Design method and device of composite wave-absorbing material based on genetic algorithm
By incorporating a genetic algorithm-based design method, weighting factors and fitness calculations are introduced to select superior individuals for crossover and mutation, thus solving the problem of low computational efficiency in the design of composite absorbing materials and achieving efficient and rapid optimization design.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING INST OF ENVIRONMENTAL FEATURES
- Filing Date
- 2025-10-24
- Publication Date
- 2026-07-21
AI Technical Summary
Existing technologies for designing composite absorbing materials suffer from low computational efficiency and long computation time, especially as computational complexity increases exponentially. Furthermore, the optimization speed is slow, resulting in unsatisfactory design results.
A genetic algorithm-based design method is adopted. An initial population is randomly generated, the reflectivity and fitness of individuals are calculated, a first weighting factor is introduced, and individuals with excellent performance and total thickness not exceeding the threshold range are selected. Crossover and mutation operations are performed until the termination condition is met to determine the design scheme of the composite absorbing material.
It improves the computational efficiency of composite absorbing materials and reduces computation time, especially significantly accelerating the calculation speed when the computational complexity is high, thus improving the design effect and approaching the global optimal solution.
Smart Images
Figure CN121439030B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of composite absorbing material optimization design technology, and in particular to a design method and apparatus for composite absorbing materials based on genetic algorithms. Background Technology
[0002] Composite absorbing materials have a significant effect on suppressing electromagnetic scattering from radar antennas. Reflectivity is an important indicator for characterizing the performance of absorbing materials.
[0003] In related technologies, when designing the structure of composite absorbing materials based on reflectivity, due to factors such as application requirements, material properties, and manufacturing processes, it is usually necessary to simultaneously limit the thickness of each layer and the total thickness, and find the optimal structure of the composite absorbing material through an ergonomic approach. However, this method suffers from low computational efficiency and long computation time. Moreover, as computational complexity increases, the computation time required for structural optimization through the ergonomic approach grows exponentially. In addition, there are also structural design methods for composite absorbing materials based on genetic algorithms. However, these methods still suffer from slow optimization speed, low computational efficiency, and unsatisfactory design results.
[0004] Therefore, to address the above shortcomings, there is a need to provide a design method for composite absorbing materials that can improve computational efficiency and reduce computation time. Summary of the Invention
[0005] The technical problem to be solved by the present invention is to provide a design method and device for composite absorbing materials based on genetic algorithms, addressing the deficiencies in the prior art.
[0006] To address the aforementioned technical problems, according to a first aspect of this disclosure, a design method for composite absorbing materials based on a genetic algorithm is provided, comprising: randomly generating an initial population, wherein the initial population includes multiple individuals, each individual representing a candidate design scheme for a composite absorbing material, the candidate design scheme including the following design variables: the material type of each layer in the multilayer structure of the composite absorbing material, and the thickness of each layer; calculating the reflectivity of each individual in the initial population at each frequency point within a specified frequency band; and determining the reflectivity of each individual based on the deviation of the total thickness of the multilayer structure of each individual from a total thickness threshold range. The first weighting factor is used; the fitness of each individual is calculated based on the difference between the reflectivity of each individual and the target reflectivity at each frequency point within the specified frequency band, and the first weighting factor corresponding to each individual; individuals are selected from the initial population based on the fitness of each individual, and crossover and mutation operations are performed on the selected individuals to generate the next generation population; fitness calculation, individual selection, and crossover and mutation operations are performed on the next generation population until the termination condition of the genetic algorithm is met; the design scheme of the composite absorbing material is determined based on the individuals in the latest generation population that meet the termination condition.
[0007] In some embodiments, when the total thickness of the multilayer structure possessed by each individual is within the total thickness threshold range, the deviation is 0; when the total thickness of the multilayer structure possessed by each individual is not within the total thickness threshold range, the deviation is greater than 0; and the first weighting factor is positively correlated with the deviation.
[0008] In some embodiments, the first weighting factor for each individual is determined according to the following formula: in, W The first weighting factor for each individual, d / (T) max -T min The degree of deviation is denoted as ). T toal The total thickness of the multi-layered structure possessed by each individual, T max This is the upper limit of the total thickness threshold range. T min is the lower limit of the total thickness threshold range, and k is a preset constant coefficient greater than 0.
[0009] In some embodiments, when the specified frequency band is a single frequency band, the fitness of each individual is calculated according to the following formula: Where F is the fitness of each individual, W is the first weighting factor corresponding to each individual, R is the reflectance of each individual at each frequency point in the specified frequency band, Ra is the target reflectance, m is the total number of frequency band units included in the specified frequency band, Δf is the frequency point step size value of the frequency band unit, and S(R) is the area enclosed by the reflectance curve formed by the reflectance of each frequency point in the specified frequency band and the ideal line segment, wherein the ideal line segment is the reflectance line segment formed when the reflectance value of each frequency point in the specified frequency band is the target reflectance.
