CN-118013853-B - Super-surface beam forming optimization method based on hybrid genetic binary dragonfly algorithm

Abstract

The invention discloses a super-surface beam forming optimization method based on a hybrid genetic binary dragonfly algorithm (GABDA), which belongs to the technical field of computer technology and application, and comprises the following steps that a super-surface array consists of 8 x 8 super-sub-units, each super-sub-unit consists of 6*6 units with the phase of all 0 degrees or all 180 degrees, the phases are relative values, namely, the phases of the two units differ by 180 degrees, the array adopts the hybrid genetic binary dragonfly algorithm to optimize a set specific waveform, and in order to quickly realize the preset waveform, the super-surface units are vertically and symmetrically distributed in a square mode and are optimized by using the hybrid genetic binary dragonfly algorithm. Compared with the traditional algorithm, the method has the advantages of fast convergence, high precision, difficult sinking into local optimum and the like, and has potential application prospect for the development of intelligent super-surfaces.

Inventors

• ZHANG CHUANLIN
• ZHANG HUAMEI

Assignees

• 南京邮电大学

Dates

Publication Date

20260505

Application Date

20240311

Claims (6)

1. 1. The super-surface beam forming optimization method based on the hybrid genetic binary dragonfly algorithm is characterized in that a super-surface array consists of 8 super-sub units, each super sub unit consists of 6*6 units with the phase of 0 degrees or 180 degrees, the phases are relative values, namely, the phases of the two units are 180 degrees different, the array adopts the hybrid genetic binary dragonfly algorithm to optimize a set specific waveform, and the method specifically comprises the following steps: Step 1, constructing an initial population describing a super-surface array and measuring an optimal value; step 2, performing one iteration based on the initial population, performing selection, crossing and mutation operations, and measuring an optimal value; Step 3, comparing the measurement result of the step 2 with the measurement result of the previous generation, returning to the step 2 if the optimal value is changed, repeatedly executing the step 2 and the step 3, outputting the optimal value and the optimal population, and carrying out the next step if the optimal value is not changed; Step 4, sorting optimal values based on the population in the step 3, iterating the binary dragonfly algorithm by the first 90%, performing genetic variation operation by the last 10%, and measuring the optimal values; Step 5, comparing the measurement result of the step 4 with the measurement result of the previous generation, if the optimal value is changed, repeatedly executing the step 2 and the step 3, outputting the optimal value and the optimal population, and if the optimal value is not changed, executing the step 4 and the step 5, and outputting the optimal value and the optimal population; the adaptability function of the hybrid genetic binary dragonfly algorithm is as follows: Wherein, the In order to adapt the function of the degree of adaptation, As the weighting coefficient(s), Respectively represent a self-set lower limit function and an upper limit function, For the index of the discrete samples, As a function of the direction-map, A function representing the normalized pattern, Respectively by Shaft and method for producing the same The positive axis direction is the pitch angle and the azimuth angle of the reference direction, As the value of the wave vector, Representing the number of rows and columns in the super-surface arrangement, Respectively represent the first The amplitude and phase corresponding to the individual cell structures, Respectively, the periods in the directions of the axes of the units.
2. 2. The method for optimizing the beamforming of the super surface based on the hybrid genetic binary dragonfly algorithm of claim 1, wherein, the method further comprises the steps of optimizing the population of the binary dragonfly algorithm by using a genetic algorithm, so that the binary dragonfly algorithm can be converged rapidly.
3. 3. The method for optimizing the beamforming of the super surface based on the hybrid genetic binary dragonfly algorithm as claimed in claim 1, wherein the binary dragonfly algorithm can improve the optimizing precision of the genetic algorithm.
4. 4. The method for optimizing the beamforming of the super surface based on the hybrid genetic binary dragonfly algorithm as claimed in claim 1, wherein the method further comprises the step of solving the problem that the binary dragonfly algorithm is in a locally optimal state in the critical domain without dragonfly by adopting the idea of variation in the genetic algorithm.
5. 5. The optimization method of the super-surface beamforming based on the hybrid genetic binary dragonfly algorithm as claimed in claim 1, wherein the size of the super-surface array is not unique, and the adaptability is strong.
6. 6. The method for optimizing the super-surface beamforming based on the hybrid genetic binary dragonfly algorithm according to claim 1, wherein the hybrid genetic binary dragonfly algorithm has a crossover probability of 0.9, a generation interval rate of 0.9, a variation probability of 0.08 in the genetic algorithm and a variation probability of 0.2 in the binary dragonfly algorithm.

