A metasurface beamforming optimization method based on a hybrid genetic binary dragonfly algorithm

Abstract

The application discloses a metasurface beamforming optimization method based on a hybrid genetic binary dragonfly algorithm (GABDA), and belongs to the technical field of computer and application, which comprises that a metasurface array is composed of 8*8 super subunits, each super subunit is composed of 6*6 units with a phase of 0° or 180°, the phase is a relative value, that is, the phase difference of two units is 180°, and the array adopts the hybrid genetic binary dragonfly algorithm to optimize a set specific waveform, in order to quickly realize a preset waveform, the metasurface units are arranged in a square vertical symmetry, and the hybrid genetic binary dragonfly algorithm is used to optimize the metasurface units. Compared with a traditional algorithm, the application has the advantages of fast convergence, high precision, and difficulty in falling into local optimization, and has a potential application prospect for the development of intelligent metasurfaces.

Description

Technical Field

[0001] This invention relates to a metasurface beamforming optimization method based on a hybrid genetic binary dragonfly algorithm, belonging to the field of computer science and technology and application technology. Background Technology

[0002] With the development of modern communication technology, the design of intelligent metasurfaces is becoming increasingly important in both military and daily life. As the size of coded metasurface arrays expands, the arrangement of the array surfaces grows exponentially, making manual design extremely limited. However, by combining intelligent optimization algorithms, it is possible to design specific and efficient metasurfaces that meet specific needs. Intelligent metasurfaces are currently a hot research topic, but as design requirements continue to increase, traditional intelligent optimization algorithms are no longer sufficient for achieving both high efficiency and high precision. Summary of the Invention

[0003] The purpose of this invention is to address the shortcomings and deficiencies of the existing technology by proposing a metasurface beamforming optimization method based on a hybrid genetic binary dragonfly algorithm. By arranging metasurface units in a square and optimizing them using the hybrid genetic binary dragonfly algorithm, this invention has advantages such as fast convergence, high accuracy, and less susceptibility to local optima compared to traditional algorithms, and has potential application prospects for the development of intelligent metasurfaces.

[0004] The solution adopted by this invention to solve the above-mentioned technical problems is: a metasurface beamforming optimization method based on a hybrid genetic binary dragonfly algorithm. The metasurface array is composed of 8*8 super sub-units, and each super sub-unit consists of 6*6 units with phases of all 0° or all 180°. Here, the phase is a relative value, that is, the phase difference between two units is 180°. The square metasurface array is beamform optimized using a hybrid genetic binary dragonfly algorithm, specifically including the following steps:

[0005] Step 1: Construct an initial population describing the metasurface array and measure the optimal values;

[0006] Step 2: Perform one iteration based on the initial population, carrying out selection, crossover, and mutation operations, and measuring the optimal value;

[0007] Step 3: Compare the measurement results of Step 2 with the measurement results of the previous generation. If the optimal value has changed, return to Step 2 and repeat Step 2 and Step 3 to output the optimal value and the optimal population. If the optimal value has not changed, proceed to the next step.

[0008] Step 4: Based on the population from Step 3 above, sort the optimal values. Perform binary dragonfly algorithm iteration on the first 90% and genetic mutation operation on the last 10% and measure the optimal value.

[0009] Step 5: Compare the measurement results of Step 4 with the measurement results of the previous generation. If the optimal value has changed, repeat Step 2 and Step 3 to output the optimal value and the optimal population. If the optimal value has not changed, perform Step 4 and Step 5 to output the optimal value and the optimal population.

[0010] Furthermore, the method also includes first optimizing the population of the binary dragonfly algorithm using a genetic algorithm, so that it can converge quickly.

[0011] Furthermore, the binary dragonfly algorithm can improve the optimization accuracy of the genetic algorithm.

[0012] Furthermore, the method also includes using the mutation concept from genetic algorithms to address the problem of the binary dragonfly algorithm getting stuck in a local optimum when there are no individuals in the neighborhood.

[0013] Furthermore, the size of the metasurface array is not unique, making it highly adaptable.

