KR102198714B1 - System and method for estimating property of noise control material using genetic algorithm - Google Patents

Abstract

The present invention relates to a system and method for estimating the physical properties of a noise control material (e.g., flow resistance, porosity, warpage, thermal length, viscous length, etc.) using a dielectric algorithm. Is constructed, the fitness of the solution group is calculated, the result of the fitness calculation and the solution group are merged to form a new array to select the parent solution, and the parent solution is coded in binary and crossed to create a child solution. It generates a mutation to increase the diversity of the parent and child solutions, and constructs an assumption by decoding the parent and child solutions into decimal numbers, calculates the fit for the assumption, compares the solution group and the assumption, and replaces the solution. The present invention relates to a system and method for estimating the property values of noise control materials using a genetic algorithm that determines a solution with goodness of fit as an estimate.

Description

A system and method for estimating the properties of noise control materials using genetic algorithms {SYSTEM AND METHOD FOR ESTIMATING PROPERTY OF NOISE CONTROL MATERIAL USING GENETIC ALGORITHM}

The present invention relates to a system and method for estimating the physical properties of a noise control material (e.g., flow resistance, porosity, warpage, thermal length, viscous length, etc.) using a dielectric algorithm. Is constructed, the fitness of the solution group is calculated, the result of the fitness calculation and the solution group are merged to form a new array to select the parent solution, and the parent solution is coded in binary and crossed to create a child solution. It generates a mutation to increase the diversity of the parent and child solutions, and constructs an assumption by decoding the parent and child solutions into decimal numbers, calculates the fit for the assumption, compares the solution group and the assumption, and replaces the solution. The present invention relates to a system and method for estimating the property values of noise control materials using a genetic algorithm that determines a solution with goodness of fit as an estimate.

In general, as an engineering static soundproofing substrate, it is largely divided into a sound absorbing material and a sound insulating material. It is different from active noise canceling, which aims to remove active noise, and it is intended to block sound through the installation of a sound insulation film to achieve a sound insulation effect. The sound-absorbing material has an effect of absorbing noise and vibration, and in the case of the sound-absorbing material, it has a characteristic of reducing transmitted noise by blocking noise. Although it depends on the place of construction, the maximum sound insulation effect can be obtained only when a sound insulation material and a sound absorption material are usually installed together.

On the other hand, in order to use the sound-absorbing material and the sound-insulating material in the right place, it is necessary to accurately grasp the physical properties thereof, and furthermore, it is necessary to estimate the physical properties of the noise control material by numerical values.

International Patent Publication WO 2009/119344 (published date: 2009.10.01)

Accordingly, the present invention has been made to solve the conventional problems as described above, and an object of the present invention is to provide a system and method for estimating the properties of a noise control material using a genetic algorithm that can accurately estimate the properties of the noise control material in numerical values. To provide.

In order to achieve the above object, the system for estimating the physical properties of the noise control material using the genetic algorithm according to the embodiment of the present invention allows the overlapping of the property values of a predetermined number (n) of the noise control material in an area where the property can exist. A solution group constructing unit configured to form a solution group in an array of m rows and n columns by randomly selecting as many as a set number (m) and making the columns become material properties; A goodness-of-fit calculation unit for calculating the fitness for all the solutions of the solution group and storing the result as an array of m rows and 1 column; By merging the calculation result of the goodness-of-fit and the solution group, a new array of m rows and n+1 columns is formed and sorted in ascending order based on column 1 (goodness-of-fit column), and two solutions (parent solutions) are duplicated in all solution groups. Parental selection unit to select by allowing; A binary encoder for encoding the parent solution selected for mating into binary numbers; A child solution generator for generating a child solution by crossing the parent solution encoded in binary numbers; A mutation generating unit that generates a mutation for increasing the diversity of the offspring; A home construction unit that decodes the parent and child solutions into decimal numbers and configures one assumption; A fitness of the assumption that calculates the fitness of the assumption by the same method as the fitness of the solution group; Compare the solution with the highest goodness of fit (recessive) in the solution group (Solution 1) and the solution with the lowest goodness of fit (dominant) in the assumption (Solution 2). The year replacement unit replacing year 1 with year 2 in; And an estimate determination unit that determines a solution having the highest fitness as an estimate.