Description

Super-surface beam forming optimization method based on hybrid genetic binary dragonfly algorithm Technical Field The invention relates to a super-surface beam forming optimization method based on a hybrid genetic binary dragonfly algorithm, and belongs to the technical field of computer science and technology and application. Background With the development of modern communication technology, the design of intelligent super-surface is more and more important in military and daily life. Along with the expansion of the coding super-surface array, the arrangement mode of the array surface grows exponentially, the artificial design has extremely limitation, and the super-surface meeting the requirement can be designed in a specific and high-efficiency way by combining an intelligent optimization algorithm. The intelligent super surface is a hot spot of research under the present condition, but with the continuous improvement of design requirements, the traditional intelligent optimization algorithm can not meet the requirements of high efficiency and high precision. Disclosure of Invention Aiming at the defects and shortcomings of the prior art, the invention provides a super-surface beam forming optimization method based on a hybrid genetic binary dragonfly algorithm, which is characterized in that the super-surface units are square arranged and optimized by using the hybrid genetic binary dragonfly algorithm. The invention provides a super-surface beam forming optimization method based on a hybrid genetic binary dragonfly algorithm, wherein the super-surface array consists of 8 x 8 super-sub-units, each super-sub-unit consists of 6*6 units with the phase of all 0 degrees or all 180 degrees, the phase is a relative value, namely the phase difference of two units is 180 degrees, and the square super-surface array adopts the hybrid genetic binary dragonfly algorithm to perform beam forming optimization on the super-surface beam forming optimization, and the method specifically comprises the following steps: Step 1, constructing an initial population describing a super-surface array and measuring an optimal value; step 2, performing one iteration based on the initial population, performing selection, crossing and mutation operations, and measuring an optimal value; Step 3, comparing the measurement result of the step 2 with the measurement result of the previous generation, returning to the step 2 if the optimal value is changed, repeatedly executing the step 2 and the step 3, outputting the optimal value and the optimal population, and carrying out the next step if the optimal value is not changed; Step 4, sorting optimal values based on the population in the step 3, iterating the binary dragonfly algorithm by the first 90%, performing genetic variation operation by the last 10%, and measuring the optimal values; And 5, comparing the measurement result in the step 4 with the measurement result of the previous generation, repeatedly executing the step 2 and the step 3 to output the optimal value and the optimal population if the optimal value is changed, and executing the step 4 and the step 5 to output the optimal value and the optimal population if the optimal value is not changed. Further, the method further comprises the step of optimizing the population of the binary dragonfly algorithm by using a genetic algorithm, so that the population can be converged rapidly. Further, the binary dragonfly algorithm can improve the optimizing precision of the genetic algorithm. Furthermore, the method also comprises the step of solving the problem that no individual falls into local optimum in the clinical domain of the binary dragonfly algorithm by adopting the idea of variation in the genetic algorithm. Further, the size of the super-surface array is not unique, and the adaptability is high. Further, the fitness function of the hybrid genetic binary dragonfly algorithm is as follows: Fitness=w1*F1+w2*F2 wherein Fitness is an Fitness function, w1 and w2 are weighting coefficients, F lowerMask、FUpperMask respectively represent self-set lower limit functions and upper limit functions, i and j are discrete sample indexes, As a function of the direction-map,A function representing the normalized pattern, θ,The pitch angle and the azimuth angle respectively take the positive directions of the y axis and the x axis as reference directions, k 0 is a wave vector value, M, N represents the number of rows and columns of super-surface arrangement, A mn,The amplitude and the phase corresponding to the (m, n) th unit structure are respectively represented, and D x、Dy represents the period of the unit coordinate axis direction. Further, the algorithm crossover probability is 0.9, the generation separation rate is 0.9, the variation probability in the genetic algorithm is 0.08, and the variation probability in the binary dragonfly algorithm is 0.2. The beneficial effects are that: 1. the invention further improves the algorithm based