CN103177143A - Artificial electromagnetic material design method - Google Patents

Abstract

Translated fromChinese

本发明公开了一种人工电磁材料设计方法，该方法包括以下步骤：初始化生成x个晶格，将每一个该晶格的S1尺寸和S2尺寸进行编码并合并成一条染色体；根据期望值建立适应度函数，其中，该适应度函数用于计算染色体的适应度值；分别计算每条染色体的适应度值；选取适应度值从高到低排列的前y条染色体作为y条“父代染色体”并进行配对且相互交换自己的部分染色体以衍生y条“子代染色体”；计算该y条“子代染色体”的适应度值，并判断适应度值是否已达到预定阈值；在判断到适应度值达到预定阈值时，选取该适应度值对应的“子代染色体”并进行反编码以产生新的二维尺寸S1’和S2’。本发明利用遗传算法的原理，较快地搜索得到最优的二维尺寸S1’和S2’，提高了设计效率。

The invention discloses a method for designing artificial electromagnetic materials. The method includes the following steps: initializing and generating x lattices, encoding the S1 size and S2 size of each lattice and merging them into a chromosome; establishing the fitness according to the expected value function, wherein the fitness function is used to calculate the fitness value of chromosomes; calculate the fitness value of each chromosome respectively; select the first y chromosomes with fitness values arranged from high to low as y "parent chromosomes" and Perform pairing and exchange some of their own chromosomes to derive y "offspring chromosomes"; calculate the fitness value of the y "offspring chromosomes", and judge whether the fitness value has reached the predetermined threshold; after judging the fitness value When the predetermined threshold is reached, the "offspring chromosome" corresponding to the fitness value is selected and reverse-coded to generate new two-dimensional sizes S1' and S2'. The invention utilizes the principle of genetic algorithm to quickly search and obtain the optimal two-dimensional sizes S1' and S2', thereby improving the design efficiency.

Description

Translated fromChinese

人工电磁材料设计方法Artificial Electromagnetic Material Design Method

技术领域technical field

本发明涉及人工电磁材料领域，特别是涉及一种人工电磁材料设计方法。The invention relates to the field of artificial electromagnetic materials, in particular to a design method for artificial electromagnetic materials.

背景技术Background technique

在设计结构参数(如二维尺寸)按特定规律变化的人工电磁材料时，针对基板上的不同位置，需要在该位置上放置特定折射率的晶格(人工电磁材料的最小单位为晶格)，因此，需要对二维尺寸不同的晶格进行筛选，使其按期望的特定二维尺寸放置在基板的对应位置上。其中，晶格二维尺寸的大小与其折射率等参数相对应。When designing artificial electromagnetic materials whose structural parameters (such as two-dimensional size) change according to specific rules, for different positions on the substrate, it is necessary to place a lattice with a specific refractive index on the position (the smallest unit of artificial electromagnetic materials is a lattice) , therefore, it is necessary to screen lattices with different two-dimensional sizes, so that they can be placed on the corresponding positions of the substrate according to the desired specific two-dimensional size. Wherein, the size of the two-dimensional dimension of the lattice corresponds to parameters such as its refractive index.

现有技术中，一般采用粒子滤波算法对人工电磁材料进行设计，比如将晶格按照其二维尺寸从小到大进行排布，然后根据每个晶格的响应值进行筛选，以求得其最优目标，此时，采用粒子滤波算法能较快的根据全局的响应值的情况实时更新粒子(对应为晶格)的运动方向，从而向全局的最优目标收敛。但是，在选择粒子的过程中往往忽略了粒子的差异性，容易出现局部收敛，而无法得到最优目标(即响应值最优的晶格)，而无法得到期望的人工电磁材料。In the prior art, particle filter algorithms are generally used to design artificial electromagnetic materials, such as arranging lattices according to their two-dimensional sizes from small to large, and then screening according to the response value of each lattice to obtain its optimal In this case, the particle filter algorithm can quickly update the movement direction of the particles (corresponding to the lattice) according to the global response value, so as to converge to the global optimal goal. However, in the process of particle selection, the difference of particles is often ignored, and local convergence tends to occur, so that the optimal target (that is, the lattice with the best response value) cannot be obtained, and the desired artificial electromagnetic material cannot be obtained.