In the system for estimating the properties of the noise control material using the dielectric algorithm according to the above embodiment, the property values may include flow resistance, porosity, thermal characteristic length, and viscous characteristic length.

In the system for estimating the property values of the noise control material using the genetic algorithm according to the above embodiment, the fitness of the solution group and the fitness of assumption calculation unit may calculate the fitness by the following [Equation 1].

[Equation 1]

는 적합도를 나타내고,

는 I번째 해를 나타내고,

는 x축 최상단 주파수를 나타내고,

는 상기 x축 최 하단 주파수를 나타내고,

비오(Biot) 이론을 활용하여 추정한 주파수에서의 흡음계수를 나타내며,

는 시험을 통해 계측한 주파수에서의 흡음계수를 나타낸다][here,

Represents the fitness,

Represents the I-th solution,

Represents the uppermost frequency of the x-axis,

Represents the lowermost frequency of the x-axis,

It represents the sound absorption coefficient at the frequency estimated using Biot theory,

Represents the sound absorption coefficient at the frequency measured through the test]

In the system for estimating the properties of the noise control material using the genetic algorithm according to the above embodiment, the parental solution selection unit allows overlapping of two solutions (parent solutions) in all solution groups using the following [Equation 2] You can choose.

[Equation 2]

는

순위의 선택확률을 나타내고, sp는 선택압을 나타내며, n은 총 해의 개수를 나타내며, i는 순위를 나타낸다][here,

Is

It represents the probability of selection of the ranking, sp represents the selection pressure, n represents the total number of solutions, and i represents the rank]

In order to achieve the above object, in the method for estimating the physical properties of the noise control material using a genetic algorithm according to another embodiment of the present invention, the solution group component has a predetermined number (n) of the material properties of the noise control material. A first step of forming a group in an array of m rows and n columns by randomly selecting as many as a set number (m) of possible areas without allowing overlapping, and making the columns become physical properties; A second step of calculating, by a fitness calculation unit of a solution group, the fitness for all solutions of the solution group and storing the result as an array of m rows and 1 column; The parental solution selection unit merges the calculation result of the goodness-of-fit and the solution group to form a new array of m rows and n+1 columns, and sorts them in ascending order based on column 1 (goodness-of-fit column), and in all solution groups, two solutions (parent A third step of selecting a solution by allowing redundancy; A fourth step of encoding, by a binary encoding unit, the parent solution selected for mating into binary numbers; A fifth step of generating a child solution by crossing the parent solution encoded as a binary number by a child solution generator; A sixth step of generating a mutation for increasing the diversity of the offspring by the mutation generating unit; A seventh step of decoding the parent solution and the child solution to a decimal number by a family construction unit and constructing one assumption; An eighth step of calculating the fitness of the assumption by the fitness calculation unit in the same manner as the fitness calculation method in the second step; If the solution replacement part has a goodness-of-fit that is less dominant than recessive by comparing the solution with the highest goodness-of-fit (recessive) in the solution group (Solution 1) and the solution with the lowest goodness-of-fit (dominant) in the assumption (Solution 2) A ninth step of replacing solution 1 with solution 2 in the solution group; A tenth step of determining whether the second to ninth steps have been repeated a set number of times, by an estimate determination unit; And an eleventh step of determining, by the estimate determination unit, a solution having the highest degree of fitness as an estimate, if the second to ninth steps are repeated a set number of times in the tenth step.

The method for estimating the property values of the noise control material using the genetic algorithm according to the other embodiment may proceed to the second step if the second to ninth steps are not repeated a set number of times in the tenth step.