[0010] In some embodiments, when the specified frequency band is multiple frequency bands, the fitness of each individual is calculated according to the following formula: Where F is the fitness of each individual, W is the first weighting factor for each individual, and w i The second weighting factor is defined for the i-th specified frequency band of each individual, where n is the total number of specified frequency bands, R is the reflectance of each individual at each frequency point within the i-th specified frequency band, Ra is the target reflectance, and m... i For the first i The total number of frequency band elements contained in a specified frequency band, where Δf is the frequency step size value of the frequency band element, S i (R) is the area enclosed by the reflectivity curve formed by the reflectivity of each frequency point in the i-th specified frequency band and the ideal line segment, wherein the ideal line segment is the reflectivity line segment formed when the reflectivity of each frequency point in the i-th specified frequency band is the target reflectivity.
[0011] In some embodiments, selecting individuals from the initial population based on the fitness of each individual includes: selecting a first set of individuals from the initial population whose total thickness is within the total thickness threshold range, and selecting the first individual with the lowest fitness from the first set; selecting a second set of individuals from the initial population whose total thickness is not within the total thickness threshold range, and selecting a second individual from the second set whose fitness is less than or equal to that of the first individual; and using the first individual and the second individual as individuals selected from the initial population.
[0012] According to a second aspect of this disclosure, a design apparatus for composite absorbing materials based on genetic algorithms is provided, characterized in that it includes a design method for composite absorbing materials based on genetic algorithms as described above.
[0013] According to a third aspect of this disclosure, an electronic device is provided, comprising: a memory; and a processor coupled to the memory, the processor being configured to execute, based on instructions stored in the memory, the design method for composite absorbing materials based on genetic algorithms as described above.
[0014] According to a fourth aspect of this disclosure, a computer-readable storage medium is provided that stores computer program instructions thereon, which, when executed by a processor, implement the design method of composite absorbing materials based on genetic algorithms as described above.
[0015] According to a fifth aspect of this disclosure, a computer program product is provided, having stored computer program instructions thereon, which, when executed by a processor, implement the design method for composite absorbing materials based on genetic algorithms as described above.
[0016] The implementation of this invention has the following beneficial effects: On the one hand, by introducing a first weighting factor in the optimization design process of composite absorbing materials based on genetic algorithms, and calculating the fitness of each individual based on the difference between the reflectivity and the target reflectivity and the first weighting factor, it is helpful to screen out individuals with excellent performance and whose total thickness does not exceed the total thickness threshold range by more than a few. These individuals, as the parent, have a high probability of producing offspring individuals with excellent performance, thereby helping to find the global optimal solution and improve the design effect of composite absorbing materials. On the other hand, the above processing flow can significantly improve computational efficiency and reduce computation time. In particular, as the computational complexity increases, its advantage of high computational efficiency becomes more obvious. Attached Figure Description
[0017] Figure 1 This is a flowchart illustrating the design method of composite absorbing materials based on genetic algorithms in some embodiments of the present invention. Figure 2 This is a flowchart illustrating the design method of composite absorbing materials based on genetic algorithms in some other embodiments of the present invention; Figure 3 This is a schematic diagram of the region enclosed by the reflectance curves calculated in some examples of this invention and the line segment containing the target reflectance; Figure 4 This is a comparative schematic diagram of the reflectance calculated in some examples of the present invention; Figure 5 This is a comparative schematic diagram of the reflectance calculated in other examples of the present invention; Figure 6 This is a structural block diagram of a design device for composite absorbing materials based on genetic algorithms in some embodiments of the present invention; Figure 7 This is a schematic diagram of the structure of an electronic device in some embodiments of the present invention. Detailed Implementation
[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0019] like Figure 1 As shown, the design method of composite absorbing material based on genetic algorithm provided in some embodiments of the present invention includes steps S11 to S14.
[0020] In step S11, an initial population is randomly generated. The initial population consists of multiple individuals. Each individual represents a candidate design scheme for a composite absorbing material. The candidate design scheme includes the following design variables: the material type of each layer in the multi-layered structure of the composite absorbing material, and the thickness of each layer.
[0021] In some embodiments, in step S11, parameter information of the composite absorbing material and preset parameters of the genetic algorithm are first obtained, and then an initial population is generated through random encoding. For example, the obtained parameter information of the composite absorbing material includes at least one of the following: optional material information for each layer of the composite absorbing material, layer number information, specified frequency band information, thickness threshold range for each layer, total thickness threshold range, and target reflectivity. The preset parameters of the genetic algorithm include at least one of the following: total number of individuals in the population, number of iterations, crossover rate, and mutation rate.
[0022] In step S12, the fitness of each individual in the initial population is calculated. Specifically, calculating the fitness of an individual may include steps A1 to A3. In step A1, the reflectance of each individual in the initial population at each frequency point within a specified frequency band is calculated.