[0014] Furthermore, the fitness function of the hybrid genetic binary dragonfly algorithm is:

[0015] Fitness = w1 * F1 + w2 * F2

[0016]

[0017]

[0018]

[0019]

[0020] Where Fitness is the fitness function, w1 and w2 are weighting coefficients, and F lowerMask F UpperMask These represent the self-defined lower bound function and upper bound function, respectively, where i and j are the discrete sample indices. For pattern functions, The function representing the normalized radiation pattern, θ, These are the pitch and azimuth angles, respectively, with the positive directions of the y-axis and x-axis as reference directions. k0 is the wave vector value, M and N represent the number of rows and columns of the metasurface arrangement, and A... mn , D represents the amplitude and phase corresponding to the (m, n)th unit structure, respectively. x D y These represent the periods in the coordinate axes of the unit cell.

[0021] Furthermore, the crossover probability of the algorithm is 0.9, the generation skipping rate is 0.9, the mutation probability in the genetic algorithm is 0.08, and the mutation probability in the binary dragonfly algorithm is 0.2.

[0022] Beneficial effects:

[0023] 1. This invention further improves upon the traditional genetic algorithm and binary dragonfly algorithm, and has the advantages of fast convergence, high accuracy, and less susceptibility to getting trapped in local optima.

[0024] 2. This invention has potential application prospects for the development of intelligent metasurfaces. Attached Figure Description

[0025] Figure 1 This is a cell arrangement diagram (0101… / 0101…) of the metasurface array when the array is not optimized according to the present invention.

[0026] Figure 2 In this invention, without array optimization, the metasurface array uses MATLAB software to calculate the dual-beam pattern when φ = 0°.

[0027] Figure 3 The dual-beamforming pattern of the metasurface array of this invention is optimized using the traditional binary dragonfly algorithm (BDA) and calculated by MATLAB simulation.

[0028] Figure 4 The metasurface array of this invention is optimized using a traditional genetic algorithm (GA), and the dual-beamforming pattern is calculated by MATLAB simulation.

[0029] Figure 5 The dual-beamforming pattern of the metasurface array of this invention is obtained through MATLAB simulation calculation using the Hybrid Genetic Binary Dragonfly Algorithm (GABDA).

[0030] Figure 6 This is a comparison chart of the iteration curves calculated by MATLAB for three different algorithms used in the metasurface array of this invention.

[0031] Figure 7 The optimal array cell layout for the metasurface array of this invention is obtained through MATLAB simulation calculation using a hybrid genetic binary dragonfly algorithm.

[0032] Figure 8 The normalized three-dimensional far-field scattering pattern of the metasurface array of this invention is obtained by using a hybrid genetic binary dragonfly algorithm for optimization and by MATLAB simulation.

[0033] Figure 9 The metasurface array of this invention is optimized using the traditional binary dragonfly algorithm, and the three-beamform pattern is calculated by MATLAB simulation.

[0034] Figure 10The three-beamform pattern of the metasurface array of this invention is obtained by optimization using a traditional genetic algorithm and simulation calculation using MATLAB.

[0035] Figure 11 The three-beamform pattern of the metasurface array of this invention is optimized using a hybrid genetic binary dragonfly algorithm and calculated by MATLAB simulation.

[0036] Figure 12 This is a comparison chart of the iteration curves calculated by MATLAB for three different algorithms used in the metasurface array of this invention.

[0037] Figure 13 The optimal array cell layout for the metasurface array of this invention is obtained through MATLAB simulation calculation using a hybrid genetic binary dragonfly algorithm.

[0038] Figure 14 The normalized three-dimensional far-field scattering pattern of the metasurface array of this invention is obtained by using a hybrid genetic binary dragonfly algorithm for optimization and by MATLAB simulation.

[0039] Figure 15 This is a flowchart of the algorithm described in this invention. Detailed Implementation

[0040] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0041] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as herein.

[0042] The invention will now be described in further detail with reference to the accompanying drawings.

