JP2007104638A - Antenna design method, antenna design program, and antenna designed by the antenna design method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an antenna design method, an antenna design program and an antenna designed by the antenna design method in which computational complexity relating to a metal arrangement pattern on an antenna element surface formed at the antenna can be suppressed while sufficiently optimizing antenna characteristics when designing a small-sized antenna. <P>SOLUTION: An antenna design method includes steps of: at each of computers 10, acquiring slave groups from an initial group that is a group of metal arrangement patterns, by applying genetic algorithm until satisfying round end conditions in accordance with setting conditions including at least the size of an antenna element surface and the size of a block and different from each other; and extracting a predetermined number of metal arrangement patterns as an initial group constitutive individual group from the slave groups acquired by the computers 10. Each of the computers 10 uses the initial group including the initial group constitutive individual group to restart a process for acquiring a slave group from the initial group. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an antenna design method, an antenna design program, and an antenna designed by the antenna design method for designing an antenna provided with antenna element surfaces divided into a predetermined number of blocks.

  Conventionally, as a method of designing an antenna having an antenna element surface divided into a plurality of blocks and having a block in which metal is arranged and a block in which metal is not arranged, a genetic algorithm (hereinafter referred to as GA) is used. There is known an antenna design method using an antenna.

  Specifically, whether or not metal is arranged in each block on the antenna element surface, that is, by calculating the metal arrangement pattern by GA, the metal arrangement pattern can be varied, and an optimal antenna can be obtained. It can be designed (for example, Non-Patent Document 1).

  Here, in the case where the number of vertical blocks on the antenna element surface is “m” and the number of horizontal blocks is “n”, the number of combinations of metal arrangement patterns is as large as 2 mn. According to the design method, the calculation amount of the metal arrangement pattern can be suppressed by GA.

As a technique for reducing the size of the linear antenna, a parallel calculation method using a GA and a vector evaluation method has also been proposed (for example, Non-Patent Document 2).
Maruyama, Cho, “Examination of automatic optimal setting using genetic algorithm for planar small multi-frequency antenna”, IEICE Letter, B, vol. J88-B No.6 June 2005 T.MARUYAMA, T.HORI, `` Vector Evaluated GA-ICT for Novel Optimum Design Method of Arbitrarily Arranged Wire Grid Model Antenna and Application of GA-ICT to Sector-antenna Downsizing Problem '', IEICE Trans. Vol.E84-B, No .11 pp.3014-3022 Nov., 2001)

  By the way, when designing a small antenna, not only the metal arrangement pattern but also the size of the antenna element surface is an important factor.

  However, in the conventional antenna design method, the metal arrangement pattern is calculated under the condition that the size of the antenna element surface is predetermined. That is, in order to reduce the size of the antenna, it is necessary to reset the size of the antenna element surface many times and calculate the metal arrangement pattern for each set size.

  On the other hand, if the calculation amount (number of generations in GA) of the metal arrangement pattern in each set size is reduced, antenna characteristics such as return loss cannot be sufficiently optimized.

  Therefore, in order to reduce the size of the antenna as much as possible, the antenna element surface size must be set a little smaller and the antenna characteristics must be fully optimized for each set size. The amount will increase.

  In addition, the method using the GA and the vector evaluation method can be applied to a linear antenna, but cannot be applied to an antenna having an antenna element surface such as a plane or an aspect.

  Therefore, the present invention provides an antenna design method capable of suppressing the amount of calculation related to the metal arrangement pattern of the antenna element surface provided in the antenna while sufficiently optimizing the antenna characteristics when designing a small antenna. An object of the present invention is to provide an antenna design program and an antenna designed by the antenna design method.

  According to a first aspect of the present invention, there is provided an antenna design method for designing an antenna provided with an antenna element surface partitioned into a predetermined number of blocks, wherein the antenna design surface includes at least the size of the antenna element surface or the size of the block. According to the conditions, a genetic algorithm is applied until one round end condition is satisfied, and an initial group that is a group of metal arrangement patterns indicating whether or not a metal is arranged in the block is changed from the initial group to the group of metal arrangement patterns. It includes at least one calculation step for obtaining a child population and the size of the antenna element surface or the size of the block, and is inherited until another round end condition is satisfied according to another setting condition different from the one setting condition. Applying another genetic algorithm to obtain the child population from the initial population; An extraction step of extracting a group of a predetermined number of the metal arrangement patterns as an initial population constituent population from the child population obtained in the calculation step and the child population obtained in the other calculation step, The initial population constituent population includes at least a part of the child population acquired in the one calculation step and a part of the child population acquired in the other calculation step, the one calculation step and The other calculation step is resumed by using the initial population including the initial population constituent individuals extracted in the extraction step.

  According to this feature, in one calculation step and another calculation step, a genetic algorithm is applied according to different setting conditions, and child populations are respectively acquired from the initial population. In addition, in the extraction step, a part (or all) of the initial group used when restarting the one calculation step and the other calculation steps from the child groups acquired in the one calculation step and the other calculation steps. ) Is an initial population constituent population.

  Therefore, diversity can be given to the metal arrangement pattern acquired by applying a genetic algorithm. Further, by providing diversity in the metal arrangement pattern, the antenna characteristics can be sufficiently optimized.

  In addition, when the one setting condition and the other setting condition include at least the size of the antenna element surface or the block size and are different from each other, the size of the antenna element surface is set again and again. Compared with the case where the antenna characteristics are sufficiently optimized for each size, it is possible to easily grasp how far the antenna element surface can be miniaturized.

  That is, the amount of calculation related to the metal arrangement pattern on the antenna element surface provided in the antenna can be suppressed.

  According to a second feature of the present invention, in the first feature of the present invention, the extraction step includes the child group acquired in the one calculation step and the child group acquired in the other calculation step. From the above, the group of the metal arrangement patterns selected in descending order of evaluation regarding the antenna characteristics of the antenna is extracted as an elite group, and the initial group constituent individual group includes the elite group.

  According to a third feature of the present invention, in the second feature of the present invention, the extraction step includes the child group acquired in the one calculation step and the child group acquired in the other calculation step. The group of the metal arrangement pattern selected at random is extracted as a random group, and the initial group constituent individual group includes the random group in addition to the elite group.

  According to a fourth aspect of the present invention, in the first aspect of the present invention, the one calculation step and the other calculation step are based on an evaluation function for evaluating the antenna characteristic of the antenna. Including the step of calculating the evaluation according to each metal arrangement pattern, the type of the evaluation function used in the one calculation step, the evaluation item based on the evaluation function, or the weight value included in the evaluation function, The summary is that the type of the evaluation function used in the calculation step is different from the evaluation value by the evaluation function or the weight value included in the evaluation function.

  A fifth feature of the present invention is summarized in that, in the fourth feature of the present invention, the one calculation step and the other calculation step are performed in parallel by different computers.

  According to a sixth aspect of the present invention, in the first aspect of the present invention, the one calculation step and the other calculation step are based on an evaluation function for evaluating the antenna characteristic of the antenna. The evaluation function used in the one calculation step is used in the other calculation step in terms of the number of frequencies for evaluating the antenna characteristics. It is different from the evaluation function, and the summary is that the one round end condition and the other round end condition are the number of generations determined according to the difference in the number of frequencies for evaluating the antenna characteristics.

  According to a seventh feature of the present invention, in the first feature of the present invention, the antenna has a configuration in which a dielectric is used, and the one calculating step includes an FDTD method in which a wavelength shortening rate is set. And evaluating the antenna characteristics of the antenna for each metal arrangement pattern, and the other calculating step includes evaluating the antenna characteristics of the antenna for each metal arrangement pattern according to a moment method. The gist.

  According to an eighth aspect of the present invention, in the first aspect of the present invention, the one calculation step and the other calculation step are based on an evaluation function for evaluating the antenna characteristic of the antenna. A crossover probability and a mutation probability applied by the genetic algorithm when the one calculation step and the other calculation step are restarted, The gist is that the weight value, the one round end condition, or the other round end condition included in the evaluation function is changed.

  A ninth feature of the present invention is that the antenna design program for designing an antenna provided with an antenna element surface divided into a predetermined number of blocks includes at least the size of the antenna element surface or the size of the block. According to the conditions, a genetic algorithm is applied until one round end condition is satisfied, and an initial group that is a group of metal arrangement patterns indicating whether or not a metal is arranged in the block is changed from the initial group to the group of metal arrangement patterns. Applying a genetic algorithm according to one calculation step of obtaining a child population and other setting conditions including at least the size of the antenna element surface or the size of the block until another round end condition is satisfied, Acquired in another calculation step of acquiring the child group from the initial group and the one calculation step. An extraction step of extracting a predetermined number of groups of the metal arrangement pattern as an initial group constituent individual group from the child group acquired in the child group and the other calculation step, and performing the one calculation The gist of the present invention is that the step and the other calculation step are resumed using the initial population including the initial population constituent individuals extracted in the extraction step.