According to the system and method for estimating the property values of noise control materials using the genetic algorithm according to the embodiments of the present invention, the property values of a predetermined number (n) of the noise control materials are randomly applied without allowing overlapping in an area where the property values exist. After selecting as many as the set number (m) and making the columns become material properties, the solution group is formed in an array of m rows and n columns, and the goodness of fit for all solutions of the solution group is calculated, and the result is calculated as an array of m rows and 1 column. Save, and merge the calculation result of the goodness-of-fit and the solution group to form a new array of m rows and n+1 columns, and sort them in ascending order based on column 1 (goodness-of-fit column), and in all solution groups, two solutions (parent solution ) Is selected by allowing duplicates, the parent solution selected for crossing is encoded in binary, the parent solution encoded in binary is crossed to generate a child solution, and mutation to increase the diversity of the child solution Is generated, the parent and child solutions are decoded into decimal numbers to form an assumption, and the fitness of the assumption is calculated in the same way as the fitness of the solution group, and the highest fitness ( Comparing the solution with recessive) (Solution 1) and the solution with the lowest goodness of fit (dominant) in the above assumption (Solution 2), and replacing the solution 1 with the solution 2 in the above solution group if the fit has a lower dominance than recessive And, by configuring the solution with the highest fit to be determined as an estimate, there is an excellent effect that the physical properties of the noise control material can be accurately estimated numerically.

1 is a block diagram of a system for estimating a property value of a noise control material using a genetic algorithm according to an embodiment of the present invention.
FIG. 2 is a flowchart illustrating a method of estimating a property value of a noise control material using the system for estimating a property value of a noise control material using the genetic algorithm of FIG. 1.
3 is a diagram showing that the father and mother chromosomes are divided into two elements for each column in step S500 of FIG. 2.
FIG. 4 is a diagram showing that the number of 1024 cases is reduced to the number of 16 cases using the orthogonal arrangement table of the experimental design method in step S500 of FIG. 2.
FIG. 5 is a diagram showing that in step S500 of FIG. 2, a child solution is finally constructed by adding the number of four cases for effective and fast calculation.
6 is a view showing that the mutation is implemented in a form in which the gene value of one of the chromosomes of the offspring is changed to the opposite one in step S600 of FIG. 2.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

1 is a block diagram of a system for estimating a property value of a noise control material using a genetic algorithm according to an embodiment of the present invention.

The system for estimating the property values of noise control materials using a genetic algorithm according to an embodiment of the present invention, as shown in FIG. 1, includes a solution group configuration unit 100, a suitability calculation unit 200, and a parent solution selection unit. (300), binary encoding unit 400, child solution generation unit 500, mutation generation unit 600, assumption construction unit 700, family suitability calculation unit 800, solution replacement unit 900, and It includes an estimated value determining unit 1000.

The solution group component 100 draws a predetermined number (n) of noise control materials (e.g., 5 columns) of physical properties (e.g., flow resistance, porosity, thermal characteristic length, viscous characteristic length, etc.) The role of forming a group in the form of an array of m rows and n columns after selecting as many as the set number (m) (e.g., 150) randomly in the area where material properties can exist, and making the columns become material properties. Do it.

Areas in which the physical properties can exist are as follows, for example.

유동저항: 1000 ~ 10000000 MKS Rayls/m

Flow Resistance: 1000 ~ 10000000 MKS Rayls/m

공극률: 0 ~ 1

Porosity: 0 to 1

뒤틀림도: 1 ~ 5

Warp degree: 1 to 5

열특성길이: (1~400)*e-6 m

Thermal characteristic length: (1~400)*e-6 m

점성특성길이: (1~400)*e-6 m

Viscous character length: (1~400)*e-6 m

The fitness of the solution group 200 calculates the fitness for all the solutions of the solution group configured in the solution group configuration unit 100 and stores the result as an arrangement of m rows and 1 column.

The fitness of the solution group 200 calculates the fitness according to the following [Equation 1].

[Equation 1]

는 적합도를 나타내고,

는 I번째 해를 나타내고,

는 x축 최상단 주파수를 나타내고,

는 상기 x축 최 하단 주파수를 나타내고,

비오(Biot) 이론을 활용하여 추정한 주파수에서의 흡음계수를 나타내며,

는 시험을 통해 계측한 주파수에서의 흡음계수를 나타낸다][here,

Represents the fitness,

Represents the I-th solution,

Represents the uppermost frequency of the x-axis,

Represents the lowermost frequency of the x-axis,

It represents the sound absorption coefficient at the frequency estimated using Biot theory,

Represents the sound absorption coefficient at the frequency measured through the test]

The fitness calculation performed by the fitness calculation unit 200 of the solution group will be described as an example as follows.