[0023] In some embodiments, the reflectivity of the composite absorbing material is calculated using a theoretical model. For a composite absorbing material with n structural layers and a backing metal reflective surface, when an electromagnetic wave is incident from free space onto an interface with an input impedance of Zi(n), part of the electromagnetic wave is reflected by the interface, and the other part enters the absorbing material. Further, in this embodiment, the reflectivity of each individual in the initial population can be calculated according to the following formula.
[0024] Where R is the reflectivity. z i(n) The input impedance of the n-layer composite absorbing material is... z 0 represents vacuum impedance. γ n Let be the complex propagation factor of the nth layer material. dn Let n be the thickness of the nth layer of material. z n Let n be the characteristic impedance of the nth layer material. μ n Let be the relative complex permeability of the nth layer material. ε n denoted as the relative permittivity, c as the speed of light in a vacuum, j as the imaginary number, f as the frequency, and n as the number of layers in the composite absorbing material, where n ≥ 1.
[0025] In this embodiment, calculating the reflectivity of an individual based on the above formula improves the accuracy of the calculated reflectivity, which in turn helps improve the accuracy of the fitness calculated subsequently. This facilitates the selection of individuals with superior performance from the population based on fitness, thus improving the design effect of the composite absorbing material.
[0026] In step A2, the first weighting factor for each individual is determined based on the degree of deviation between the total thickness of the multi-layer structure possessed by each individual and the total thickness threshold range.
[0027] There are several ways to determine the first weighting factor. Two implementation methods are illustrated below. In the first implementation, the first weighting factor is positively correlated with the degree of deviation. That is, the greater the degree of deviation, the larger the first weighting factor; the smaller the degree of deviation, the smaller the first weighting factor. The degree of deviation can satisfy the following: when the total thickness of the multilayer structure possessed by each individual is within the total thickness threshold range, the degree of deviation is 0; when the total thickness of the multilayer structure possessed by each individual is not within the total thickness threshold range, the degree of deviation is greater than 0. Furthermore, in the first implementation, the fitness of an individual can be positively correlated with the first weighting factor, and the minimum fitness can be used as one of the conditions for subsequently selecting individuals from the population.
[0028] Furthermore, in the first implementation, the first weighting factor corresponding to each individual can be determined according to the following exemplary formula: in, W The first weighting factor for each individual, d / (T) max -T min () represents the degree of deviation. Ttoal The total thickness of the multi-layered structure possessed by each individual, T max This represents the upper limit of the total thickness threshold range. T min is the lower limit of the total thickness threshold range, and k is a preset constant coefficient greater than 0.
[0029] In the second implementation, the first weighting factor is negatively correlated with the degree of deviation. That is, the greater the degree of deviation, the smaller the first weighting factor; the smaller the degree of deviation, the larger the first weighting factor. The degree of deviation can satisfy the following: when the total thickness of the multilayer structure possessed by each individual is within the total thickness threshold range, the degree of deviation is 0; when the total thickness of the multilayer structure possessed by each individual is not within the total thickness threshold range, the degree of deviation is greater than 0. Furthermore, in the second implementation, the fitness of an individual can be negatively correlated with the first weighting factor, and the minimum fitness can be used as one of the conditions for subsequently selecting individuals from the population.
[0030] Furthermore, in the second implementation, the first weighting factor corresponding to each individual can be determined according to the following exemplary formula: in, W The first weighting factor for each individual, d / (T) max -T min () represents the degree of deviation. T toal The total thickness of the multi-layered structure possessed by each individual, T max This represents the upper limit of the total thickness threshold range. T min is the lower limit of the total thickness threshold range, and k is a preset constant coefficient greater than 0.
[0031] In step A3, the fitness of each individual is calculated based on the difference between the reflectance of each individual at each frequency point in the specified frequency band and the target reflectance, as well as the first weighting factor corresponding to each individual.
[0032] There are several ways to calculate fitness. The following describes two examples using different implementation methods.
[0033] In the first implementation, the fitness of an individual is positively correlated with the absolute value of the difference between its reflectivity and the target reflectivity, and the fitness of an individual is also positively correlated with a first weighting factor. In the second implementation, minimum fitness is used as one of the screening criteria for selecting individuals from the population. This way, when selecting individuals from the population using minimum fitness as the screening rule, individuals with excellent performance and whose total thickness does not significantly exceed the total thickness threshold range can be selected. After crossover and mutation operations on the selected individuals as parent materials, there is a high probability of generating offspring individuals with excellent performance whose total thickness is within the total thickness threshold range, thereby contributing to obtaining the optimal design scheme for the composite absorbing material.
[0034] In the first implementation, when the specified frequency band is one frequency band, the fitness of each individual can be calculated according to the following exemplary formula.
[0035] Where F is the fitness of each individual, W is the first weighting factor corresponding to each individual, R is the reflectance of each individual at each frequency point in the specified frequency band, Ra is the target reflectance, m is the total number of frequency band units included in the specified frequency band, Δf is the frequency point step size value of the frequency band unit, S(R) is the area enclosed by the reflectance curve formed by the reflectance of each frequency point in the specified frequency band and the ideal line segment, and the ideal line segment is the reflectance line segment formed when the reflectance value of each frequency point in the specified frequency band is the target reflectance.