[0043] This invention proposes a metasurface beamforming optimization method based on a hybrid genetic binary dragonfly algorithm. To quickly achieve a predefined waveform, the metasurface units are arranged in a square pattern and optimized using the hybrid genetic binary dragonfly algorithm. Compared with traditional algorithms, this invention has advantages such as fast convergence, high accuracy, and less susceptibility to local optima, and has potential application prospects for the development of intelligent metasurfaces.

[0044] This invention discloses a metasurface beamforming optimization method based on a hybrid genetic binary dragonfly algorithm. The metasurface array consists of 8*8 super sub-units, and each super sub-unit comprises 6*6 units with phases of either all 0° or all 180°. Here, the phase is a relative value, meaning that the phase difference between two units is 180°. Figure 15 As shown, the hybrid genetic binary dragonfly algorithm is used to optimize metasurface beamforming, specifically including the following steps:

[0045] Step 1: Construct an initial population describing the metasurface array and measure the optimal values;

[0046] Step 2: Perform one iteration based on the initial population, carrying out selection, crossover, and mutation operations, and measuring the optimal value;

[0047] Step 3: Compare the measurement results of Step 2 with the measurement results of the previous generation. If the optimal value has changed, return to Step 2 and repeat Step 2 and Step 3 to output the optimal value and the optimal population. If the optimal value has not changed, proceed to the next step.

[0048] Step 4: Based on the population from Step 3 above, sort the optimal values. Perform binary dragonfly algorithm iteration on the first 90% and genetic mutation operation on the last 10% and measure the optimal value.

[0049] Step 5: Compare the measurement results of Step 4 with the measurement results of the previous generation. If the optimal value has changed, repeat Step 2 and Step 3 to output the optimal value and the optimal population. If the optimal value has not changed, perform Step 4 and Step 5 to output the optimal value and the optimal population.

[0050] In this invention, an optimization method is adopted that integrates the binary dragonfly algorithm and the genetic algorithm. First, the population of the binary dragonfly algorithm is optimized by the genetic algorithm. In the early stage of the algorithm, the genetic algorithm is used to achieve fast convergence. When the optimal value has not changed, the binary dragonfly algorithm is connected. The idea of ​​genetic mutation is used to solve the problem of no individuals in the neighborhood of the binary dragonfly algorithm. This method accelerates the convergence speed of the algorithm and improves the optimization accuracy of the genetic algorithm.

[0051] In this invention, the idea of ​​mutation in genetic algorithms is adopted to solve the problem of the binary dragonfly algorithm getting stuck in a local optimum when there are no dragonflies in the neighborhood.

[0052] The size of the metasurface array in this invention is not fixed, making it highly adaptable.

[0053] The fitness function of the hybrid genetic binary dragonfly algorithm of this invention is:

[0054] Fitness = w1 * F1 + w2 * F2

[0055]

[0056]

[0057]

[0058]

[0059] Where Fitness is the fitness function, w1 and w2 are weighting coefficients, and F lowerMask F UpperMask These represent the self-defined lower bound function and upper bound function, respectively, where i and j are the discrete sample indices. For pattern functions, The function representing the normalized radiation pattern, θ, These are the pitch and azimuth angles, respectively, with the positive directions of the y-axis and x-axis as reference directions. k0 is the wave vector value, M and N represent the number of rows and columns of the metasurface arrangement, and A... mn , D represents the amplitude and phase corresponding to the (m, n)th unit structure, respectively. x D y These represent the periods in the coordinate axes of the unit cell.

[0060] Specifically, the hybrid genetic binary dragonfly algorithm of this invention has a crossover probability of 0.9, a generation skipping rate of 0.9, a mutation probability of 0.08 in the genetic algorithm, and a mutation probability of 0.2 in the binary dragonfly algorithm.

[0061] Example 1

[0062] like Figure 1 As shown, without optimization of the cell arrangement, the metasurface array consists of 64 super sub-cells, each of which comprises 6*6 phase cells of either 0° or 180°. The super sub-cells are arranged in the pattern 10101010 / 10101010 / …, with phase gradients only in the horizontal direction.