  A tenth feature of the present invention is an antenna design method for designing an antenna, which applies a genetic algorithm according to a first setting condition until a first round end condition is satisfied, Applying a genetic algorithm until a second round end condition is satisfied in accordance with a first child group acquisition step of acquiring a child group of the first and a second setting condition different from the first setting condition; The second child group acquisition step for acquiring the second child group from the first child group acquisition step and the first child group acquisition step include the first child group acquired in the first child group acquisition step or the second child group acquisition step. The second child group acquired in the child group acquisition step is resumed using the initial group, and the second child group acquisition step includes the first child group acquisition step. And gist to be resumed by using the child set as the initial population.

  According to an eleventh feature of the present invention, in the tenth feature of the present invention, at least one of the first child group acquisition step or the second child group acquisition step ends one round according to one set condition. The genetic algorithm is applied until the condition is satisfied to obtain a child population from the initial population, and the genetic condition is satisfied until another round end condition is satisfied according to another setting condition different from the one setting condition. Applying another algorithm to obtain a child population from an initial population, the child population obtained in the one computation step and the child population obtained in the other computation step, The child group acquired by the extraction step of extracting a predetermined number of initial population constituent individuals is the first or second child population, and the initial population constituent individuals include The child group acquired in the one calculation step and at least a part of the child group acquired in the other calculation step are included, and the one calculation step and the other calculation step include the extraction The gist is to resume using the initial population including the initial population constituent population extracted in the step.

  A twelfth feature of the present invention is an antenna design method for designing an antenna provided with an antenna element surface divided into a predetermined number of blocks, and includes at least the size of the antenna element surface or the size of the block. Applying the genetic algorithm in accordance with a set condition of 1 until a first round end condition is satisfied, the metal placement from an initial population that is a population of metal placement patterns indicating whether or not metal is placed in the block A first child group acquisition step of acquiring a first child group, which is a group of patterns, and a second setting condition different from the first setting condition, including at least the size of the antenna element surface or the size of the block And applying a genetic algorithm until the second round termination condition is satisfied, from the initial population, the metal placement pattern A second child group acquisition step of acquiring a second child group that is a group of the second child group, and the first child group acquisition step include the second child group acquired in the second child group acquisition step The second child population acquisition step is resumed using the first child population acquired in the first child population acquisition step as the initial population. The gist.

  According to this invention, it is possible to design an antenna that simultaneously satisfies desired characteristics even when a single antenna is used with different shapes, structures, and the like.

  According to a thirteenth aspect of the present invention, in the twelfth aspect of the present invention, at least one of the first child group acquisition step or the second child group acquisition step ends one round according to one setting condition. A calculation step of applying a genetic algorithm until a condition is satisfied to obtain a child group that is a group of the metal arrangement patterns from the initial group, and at least the size of the antenna element surface or the size of the block Another calculation step of obtaining the child population from the initial population by applying a genetic algorithm until another round end condition is satisfied according to another setting condition different from the one setting condition; A predetermined number of the metal arrangement patterns are selected from the subpopulations acquired in the calculation step and the subpopulations acquired in the other calculation step. The child group acquired by the extraction step of extracting the group of the first group as an initial group constituent individual is the first or second subpopulation, and the initial group constituent individual is acquired by the one calculation step. At least a part of the child group obtained in the other calculation step, and the one calculation step and the other calculation step include the initial step extracted in the extraction step. The gist is that the initial group including the group constituent population is resumed.

  A fourteenth feature of the present invention is an antenna design method for designing an antenna provided with an antenna element surface divided into a predetermined number of blocks, wherein the size of the antenna element surface or the size of the block is stored in a computer. From an initial population, which is a population of metal placement patterns indicating whether or not metal is placed in the block, by applying a genetic algorithm according to at least a first setting condition that includes, until a first round end condition is met A first child group acquisition step of acquiring a first child group that is a group of the metal arrangement patterns; and a second different from the first setting condition, including at least a size of the antenna element surface or a size of the block And applying a genetic algorithm until the second round end condition is satisfied, from the initial population And a second child group acquisition step of acquiring a second child group that is a group of metal arrangement patterns, wherein the first child group acquisition step is acquired in the second child group acquisition step The second child group is restarted using the second child group as the initial group, and the second child group acquisition step uses the first child group acquired in the first child group acquisition step as the initial group. The main point is that the antenna design program is resumed.

  According to a fifteenth feature of the present invention, in the tenth to thirteenth features of the present invention, the first child acquired in the first child group acquisition step before the first round end condition is resumed. It is higher than the evaluation in the group, and the second round end condition is higher than the evaluation in the second child group acquired in the second child group acquisition step before being resumed. This is the gist.

  The sixteenth feature of the present invention is summarized in that the antenna is designed by the antenna design method described in any one of the first to eighth features and the tenth to thirteenth features of the present invention.

  According to the present invention, when designing a small antenna, an antenna design method capable of suppressing the calculation amount related to the metal arrangement pattern of the antenna element surface provided in the antenna while sufficiently optimizing the antenna characteristics. An antenna design program and an antenna designed by the antenna design method can be provided.

<First Embodiment>
(Configuration of antenna design system)
The configuration of the antenna design system according to the first embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing a configuration of an antenna design system according to the first embodiment of the present invention.

  The antenna designed by the antenna design system is an antenna provided with antenna element surfaces divided into a predetermined number of blocks. An antenna is an antenna having a curved or flat antenna element surface. For example, the antenna is a patch antenna, an inverted F antenna, an inverted L antenna, or a meander line antenna. The antenna element surface is a surface on which a metal (metal patch) that functions as an antenna element is disposed.

  As shown in FIG. 1, the antenna design system includes a plurality of computers 10 (computers 10a to 10h). Each computer 10 applies a genetic algorithm according to setting conditions including at least the size of the antenna element surface and the size of the block, and from the “initial group” that is a group of metal patterns (metal arrangement patterns) arranged in the block. Acquire "child group". Also, different setting conditions are set for each computer 10.

  Specifically, as will be described later, the size of the antenna element surface is determined by the height of the antenna element surface and the width of the antenna element surface, and the setting conditions set for each computer 10 are different from each other. (Or the width of the antenna element surface).

  The size of the block is determined by the height of the reference point block and the width of the reference point block, which will be described later, and the setting conditions set for each computer 10 include the heights of different reference point blocks (or reference points). Block width).

  Here, when the height of the reference point block (or the width of the reference point block) is changed, the height of the line (or the width of the line) is also changed along with this change. In addition, as an example of changing the block size, the size of each block is changed uniformly while maintaining the size ratio between the blocks, and the size ratio of each block is changed by changing the size ratio between the blocks. And the case where the size is changed unevenly.

  Details of the size of the antenna element surface and the size of the block will be described later (FIGS. 5 and 6).

(Computer configuration)
The configuration of the computer according to the first embodiment of the present invention will be described below with reference to the drawings. FIG. 2 is a block diagram showing the computer 10 according to the first embodiment of the present invention. Since the computers 10a to 10h have the same configuration, in the following, these will be collectively described as the computer 10.

  As shown in FIG. 2, the computer 10 includes a communication unit 11, a GA processing unit 12, a setting condition storage unit 13, and an evaluation result calculation unit 14.

  The communication unit 11 is an interface that communicates with another computer 10. Specifically, the communication unit 11 outputs a group of metal arrangement patterns (individuals) acquired by the GA processing unit 12 by applying a genetic algorithm (child group described later) to another computer 10. The communication unit 11 acquires a group of metal arrangement patterns (individuals) acquired by another computer 10 by applying a genetic algorithm (child group described later).

  The GA processing unit 12 calculates an optimal metal arrangement pattern on the antenna element surface based on a genetic algorithm (GA).

  Specifically, the GA processing unit 12 randomly generates individuals having a bit string indicating a metal arrangement pattern, and acquires a predetermined number of individuals as an “initial population”. Further, the GA processing unit 12 selects a plurality of individuals from the individuals included in the “initial population” in order to generate a next generation individual.