如何避免筛选过程中由于出现局部收敛，导致无法获得响应值最优的二维尺寸的情况，是本技术领域亟需解决的技术问题。How to avoid the situation that the two-dimensional size with the optimal response value cannot be obtained due to local convergence in the screening process is a technical problem that needs to be solved urgently in this technical field.

发明内容Contents of the invention

本发明主要提供一种人工电磁材料设计方法，能够有效解决上述技术问题。The present invention mainly provides a method for designing artificial electromagnetic materials, which can effectively solve the above-mentioned technical problems.

为解决上述技术问题，本发明采用的一个技术方案是：提供一种人工电磁材料设计方法，该人工电磁材料包括晶格，该晶格的二维尺寸为S1尺寸和S2尺寸，该方法包括以下步骤：初始化生成x个晶格，将每一个该晶格的S1尺寸和S2尺寸进行编码并合并成一条染色体，以此形成x条染色体；根据期望值建立适应度函数，其中，该适应度函数用于计算染色体的适应度值；分别计算每条染色体的适应度值；选取适应度值从高到低排列的前y条染色体作为y条“父代染色体”并进行配对且相互交换自己的部分染色体以衍生y条“子代染色体”；返回该分别计算每条染色体的适应度值的步骤，以计算该y条“子代染色体”的适应度值，并判断适应度值是否已达到预定阈值；在判断到适应度值达到预定阈值时，选取该适应度值对应的“子代染色体”并进行反编码以产生新的二维尺寸S1’和S2’。In order to solve the above-mentioned technical problems, a technical solution adopted in the present invention is: provide a kind of artificial electromagnetic material design method, this artificial electromagnetic material comprises crystal lattice, and the two-dimensional dimension of this lattice is S1 dimension and S2 dimension, and this method comprises the following Steps: Initialize and generate x lattices, code the S1 size and S2 size of each lattice and merge them into one chromosome to form x chromosomes; establish a fitness function according to the expected value, where the fitness function is used It is used to calculate the fitness value of chromosomes; calculate the fitness value of each chromosome separately; select the first y chromosomes with fitness values arranged from high to low as y "parent chromosomes" and pair them and exchange some of their chromosomes Deriving y "progeny chromosomes"; returning to the step of calculating the fitness value of each chromosome separately, to calculate the fitness value of the y "progeny chromosomes", and judging whether the fitness value has reached a predetermined threshold; When it is judged that the fitness value reaches the predetermined threshold, the "progeny chromosome" corresponding to the fitness value is selected and reverse-coded to generate new two-dimensional sizes S1' and S2'.

其中，该y条“父代染色体”进行配对且相互交换自己的部分染色体以衍生y条“子代染色体”的步骤之后还包括：该y条“子代染色体”分别进行变异，接着，返回该利用该分别计算每条染色体的适应度值的步骤，以计算该经过变异的y条“子代染色体”的适应度值。Wherein, after the step of pairing the y "parent chromosomes" and exchanging part of their own chromosomes to derive y "offspring chromosomes", it also includes: mutating the y "offspring chromosomes" respectively, and then returning to the The step of separately calculating the fitness value of each chromosome is used to calculate the fitness value of the mutated y "progeny chromosomes".

其中，在该y条“子代染色体”分别进行变异的步骤中包括：该“子代染色体”在某节点上的数值进行变异，和/或更换不同节点上对应的数值。Wherein, the step of mutating the y "progeny chromosomes" respectively includes: mutating the value of the "progeny chromosomes" on a certain node, and/or replacing the corresponding values on different nodes.

其中，在该y条“父代染色体”进行配对且相互交换自己的部分染色体以衍生y条“子代染色体”的步骤中包括：该y条“父代染色体”配对时，配对的两条该“父代染色体”在该“适应度值从高到低排列的前y条染色体”中随机组合。Wherein, the step of pairing the y "parent chromosomes" and exchanging part of their own chromosomes to derive y "offspring chromosomes" includes: when the y "parent chromosomes" are paired, the paired two The "parent chromosome" is randomly combined in the "first y chromosomes arranged from high to low fitness value".