[0036] In this embodiment of the disclosure, the fitness is calculated using the above formula. On the one hand, it can better characterize the candidate design scheme of the composite absorbing material represented by the individual from two aspects: whether it is close to the target reflectivity and whether the total thickness is within the total thickness threshold range. This helps to improve the design effect of the composite absorbing material. On the other hand, by evaluating the difference between the reflectivity and the target reflectivity with S(R) / Δf, and using the product of the first weighting factor and S(R) / Δf as the fitness, it helps to reduce the amount of computation based on the genetic algorithm to determine the design scheme of the composite absorbing material and speed up the calculation.
[0037] In the first implementation, when there are multiple frequency bands specified, the fitness of each individual can be calculated according to the following exemplary formula.
[0038] Where F is the fitness of each individual, W is the first weighting factor for each individual, and w iThe second weighting factor is defined for the i-th specified frequency band of each individual, where n is the total number of specified frequency bands, R is the reflectance of each individual at each frequency point within the i-th specified frequency band, Ra is the target reflectance, and m... i For the first i The total number of frequency band elements contained in a specified frequency band, where Δf is the frequency step size of the frequency band element, S i (R) is the area enclosed by the reflectivity curve formed by the reflectivity of each frequency point in the i-th specified frequency band and the ideal line segment. The ideal line segment is the reflectivity line segment formed when the reflectivity of each frequency point in the i-th specified frequency band is the target reflectivity.
[0039] The second weighting factor can be preset. For example, different second weighting factors can be preset for different frequency bands. The second weighting factor can be a value greater than 0 and less than 1. For example, assuming that each individual has three designated frequency bands, specifically frequency bands 1 to 3, the second weighting factor for frequency band 1 can be preset to 0.1, the second weighting factor for frequency band 2 to 0.3, and the second weighting factor for frequency band 3 to 0.6.
[0040] In this embodiment, the fitness calculated using the above formula can better characterize the merits of candidate design schemes for composite absorbing materials from three aspects: whether the reflectivity is close to the target, whether the total thickness is within the total thickness threshold range, and the importance of each frequency band. This helps improve the design performance of composite absorbing materials. Furthermore, using the product of the first weighting factor and the weighted sum of the reflectivity differences of each frequency band as the fitness helps reduce the computational load in determining the design scheme of composite absorbing materials based on genetic algorithms, thus accelerating the computation speed. In addition, in specific implementations, besides using the above exemplary formula to calculate the fitness, other formulas can also be used to calculate the fitness.
[0041] In the second implementation, the fitness of an individual is negatively correlated with the absolute value of the difference between its reflectivity and the target reflectivity, and the fitness of an individual is also negatively correlated with the first weighting factor. Furthermore, in this second implementation, maximum fitness is used as one of the screening criteria for selecting individuals from the population. This way, when selecting individuals from the population using maximum fitness as the screening rule, individuals with excellent performance and whose total thickness does not significantly exceed the total thickness threshold range can be selected. Subsequently, after crossover and mutation operations on the selected individuals as parent materials, there is a high probability of generating offspring individuals with excellent performance whose total thickness is within the total thickness threshold range, thereby contributing to obtaining the optimal design scheme for the composite absorbing material.
[0042] Furthermore, in the second embodiment, when the specified frequency band is a single frequency band, the fitness of each individual can be calculated according to the following exemplary formula.
[0043] Where F is the fitness of each individual, W is the first weighting factor corresponding to each individual, R is the reflectance of each individual at each frequency point in the specified frequency band, Ra is the target reflectance, m is the total number of frequency band units included in the specified frequency band, Δf is the frequency point step size value of the frequency band unit, and S(R) is the area enclosed by the reflectance curve formed by the reflectance of each frequency point in the specified frequency band and the ideal line segment, wherein the ideal line segment is the reflectance line segment formed when the reflectance value of each frequency point in the specified frequency band is the target reflectance.
[0044] In the second implementation, when there are multiple frequency bands specified, the fitness of each individual can be calculated according to the following exemplary formula.
[0045] Where F is the fitness of each individual, and W is the first weighting factor for each individual. w i The first for each individual i The second weighting factor corresponds to each specified frequency band, where n is the total number of specified frequency bands, and R is the reflectance of each individual at each frequency point within the i-th specified frequency band. R a For target reflectivity, m i S represents the total number of band elements contained in the i-th specified frequency band, Δf represents the frequency step size of the band element, and S i (R) is the area enclosed by the reflectivity curve formed by the reflectivity of each frequency point in the i-th specified frequency band and the ideal line segment. The ideal line segment is the reflectivity line segment formed when the reflectivity of each frequency point in the i-th specified frequency band is the target reflectivity.
[0046] Back Figure 1 In step S13, individuals are selected from the initial population based on fitness, and the selected individuals are crossovered and mutated to generate the next generation population.
[0047] In step S13, individuals can be selected from the initial population based on various methods. Two implementation methods are illustrated below.
