adding a penalty term to the objective function, and transform the constrained optimization problem into an unconstrained optimization problem for processing. The objective The solution of the function is called a feasible solution, which is within the range of constraints, and the solution of the penalty function is called an infeasible solution, which is outside the range of constraints.

Step 204: Preset comparison parameters.

The comparison parameter in step 204 is preset, denoted as P _f |P _f ∈ U(0, 1), which means that in the infeasible domain, only the objective function is used to compare the probability of individuals, that is, the comparison parameter P _f will decide whether to use Objective function or degree of constraint violation to compare individuals. In this embodiment, P _f takes a value of 0.45. In other embodiments, any value from 0 to 1 can be selected as required.

Step 205: Generate random parameters.

In this embodiment, the random parameter in step 205 is randomly generated, denoted as Every time step 205 is performed, a random parameter will be regenerated.

Step 206: sort the two adjacent individuals according to the comparison result of the penalty function values of the two adjacent individuals or the comparison result between the random parameter and the comparison parameter.

Step 206 includes: for two adjacent individuals x _i and x _i+1 , when the penalty function values of x _i and x _i+1 φ( _xi )=φ( _xi+1 )=0, obtain x Both _i and x _i+1 are feasible solutions of the augmented objective function, and the probability of comparing their objective functions is 1, that is, comparing f( _xi ) and f( _xi+1 ), according to f( _xi ) and f The size of ( _xi+1 ) swaps the order of x _i and x _i+1 .

When the penalty function values of x _i and x _i+1 do not satisfy φ( _xi )=φ( _xi+1 )=0, compare the size of the random parameter and the comparison parameter. , if u<P _f , then compare φ(x _i ) and φ(x _i+1 ), and exchange order of _xi and _xi+1 according to the size of φ(x _i ) and φ(x _i+1 ); If u>P _f , go to step 207 .

Step 206 further includes: sorting _xi and xi ₊ 1 in ascending order according to the size of f( _xi ) and f( _xi+1 ), that is, when f( _xi )>f( _xi+1 ) or φ When (x _i )>φ(x _i+1 ), exchange the order of x _i and x _i+1 , and after the sorting is completed, execute step 207; for x _i and x _i+1 according to f(x _i ) and The size of f(x _i+1 ) is sorted in descending order, that is, when f(x _i )<f(x _i+1 ) or φ(x _i )<φ(x _i+1 ), exchange _xi and _{xi +1} , and after the sorting is complete, go to step 207.

Step 207: Repeat the above steps of generating random parameters and sorting two adjacent individuals until all individuals in the population are sorted.

In this embodiment, after exchanging _xi and _xi+1 , continue to regenerate a random number u, and sort the next pair of adjacent individuals _xi+1 and _xi+2 according to the aforementioned sorting method, Until all individuals are sorted, no further details will be given here.

Step 208: Obtain the optimal data satisfying the constraints of the objective function according to the sorting results of the above repeated steps.

The optimal data in the simulation data is not necessarily distributed in the feasible region, because the curve fitting of the simulation data includes all the data in the feasible region and the infeasible region, and some infeasible solutions in the infeasible region are not feasible. The row solution is closer to the optimal data of the simulation data than the feasible solution, but not all infeasible solutions have to be processed, which increases the workload, and the optimal data may not be obtained, which is not worth the loss. By setting a comparison parameter to indicate the probability of comparing individuals in the infeasible domain, it turns whether to deal with infeasible solutions into a probability problem. In order to find the optimal data, on the basis of using the penalty function method to deal with the constrained optimization problem, this method compares with the comparison parameters by setting random parameters, and judges whether to sort the individuals in the population according to the comparison results.

In this embodiment, the optimal data satisfying the constraints of the objective function will be obtained from the above sorting results in ascending order or in descending order.

The optimization method of artificial electromagnetic material design data of the present invention is on the basis of applying the penalty function method to deal with the constraint optimization problem of artificial electromagnetic material design data, and proposes a random sorting method to process data results, which can simplify the processing process of artificial electromagnetic material design data optimization, Being able to quickly and accurately obtain the optimal data that meets the conditions in the design data of artificial electromagnetic materials has a very positive effect on promoting the industrialization of artificial electromagnetic materials.

The above is only an embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the description of the present invention and the contents of the accompanying drawings, or directly or indirectly used in other related technologies fields, all of which are equally included in the scope of patent protection of the present invention.