其中，在该将每一个该晶格的S1尺寸和S2尺寸进行编码并合并成一条染色体的步骤中包括：通过十六进制对该S1尺寸和S2尺寸进行编码。Wherein, the step of encoding and merging the S1 size and S2 size of each lattice into one chromosome includes: encoding the S1 size and S2 size by hexadecimal notation.

其中，在该将每一个该晶格的S1尺寸和S2尺寸进行编码并合并成一条染色体的步骤中包括：该尺寸S1和尺寸S2分别编码成16位的十六进制数值，该S1尺寸和S2尺寸进行编码并合并成一条染色体后，该染色体为32位的十六进制数值。Wherein, the step of encoding and merging the S1 size and S2 size of each lattice into one chromosome includes: the size S1 and the size S2 are respectively encoded into 16-bit hexadecimal values, and the S1 size and After the S2 dimension is encoded and merged into a chromosome, the chromosome is a 32-bit hexadecimal value.

其中，在该判断适应度值是否已达到预定阈值的步骤后包括：若判断到适应度值未达到预定阈值时，再次选取适应度值从高到低排列的前y’条染色体作为y’条“父代染色体”并进行配对且相互交换自己的部分染色体以衍生y’条“子代染色体”，计算该y’条“子代染色体”的适应度值，并再次判断适应度值是否已达到预定阈值。Wherein, after the step of judging whether the fitness value has reached the predetermined threshold includes: if it is judged that the fitness value has not reached the predetermined threshold, selecting the first y' chromosomes with the fitness values arranged from high to low as the y' again The "parent chromosomes" are paired and part of their chromosomes are exchanged with each other to derive y' "offspring chromosomes", the fitness value of the y' "offspring chromosomes" is calculated, and it is judged again whether the fitness value has reached predetermined threshold.

其中，该y’条“父代染色体”并进行配对且相互交换自己的部分染色体以衍生y’条“子代染色体”步骤后还包括：该y’条“子代染色体”分别进行变异，接着，返回该分别计算每条染色体的适应度值的步骤，以计算该经过变异的y’条“子代染色体”的适应度值。Wherein, after the step of pairing the y' "parental chromosomes" and exchanging part of their own chromosomes to derive y' "offspring chromosomes", it also includes: the y' "offspring chromosomes" are mutated respectively, and then , returning to the step of calculating the fitness value of each chromosome separately, so as to calculate the fitness value of the mutated y' "progeny chromosomes".

其中，在进行反编码以产生新的二维尺寸S1’和S2’的步骤之后，还包括：记录该二维尺寸S1’和S2’，以备选取该二维尺寸S1’和S2’对应的晶格。Wherein, after the step of reverse encoding to generate new two-dimensional sizes S1' and S2', it also includes: recording the two-dimensional sizes S1' and S2', so as to prepare for selecting the corresponding two-dimensional sizes S1' and S2' lattice.

本发明的有益效果是：区别于现有技术的情况，本发明人工电磁材料设计方法利用遗传算法的原理，较快地搜索得到适应度值最优的二维尺寸S1’和S2’。本发明有效地解决了粒子滤波算法中由于粒子的差异性而出现局部收敛的问题，提高了设计效率，从而可实现大规模产业化生产。The beneficial effect of the present invention is: different from the situation of the prior art, the artificial electromagnetic material design method of the present invention utilizes the principle of the genetic algorithm to quickly search and obtain the two-dimensional dimensions S1' and S2' with optimal fitness values. The invention effectively solves the problem of local convergence due to the difference of particles in the particle filter algorithm, improves the design efficiency, and thus can realize large-scale industrial production.

附图说明Description of drawings

图1是本发明人工电磁材料设计方法的第一实施例流程示意图；以及Fig. 1 is a schematic flow chart of the first embodiment of the artificial electromagnetic material design method of the present invention; and

图2是本发明人工电磁材料设计方法的第二实施例流程示意图。Fig. 2 is a schematic flowchart of the second embodiment of the artificial electromagnetic material design method of the present invention.