  Further, the GA processing unit 12 generates a next generation individual (child) from the selected plurality of individuals (parent). Specifically, the GA processing unit 12 crosses the bits included in the parent individual with a predetermined crossover probability. Note that the number of crossed bits may be one (one point crossing) or two (two point crossing). In addition, the GA processing unit 12 inverts the bit included in the individual that becomes a child with a predetermined mutation probability.

  Further, the GA processing unit 12 repeats selection, crossover, and mutation until a predetermined round end condition is satisfied, with “one generation” being the period from generation of a parent individual to generation of a child individual.

  The predetermined round end condition is that the number of generations processed has become a predetermined number of generations, that the time elapsed since the start of processing has become a predetermined time, and an evaluation to be described later For example, the evaluation related to the antenna characteristic calculated by the result calculation unit 14 is a predetermined evaluation.

  Furthermore, when a predetermined round end condition is satisfied, the GA processing unit 12 instructs the communication unit 11 to output an individual group (hereinafter, “child group”) obtained by calculation by the own computer.

  Further, the GA processing unit 12 selects a predetermined number of individuals from the “child population” acquired by the calculation by the own computer and the “child population” acquired by the calculation by another computer as the “initial population constituent individual group”. Extract as Here, the “initial population constituent population” includes at least a part of “child population” acquired by calculation by the own computer and “child population” acquired by calculation by another computer.

  Further, the GA processing unit 12 resumes selection, crossover, and mutation using the “initial population” including the extracted “initial population constituent population”.

  In addition, the GA processing unit 12 extracts a group of individuals selected in descending order of evaluation regarding antenna characteristics as an “elite group”, and includes “elite group” in the “initial group constituent individual group”.

  The GA processing unit 12 of each computer 10 extracts an “initial group” for the computer 10 in cooperation with each other. Details of this processing will be described later (FIGS. 8 to 10).

  The setting condition storage unit 13 stores the antenna element surface size and the block size as setting conditions. As will be described later, the size of the antenna element surface is the height (HA) of the antenna element surface, and the block size is the height of the reference point block (RZj).

  In addition, the setting condition storage unit 13 stores, as setting conditions, the type of evaluation function for calculating the evaluation related to the antenna characteristics, the evaluation item based on the evaluation function, the weight value included in the evaluation function, and the like.

  The setting condition storage unit 13 of each computer 10 stores different setting conditions, which will be described later (see Tables 1 and 2).

  The evaluation result calculation unit 14 calculates the evaluation related to the antenna characteristics for each individual (metal arrangement pattern) generated by the GA processing unit 12 according to the setting conditions stored in the setting condition storage unit 13.

  Here, an example of the evaluation function described above will be described. In the following description, a case where return loss characteristics and gain characteristics at three frequencies f1, f2, and f3 are calculated as antenna characteristics will be described as an example.

First, the return loss characteristic will be described. Return loss (RL) is expressed by equation (1).

Here, Γ represents a reflection coefficient and is represented by Expression (2).

  Here, Zin is the input impedance, and Z0 is the characteristic impedance of the transmission line.

  The return loss (RLOSS) is a positive value. When the reflection coefficient (Γ) is small, the return loss (RLOSS) increases. When the reflection coefficient (Γ) is large, the return loss (RLOSS) Becomes smaller.

  Next, an evaluation function using return loss characteristics and gain characteristics will be described. In the following, the “return loss” of the three frequencies f1, f2, and f3 is RLf1, RLf2, and RLf3, respectively, and the “gain” of the three desired frequencies f1, f2, and f3 is Gf1, respectively. , Gf2 and Gf3.

The evaluation function (EVAL) is expressed by Expression (3).

  Here, w1 to w6 are weight values multiplied by the return loss or gain.

  For example, when designing a multi-frequency antenna, when optimization is performed for a plurality of mutually related conditions, if one evaluation item satisfies a desired characteristic, the evaluation does not satisfy the other desired characteristic. It is effective to set the weight value so that the weight value becomes heavier for the item.

Furthermore, the other evaluation function (O (x)) is expressed by Expression (4). In the evaluation function (O (x)), the desired resonance frequencies are 900 MHz (fl), 1.5 GHz (fm), and 1.9 GHz (fh), the desired return loss is 10 dB or more, The gain is 3 dBi or more.

  Here, RLfk (x) is a return loss at the frequency fk (k: l, m, h) of the individual x, and Gfk (x) is at the frequency fk (k: l, m, h) of the individual x. It is a gain, and wi (i: 1 to 11) is a weighted value. DRL is a desired return loss, and DGAIN is a desired gain.

  Furthermore, RLflεl (x), RLfmεm (x), and RLfhεh (x) are return losses at frequencies that are εl, εm, and εh away from the desired frequencies (fl, fm, and fh).

  In this way, weighting is performed so that the return loss at a frequency close to a desired frequency is also reduced, thereby avoiding that a desired characteristic appears only at a singular point.

  Also, in the first and second terms, the highest return loss and gain among the three waves (fl, fm, and fh) are given the strongest weight among the three waves (fl, fm, and fh). We choose and weight low return loss and gain.

  As for the third term to the eighth term, the evaluation becomes higher as the return loss and gain at each frequency approach the desired value, and after reaching the desired value, even if the return loss and gain become the desired value or more. It is defined so that the evaluation does not change.

(Antenna configuration)
Hereinafter, the configuration of the antenna according to the first embodiment of the present invention will be described with reference to the drawings. FIG. 3 is a schematic diagram showing the configuration of the antenna 100 according to an embodiment of the present invention.

  As shown in FIG. 3, the antenna 100 includes a scroll-shaped antenna element surface 110 and a plate-shaped ground plane 120. The scroll-shaped antenna element surface 110 is in contact with the ground plane 120 at one side, and is connected to the ground plane 120 by a feeding point 101.

  The antenna element surface 110 is not limited to the scroll shape, but the shape of an inverted F antenna, the shape of an inverted L antenna, a columnar shape, a shape in which the antenna element surface 110 continues on the same plane as the ground plane 120, and the like. It may be.

  FIG. 4 is a diagram showing a final shape of the antenna 100 according to the first embodiment of the present invention. As shown in FIG. 4, the antenna element surface 110 includes a block in which metal is disposed and a block in which metal is not disposed. Note that the metal pattern arranged on the antenna element surface 110 is calculated based on a genetic algorithm, and a method for assigning chromosomes and a calculation method will be described later.

(Configuration of antenna element surface)
Hereinafter, the configuration of the antenna element surface according to the first embodiment of the present invention will be described with reference to the drawings. FIG. 5 is a development view showing the antenna element surface 110 according to the embodiment of the present invention.

  As shown in FIG. 5, the antenna element surface 110 is divided into a plurality of blocks, and each block is in contact with another block on one side instead of a point. Also, on the antenna element surface 110, blocks (hereinafter referred to as reference point blocks) that are used as a reference when the genetic algorithm is applied are arranged, and each reference point block has an identification number of 1 to n. Allocated.

  In addition, as will be described later, a bit string indicating which block metal is to be removed among blocks located around the reference point block (hereinafter referred to as peripheral blocks) is assigned to the reference point block as a chromosome.

  In this embodiment, the peripheral blocks are a block adjacent to the right side of the reference point block, a block on the upper right side of the reference point block, and a block adjacent to the upper side of the reference point block.

  In the present embodiment, a bit string (chromosome) is assigned to the reference point block so that the metal is removed by default, and the metal is not removed by default for the upper right block of the reference point block. A bit string (chromosome) is assigned to. On the other hand, for the block adjacent to the upper side of the reference point block and the block adjacent to the right side of the reference point block, whether or not the metal is removed is determined according to the bit string.

  Thus, the metal arrangement pattern of the antenna element surface 110 is determined by the identification number and the bit string assigned to the reference point block.

  Further, as described above, the antenna element surface 110 is grounded to the ground plane 120 at one side (ground plane grounding area). When there is a block from which the metal is not removed among the lowermost blocks of the ground plane grounding area, the block functions as a short-circuit element that short-circuits between the antenna element surface 110 and the ground plane 120.

  FIG. 6 is a development view showing another antenna element surface 110 according to the first embodiment of the present invention. As shown in FIG. 6, when calculating a metal arrangement pattern based on a genetic algorithm, the setting of the height (RZj) of the reference point block may be freely changed. The line height (LZj) is also changed in accordance with the change in the height (RZj) of the reference point block.

  Further, as with the height of the reference point block, the setting of the width (RXj, RYj) of the reference point block may be freely changed. Note that the line widths (LXj, LYj) are also changed in accordance with the change of the width (RZj) of the reference point block.

  Hereinafter, the bit string (chromosome) assigned to the reference point block according to the first embodiment of the present invention will be described. FIG. 7 is a diagram showing a bit string (chromosome) assigned to the reference point block according to the first embodiment of the present invention.