[0063] Figure 2 This is a dual-beam pattern of the metasurface with φ = 0° calculated by MATLAB based on the unoptimized array arrangement. The beam orientation is ±30°, the left and right sidelobe levels are -16.85dB, and the middle sidelobe level is -10.52dB.

[0064] Figure 3 This is the dual-beam pattern of the metasurface at φ=0°, calculated by MATLAB using the traditional Binary Dragonfly Algorithm (BDA) for the array. The -3dB width is limited to 6°, the beam is limited to ±30°, the left and right sidelobe levels are limited to -22dB, and the middle sidelobe level is limited to -15dB. It can be seen that the pattern fails to meet the constraints around ±40°.

[0065] Figure 4 This is a dual-beam pattern of the metasurface at φ=0°, calculated by MATLAB using a traditional genetic algorithm (GA) based on the array. The -3dB width is limited to 6°, the beam direction is limited to ±30°, the left and right sidelobe levels are limited to -22dB, and the middle sidelobe level is limited to -15dB. It can be seen that the pattern fails to meet the constraints within the range of -20° to 20°.

[0066] Figure 5 This is the dual-beam radiation pattern of the metasurface at φ=0°, calculated by MATLAB using the Hybrid Genetic Binary Dragonfly Algorithm (GABDA) for the array. The -3dB width is limited to 6°, the beam direction is limited to ±30°, the left and right sidelobe levels are limited to -22dB, and the middle sidelobe level is limited to -15dB. It can be seen that the radiation pattern meets all the set conditions.

[0067] Figure 6 This is a comparison of fitness iteration curves obtained from three algorithms using MATLAB. It can be seen that the genetic algorithm converges after 130 generations, but its accuracy is not high. While the binary dragonfly algorithm has higher accuracy than the genetic algorithm, its convergence speed is slow, only converging around the 310th generation. The hybrid genetic binary dragonfly algorithm used in this invention combines the advantages of the first two algorithms, achieving fast convergence, high accuracy, and reducing the left and right sidelobe levels to -22.8dB and the middle level to -15.67dB, which is superior to the first two algorithms. The optimal array arrangement is as follows: Figure 7 As shown, the normalized scattering pattern of the three-dimensional far-field is as follows: Figure 8 As shown.

[0068] Example 2

[0069] Figure 9 This is a three-beam pattern of the metasurface at φ=0° calculated by MATLAB using the traditional binary dragonfly algorithm. The -3dB width is limited to 6°, the beam direction is limited to ±30° and 0°, the left and right sidelobe levels are limited to -21dB, and the middle sidelobe level is limited to -15dB. The iterative curve is shown below. Figure 11 As shown, it can be seen that the design conditions are basically met, but the convergence speed is too slow.

[0070] Figure 10 This is the three-beam radiation pattern of the metasurface at φ=0°, calculated by MATLAB using a traditional genetic algorithm based on the array. The -3dB width is limited to 6°, the beams are limited to ±30° and 0°, the left and right sidelobe levels are limited to -21dB, and the middle sidelobe level is limited to -15dB. It can be seen that the radiation pattern fails to meet the design requirements near ±40° and between -10° and 10°.

[0071] Figure 11This is the three-beam radiation pattern of the metasurface at φ=0°, calculated by MATLAB using a hybrid genetic binary dragonfly algorithm. The -3dB width is limited to 6°, the beam direction is limited to ±30° and 0°, the left and right sidelobe levels are limited to -21dB, and the middle sidelobe level is limited to -15dB. It can be seen that the radiation pattern meets all design requirements.

[0072] Figure 12 This is a comparison of fitness iteration curves obtained from three algorithms using MATLAB. It can be seen that the genetic algorithm converges after 40 generations, but its accuracy is low and it does not meet the design conditions. While the binary dragonfly algorithm has higher accuracy than the genetic algorithm, its convergence speed is slow, reaching its optimal convergence around 350 generations. The hybrid genetic binary dragonfly algorithm adopted in this invention combines the advantages of the first two algorithms, achieving fast convergence (reaching optimal convergence around 110 generations), high accuracy, and meeting the design conditions, thus outperforming the first two algorithms. The optimal array arrangement is as follows: Figure 13 As shown, the normalized scattering pattern of the three-dimensional far-field is as follows: Figure 14 As shown.