具体实施方式Detailed ways

本发明人工电磁材料设计方法由于采用的是遗传算法的原理，因此在搜索最优解的过程中，将引用遗传算法的定义进行描述，譬如将十六进制数值定义为一条染色体，数值的改变和调换则对应定义为变异等，其他内容亦根据该原则进行定义。下面结合其具体实施例对本发明人工合成材料设计方法作详细的描述。Since the artificial electromagnetic material design method of the present invention adopts the principle of genetic algorithm, in the process of searching for the optimal solution, the definition of genetic algorithm will be cited for description, for example, a hexadecimal value is defined as a chromosome, and the change of value And transposition is defined as variation, etc., and other contents are also defined according to this principle. The method for designing artificial synthetic materials of the present invention will be described in detail below in conjunction with its specific examples.

请参阅图1，本发明第一实施例。Please refer to FIG. 1 , the first embodiment of the present invention.

人工电磁材料一般包括晶格，该晶格的二维尺寸为S1尺寸和S2尺寸，该人工电磁材料设计方法包括：Artificial electromagnetic materials generally include a lattice, and the two-dimensional dimensions of the lattice are S1 size and S2 size. The artificial electromagnetic material design method includes:

步骤A101，初始化生成x个晶格，将每一个该晶格的S1尺寸和S2尺寸进行编码并合并成一条染色体，以此形成x条染色体；Step A101, initializing and generating x lattices, encoding the S1 and S2 dimensions of each lattice and merging them into one chromosome to form x chromosomes;

步骤A102，根据期望值建立适应度函数，其中，该适应度函数用于计算染色体的适应度值；Step A102, establishing a fitness function according to the expected value, wherein the fitness function is used to calculate the fitness value of the chromosome;

步骤A103，分别计算每条染色体的适应度值；Step A103, calculating the fitness value of each chromosome respectively;

步骤A104，选取适应度值从高到低排列的前y条染色体作为y条“父代染色体”并进行配对且相互交换自己的部分染色体以衍生y条“子代染色体”；Step A104, select the first y chromosomes arranged from high to low fitness values as y "parent chromosomes" and perform pairing and exchange part of their own chromosomes to derive y "child chromosomes";

步骤A105，返回该利用该适应度函数计算染色体的适应度值的步骤，以计算该y条“子代染色体”的适应度值，并判断适应度值是否已达到预定阈值；Step A105, return to the step of using the fitness function to calculate the fitness value of the chromosome, to calculate the fitness value of the y "progeny chromosomes", and judge whether the fitness value has reached a predetermined threshold;

步骤A106，在判断到适应度值达到预定阈值时，选取该适应度值对应的“子代染色体”并进行反编码以产生新的二维尺寸S1’和S2’；Step A106, when it is judged that the fitness value reaches the predetermined threshold, select the "progeny chromosome" corresponding to the fitness value and perform reverse coding to generate new two-dimensional sizes S1' and S2';

其中，x和y为自然数。Among them, x and y are natural numbers.

下面结合实例对本发明人工电磁材料设计方法作进一步描述。The method for designing artificial electromagnetic materials of the present invention will be further described below in conjunction with examples.

在步骤A101中，假设某一个晶格的二维尺寸S1＝3.13，S2＝2.21，将其按照十六进制等将其编码成一组具有一定长度的字符串，譬如S1编码后为40090a3d70a3d70a，S2为4001ae147ae147ae，最后，该晶格的二维尺寸合成的“染色体”为：40090a3d70a3d70a4001ae147ae147ae或4001ae147ae147ae40090a3d70a3d70a。当然，还可以用其他进制数进行表达，比如二进制和八进制等，在此不作限定。In step A101, assuming that the two-dimensional size of a certain lattice is S1=3.13, S2=2.21, encode it into a group of character strings with a certain length according to hexadecimal system, for example, after encoding S1, it is 40090a3d70a3d70a, S2 is 4001ae147ae147ae, and finally, the synthesized "chromosome" of the two-dimensional size of the lattice is: 40090a3d70a3d70a4001ae147ae147ae or 4001ae147ae147ae40090a3d70a3d70a. Of course, other base numbers can also be used for expression, such as binary and octal, etc., which are not limited here.