  As shown in FIGS. 7A to 7D, a 2-bit long bit string is assigned to the reference point block.

  As shown in FIG. 7A, when “00” is assigned to the reference point block, the metal of the adjacent blocks on the upper side and the right side of the reference point block is removed, and as shown in FIG. 7B. When “10” is assigned to the reference point block, the metal of the block adjacent to the right side of the reference point block is removed.

  Similarly, as shown in FIG. 7C, when “01” is assigned to the reference point block, the metal of the block adjacent to the upper side of the reference point block is removed, as shown in FIG. 7D. In addition, when “11” is assigned to the reference point block, the metal of any peripheral block is not removed.

  In this embodiment, since the reference point block is included in the lowermost block of the ground plane grounding area, the block that short-circuits the antenna element surface 110 and the ground plane 120 is calculated based on the genetic algorithm. Therefore, the antenna element surface 110 and the ground plane 120 can be varied with respect to the position and number of short circuits.

(Antenna design method)
Hereinafter, an antenna design method according to a first embodiment of the present invention will be described with reference to the drawings. FIG. 8 is a diagram illustrating an antenna design method according to the first embodiment of the present invention. Here, as shown in FIG. 8, the computers 10a to 10h process selection, crossover, and mutation in parallel.

  Further, since the processing of the computer 10a is the same as the processing of the computer 10b to the computer 10h, step 10a to step 60a will be described as an example.

  In step 10a, the computer 10a randomly generates a bit string indicating a metal arrangement pattern, that is, a combination of bit strings assigned to the reference point block described above as an individual, and sets a predetermined number of individuals as an “initial group a”. get.

  In step 20a, the computer 10a determines whether or not a condition for ending the calculation of the antenna design method (hereinafter, calculation end condition) is satisfied.

  Here, the calculation end condition is that the number of times of step 70 (extraction of “initial population constituent population”) to be described later is a predetermined number of times, and the time that has elapsed since the start of processing is predetermined. And the evaluation related to the antenna characteristics calculated by the evaluation result calculation unit 14 to be described later is a predetermined evaluation.

  The computer 10a ends the process related to the antenna design method when the calculation end condition is satisfied, and proceeds to the process of step 30a when the calculation end condition is not satisfied.

  In step 30a, the computer 10a selects a plurality of individuals (parent individuals) from the individuals included in the “initial population a” in order to generate the next generation individuals (child individuals). Here, when the computer 10a selects a parent individual, it is preferable to select the individual in descending order of evaluation related to antenna characteristics.

  In step 40a, the computer 10a crosses the bits included in the parent individual with a predetermined crossover probability to generate a child individual. As described above, the number of bits to be crossed may be one (one point crossing) or two (two point crossing).

  In step 50a, the computer 10a inverts the bit included in the individual that becomes a child with a predetermined mutation probability.

  In step 60a, the computer 10a determines whether or not the round end condition is satisfied.

  Here, the round end condition is that, as described above, the number of generations processed has become a predetermined number of generations, and the time elapsed since the start of processing has become a predetermined time. The evaluation related to the antenna characteristics calculated by the evaluation result calculation unit 14 described later is a predetermined evaluation.

  Further, the computer 10a proceeds to the process of Step 70 when the round end condition is satisfied, and returns to the process of Step 30a when the round end condition is not satisfied.

  In step 70, the computer 10 a selects “child group a” acquired by the calculation by the own computer 10 a and “child group” (“child group b to child group h” acquired by calculation by the other computer 10. A predetermined number of individuals are extracted as “initial group constituent individual group a.” Further, the computer 10a regenerates “initial group a” including “initial group constituent individual group a” and returns to the processing of step 20a.

(Method of extracting “initial population constituent population”)
Hereinafter, a method for extracting the “initial population constituent population” according to the first embodiment of the present invention will be described with reference to the drawings. 9 and 10 are diagrams showing a method of extracting the “initial population constituent population” described above.

  First, an example of a method for extracting the “initial population constituent population” will be described with reference to FIG.

  As shown in FIG. 9, each computer 10 repeats the processes related to selection, crossover, and mutation until the round end condition is satisfied, whereby “initial group” (“initial group a”) that is a group of metal arrangement patterns. To “initial group h”), a “child group” (“child group a” to “child group h”), which is a group of metal arrangement patterns, is acquired.

  Next, the individual (metal arrangement pattern) selected from the “child population” (“child population a” to “child population h”) acquired by each computer 10 in descending order of evaluation related to the antenna characteristics of the antenna 100. ) Are extracted as “elite group” (“elite group a” to “elite group h”).

  Here, the computer 10 a will be described as an example. The computer 10 a is an “elite group” extracted from each “child group” (“child group a” to “child group h”) acquired by each computer 10. (“Elite group a” to “elite group h”) are collected, and “initial group constituent individual group a” is extracted from the collected “elite group”. In addition, the computer 10a restarts the processes related to selection, crossover, and mutation using the “initial group a” including the “initial group constituent individual group a”.

  Similarly to the computer 10a, the computers 10b to 10h have the "initial population constituent population b" to "initial population constituent population" from the "elite population" ("elite population a" to "elite population h"). h ”is extracted, and“ initial population b ”to“ initial population h ”each including“ initial population constituent population b ”to“ initial population constituent population h ”are used for selection, crossover, and mutation. Such processing is resumed.

  Next, another example of the method for extracting the “initial population constituent population” will be described with reference to FIG.

  As shown in FIG. 10, each computer 10 repeats the processes related to selection, crossover, and mutation until the round end condition is satisfied, whereby “initial group” (“initial group a”) that is a group of metal arrangement patterns. To “initial group h”), a “child group” (“child group a” to “child group h”), which is a group of metal arrangement patterns, is acquired.

  Next, the individual (metal arrangement pattern) selected from the “child population” (“child population a” to “child population h”) acquired by each computer 10 in descending order of evaluation related to the antenna characteristics of the antenna 100. ) Are extracted as “elite group” (“elite group a” to “elite group h”).

  In addition, a group of individuals (metal arrangement patterns) randomly selected from “child groups” (“child group a” to “child group h”) acquired by each computer 10 is a “random group” (“random group”). Group a ”to“ random group h ”).

  Here, the computer 10 a will be described as an example. The computer 10 a is an “elite group” extracted from each “child group” (“child group a” to “child group h”) acquired by each computer 10. (“Elite group a” to “elite group h”) and “random group” (“random group a” to “random group h”) are collected, and the collected “elite group” and “random group” are selected. “Initial population constituent population a” is extracted. In addition, the computer 10a restarts the processes related to selection, crossover, and mutation using the “initial group a” including the “initial group constituent individual group a”.

  Similarly to the computer 10a, the computers 10b to 10h are “elite group” (“elite group a” to “elite group h”) and “random group” (“random group a” to “random group h”). "Initial group constituent individual b" to "Initial group constituent individual h" are respectively extracted from "Initial group constituent individual b" to "Initial group b" The processing relating to selection, crossover, and mutation is resumed using “-“ initial population h ”.

  Note that the number of individuals included in the “random population” extracted from the “child population” acquired by each computer 10 may be the same as or different from each other.

  Further, as the “initial population constituent population”, an “initial population constituent population” common to each computer 10 may be extracted, or a different “initial population constituent population” may be extracted for each computer 10. .

  Furthermore, the “initial population” used when the process related to selection, crossover, and mutation is resumed may be configured only by the “initial population constituent population”, in addition to the “initial population constituent population”. Individuals other than the “initial population constituent population” may be included.

(Setting conditions)
Hereinafter, an example of setting conditions set in each computer 10 will be described with reference to a table. Tables 1 and 2 are tables showing examples of setting conditions according to an embodiment of the present invention.

  As shown in Table 1, in each computer 10, the height (HA) of the antenna element surface 110 is set as the size of the antenna element surface, and the height (RZj) of the reference point block is set as the block size. Yes. Each computer 10 is set with the number of frequencies for evaluating antenna characteristics and the round end condition.

  Further, as shown in Table 2, each computer 10 is set with an evaluation function (for example, the above-described evaluation function (O (x))), and a crossover probability used in calculation by a genetic algorithm. , Mutation probability and random seed are set.

  In this way, different setting conditions are set for each computer 10. Of course, some of the setting conditions set for each computer 10 may be common.