For a total of 150 solutions, the sound absorption coefficient curve is estimated using Biot theory. The y-axis of the sound absorption coefficient curve is given as the sound absorption coefficient and the x-axis is the frequency. At this time, the y-axis has a value between 0 and 1, the x-axis has a value between 16 Hz and 16000 Hz, and the interval for each frequency follows a 1/3 octave band. The goodness of fit is defined as the sum of the absolute values of errors for each frequency between the sound-absorbing coefficient curve estimated based on a certain solution and the sound-absorbing coefficient curve obtained from the test result on the entire x-axis (see Equation 1). The "smaller" of fit is in common with the meaning of getting closer to the correct answer (test result).

For all solutions, the fit is calculated in the same way as above, and the result is stored as an array of 150 rows and 1 column.

The parental solution selection unit 300 merges the calculation result of the fitness of the solution group and the solution group configured by the solution group construction unit 100 to form a new array of m rows and n+1 columns. It sorts in ascending order based on column 1 (conformity column), and selects two solutions (parent solutions) from all solution groups by allowing overlapping.

The parental solution selection unit 300 selects two solutions (parent solutions) by allowing overlapping in all solution groups using the following [Equation 2].

[Equation 2]

는

순위의 선택확률을 나타내고, sp는 선택압을 나타내며, n은 총 해의 개수를 나타내며, i는 순위를 나타낸다][here,

Is

It represents the probability of selection of the ranking, sp represents the selection pressure, n represents the total number of solutions, and i represents the rank]

A process of selecting two solutions (parent solutions) by allowing overlapping by the parental solution selection unit 300 will be described as follows.

The result of calculating the fitness of the solution group (data in 150 rows and 1 column) and the solution group (data in 150 rows and 5 columns) are merged to form a new array of 150 rows and 6 columns (column 1 is goodness-of-fit). Sort by ascending order. In this arrangement, row 1 is the "dominant" solution with the lowest goodness of fit, and row 150 is the "recessive" solution with the highest goodness of fit. The roulette-wheel operation is used to ensure that the dominant solution is selected with a higher probability than the recessive solution probabilistically, but the selection probability of each solution is determined based on each fitness ranking. At this time, the order of fitness is the number of the row. In other words, row 1 is the highest with first priority and row 150 is the lowest with 150th. In this way, the selection probability is given according to the ranking, but pressure is applied to the selection so that the selection probability of the first-ranked solution does not exceed 4 times the selection probability of the 150-ranked solution. Here, the selection pressure (sp) is expressed as 4. According to [Equation 2], two solutions are selected by allowing duplicates from 150 solution groups.

The binary encoding unit 400 serves to encode a parent solution selected for mating by the parent solution selection unit 300 into binary numbers.

The coding process by the binary encoder 400 will be described in detail as follows.

Each element of the two solutions selected as parent solutions (row 2, column 5) is encoded from decimal to binary. Here, the flow resistance corresponding to the first row is 10 bits, the porosity corresponding to the second row is 10 bits, the twisting degree corresponding to the third row is 9 bits, the column characteristic length corresponding to the fourth row is 16 bits, and the fifth row is The corresponding viscous characteristic length is encoded as 16 bits. The row composed of binary numbers in this way is called a chromosome, and one element constituting the chromosome is called a gene.

For example, for the binary number 1011,

1011 is a 4 bit binary number,

1011 is a chromosome,

Each of 1, 0, 1, and 1 constituting a chromosome is a gene.

The child solution generator 500 serves to generate a child solution by crossing parent solutions encoded in binary numbers by the binary encoding unit 400.

A process of generating a child solution by the child solution generation unit 500 will be described as an example.

For mating of the parental solution, a point is randomly selected inside each chromosome. The chromosomes of father (row 1) and mother (row 2) are divided into two elements for each column, as shown in FIG. 3.