在步骤A102中，该期望值即为前述的表示其折射率或其他参数的期望值，比如希望找到折射率为1.2的晶格，其对应存在一个或者多个二维尺寸符合其要求的染色体，所以需要建立仿真以求得其适应度，再选择适应度值较高的染色体所对应的二维尺寸进行选择晶格。同时，也存在另外一个可能，就是对S1尺寸和S2尺寸进行第一次编码后，即获得适应度值最高的染色体，则步骤104和步骤105可以省略，然而为了确保结果的正确性，后续的步骤104和步骤105可以作为检验过程，具体而言，如果第一次编码即产生最优的染色体，其衍生的“子代染色体”也会逐渐返回到该最优的染色体的位置，以再次得到该最优的染色体。In step A102, the expected value is the aforementioned expected value representing its refractive index or other parameters. For example, it is desired to find a lattice with a refractive index of 1.2, which corresponds to the existence of one or more chromosomes whose two-dimensional size meets its requirements, so it is necessary to Establish a simulation to obtain its fitness, and then select the two-dimensional size corresponding to the chromosome with a higher fitness value to select the lattice. At the same time, there is another possibility, which is to obtain the chromosome with the highest fitness value after encoding the S1 size and S2 size for the first time, then step 104 and step 105 can be omitted, but in order to ensure the correctness of the result, the subsequent Steps 104 and 105 can be used as a verification process. Specifically, if the first coding produces the optimal chromosome, the derived "offspring chromosome" will gradually return to the position of the optimal chromosome to obtain the optimal chromosome again. The optimal chromosome.

在步骤A104中，若其中一条适应度值较高的“父代染色体”为40070a3d70a3d70a4001ae147ae147ae，另一条为适应度值较高的“父代染色体”为40070a3d70a3d70a4002bc258ac139ee，进行配对并交换对等的某一段，比如将第一条的“4001ae147ae”与第二条的“4002bc258ac”进行交换，则交换后，对应衍生的第一条“子代染色体”为40070a3d70a3d70a4002bc258ac147ae，相对地，衍生的第二条“子代染色体”为40070a3d70a3d70a4001ae147ae139ee。当然，并不是说适应度值较高的染色体必然成为“父代染色体”，只是概率相对比较大，类似于tournament原则(竞赛原则，或物竞天择原则)进行选择；而其他适应度值较低的染色体也存在成为“父代染色体”的可能性，以尽可能地找到较优的染色体并进行衍生的动作。同时，两条“父代染色体”进行配对交换的过程，其交换的节点是随机的，而且在配对时，配对的两条父代染色体也是在该“适应度值从高到低排列的前y条染色体”中随机组合的任意两条父代染色体。In step A104, if one of the "parent chromosomes" with a higher fitness value is 40070a3d70a3d70a4001ae147ae147ae, and the other "parent chromosome" with a higher fitness value is 40070a3d70a3d70a4002bc258ac139ee, perform pairing and exchange an equivalent segment, for example Exchange the "4001ae147ae" of the first article with the "4002bc258ac" of the second article, then after the exchange, the corresponding derived first "offspring chromosome" is 40070a3d70a3d70a4002bc258ac147ae, relatively, the derived second "offspring chromosome" is 40070a3d70a3d70a4001ae147ae139ee. Of course, it does not mean that the chromosome with higher fitness value must become the "parent chromosome", but the probability is relatively high, which is similar to the principle of tournament (the principle of competition, or the principle of natural selection); while other chromosomes with higher fitness value Low chromosomes also have the possibility to become "parent chromosomes", so as to find better chromosomes and perform derivative actions as much as possible. At the same time, in the process of pairing and exchanging two "parent chromosomes", the exchanged nodes are random, and when pairing, the paired two parent chromosomes are also in the first y of the "fitness values" arranged from high to low. Any two parental chromosomes randomly combined in "one chromosome".

进一步地，“父代染色体”和“子代染色体”也存在随机进行自变异的可能性，譬如某一条染色体为，接着，其“40080a3”部分变异为“41280a0”；或者，其“a3e60a3”部分与“147ae13”部分调换位置。当然，这只是对变异的一些举例，还包括其他方式，在本技术领域人员理解的范围内，不作赘述。Furthermore, the "parent chromosome" and "offspring chromosome" also have the possibility of random self-mutation. For example, a certain chromosome is, and then, its "40080a3" part mutates into "41280a0"; or, its "a3e60a3" part Swapped parts with "147ae13". Of course, these are just some examples of variation, and other ways are also included, which are within the scope of understanding of those skilled in the art, and will not be described in detail.