(Action and effect)
According to the antenna design method according to the first embodiment described above, the genetic algorithm is applied according to different setting conditions including at least the size of the antenna element surface and the size of the block. Is acquired. In addition, an “initial group constituent solid group” is extracted from the “child groups” acquired by the respective computers 10. Further, each computer 10 restarts the process related to the genetic algorithm using the “initial population” including the “initial population constituent solid group”.

  Accordingly, diversity can be maintained in the metal arrangement pattern obtained by applying the genetic algorithm.

  Further, by providing diversity in the metal arrangement pattern, the antenna characteristics can be sufficiently optimized.

  Further, the setting conditions set for each computer 10 include at least the size of the antenna element surface and the size of the block, and are different from each other, so that the size of the antenna element surface is set again and again. Compared with the case where the antenna characteristics are sufficiently optimized for each size, it is possible to easily grasp how far the antenna element surface can be reduced.

  That is, the amount of calculation related to the metal arrangement pattern on the antenna element surface provided in the antenna can be suppressed.

  Moreover, when each computer 10 advances a process in parallel, the time until the metal arrangement pattern which has a desired antenna characteristic is calculated can be shortened.

  Furthermore, by suppressing the amount of calculation related to the metal arrangement pattern on the antenna element surface, even if the FDTD (Finite Difference Time Domain) method, which takes time to calculate, is used as an antenna characteristic evaluation method, the metal having the desired antenna characteristics It is possible to prevent the convergence of calculation of the arrangement pattern from becoming extremely slow.

(Example of change)
It is preferable that the type of evaluation function used in each computer 10, the evaluation item based on the evaluation function, and the weight value included in the evaluation function are different from each other. These values used in each computer 10 are different from each other, and by extracting the “initial population constituent population” from the “child population” acquired by each computer 10, a variety of individuals (metal arrangement patterns) can be obtained. Can keep sex.

  In addition, when the round end condition is that the number of processed generations becomes a predetermined number of generations, the predetermined number of generations is preferably determined according to the number of frequencies for evaluating antenna characteristics.

  Since the predetermined number of generations is determined according to the number of frequencies for evaluating the antenna characteristics, it is possible to synchronize the timing at which each computer 10 acquires the “child group” from the “initial group”. When the “initial population constituent population” is extracted from the “child population” acquired by the above, the waiting time of each computer 10 can be shortened.

  Furthermore, when the antenna 100 has a configuration using a dielectric, the FDTD method and the moment method in which the wavelength shortening rate is set are used as the antenna characteristic evaluation method applied to each computer 10. Also good.

  In addition, the crossover and mutation probabilities applied in the genetic algorithm, and the weight value used in the evaluation function are the “children” obtained from the “initial population” in order to maintain the diversity of individuals (metal placement patterns). It may be changed each time “initial population constituent population” is extracted from “population”.

  Further, the antenna design program may cause the computer to execute the antenna design method described above.

  Further, in the above-described embodiment, calculation of individual individuals (metal arrangement patterns) that are children is performed in parallel by a plurality of computers 10, but the present invention is not limited to this, and the calculation result is temporarily stored in a memory or the like. By storing, it may be performed sequentially by one computer 10.

  Further, the round end condition may be common to each computer 10 or may be different for each computer 10.

  Further, the bit string assignment method (chromosome assignment method) indicating whether or not the metal is to be removed is not limited to the above-described embodiment, and various chromosome assignment methods can be applied to one embodiment of the present invention. You may apply to the antenna design method concerning. For example, a 1-bit bit string indicating whether or not to remove metal may be assigned to all blocks.

Example 1
Hereinafter, Example 1 of the present invention will be described with reference to the drawings. FIG. 11 is a diagram illustrating setting conditions set in the computers 10a to 10h.

  In the first embodiment, the number of blocks on the antenna element surface shown in FIGS. 5 and 6 and the bit string assigned to the reference point block shown in FIG.

  As shown in FIG. 11, in each computer 10, the height (HA) of the antenna element surfaces different from each other within the range of 14 mm to 20 mm was set as the size of the antenna element surface.

  In each computer 10, the heights (RZj) of reference point blocks different from each other within the range of the line width (LZj) in the range of 0.85 to 1.25 are set as the block sizes.

  That is, the height (RZj) of the reference point block is expressed as a ratio to the line width (LZj), and the heights (RZj) of the reference point blocks of the computer 10c, the computer 10e, and the computer 10f are actually mutually different. Different.

  The round end condition is a common condition that the number of generations processed is five, and the timing at which each computer 10 acquires the “child group” from the “initial group” is synchronized.

  Further, the crossover probability, mutation probability, and weighting value included in the evaluation function are different for each computer 10 in order to maintain individual diversity.

  FIG. 12 and FIG. 13 are diagrams showing the return loss in each generation for each computer 10. As shown in FIGS. 12 and 13, the return loss in each generation is close to the optimum value even when it exceeds 100 generations, and it has been confirmed that the diversity of calculation by each computer 10 is maintained.

  In addition to this, as shown in FIG. 12, the return loss of the individual (metal arrangement pattern) calculated by the computer 10a in which the largest height (20 mm) is set as the height (HA) of the antenna element surface is It was confirmed that the desired level (−10 dB or less) was reached from the beginning of the genetic algorithm.

  The return loss of the individual (metal arrangement pattern) calculated by the computer 10b in which the second largest height (18 mm) is set as the height (HA) of the antenna element surface is 75 generations after the genetic algorithm is started. It was confirmed that the desired level (−10 dB or less) was reached.

  The return loss of the individual (metal arrangement pattern) calculated by the computer 10c in which the third largest height (16 mm) is set as the height (HA) of the antenna element surface is 270 generations after the genetic algorithm is started. It was confirmed that the desired level (−10 dB or less) was reached.

  The return loss of the individual (metal arrangement pattern) calculated by the computer 10d in which the smallest height (14mm) is set as the height (HA) of the antenna element surface exceeds 300 generations after the genetic algorithm is started. However, it was confirmed that the desired level (−10 dB or less) was not reached.

  On the other hand, as shown in FIG. 13, in the computer 10e in which the same height (20 mm) as that of the computer 10a is set as the height (HA) of the antenna element surface, the height of the reference point block (RZj) is different from that of the computer 10a. It was confirmed that the return loss of the individual (metal arrangement pattern) did not reach the desired level (−10 dB or less) even after 300 generations.

  In the computer 10f in which the same height (18 mm) as that of the computer 10b is set as the height (HA) of the antenna element surface, even if the height of the reference point block (RZj) is different from the computer 10b, the individual (metal arrangement pattern) ) Was confirmed to reach a desired level (−10 dB or less) in about 75 generations after the genetic algorithm was started.

  In the computer 10g in which the same height (16 mm) as the computer 10c is set as the height (HA) of the antenna element surface, if the height of the reference point block (RZj) is different from that of the computer 10c, the individual (metal arrangement pattern) ) Return loss does not reach the desired level (−10 dB or less) even after 300 generations.

  In the computer 10h in which the same height (14mm) as that of the computer 10d is set as the height (HA) of the antenna element surface, even if the height of the reference point block (RZj) is different from the computer 10d, the individual (metal arrangement pattern) ) Return loss does not reach the desired level (−10 dB or less) even after 300 generations.

  FIG. 14 is a diagram showing the return loss of an individual (metal arrangement pattern) calculated by the computer 10c. As shown in FIG. 12, it was confirmed that the return loss at the desired three frequencies reached a desired level (−10 dB or less).

  As described above, among the computers in which the return loss of the individual (metal arrangement pattern) has reached a desired level (−10 dB or less), the smallest height is set as the height (HA) of the antenna element surface. The computer was confirmed to be the computer 10c.

  In this way, even in a situation where it is not known how far the size of the antenna element surface can be reduced, different antenna element surface sizes and different reference point block heights (RZj) By setting to the computer 10, an optimal (minimum) antenna element surface having desired characteristics could be confirmed.

  This is because the “child population” (“child population a” to “child population h”) acquired by each computer 10 in which different values are set as the size of the antenna element surface and the height (RZj) of the reference point block. It is considered that a variety of calculations by each computer 10 is achieved by restarting the calculation related to the genetic algorithm using the “initial population constituent population” including a part.

(Example 2)
Hereinafter, Example 2 of the present invention will be described with reference to the drawings. FIG. 15 is a diagram illustrating setting conditions set in the computers 10a to 10h.

  In the second embodiment, the number of blocks on the antenna element surface shown in FIGS. 5 and 6 and the bit string assigned to the reference point block shown in FIG.

  As shown in FIG. 15, a common value (20 mm) was set as the height (HA) of the antenna element surface for each computer 10. In each computer 10, the height (RZj) of different reference point blocks within the range of 0.85 to 1.25 in the index display is set as the block size.