[0073] This invention discloses a metasurface beamforming optimization method based on a hybrid genetic binary dragonfly algorithm, belonging to the field of computer technology and applications. To quickly achieve a predefined waveform, the metasurface units are arranged in a square pattern and optimized using the hybrid genetic binary dragonfly algorithm. Compared with traditional algorithms, this invention has advantages such as fast convergence, high accuracy, and less susceptibility to local optima, and has potential application prospects for the development of intelligent metasurfaces.

[0074] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as herein.

[0075] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any transformations or substitutions that can be conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A metasurface beamforming optimization method based on a hybrid genetic binary dragonfly algorithm, characterized in that, The metasurface array consists of 8*8 super sub-units, each of which comprises 6*6 units with phases of either 0° or 180°. The phase is a relative value, meaning two units are 180° out of phase. The array employs a hybrid genetic binary dragonfly algorithm to optimize a specific waveform, specifically including the following steps: Step 1: Construct an initial population describing the metasurface array and measure the optimal values; Step 2: Perform one iteration based on the initial population, carrying out selection, crossover, and mutation operations, and measuring the optimal value; Step 3: Compare the measurement results of Step 2 with the measurement results of the previous generation. If the optimal value has changed, return to Step 2 and repeat Step 2 and Step 3 to output the optimal value and the optimal population. If the optimal value has not changed, proceed to the next step. Step 4: Based on the population from Step 3 above, sort the optimal values. Perform binary dragonfly algorithm iteration on the first 90% and genetic mutation operation on the last 10% and measure the optimal value. Step 5: Compare the measurement results of Step 4 with the measurement results of the previous generation. If the optimal value has changed, repeat Step 2 and Step 3 to output the optimal value and the optimal population. If the optimal value has not changed, perform Step 4 and Step 5 to output the optimal value and the optimal population. The fitness function of the hybrid genetic binary dragonfly algorithm is: in, For the fitness function, These are weighting coefficients. These represent the self-defined lower bound function and upper bound function, respectively. For discrete sample indexing, For pattern functions, A function representing the normalized radiation pattern. They are respectively based on shaft and The pitch and azimuth angles are referenced in the positive direction of the axis. The wave vector value, This indicates the number of rows and columns in the metasurface arrangement. They represent the first The amplitude and phase corresponding to each unit structure These represent the periods in the coordinate axes of the unit cell.

2. The metasurface beamforming optimization method based on a hybrid genetic binary dragonfly algorithm according to claim 1, characterized in that, The method also includes first optimizing the population of the binary dragonfly algorithm using a genetic algorithm, so that it can converge quickly.

3. The metasurface beamforming optimization method based on a hybrid genetic binary dragonfly algorithm according to claim 1, characterized in that, The binary dragonfly algorithm can improve the optimization accuracy of genetic algorithms.

4. The metasurface beamforming optimization method based on a hybrid genetic binary dragonfly algorithm according to claim 1, characterized in that, The method also includes using the mutation concept from genetic algorithms to address the problem of the binary dragonfly algorithm getting stuck in a local optimum when there are no dragonflies in the vicinity.

5. The metasurface beamforming optimization method based on a hybrid genetic binary dragonfly algorithm according to claim 1, characterized in that, The size of the metasurface array is not unique, and it is highly adaptable.

6. The metasurface beamforming optimization method based on a hybrid genetic binary dragonfly algorithm according to claim 1, characterized in that, The hybrid genetic binary dragonfly algorithm has a crossover probability of 0.9, a generation skipping rate of 0.9, a mutation probability of 0.08 in the genetic algorithm, and a mutation probability of 0.2 in the binary dragonfly algorithm.

Images