Claims (8)

1. an optimization method of artificial electromagnetic material design data, is characterized in that, the method comprises the steps: Extract the simulation data of artificial electromagnetic material design to form a population; Carry out curve fitting on the individuals in the population to form an objective function and give constraints; Adding a penalty factor and a penalty function to the objective function to construct an augmented objective function; Preset comparison parameters; the comparison parameters are preset, indicating the probability of using only the objective function to compare individuals in the infeasible domain, that is, the comparison parameters will determine whether to use the objective function or the degree of constraint violation to compare individuals; generate random parameters; sorting the two adjacent individuals according to a comparison result of the penalty function values of the two adjacent individuals or a comparison result of the size of the random parameter and the comparison parameter; Repeat the above steps of generating random parameters and sorting two adjacent individuals until all individuals in the population are sorted; Obtaining the optimal data satisfying the constraints of the objective function according to the sorting results of the above repeated steps; The step of sorting the two adjacent individuals according to the comparison result of the penalty function value of the two adjacent individuals or the comparison result of the size of the random parameter and the comparison parameter comprises: When the penalty function values of the two adjacent individuals are equal and equal to zero, the two adjacent individuals are both feasible solutions of the augmented objective function, and the objective function values of the two adjacent individuals are compared , exchanging the order of the two adjacent individuals according to the value of the objective function; When the penalty function values of the two adjacent individuals are not equal and equal to zero, compare the size of the random parameter and the comparison parameter, if the random parameter is smaller than the comparison parameter, then compare the two phases The penalty function value of the adjacent individual, according to the size of the penalty function value, the order of the two adjacent individuals is exchanged; if the random parameter is greater than the comparison parameter, then perform the repetition of the above generated random parameter and the above pair of two The step of sorting adjacent individuals. 2. method according to claim 1, is characterized in that, the formula of the objective function formed by curve fitting is carried out to the individual in described population is as follows: minf(x), where x ₁ ≤ x ≤ x _n $\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; R</span>}}^{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&ForAll;</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \}</span>$ Among them, f(x) is the objective function, x=(x ₁ ,...,x _n )∈R ⁿ is the population, x ₁ ,...,x _n are the individuals in the population, n and m are A positive integer, given that x _min ≤ x ≤ x _max is the constraint condition of the objective function, x _min is the minimum value of x, x _max is the maximum value of x, g _j (x) is the constraint function of x, and F is Feasible region of the objective function. 3. method according to claim 2, is characterized in that, described objective function increases the augmented objective function formula that penalty factor and penalty function are constructed are as follows: Ψ(x)=f(x)+r _g φ(g _j (x); j=1,...,m) $\mathrm{<span\; class="notranslate">\; \&phi;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ;</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; max</span>}{<span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \}</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}$ Among them, r _g is the penalty factor, and φ is the penalty function. 4. The method according to claim 1, wherein the comparison parameter and the random parameter are between 0 and 1. 5. The method according to claim 4, characterized in that, the value of the comparison parameter is preset as 0.45. 6. The method according to claim 1, characterized in that, the comparison result of the penalty function value according to two adjacent individuals or the comparison result of the size of the random parameter and the comparison parameter is to the two The step of sorting adjacent individuals further includes: According to the size of the objective function value or the penalty function value, the two adjacent individuals are sorted in ascending order, that is, the individual with the smaller objective function value or penalty function value among the two adjacent individuals is arranged in the two adjacent individuals. Before another individual with a larger objective function value or penalty function value in the adjacent individual, and after the sorting is completed, repeat the above steps of generating random parameters and the above steps of sorting two adjacent individuals. 7. The method according to claim 1, characterized in that, the comparison result of the penalty function value according to two adjacent individuals or the comparison result of the size of the random parameter and the comparison parameter is to the two The step of sorting adjacent individuals further includes: According to the size of the objective function value or penalty function value, the two adjacent individuals are sorted in descending order, that is, the individual with a larger objective function value or penalty function value among the two adjacent individuals is arranged in the two adjacent individuals. The objective function value or penalty function in the adjacent individual. Before another individual with a small value, and after the sorting is completed, perform the steps of repeating the above steps of generating random parameters and sorting the two adjacent individuals. 8. The method according to claim 1, wherein the step of obtaining the optimal data satisfying the constraints of the objective function according to the sorting results of the above-mentioned repeated steps comprises: The optimal data satisfying the constraint condition of the objective function is obtained according to the ascending sorting result or the descending sorting result.