  Further, in the computers 10a to 10d, “3” is set as the number of frequencies for evaluating antenna characteristics, and the fact that the number of processed generations is 10 generations is set as a round end condition.

  On the other hand, in the computers 10e to 10h, “6” is set as the number of frequencies for evaluating the antenna characteristics, and the fact that the number of processed generations is 5 is set as a round end condition.

  In this way, each computer 10 obtains a “child population” from the “initial population” by determining the number of generations for repeating the processes related to selection, crossover, and mutation in accordance with the number of frequencies for evaluating antenna characteristics. The timing was synchronized.

  FIG. 16 is a diagram showing the return loss of the optimum individual among the individual (metal arrangement pattern) calculated by each computer. As shown in FIG. 16, it was confirmed that the return loss at the desired three frequencies reached a desired level (−10 dB or less). Thus, in the second embodiment, the height (RZj) of the reference point block at which the return loss at the desired three frequencies reaches a desired level (−10 dB or less) can be found.

<Second Embodiment>
A second embodiment of the present invention will be described with reference to FIGS. Hereinafter, the second embodiment of the present invention will be described mainly with respect to differences from the first embodiment described above.

  The antenna design method according to the present embodiment includes an “open state (that is, a state where the metal plates 120A and 120B of the mobile communication terminal are opened)” as shown in FIG. 17A and a state shown in FIG. The antenna for a mobile communication terminal that is assumed to be used in a different state such as a “close state (that is, a state where the metal plates 120A and 120B of the mobile communication terminal are closed)” is designed to satisfy the design conditions. Is to do.

  In addition, the antenna design method according to the present embodiment is a design condition for an antenna for a mobile communication terminal that is assumed to be used in different shapes such as individual variations resulting from differences in soldering methods and materials. Is designed to satisfy the above.

  18, in the mobile communication terminal in which the antenna 100 is installed near the hinge 5, as shown in FIG. 17, the return loss (open state and close state) in the antenna 100 having the metal arrangement pattern calculated by the conventional antenna design method. For each).

  As shown in FIG. 18, in the antenna 100 having the metal arrangement pattern calculated by the conventional antenna design method, there is a shift between the resonance frequency in the “open state” and the resonance frequency in the “close state”. Can be confirmed.

  As shown in FIG. 19, the antenna design method according to the present embodiment designs an antenna 100 provided with antenna element surfaces divided into a predetermined number of blocks.

  In the antenna design method according to the present embodiment, the “first child group” calculated and acquired as the mobile communication terminal being in the “open state”, and the mobile communication terminal being in the “close state” The “second child group” calculated and acquired as the mobile communication terminal being in the “colle state” is evaluated as the mobile communication terminal being in the “open state”. Then, by repeating this operation, the antenna 100 that satisfies the desired characteristics is designed in both the “open state” and the “colle state”.

  Hereinafter, an antenna design method according to a second embodiment of the present invention will be described with reference to the drawings. FIG. 20 is a diagram illustrating an antenna design method according to the second embodiment of the present invention. Here, as shown in FIG. 20, the computer 10a and the computer 10b process selection, crossover, and mutation in parallel. The number of computers 10 that perform processing in parallel may be two or more.

  In step 10a, the computer 10a randomly generates a bit string indicating a metal arrangement pattern, that is, a combination of bit strings assigned to the reference point block described above as an individual, and sets a predetermined number of individuals as an “initial group a”. get.

  In step 20a, the computer 10a determines whether or not a condition for ending the calculation of the antenna design method (hereinafter, calculation end condition) is satisfied.

  Here, the calculation end condition is that the number of times of step 71a (extraction of “child group a”), which will be described later, is a predetermined number of times, and the time that has elapsed since the start of the processing is a predetermined time. That the evaluation related to the antenna characteristics calculated by the evaluation result calculation unit 14 described later is a predetermined evaluation.

  The computer 10a ends the process related to the antenna design method when the calculation end condition is satisfied, and proceeds to the process of step 30a when the calculation end condition is not satisfied.

  In step 30a, the computer 10a selects a plurality of individuals (parent individuals) from the individuals included in the “initial population a” in order to generate the next generation individuals (child individuals). Here, when the computer 10a selects a parent individual, it is preferable to select the individual in descending order of evaluation related to antenna characteristics.

  In step 40a, the computer 10a crosses the bits included in the parent individual with a predetermined crossover probability to generate a child individual. As described above, the number of bits to be crossed may be one (one point crossing) or two (two point crossing).

  In step 50a, the computer 10a inverts the bit included in the individual that becomes a child with a predetermined mutation probability.

  In step 60a, the computer 10a determines whether or not the round end condition is satisfied.

  Here, the round end condition is that, as described above, the number of generations processed has become a predetermined number of generations, and the time elapsed since the start of processing has become a predetermined time. The evaluation related to the antenna characteristics calculated by the evaluation result calculation unit 14 described later is a predetermined evaluation.

  If the round end condition is satisfied, the computer 10a proceeds to step 71a. If the round end condition is not satisfied, the computer 10a returns to step 30a.

  In step 71a, the computer 10a acquires the “child group a” acquired by the calculation by the own computer 10a.

  In step 10b, the computer 10b performs steps 10b to 71b using the first child group acquired by the calculation by the computer 10a as the initial group b, and calculates and acquires the second child group. .

  Thereafter, in step 10a, the computer 10a restarts steps 10a to 71a using the second child group acquired by the calculation by the computer 10b as the initial group a.

  In step 10a, the computer 10a uses the first child group acquired by the calculation by the computer 10a, not the second child group acquired by the calculation by the computer 10b, as the initial group a. 71a may be resumed (see FIGS. 27 to 31).

  That is, in the antenna design method according to the present embodiment, as shown in FIG. 21, in round 1, the computer 10a uses the initial group a generated randomly, and the antenna 100 in the mobile communication terminal in the “close” state. (That is, the child group a estimated to be an optimal metal arrangement pattern is obtained), and the computer 10b uses the randomly generated initial group b to carry the mobile phone in the “open” state. The antenna 100 in the communication terminal is optimized (that is, a child group b estimated to be an optimal metal arrangement pattern is acquired).

  Specifically, as in the first embodiment described above, in round 1, the computer 10a determines that the end condition (first step) of round 1 is in accordance with the first setting condition including at least the size of the antenna element surface or the size of the block. A genetic algorithm is applied until the first round end condition) is satisfied, and an initial group a, which is a group of metal arrangement patterns indicating whether or not metal is arranged in such a block, A certain first child group is acquired.

  Similarly, in round 1, the computer 10b includes at least the size of the antenna element surface or the size of the block, and according to a second setting condition different from the first setting condition described above, the end condition of the round 1 (second round A genetic algorithm is applied until the end condition is satisfied, and a second child group that is a group of metal arrangement patterns is obtained from the initial group b.

  In round 2, the computer 10a uses the child group b acquired in round 1 as the initial group a to optimize the antenna 100 in the mobile communication terminal in the “close” state (that is, with an optimal metal arrangement pattern). And the computer 10b uses the child group a acquired in round 1 as the initial group b to optimize the antenna 100 in the mobile communication terminal in the “open” state. (In other words, the child group b estimated to be the optimum metal arrangement pattern is acquired).

  That is, in round 2, the computer 10a uses the child group b acquired in round 1 as the initial group a, and resumes the same operation as in round 1, and the computer 10b acquires the child group a acquired in round 1. Is used as the initial group b, and the same operation as in round 1 is resumed.

  In round 3 and later, by repeating this operation until the calculation termination condition in the antenna design method according to the present embodiment is satisfied, individuals (metals) satisfying desired characteristics in different states such as “open state” and “close state”, respectively. Arrangement pattern) can be acquired.

  In each round, at least one of the computer 10a or the computer 10b acquires the optimum metal arrangement pattern acquired by the antenna design method according to the first embodiment as the child group a or the child group b. It may be configured.

  That is, in each round, at least one of the computer 10a or the computer 10b applies a genetic algorithm in step (1) according to one set condition until one round end condition is satisfied, and A child group that is a group of metal arrangement patterns is acquired, and in step (2), another round end condition is determined according to another setting condition that includes at least the size of the antenna element surface or the block size and is different from the one setting condition. Applying another genetic algorithm until satisfied to obtain a child population from the initial population, and in step (3), the child population obtained in step (1) and obtained in step (2) A group of a predetermined number of metal arrangement patterns is extracted from the child group as an initial group constituent individual group. Child population acquisition it may be above the child population a (first child set) or child population b (second child population).