[0048] In the first implementation, a lower fitness indicates a superior individual. In this implementation, the individual with the lowest fitness can be selected from the population that satisfies the constraints using a roulette wheel algorithm. For example, from the population that satisfies both the single-layer thickness constraint (i.e., the thickness of each layer corresponding to the individual is within the threshold range for each layer thickness) and the total thickness constraint (i.e., the total thickness corresponding to the individual is within the threshold range for the total thickness), the individual with the lowest fitness can be selected using a roulette wheel algorithm and used as the parent (or mother generation) individual to generate the next generation of the population.
[0049] In the second implementation, a lower fitness indicates a superior individual. In this implementation, for each generation from the initial population to the (P-1)th generation, individuals can be selected from the population as follows: a first set of individuals with a total thickness within a threshold range is selected, and the first individual with the lowest fitness is selected from this first set; a second set of individuals with a total thickness not within the threshold range is selected, and a second individual with a fitness less than or equal to that of the first individual is selected from this second set; the first and second individuals are used as the individuals selected from the initial population, and these selected individuals are used as the parent individuals for the next generation. Here, P represents the last generation, and P-1 represents the penultimate generation. By selecting parent individuals from the population using the above method, more high-performing individuals can be retained as parent individuals during the genetic process, thereby increasing the likelihood of finding the optimal design scheme for the composite absorbing material.
[0050] In addition to the two methods described above, other methods can also be used to select parent individuals from the population. For example, when higher fitness indicates a superior individual, the roulette wheel algorithm can be used to select the individual with the highest fitness from those individuals that meet the constraints. The specific method for selecting the individual with the highest fitness can be similar to the method for selecting the individual with the lowest fitness, and will not be elaborated upon here.
[0051] After selecting parent individuals from the initial population, crossover and mutation operations can be performed on these parent individuals to generate the next generation population. Specifically, existing crossover operators or custom crossover operators can be selected for crossover operations, and existing mutation operators or custom mutation operators can be selected for mutation operations. For example, a multi-point crossover operator can be selected for crossover operations on parent individuals, and a basic position mutation operator can be selected for mutation operations on parent individuals.
[0052] In step S14, the fitness calculation, individual selection, and crossover and mutation operations on the selected individuals are performed iteratively until the termination condition of the genetic algorithm is met.
[0053] There are several possible termination conditions for genetic algorithms. In some examples, the termination condition can be that the number of iterations reaches a specified number. For instance, if the preset total number of iterations is 150, the termination condition is met when the number of iterations reaches 150. In other examples, the termination condition is that there are individuals in the population with a fitness value lower than a preset fitness threshold.
[0054] If the termination condition is met, no new generation of population will be generated, and step S15 will be executed instead.
[0055] In step S15, the design scheme of the composite absorbing material is determined based on the individuals in the latest generation of the population.
[0056] In some embodiments, a lower fitness indicates a better individual. In these embodiments, the individual with the lowest fitness can be selected from the latest generation of the population that satisfies the single-layer thickness constraint (i.e., the thickness of each layer corresponding to the individual is within the threshold range of each layer thickness) and the total thickness constraint (i.e., the total thickness corresponding to the individual is within the threshold range of the total thickness). The design scheme represented by this individual is then used as the design scheme of the composite absorbing material.
[0057] In other embodiments, a higher fitness indicates a superior individual. In these embodiments, the individual with the highest fitness can be selected from the latest generation of the population that satisfies both the single-layer thickness constraint (i.e., the thickness of each layer corresponding to the individual is within the threshold range of each layer thickness) and the total thickness constraint (i.e., the total thickness corresponding to the individual is within the threshold range of the total thickness). The design scheme represented by this individual is then used as the design scheme for the composite absorbing material.
[0058] In the embodiments disclosed herein, the above methods can improve the computational efficiency of the design method for composite absorbing materials and improve the design effect.
[0059] like Figure 2 As shown, the design method of composite absorbing material based on genetic algorithm provided in some other embodiments of this disclosure includes steps S201 to S211.
[0060] In step S201, the parameters of the genetic algorithm, the target reflectivity, and the material data are input.
[0061] In some embodiments, a user can input parameters of a genetic algorithm, target reflectivity, and material data through a first device, and then send this data to a second device. Upon receiving this data, the second device automatically executes the design process for the composite microwave absorbing material. The first and second devices are different devices. For example, the first device could be a terminal device, and the second device a server.
[0062] In other embodiments, the user can input parameters of the genetic algorithm, target reflectivity, and material data through an interactive interface on the device (e.g., a terminal device or server). After receiving this data, the device automatically executes the design process for the composite absorbing material.
[0063] The parameters of a genetic algorithm may include the population size, the number of iterations, and the crossover rate. For example, the parameters of a genetic algorithm may include a population size of 300, a number of iterations of 150, and a crossover rate of Pc = 0.3.
[0064] The target reflectivity may include the expected reflectivity for each specified frequency band. For example, the expected reflectivity for the 4-8 GHz band is equal to -18 dB.