在步骤105中，该预定阈值可以是人为设定的，也可以根据实际而进行更改，进一步来说，也可以不设定预定阈值，而是通过设定步骤103和步骤104之间进行循环，当循环到一定次数(比如10次)之后，步骤105只要是选择适应度值最高的染色体即可。In step 105, the predetermined threshold can be set artificially, or can be changed according to the actual situation. Furthermore, the predetermined threshold can not be set, but looped between step 103 and step 104 by setting, After the cycle reaches a certain number of times (for example, 10 times), step 105 only needs to select the chromosome with the highest fitness value.

在步骤106中，譬如染色体“40080a3e60a3d80a4021ae147ae139ee”为适应度值超过预定阈值的最优的染色体，对其进行反编码后，S1’尺寸为2.89，S2尺寸为3.11，则将获取该二维尺寸的晶格以放置到人工合成材料的对应基板的位置上。In step 106, for example, if the chromosome "40080a3e60a3d80a4021ae147ae139ee" is the optimal chromosome whose fitness value exceeds the predetermined threshold, after inverse encoding, the size of S1' is 2.89, and the size of S2 is 3.11, then the crystal of the two-dimensional size will be obtained. grid to be placed on the position of the corresponding substrate of the artificial synthetic material.

通过本实施例，利用遗传算法的原理，较快地搜索得到适应度值最优的二维尺寸S1’和S2’，提高了设计效率，从而可实现大规模产业化生产。Through this embodiment, the two-dimensional sizes S1' and S2' with the best fitness values can be searched quickly by using the principle of genetic algorithm, which improves the design efficiency and realizes large-scale industrial production.

请参阅图2，本发明第二实施例。Please refer to FIG. 2 , the second embodiment of the present invention.

在本实施例中，该人工电磁材料设计方法包括：In this embodiment, the artificial electromagnetic material design method includes:

步骤A201，初始化生成x个晶格，通过十六进制将每一个该晶格的S1尺寸和S2尺寸进行编码并合并成一条染色体，以此形成x条染色体；Step A201, initializing and generating x lattices, encoding the S1 and S2 dimensions of each lattice in hexadecimal and merging them into one chromosome to form x chromosomes;

步骤A202，根据期望值建立适应度函数，利用该适应度函数计算染色体的适应度值；Step A202, establishing a fitness function according to the expected value, and using the fitness function to calculate the fitness value of the chromosome;

步骤A203，判断适应度值是否已达到预定阈值，若“是”，执行步骤A204，若“否”则执行步骤A206；Step A203, judging whether the fitness value has reached a predetermined threshold, if "yes", execute step A204, if "no", execute step A206;

步骤A204，选取适应度值对应的染色体并进行反编码以产生新的二维尺寸S1’和S2’；Step A204, select the chromosome corresponding to the fitness value and perform reverse encoding to generate new two-dimensional sizes S1' and S2';

步骤A205，记录该二维尺寸S1’和S2’，以备选取该二维尺寸S1’和S2’对应的晶格，流程结束。Step A205, record the two-dimensional dimensions S1' and S2', in preparation for selecting the lattice corresponding to the two-dimensional dimensions S1' and S2', and the process ends.

步骤A206，选取适应度值从高到低排列的前y条染色体作为y条“父代染色体”并进行配对且相互交换自己的部分染色体以衍生y条“子代染色体”，接着，返回步骤A201或执行步骤A207；Step A206, select the first y chromosomes arranged from high to low fitness values as y "parent chromosomes" and pair them up and exchange some of their own chromosomes to derive y "offspring chromosomes", then return to step A201 Or perform step A207;

步骤A207，该y条“子代染色体”分别进行变异，返回步骤A202。In step A207, the y "progeny chromosomes" are mutated respectively, and return to step A202.