  Here, the initial group constituent population includes at least a part of the child group acquired in step (1) and a part of the child group acquired in step (2). ) Is resumed using the initial population including the initial population constituent population extracted in step (3).

  Further, in each round, if the round end condition in steps 60a and 60b in FIG. 20 is determined by the number of generations and the number of generations is set to a small value, it is possible to quickly create an individual (child population) in each round. However, if the number of generations is too small, data will only be exchanged between computers, and optimization of the child population may not proceed (it will not evolve).

  Therefore, as a means for eliminating this point, in each round, the evaluation value of the child group in each computer is improved as compared with the evaluation value in the previous round as a round end condition in steps 60a and 60b in FIG. Also good.

(Example of change)
The antenna design method described above is not limited to designing an antenna provided with antenna element surfaces divided into a predetermined number of blocks, and can be applied to designing an antenna having an arbitrary structure.

  For example, the antenna design method described above can also be applied to designing an antenna whose structure is changed so that the mobile communication terminal opens and closes and the antenna of the mobile communication terminal expands and contracts.

(Example 3)
Next, Example 3 will be described with reference to FIGS.

  That is, as shown in FIG. 22, Example 3 in the case where the antenna 100 designed using the antenna design method according to the second embodiment described above is installed at the top position of the mobile communication terminal will be described.

  FIG. 23 shows the case of the “open state” and the “close state” for the antenna 100 having the antenna element surface of the final metal arrangement pattern calculated by the antenna design method according to the second embodiment of the present invention. The frequency characteristics of the return loss are obtained.

  In the example of the antenna according to the prior art shown in FIG. 18, the characteristics of the “close state” are deteriorated, whereas in the case of the antenna designed by the antenna design method according to the present embodiment shown in FIG. Satisfies the desired characteristic of “return loss is −7 dB or less” in both the “open state” and the “close state”.

  Here, the frequency ratio was set to 857.5: 1795: 2045 as in the case of FIG.

  FIG. 24 shows a change in return loss for each generation for an antenna designed by the antenna design method according to the second embodiment of the present invention.

  Here, in the antenna design method according to the second embodiment of the present invention, the calculation in the “open state” that satisfies the desired characteristics at the three desired frequencies is started as the initial group in the “close state”. The number of groups (population) was “30”, and the number of generations in each round was “2”.

  Originally, individuals (metal arrangement patterns) that are highly evaluated in the “open state” are used as the initial group, so the characteristics of the antenna increase or decrease every two generations, that is, each time the round changes.

  That is, when the excellent individuals in the “close state” and the “open state” are substituted into the “close state” and the “open state” in the next round, the characteristics deteriorate.

  It can be seen that the worst value of the return loss gradually decreases by repeating this operation many times.

  When the worst value of the return loss in both the “close state” and the “open state” became −6 dB or less, the calculation was completed in about 30 generations.

  Here, since the population for each computer is “30”, the individual with the highest evaluation value of the final generation is not necessarily the same in the two computers.

  Note that the data in FIG. 23 is obtained by calculating the characteristics in the “close state” and the “open state” by adopting the best individual (metal arrangement pattern) in the “close state” on the antenna element surface. It is a thing.

Example 4
Next, Example 4 will be described with reference to FIGS. 25 and 26.

  That is, as shown in FIG. 17, a description will be given of a fourth embodiment in which the antenna 100 designed using the antenna design method according to the second embodiment described above is installed near the hinge 5 in the mobile communication terminal.

  FIG. 25 shows the antenna design method according to the second embodiment described above, in both the “close state” and the “open state” for the antenna 100 installed near the hinge 5 as shown in FIG. This is an example when applied to a design for satisfying the desired characteristics, and shows how the return loss changes for each generation.

  In Example 4, the calculation in each round was 10 generations. A setting was adopted in which an individual group that satisfies a desired characteristic in the “open state” is set as an initial group in the “close state”.

  Focusing on the return loss characteristics in the “close state”, each round operates until the round end condition is reached (every 10 generations), so that the worst value of the return loss at the three frequencies is reduced. If changed, the data optimized in the “open state” is replaced, so that the worst value of the return loss is pulled back to the worse one.

  However, as the rounds are repeated, it can be confirmed that the worst value of the return loss gradually pulled back converges to become smaller.

  On the other hand, paying attention to the return loss characteristic in the “open state”, since the initial group in the “close state” is originally the one with high characteristics in the “open state”, the initial group evolved in round 1 Since an object is included as an initial group in the “open state”, the characteristics at the beginning are good, and the characteristics are slightly deteriorated once by being merged with the excellent value of the return loss in the “close state”.

  In FIG. 26, the antenna 100 is formed using the final individual in the “close state” in the fourth embodiment, and the return loss of the antenna 100 is compared between the case of the “open state” and the case of the “close state”. Shows frequency characteristics.

It can be confirmed that the area is obtained.

  In the case of the antenna designed by the antenna design method according to the present embodiment shown in FIG. 26, the desired characteristic that “return loss is −7 dB or less” in both the “open state” and the “close state”. Meet.

It is a figure which shows the structure of the antenna design system which concerns on the 1st Embodiment of this invention. It is a block diagram showing computer 10 concerning a 1st embodiment of the present invention. It is a figure which shows the antenna 100 which concerns on the 1st Embodiment of this invention (the 1). It is a figure which shows the antenna 100 which concerns on the 1st Embodiment of this invention (the 2). It is a development view showing antenna element surface 110 concerning a 1st embodiment of the present invention (the 1). It is a development view showing antenna element surface 110 concerning a 1st embodiment of the present invention (the 2). It is a figure which shows the bit string allocated to the reference point block which concerns on the 1st Embodiment of this invention. It is a figure which shows the antenna design method which concerns on the 1st Embodiment of this invention. It is a figure which shows the extraction method of the "initial population constituent population" which concerns on the 1st Embodiment of this invention (the 1) It is a figure which shows the extraction method of the "initial population constituent individual population" which concerns on the 1st Embodiment of this invention (the 2). It is a figure which shows the setting conditions set to the computer 10a-the computer 10h which concern on Example 1. FIG. It is a figure which shows the return loss in each generation which concerns on Example 1 for every computer 10 (the 1). It is a figure which shows the return loss in each generation which concerns on Example 1 for every computer 10 (the 2). It is a figure which shows the return loss of the individual | organism | solid (metal arrangement pattern) calculated by the computer 10c which concerns on Example 1. FIG. It is a figure which shows the setting conditions set to the computer 10a-the computer 10h which concern on Example 2. FIG. It is a figure which shows the return loss of the individual | organism | solid (metal arrangement | positioning pattern) calculated by the computer 10c which concerns on Example 2, and the computer 10g. 1 is an overall configuration diagram of a general mobile communication terminal including an antenna. It is a figure which shows the return loss of the individual | organism | solid (metal arrangement | positioning pattern) calculated by the conventional computer 10c in the portable communication terminal which installed the antenna near the hinge. It is a figure for demonstrating the concept of the antenna design method which concerns on the 2nd Embodiment of this invention. It is a figure which shows the antenna design method which concerns on the 2nd Embodiment of this invention (the 1). It is a figure which shows the antenna design method which concerns on the 2nd Embodiment of this invention (the 2). FIG. 6 is an overall configuration diagram of a general mobile communication terminal including an antenna according to a third embodiment. It is a figure which shows the return loss of the individual | organism | solid (metal arrangement | positioning pattern) calculated by the computer 10c which concerns on Example 3. FIG. It is a figure which shows the return loss in each generation which concerns on Example 3 for every computer 10. FIG. It is a figure which shows the return loss in each generation which concerns on Example 4 for every computer 10. FIG. It is a figure which shows the return loss of the individual | organism | solid (metal arrangement pattern) calculated by the computer 10c which concerns on Example 4. FIG. It is a figure for demonstrating the concept of the antenna design method which concerns on the example of a change of this invention. It is a figure for demonstrating the concept of the antenna design method which concerns on the example of a change of this invention. It is a figure for demonstrating the concept of the antenna design method which concerns on the example of a change of this invention. It is a figure for demonstrating the concept of the antenna design method which concerns on the example of a change of this invention. It is a figure for demonstrating the concept of the antenna design method which concerns on the example of a change of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Computer 11 ... Communication part 12 ... GA process part 13 ... Setting condition memory | storage part 14 ... Evaluation result calculation part 100 ... Antenna 101 ... Feeding point 110 ... Antenna element surface 120 ... Ground plane

Claims (16)