[0065] The material parameters may include the type of material, the frequency band of the material, the frequency step size, the number of layers of the composite absorbing material, the thickness threshold range of each layer, and the total thickness threshold range of the composite absorbing material. For example, the material parameters of the composite absorbing material include 6 types of materials, a frequency band of 2-10 GHz, a step size of 0.1 GHz, a 3-layer structure of the composite absorbing material, a thickness of 1-3 mm for each layer, and a total thickness of 4-5 mm. In step S202, let the algebraic number = 0.
[0066] In step S203, M individuals are randomly generated (i.e., the initial population).
[0067] In some embodiments, an initial population is generated through random encoding. For example, the data input by the user in step S201 is binary encoded to obtain a representative matrix of the initial population. For example, assuming the population contains 300 individuals, a 300*a binary matrix can be randomly generated, where each row of the matrix represents an individual, and 'a' is the number of bits in the binary encoding of the individual.
[0068] In step S204, decoding is performed.
[0069] By decoding the binary codes of each individual obtained in step 203, information such as the material type of each individual can be obtained.
[0070] In step S205, the reflectance of each individual is calculated.
[0071] In practice, the reflectance of each individual in the initial population can be calculated based on the reflectance calculation method described in the previous embodiments. Alternatively, reflectance can be calculated using other methods.
[0072] In step S206, the fitness of each individual is calculated.
[0073] In some embodiments, fitness can be calculated according to steps B1 to B3. In step B1, an objective function is constructed. In step B2, a fitness function is constructed based on the objective function for each individual. In step B3, the fitness value for each individual is calculated based on the fitness function.
[0074] In some examples, in step B1, an objective function can be constructed based on the integral of the difference between the individual's reflectance and the target reflectance. For a single computational target (e.g., a single frequency band), the objective function is equivalent to the area enclosed by the calculated reflectance curve and the corresponding target reflectance line segment on that frequency band. For example, as... Figure 2 As shown, individuals in frequency band [f a f b The area S is the region enclosed by the line segment containing the calculated reflectance curve and the target reflectance Ra. The smaller the area of this region, the closer the calculated reflectance is to the target reflectance.
[0075] Furthermore, in step B1, the constructed objective function can be: Where S(R) is the objective function value, representing the integral of the difference between the individual's reflectance and the target reflectance; R represents the individual's reflectance in the frequency band [f] calculated by the algorithm. a f b Reflectivity at frequencies within ]; R a This indicates that the individual is in the frequency band [f] a f b The target reflectivity at a frequency point within the specified range. The objective function constructed using the above formula can reflect the overall optimal solution for a frequency band.
[0076] Alternatively, in step B1, when the frequency step size values of the specified frequency band are all In this case, the constructed objective function can also be: Where S(R) is the objective function value, representing the integral of the difference between the individual's reflectance and the target reflectance; For frequency band [f a f b The frequency step size value of ]; Ra is the target reflectivity; R is the calculated value of the individual in the frequency band [f a ,f b The reflectivity of the frequency points within the specified frequency band; m is the total number of frequency band elements (or frequency points) contained in the specified frequency band. Furthermore, when the frequency step size for a specified frequency band is multiple different step size values, the step size can be unified by fitting material data. Then calculate the objective function according to the above formula.
[0077] In step B2, the fitness function can be constructed using various implementation methods. Two implementation methods are illustrated below. In the first implementation, the objective function is used as the fitness function. The smaller S(R) is, the better the optimization effect; therefore, the minimum S(R) can be directly used as the selection criterion for choosing mother individuals from the population.
[0078] In the second implementation, for a single computational target (e.g., a single frequency band), a fitness function is constructed based on the objective function and a first weighting factor. This fitness function satisfies: Where F is the fitness of each individual. W The first weighting factor for each individual, d / (T) max -T min () represents the degree of deviation. T toal The total thickness of the multi-layered structure possessed by each individual, T max This represents the upper limit of the total thickness threshold range. T min S(R) is the lower limit of the total thickness threshold range, k is a preset constant coefficient greater than 0, R is the reflectivity of each individual at each frequency point in the specified frequency band, Ra is the target reflectivity, m is the total number of frequency band units included in the specified frequency band, Δf is the frequency point step size value of the frequency band unit, and S(R) is the objective function.
[0079] Furthermore, in the second implementation, for multiple computational objectives (e.g., multiple frequency bands), a fitness function is constructed based on the objective function, a first weighting factor, and a second weighting factor. This fitness function satisfies: Where F is the fitness of each individual, W is the first weighting factor for each individual, and w i The second weighting factor is defined for the i-th specified frequency band of each individual, where n is the total number of specified frequency bands, R is the reflectance of each individual at each frequency point within the i-th specified frequency band, Ra is the target reflectance, and m... i For the first i The total number of frequency band elements contained in a specified frequency band, where Δf is the frequency step size of the frequency band element, S i (R) is the objective function corresponding to the i-th specified frequency band.