如前所述，在本实施例中，x和y均属自然数，且x≥y。As mentioned above, in this embodiment, both x and y are natural numbers, and x≥y.

在该步骤A201中，假设某晶格的二维尺寸S1＝3.13，S2＝2.21，将其按照十六进制等将其编码成一组具有一定长度的字符串，譬如S1编码后为40090a3d70a3d70a，S2为4001ae147ae147ae，最后，该晶格的二维尺寸合成的“染色体”为：40090a3d70a3d70a4001ae147ae147ae或4001ae147ae147ae40090a3d70a3d70a的十六进制数值。当然，还可以用其他进制数进行表达，比如二进制和八进制等，在此不作限定。In this step A201, assuming that the two-dimensional size of a certain lattice is S1=3.13, S2=2.21, it is encoded into a set of character strings with a certain length according to hexadecimal, for example, after S1 is encoded, it is 40090a3d70a3d70a, S2 is 4001ae147ae147ae, and finally, the “chromosome” synthesized by the two-dimensional size of the lattice is: the hexadecimal value of 40090a3d70a3d70a4001ae147ae147ae or 4001ae147ae147ae40090a3d70a3d70a. Of course, other base numbers can also be used for expression, such as binary and octal, etc., which are not limited here.

与第一实施例相比，本实施例的优点在于，在计算第一次适应度值之后，进入步骤A203以判断适应度值是否达到预定阈值，若已经达到预定阈值，则可以省略掉步骤A206和步骤A207，提高了设计效率。Compared with the first embodiment, the advantage of this embodiment is that, after calculating the fitness value for the first time, enter step A203 to judge whether the fitness value has reached the predetermined threshold, and if it has reached the predetermined threshold, step A206 can be omitted and step A207, improving design efficiency.

在步骤A203到步骤A207的循环过程中，可能出现如下情况：During the cycle from step A203 to step A207, the following situations may occur:

若判断到适应度值未达到预定阈值时，再次选取适应度值从高到低排列的前y’条染色体作为y’条“父代染色体”并进行配对且相互交换自己的部分染色体以衍生y’条“子代染色体”，计算该y’条“子代染色体”的适应度值，并再次判断适应度值是否已达到预定阈值；进一步地，若该y’条“子代染色体”分别进行变异，接着，返回该利用该适应度函数计算染色体的适应度值的步骤，以计算该经过变异的y’条“子代染色体”的适应度值，其中，y’是自然数且y’≤y。当然，这个循环过程如果达到一定的次数，可智能地中断，并从新执行步骤A201，这仅是举例说明而并非限定，在此不作赘述。If it is judged that the fitness value does not reach the predetermined threshold, the first y' chromosomes arranged from high to low fitness values are selected again as y' "parent chromosomes" and paired, and some of their chromosomes are exchanged with each other to derive y 'offspring chromosomes', calculate the fitness value of the y' 'offspring chromosomes', and judge again whether the fitness value has reached the predetermined threshold; further, if the y' 'offspring chromosomes' are respectively Mutation, then, return to the step of calculating the fitness value of the chromosome by using the fitness function to calculate the fitness value of the mutated y' "offspring chromosomes", where y' is a natural number and y'≤y . Of course, if this looping process reaches a certain number of times, it can be interrupted intelligently, and step A201 will be executed again. This is only an example and not a limitation, and will not be repeated here.

另外，本发明的具体实施过程其工作原理请参阅第一实施例所述的工作原理，在本技术领域人员理解的范围内，不再赘述。In addition, for the working principle of the specific implementation process of the present invention, please refer to the working principle described in the first embodiment, and will not be repeated within the scope of understanding of those skilled in the art.

通过本实施例，利用遗传算法的原理，较快地搜索得到适应度值最优的二维尺寸S1’和S2’，提高了设计效率，从而可实现大规模产业化生产。Through this embodiment, the two-dimensional sizes S1' and S2' with the best fitness values can be searched quickly by using the principle of genetic algorithm, which improves the design efficiency and realizes large-scale industrial production.