Referring to FIG. 3, the number of cases in which child solutions for the flow resistance exist are four in total: CF1CF2, CF1CM2, CM1CF2, and CM1CM2. This means that for 5 properties (control variables) such as flow resistance, each has 4 variables (control levels). If all 5 control variables have a control level of 4, the number of possible child chromosome cases is 4 ⁵ (=1024). Since it is difficult and inefficient to consider the number of all cases, the number of 1024 cases using an orthogonal array of Design Of Experiment (DOE) that has almost the same effect as the case of considering all cases. Is reduced to only 16 cases (see [Equation 3] and Fig. 4 below)

[Equation 3]

로 정의된다][Here, L means an orthogonal array table, V represents the control level, m represents the number of control variables, and n represents the number of tests.

Is defined as]

At this time, for an effective and fast calculation, focusing on the point at which the distortion of the 3rd column and the viscous characteristic length of the 5th column converge at the last, as shown in Fig. 5, the number of 4 cases is added to finally construct a child solution. do.

The mutation generating unit 600 serves to generate a mutation to increase the diversity of the child solutions constituted by the child solution generating unit 500.

The process of generating the mutation will be described in detail with an example.

If the initial solution group is all selected in a region that is far from the correct answer, there is a problem of local converging that cannot converge to the correct answer no matter how much they are crossed, so mutations are generated for some of the child solutions to escape convergence to the local solution. . From the 1st generation to the 150th generation, which converges to some extent as the correct answer, a 2% mutation probability is used to preserve the parent's genetic information as much as possible, and after the 150th generation when the correct answer converges to some extent, a high mutation probability of 60% is basically given. Maximize the diversity of the year. Mutation, as shown in Fig. 6, is implemented in the form of randomly changing the value of one of the chromosomes of the offspring to the opposite one.

The home configuration unit 700 decodes the parent solution and the child solution generated by the parent solution selection unit 300 and the child solution generation unit 500 into decimal numbers, and configures one home.

For example, two parent solutions and 20 child solutions can be decoded into a decimal number to form an assumption (row 22, column 5).

The fitness of the assumption calculation unit 800 serves to calculate the fitness of the assumption in the same manner as the fitness calculation method of the fitness calculation unit 200 of the group.

The solution replacement unit 900 is a solution (solution 1) having the highest fitness (recessiveness) in the solution group calculated by the fitness calculation unit 200 of the solution group and an assumption calculated by the fitness calculation unit 800 The solution (Solution 2) with the lowest fit (dominant) in is compared, and if the dominant fit has a lower fit than the recessive, it plays a role of replacing Solution 1 with Solution 2.

The estimate determination unit 1000 serves to determine a solution having the highest fitness as an estimate.

Hereinafter, a method for estimating the properties of the noise control material using the system for estimating the properties of the noise control material using the genetic algorithm according to an embodiment of the present invention constructed as described above will be described.

FIG. 2 is a flowchart illustrating a method of estimating a property value of a noise control material using the system for estimating property value of a noise control material using the genetic algorithm of FIG. 1, where S represents a step.

First, the solution group component 100 randomly selects a predetermined number (n) of material properties of the noise control material as much as a set number (m) in an area where the material properties can exist without allowing overlapping, So that the group is formed in an array of m rows and n columns (S100)

Subsequently, the fitness calculation unit 200 of the solution group calculates the fitness for all the solutions of the solution group configured in step S100, and stores the result as an arrangement of m rows and 1 column (S200).

Thereafter, the parental solution selection unit 300 merges the calculation result of the fitness at step S200 and the solution group configured at step S100 to form a new array of m rows and n+1 columns, ), and selects by allowing overlapping of two solutions (parent solutions) in all solution groups (S300).

Then, the binary encoding unit 400 encodes the parent solution selected for crossing in the step (S300) into binary numbers (S400), and the child solution generation unit 500 crosses the parent solution encoded in binary numbers, A solution is generated (S500).

Subsequently, the mutation generator 600 generates a mutation to increase the diversity of the child solutions generated in the step S500 (S600), and the home construction unit 700 decodes the parent and child solutions into decimal numbers. And constitutes one assumption (S700).

Thereafter, the fitness of the assumption calculation unit 800 calculates the fitness of the assumption configured in the step S700 in the same manner as the calculation method of the fitness in the step S200 (S800).