[0080] In step S207, it is determined whether the termination condition is met.
[0081] In some examples, the termination condition is that the number of iterations reaches a specified number. For example, the termination condition is determined to be met when the number of iterations reaches a preset 150.
[0082] If the termination condition is met, proceed to step S211; otherwise, proceed to steps S208 to S210.
[0083] In step S208, individuals are selected based on fitness.
[0084] For details on how to select individuals from the population as mothers based on fitness, please refer to the relevant descriptions in the examples described above.
[0085] In step S209, the selected individuals are crossovered and mutated to obtain the next generation population.
[0086] In step S210, let algebra = algebra + 1.
[0087] After step S210, steps 204 to 210 are executed iteratively until the termination condition is met.
[0088] In step S211, the design scheme of the composite absorbing material is determined based on the individuals in the latest generation of the population.
[0089] In some examples, the design scheme represented by the individual in the latest generation of the population that satisfies the total thickness constraint, the thickness constraint per layer, and has the lowest fitness is taken as the design scheme of the composite absorbing material.
[0090] In the embodiments disclosed herein, the above process can improve the computational efficiency of the design method for composite absorbing materials and improve the design effect.
[0091] The following two specific examples will be used to compare the design method of composite absorbing materials based on genetic algorithms and the ergonomic method provided in this disclosure.
[0092] In the first example, the composite absorbing material was set as follows: 6 materials, with a frequency band of 2-10 GHz and a step size of 0.1 GHz; the constraints were set as follows: a 3-layer structure, with each layer having a thickness of 1-3 mm and a total thickness of 4-5 mm; the optimization objective was set as: a reflectivity of -18 dB in the 4-8 GHz range; and the genetic algorithm's calculation parameters were set as follows: population size 300, number of iterations 150, crossover rate Pc=0.3, and mutation rate... P m (k) The method of variable crossover rate is adopted, and the formula for calculating the mutation rate satisfies: Where k is the current iteration number,L d The length of the binary code is 20.
[0093] The calculation parameters for the composite absorbing material design based on the ergodic method were set as follows: the thickness step of each material layer was 0.1 mm. The calculation results obtained from these two methods are shown in Table 1, and the comparison graph of reflectivity obtained from these two methods is shown below. Figure 4 As shown.
[0094] Table 1 In the second example, the composite absorbing material was changed to nine types, and the constraint condition was changed to a four-layer structure, with each layer having a thickness of 0.5-2 mm and a total thickness of 4-5 mm. Other parameters in this example are the same as in the first example. The calculation results obtained based on the algorithm and ergonomic method of this disclosure are shown in Table 2, and the comparison graph of reflectivity obtained based on these two methods is shown below. Figure 5 As shown.
[0095] Table 2 As can be seen from the fitness F, the optimization result of the improved genetic algorithm proposed in this disclosure is very close to the optimal solution, differing by only about 1%, which has almost no impact on actual processing, while the computation speed is much faster than the traversal method. Moreover, as the number of computation layers and the types of materials increase, the computation time of the genetic algorithm increases linearly, while the computation time of the traversal method increases exponentially. When calculating complex problems, the computation speed advantage of the improved genetic algorithm provided in this disclosure becomes even more obvious. As can be seen from the computation time, in the first example, the computation speed based on the genetic algorithm provided in this disclosure is nearly 18 times that of the traversal method, and in the second example, the computation speed based on the genetic algorithm provided in this disclosure is nearly 2000 times that of the traversal method. It is evident that using this method can significantly increase computational efficiency and reduce computation time. At the same time, the higher the computational complexity, the more obvious the advantage of the method provided in this disclosure in terms of computational efficiency becomes.
[0096] like Figure 6 As shown, the design apparatus 60 for composite absorbing materials based on genetic algorithms provided in some embodiments of this disclosure includes a generation module 61, a fitness calculation module 62, a genetic module 63, and a determination module 64.
[0097] The generation module 61 is configured to randomly generate an initial population. The initial population includes multiple individuals, each representing a candidate design scheme for a composite absorbing material. The candidate design scheme includes the following design variables: the material type of each layer in the multi-layered structure of the composite absorbing material, and the thickness of each layer.
[0098] The fitness calculation module 62 is configured to calculate the reflectance of each individual in the initial population at each frequency point within a specified frequency band; determine the first weighting factor corresponding to each individual based on the deviation of the total thickness of the multilayer structure possessed by each individual from the total thickness threshold range; and calculate the fitness of each individual based on the difference between the reflectance of each individual at each frequency point within the specified frequency band and the target reflectance, as well as the first weighting factor corresponding to each individual.
[0099] Genetic module 63 is configured to select individuals from the initial population based on the fitness of each individual, and perform crossover and mutation operations on the selected individuals to generate the next generation population; iteratively performing fitness calculation, individual selection, and crossover and mutation operations on the selected individuals until the termination condition of the genetic algorithm is met.