在本发明的实施例中，都是针对人工合成材料的晶格的二维尺寸进行寻求最优的解，但是也可以针对其他参数，譬如损耗、折射率和体积等，其具体工作原理过程基本一样，本发明具体实施例只是针对二维尺寸进行描述，但不限于此。In the embodiments of the present invention, the optimal solution is sought for the two-dimensional size of the crystal lattice of artificial synthetic materials, but it can also be aimed at other parameters, such as loss, refractive index and volume, etc. The specific working principle process is basically Likewise, the specific embodiments of the present invention are only described for two-dimensional dimensions, but are not limited thereto.

以上所述仅为本发明的实施例，并非因此限制本发明的专利范围，凡是利用本发明说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本发明的专利保护范围内。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 (9)

1. artificial electromagnetic material method for designing, described artificial electromagnetic material comprises lattice, the two-dimensional of described lattice is S1 size and S2 size, it is characterized in that, said method comprising the steps of:

Initialization generates x lattice, and S1 size and the S2 size of each described lattice are encoded and be merged into item chromosome, forms x bar chromosome with this;

Set up fitness function according to expectation value, wherein, described fitness function is used for calculating chromosomal fitness value;

Calculate respectively every chromosomal fitness value;

Choose front y bar chromosome that fitness value arranges from high to low as y bar " parent chromosome " and match and mutually the own chromosome dyad of exchange to derive y bar " child chromosome ";

Return to the described step of calculating respectively every chromosomal fitness value, calculating the fitness value of described y bar " child chromosome ", and judge whether fitness value has reached predetermined threshold;

When determining fitness value and reach predetermined threshold, choose described fitness value corresponding " child chromosome " and carry out Gray code to produce new two-dimensional S1 ' and S2 '.

2. method according to claim 1, is characterized in that, described y bar " parent chromosome " also comprises after matching and mutually exchange the step of chromosome dyad with derivative y bar " child chromosome " of oneself:

Described y bar " child chromosome " makes a variation respectively, then, returns to the described step of calculating respectively every chromosomal fitness value, to calculate the described fitness value that passes through the y bar " child chromosome " of variation.

3. method according to claim 2, is characterized in that, the step that makes a variation respectively at described y bar " child chromosome " comprises:

Described " child chromosome " numerical value on certain node makes a variation, and/or changes numerical value corresponding on different nodes.

4. method according to claim 1, is characterized in that, described y bar " parent chromosome " and match and mutually the own chromosome dyad of exchange comprise with the step that derives y bar " child chromosome ":

During described y bar " parent chromosome " pairing, two described " parent chromosomes " of pairing are random combine in described " the front y bar chromosome that fitness value is arranged from high to low ".

5. method according to claim 1, is characterized in that, encodes and the step that is merged into item chromosome comprises in described S1 size with each described lattice and S2 size:

By sexadecimal, described S1 size and S2 size are encoded.

6. method according to claim 5, is characterized in that, encodes and the step that is merged into item chromosome comprises in described S1 size with each described lattice and S2 size:

Described size S1 and size S2 are encoded into respectively the sexadecimal numerical value of 16, and after described S1 size and S2 size were encoded and be merged into item chromosome, described chromosome was the sexadecimal numerical value of 32.

7. method according to claim 1, is characterized in that, described judge after whether fitness value has reached the step of predetermined threshold comprise:

If when determining fitness value and not reaching predetermined threshold, again choose front y ' bar chromosome that fitness value arranges from high to low as y ' bar " parent chromosome " and match and mutually the own chromosome dyad of exchange to derive y ' bar " child chromosome ", calculate the fitness value of described y ' bar " child chromosome ", and judge again whether fitness value has reached predetermined threshold.

8. method according to claim 7, is characterized in that, described y ' bar " parent chromosome " and match and mutually the own chromosome dyad of exchange also comprise after deriving y ' bar " child chromosome " step:

Described y ' bar " child chromosome " makes a variation respectively, then, returns to the described step of calculating respectively every chromosomal fitness value, to calculate the described fitness value that passes through the y ' bar " child chromosome " of variation.

9. method according to claim 1, is characterized in that, carry out Gray code with the step that produces new two-dimensional S1 ' and S2 ' after, also comprise:

Record described two-dimensional S1 ' and S2 ', in order to choosing lattice corresponding to described two-dimensional S1 ' and S2 '.

Images