Subsequently, the solution replacement unit 900 has the highest fitness (recessive) in the solution group calculated in step S200 (solution 1) and the lowest fitness (dominant) in the assumption calculated in step S800 By comparing the solution with (Solution 2), it is determined whether the dominance has a lower fit than the recessive (S900).

In the step S900, if the dominance is lower than the recessive fit (YES), the solution 1 is replaced by the solution 2 in the solution group (S1000).

On the other hand, in the step S900, if the dominance does not have a degree of fitness lower than the recessive (NO), the step S900 is repeated.

Next, the estimate determination unit 1000 determines whether the steps S200 to S1000 are repeated a set number of times (S1100).

If the steps S200 to S1000 are repeated a set number of times (YES) in the step S1100, the solution having the highest degree of fit is determined by the estimate determining unit 1000 as an estimate value (S1200).

On the other hand, if the steps S200 to S1000 are not repeated the set number of times (NO) in the step S1100, the process proceeds to the step S200.

According to the system and method for estimating the property values of noise control materials using a genetic algorithm according to an embodiment of the present invention, the property values of a predetermined number (n) of the noise control materials are randomly obtained without allowing overlapping in an area where the property values exist. After selecting as many as the set number (m), configure the solution group in the form of an array of m rows and n columns by making the columns become material properties, calculate the fit for all solutions of the solution group, and store the result as an array of m rows and 1 column. And, by merging the calculation result of the goodness of fit and the solution group, a new array of m rows and n+1 columns is formed and sorted in ascending order based on column 1 (the fitness column), and two solutions (parent solutions) in all solution groups The parent solution selected for mating is selected by allowing duplication, encoding the parent solution selected for mating in binary, generating a child solution by crossing the parent solution coded in binary, and mutating mutations to increase the diversity of the child solution. The parent and child solutions are decoded into decimal numbers to form an assumption, and the fitness of the assumption is calculated in the same way as the fitness of the solution group, and the highest fitness (recessive) is calculated in the solution group. ) With the solution (Solution 1) and the solution with the lowest goodness of fit (dominant) in the assumption (Solution 2), and if the fit has a lower dominance than the recessive, the solution 1 is replaced by the solution 2 In addition, by configuring the solution with the highest fitness to be determined as an estimate, the properties of the noise control material can be accurately estimated numerically.

In the drawings and specification, the best embodiments have been disclosed, and specific terms are used, but these are only used for the purpose of describing the embodiments of the present invention, and are used to limit the meaning or the scope of the present invention described in the claims. Was not done. Therefore, those of ordinary skill in the art will appreciate that various modifications and equivalent other embodiments are possible therefrom. Therefore, the true technical scope of the present invention should be determined by the technical spirit of the appended claims.

100: sea group composition
200: fit calculation unit of solution group
300: parental choice
400: binary encoding unit
500: child year generator
600: mutation generator
700: home building block
800: family fitness calculation unit
900: Sun replacement
1000: estimate determination unit

Claims (9)

As a system for estimating the properties of the noise control material using a genetic algorithm that determines the solution with the highest fitness as an estimate among the solutions for the properties of the noise control material using the genetic algorithm,
The mass properties of a predetermined number (n) of the noise control material are randomly selected as many as the set number (m) without allowing duplication in the area where the properties can exist, and the columns are set to be the mass properties, so that they are arranged in m rows and n columns. A group composition unit that constitutes a group;
A goodness-of-fit calculation unit for calculating the fitness for all the solutions of the solution group and storing the result as an array of m rows and 1 column;
By merging the calculation result of the goodness-of-fit and the solution group, a new array of m rows and n+1 columns is formed and sorted in ascending order based on column 1 (goodness-of-fit column), and two solutions (parent solutions) are duplicated in all solution groups. Parental selection unit to select by allowing;
A binary encoder for encoding the parent solution selected for mating into binary numbers;
A child solution generator for generating a child solution by crossing the parent solution encoded in binary numbers;
A mutation generating unit that generates a mutation for increasing the diversity of the offspring;
A home construction unit that decodes the parent and child solutions into decimal numbers and configures one assumption;
A fitness of the assumption that calculates the fitness of the assumption by the same method as the fitness of the solution group;
Compare the solution with the highest goodness of fit (recessive) in the solution group (Solution 1) and the solution with the lowest goodness of fit (dominant) in the assumption (Solution 2). The year replacement unit replacing year 1 with year 2 in; And
Including; an estimate determination unit for determining a solution having the highest fitness as an estimate,
A system for estimating a property value of a noise control material using a genetic algorithm, which calculates the fitness according to the following [Equation 1].