[0100] The determination module 64 is configured to determine the design scheme of the composite absorbing material based on individuals in the latest generation of the population that meet the termination conditions.
[0101] In the embodiments disclosed herein, the above-mentioned device can significantly improve computational efficiency and reduce computation time. In particular, the advantage of its high computational efficiency becomes more and more obvious as the computational complexity increases.
[0102] like Figure 7 As shown, in some embodiments of this disclosure, the electronic device 70 includes a memory 71 and a processor 72 coupled to the memory 71. The memory 71 is used to store instructions for executing embodiments of the design method for composite absorbing materials based on genetic algorithms as described above. The processor 72 is configured to execute the design method for composite absorbing materials based on genetic algorithms in any of the embodiments of this disclosure based on the instructions stored in the memory 71.
[0103] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A design method for composite absorbing materials based on genetic algorithms, characterized in that, The method includes: An initial population is randomly generated, wherein the initial population includes multiple individuals, each individual representing a candidate design scheme for a composite absorbing material, and the candidate design scheme includes the following design variables: the material type of each layer in the multi-layer structure of the composite absorbing material, and the thickness of each layer; Calculate the reflectance of each individual in the initial population at each frequency point within a specified frequency band; determine the first weighting factor corresponding to each individual based on the deviation of the total thickness of the multilayer structure possessed by each individual from the total thickness threshold range; calculate the fitness of each individual based on the difference between the reflectance of each individual at each frequency point within the specified frequency band and the target reflectance, and the first weighting factor corresponding to each individual. When the specified frequency band is a single band, the fitness of each individual is calculated according to the following formula: Where F is the fitness of each individual, W is the first weighting factor corresponding to each individual, R is the reflectance of each individual at each frequency point in the specified frequency band, Ra is the target reflectance, m is the total number of frequency band units included in the specified frequency band, Δf is the frequency point step size value of the frequency band unit, and S(R) is the area enclosed by the reflectance curve formed by the reflectance of each frequency point in the specified frequency band and the ideal line segment, wherein the ideal line segment is the reflectance line segment formed when the reflectance value of each frequency point in the specified frequency band is the target reflectance. When the specified frequency band is multiple frequency bands, the fitness of each individual is calculated according to the following formula: Where F is the fitness of each individual, W is the first weighting factor for each individual, and w i The second weighting factor is defined for the i-th specified frequency band of each individual, where n is the total number of specified frequency bands, R is the reflectance of each individual at each frequency point within the i-th specified frequency band, Ra is the target reflectance, and m... i For the first i The total number of frequency band elements contained in a specified frequency band, where Δf is the frequency step size value of the frequency band element, S i (R) is the area enclosed by the reflectivity curve formed by the reflectivity of each frequency point in the i-th specified frequency band and the ideal line segment, wherein the ideal line segment is the reflectivity line segment formed when the reflectivity of each frequency point in the i-th specified frequency band is the target reflectivity; Individuals are selected from the initial population based on the fitness of each individual, and crossover and mutation operations are performed on the selected individuals to generate the next generation population. From the initial population, a first set is selected consisting of individuals whose total thickness is within the threshold range of the total thickness, and from the first set, the first individual with the lowest fitness is selected. A second set is formed by selecting individuals from the initial population whose total thickness is not within the total thickness threshold range, and a second set is formed by selecting individuals from the second set whose fitness is less than or equal to that of the first individual. The first individual and the second individual are selected from the initial population; The process iteratively calculates fitness, selects individuals, and performs crossover and mutation operations on the selected individuals until the termination condition of the genetic algorithm is met. The design scheme of the composite absorbing material is determined based on the individuals in the latest generation of the population that meet the termination conditions.
2. The method according to claim 1, characterized in that: When the total thickness of the multilayer structure possessed by each individual is within the total thickness threshold range, the deviation is 0; when the total thickness of the multilayer structure possessed by each individual is not within the total thickness threshold range, the deviation is greater than 0; and the first weighting factor is positively correlated with the deviation.
3. The method according to claim 2, characterized in that, The first weighting factor for each individual is determined according to the following formula: in, W The first weighting factor for each individual, d / (T) max -T min The degree of deviation is denoted as ). T toal The total thickness of the multi-layered structure possessed by each individual, T max This is the upper limit of the total thickness threshold range. T min is the lower limit of the total thickness threshold range, and k is a preset constant coefficient greater than 0.
4. A design device for composite absorbing materials based on genetic algorithms, characterized in that: Includes a design method for composite absorbing materials based on genetic algorithms as described in any one of claims 1-3.
5. An electronic device, comprising: Memory; as well as A processor coupled to the memory, the processor being configured to execute the design method of the composite absorbing material based on a genetic algorithm as described in any one of claims 1 to 3, based on instructions stored in the memory.
6. A computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the design method of composite absorbing material based on genetic algorithm as described in any one of claims 1 to 3.
7. A computer program product having stored computer program instructions thereon, which, when executed by a processor, implement the design method for composite absorbing materials based on genetic algorithms as described in any one of claims 1 to 3.