An antenna design method for designing an antenna provided with an antenna element surface divided into a predetermined number of blocks,
According to one setting condition including at least the size of the antenna element surface or the size of the block, a genetic algorithm is applied until one round end condition is satisfied to indicate whether or not metal is arranged in the block One calculation step of acquiring a child group which is a group of the metal arrangement pattern from an initial group which is a group of metal arrangement patterns;
Applying a genetic algorithm from the initial population until the other round end condition is satisfied, according to another set condition different from the one set condition, including at least a size of the antenna element surface or the block. Other calculation steps to get the child population,
An extraction step of extracting a predetermined number of groups of the metal arrangement patterns as an initial population constituent population from the child population acquired in the one calculation step and the child population acquired in the other calculation step; Including
The initial population constituent population includes at least a part of the child population obtained in the one calculation step and the child population obtained in the other calculation step,
The antenna design method according to claim 1, wherein the one calculation step and the other calculation step are resumed by using the initial population including the initial population constituent individuals extracted in the extraction step. In the extraction step, the metal selected in descending order of evaluation regarding the antenna characteristics of the antenna from the subpopulation acquired in the one calculation step and the subpopulation acquired in the other calculation step. Extract a group of placement patterns as an elite group,
The antenna design method according to claim 1, wherein the initial population constituent population includes the elite population. In the extraction step, a group of metal arrangement patterns selected at random is extracted as a random group from the child group acquired in the one calculation step and the child group acquired in the other calculation step. And
The antenna design method according to claim 2, wherein the initial population constituent population includes the random population in addition to the elite population. The one calculating step and the other calculating step include a step of calculating an evaluation related to the antenna characteristic for each metal arrangement pattern based on an evaluation function for evaluating the antenna characteristic of the antenna,
The type of the evaluation function used in the one calculation step, the evaluation item based on the evaluation function, or the weight value included in the evaluation function depends on the type of the evaluation function used in the other calculation step and the evaluation function. The antenna design method according to claim 1, wherein the antenna design method is different from an evaluation item or a weight value included in the evaluation function.   The antenna design method according to claim 1, wherein the one calculation step and the other calculation step are performed in parallel by different computers. The one calculating step and the other calculating step include a step of calculating an evaluation related to the antenna characteristic for each metal arrangement pattern based on an evaluation function for evaluating the antenna characteristic of the antenna,
The evaluation function used in the one calculation step is different from the evaluation function used in the other calculation step in terms of the number of frequencies for evaluating the antenna characteristics.
6. The antenna design method according to claim 5, wherein the one round end condition and the other round end condition are generation numbers determined according to a difference in the number of frequencies for evaluating the antenna characteristics. The antenna has a configuration in which a dielectric is used,
The one calculating step includes a step of evaluating the antenna characteristics of the antenna for each metal arrangement pattern according to the FDTD method in which a wavelength shortening rate is set,
The antenna design method according to claim 1, wherein the other calculation step includes a step of evaluating antenna characteristics of the antenna for each metal arrangement pattern according to a moment method. The one calculating step and the other calculating step include a step of calculating an evaluation related to the antenna characteristic for each metal arrangement pattern based on an evaluation function for evaluating the antenna characteristic of the antenna,
When the one calculation step and the other calculation step are restarted, the crossing probability and mutation probability applied by the genetic algorithm, the weight value included in the evaluation function, the one round end condition or The antenna design method according to claim 1, wherein the other round end condition is changed. An antenna design program for designing an antenna provided with an antenna element surface divided into a predetermined number of blocks,
According to one setting condition including at least the size of the antenna element surface or the size of the block, a genetic algorithm is applied until one round end condition is satisfied to indicate whether or not metal is arranged in the block One calculation step of acquiring a child group which is a group of the metal arrangement pattern from an initial group which is a group of metal arrangement patterns;
Applying a genetic algorithm from the initial population until the other round end condition is satisfied, according to another set condition different from the one set condition, including at least a size of the antenna element surface or the block. Other calculation steps to get the child population,
An extraction step of extracting a predetermined number of groups of the metal arrangement patterns as an initial population constituent population from the child population acquired in the one calculation step and the child population acquired in the other calculation step; And execute
The initial population constituent population includes at least a part of the child population obtained in the one calculation step and the child population obtained in the other calculation step,
The antenna design program, wherein the one calculation step and the other calculation step are resumed by using the initial group constituent population extracted in the extraction step. An antenna design method for designing an antenna,
Applying a genetic algorithm according to a first setting condition until a first round end condition is satisfied to obtain a first child population from the initial population;
According to a second setting condition that is different from the first setting condition, a second child population acquisition that acquires a second child population from an initial population by applying a genetic algorithm until a second round end condition is satisfied Steps,
In the first child group acquisition step, the first child group acquired in the first child group acquisition step or the second child group acquired in the second child group acquisition step is used as the initial stage. Resumed as a group,
The antenna design method, wherein the second child group acquisition step is restarted using the first child group acquired in the first child group acquisition step as the initial group. At least one of the first child group acquisition step or the second child group acquisition step is:
Applying a genetic algorithm according to one set condition until a round end condition is met to obtain a child population from the initial population;
Another calculation step of obtaining a child population from the initial population by applying a genetic algorithm in accordance with another setting condition different from the one setting condition until another round end condition is satisfied;
The child population obtained by extracting a predetermined number of initial population constituent individuals from the child population obtained in the one calculation step and the child population obtained in the other calculation step. As the first or second child group,
The initial population constituent population includes at least a part of the child population obtained in the one calculation step and the child population obtained in the other calculation step,
The antenna design according to claim 10, wherein the one calculation step and the other calculation step are resumed by using the initial population including the initial population constituent population extracted in the extraction step. Method. An antenna design method for designing an antenna provided with an antenna element surface divided into a predetermined number of blocks,
Whether a metal is arranged in the block by applying a genetic algorithm according to a first setting condition including at least a size of the antenna element surface or a size of the block until a first round end condition is satisfied A first child group acquisition step of acquiring a first child group that is a group of the metal arrangement pattern from an initial group that is a group of metal arrangement patterns indicating:
Applying a genetic algorithm until a second round end condition is satisfied according to a second setting condition different from the first setting condition, including at least a size of the antenna element surface or the block; A second child group acquisition step of acquiring a second child group that is a group of the metal arrangement pattern from a group;
In the first child group acquisition step, the first child group acquired in the first child group acquisition step or the second child group acquired in the second child group acquisition step is used as the initial stage. Resumed as a group,
The antenna design method, wherein the second child group acquisition step is restarted using the first child group acquired in the first child group acquisition step as the initial group. At least one of the first child group acquisition step or the second child group acquisition step is:
According to one set condition, applying a genetic algorithm until one round end condition is satisfied to obtain a child population that is a group of the metal arrangement pattern from the initial population;
Applying a genetic algorithm from the initial population until the other round end condition is satisfied, according to another set condition different from the one set condition, including at least a size of the antenna element surface or the block. Other calculation steps to get the child population,
An extraction step of extracting a predetermined number of groups of the metal arrangement patterns as an initial population constituent population from the child population acquired in the one calculation step and the child population acquired in the other calculation step; The child population obtained by the above-mentioned first or second child population,
The initial population constituent population includes at least a part of the child population obtained in the one calculation step and the child population obtained in the other calculation step,
The antenna design according to claim 12, wherein the one calculation step and the other calculation step are resumed using the initial population including the initial population constituent population extracted in the extraction step. Method. An antenna design method for designing an antenna provided with an antenna element surface divided into a predetermined number of blocks, the computer,
Whether a metal is arranged in the block by applying a genetic algorithm according to a first setting condition including at least a size of the antenna element surface or a size of the block until a first round end condition is satisfied A first child group acquisition step of acquiring a first child group that is a group of the metal arrangement pattern from an initial group that is a group of metal arrangement patterns indicating:
Applying a genetic algorithm until a second round end condition is satisfied according to a second setting condition different from the first setting condition, including at least a size of the antenna element surface or the block; A second child group acquisition step of acquiring a second child group that is a group of the metal arrangement pattern from a group;
The first child group acquisition step is resumed using the second child group acquired in the second child group acquisition step as the initial group,
The antenna design program, wherein the second child group acquisition step is resumed using the first child group acquired in the first child group acquisition step as the initial group. The first round end condition is higher than the evaluation in the first child group acquired in the first child group acquisition step before being resumed,
14. The second round end condition is higher than the evaluation in the second child group acquired in the second child group acquisition step before resumption. The antenna design method according to any one of the above. An antenna designed by the antenna design method according to any one of claims 1 to 8 and claims 10 to 14.

Images