[Equation 1]

[here,

Represents the fitness,

Represents the I-th solution,

Represents the uppermost frequency of the x-axis,

Represents the lowest frequency of the x-axis,

It represents the sound absorption coefficient at the frequency estimated using Biot theory,

Represents the sound absorption coefficient at the frequency measured through the test]
The method of claim 1,
The above properties are
A system for estimating the properties of a noise control material using a dielectric algorithm, including flow resistance, porosity, thermal characteristic length, and viscous characteristic length.
delete The method of claim 1,
The parental solution selection unit selects two solutions (parent solutions) in all solution groups using the following [Equation 2] by allowing overlapping, a system for estimating material properties of a noise control material using a genetic algorithm.

[Equation 2]

[here,

Is

It represents the probability of selection of the ranking, sp represents the selection pressure, n represents the total number of solutions, and i represents the rank]
As a method for estimating the properties of a noise control material using the system for estimating properties of the noise control material using the genetic algorithm described in claim 1,
The solution group composition part randomly selects a predetermined number (n) of material properties of the noise control material as much as the set number (m) without allowing overlapping in the area where the material properties can exist. A first step of constructing a solution group in an array form;
A second step of calculating, by a fitness calculation unit of a solution group, the fitness for all solutions of the solution group and storing the result as an array of m rows and 1 column;
The parental solution selection unit merges the calculation result of the goodness-of-fit and the solution group to form a new array of m rows and n+1 columns, and sorts them in ascending order based on column 1 (goodness-of-fit column), and in all solution groups, two solutions (parent A third step of selecting a solution by allowing redundancy;
A fourth step of encoding, by a binary encoding unit, the parent solution selected for mating into binary numbers;
A fifth step of generating a child solution by crossing the parent solution encoded as a binary number by a child solution generator;
A sixth step of generating a mutation for increasing the diversity of the offspring by the mutation generating unit;
A seventh step of decoding the parent solution and the child solution to a decimal number by a family construction unit and constructing one assumption;
An eighth step of calculating the fitness of the assumption by the fitness calculation unit in the same manner as the fitness calculation method in the second step;
If the solution replacement part has a goodness-of-fit that is less dominant than recessive by comparing the solution with the highest goodness-of-fit (recessive) in the solution group (Solution 1) and the solution with the lowest goodness-of-fit (dominant) in the assumption (Solution 2) A ninth step of replacing solution 1 with solution 2 in the solution group;
A tenth step of determining whether the second to ninth steps have been repeated a set number of times, by an estimate determining unit; And
In the tenth step, if the second to ninth steps are repeated a set number of times, an eleventh step of determining, by the estimate determination unit, a solution having the highest fitness as an estimate value; estimating the physical properties of the noise control material, including: Way.
The method of claim 5,
In the tenth step, if the second to ninth steps are not repeated a set number of times, the method proceeds to the second step.
The method of claim 5,
The above properties are
A method for estimating properties of a noise control material, including flow resistance, porosity, thermal characteristic length, and viscous characteristic length.
The method of claim 5,
The goodness-of-fit calculation unit of the solution group and the goodness-of-fit calculation unit of the assumption calculate the fitness by the following [Equation 1].

[Equation 1]

[here,

Represents the fitness,

Represents the I-th solution,

Represents the uppermost frequency of the x-axis,

Represents the lowermost frequency of the x-axis,

It represents the sound absorption coefficient at the frequency estimated using Biot theory,

Represents the sound absorption coefficient at the frequency measured through the test]
The method of claim 5,
The parental solution selection unit selects two solutions (parent solutions) in all solution groups by allowing overlapping using the following [Equation 2], a method for estimating material properties of a noise control material.

[Equation 2]

[here,

Is

It represents the probability of selection of the ranking, sp represents the selection pressure, n represents the total number of solutions, and i represents the rank]

Images