Band-pass filter design method and system and storage medium

Abstract

The invention provides a band-pass filter design method and system and a storage medium, and the method comprises the steps: randomly generating a plurality of film system parameter matrixes, taking the plurality of film system parameter matrixes as an initial population, and enabling the film system parameter matrixes to be constructed according to the material types and thicknesses corresponding to a plurality of film layers of a band-pass filter; and performing iterative optimization on the initial population based on an unbalanced multi-modal multi-objective optimization algorithm of two-stage evolution, obtaining a band-pass curve corresponding to the membrane system parameter matrix, obtaining band-pass performance correspondingly embodied according to the band-pass curve, performing iterative optimization on the population according to the band-pass performance, and obtaining the membrane system parameter matrix. Outputting and obtaining a plurality of film system parameter matrix optimal solutions until a preset number of iterations is met; and randomly selecting one from the optimal solutions of the plurality of film system parameter matrixes, determining material types and thicknesses corresponding to a plurality of film layers, and designing to obtain the band-pass filter. The design efficiency of the band-pass filter is improved.

Description

Technical Field

[0001] The present invention is applicable to the field of optical optimization technology, and in particular relates to a bandpass filter design method, system and storage medium. Background Art

[0002] Bandpass filters are widely used in various optical scenarios. Within the working band, their characteristic is that they only allow light in the specified band to pass through and cut off light in other bands.

[0003] Bandpass filters can be fabricated by applying multilayer coatings of various materials, such as dielectrics and metals, at varying film thicknesses onto a transparent substrate. Performance indicators such as the center wavelength, half-width (FWHM), transmittance, and optical density of a bandpass filter are determined by the materials and thicknesses of each film layer during fabrication. Once the film system parameters—that is, the materials and thicknesses of each film layer—are determined, the bandpass curve corresponding to these parameters can be calculated using the finite-difference time-domain (FDTD) method. Therefore, when fabricating a bandpass filter using multilayer coating, the first step is to design the film system parameters based on the four desired bandpass specifications: center wavelength, half-width (FWHM), transmittance, and optical density. This ensures that the corresponding bandpass curve meets these desired bandpass specifications.

[0004] It can be seen that the design of film parameters is essentially a multi-objective optimization problem. Due to the highly complex mapping relationship between film parameters and bandpass curves, this problem is multimodal and a multi-modal, multi-objective optimization problem. Furthermore, due to the specific requirements for the materials used in specific film layers in practice, this multi-modal problem is unbalanced. Therefore, the design of film parameters is a complex, unbalanced, multi-modal, multi-objective optimization problem.

[0005] During implementation, the design of membrane parameters is an NP-problem, meaning that existing mathematical methods cannot provide an exact solution. A common approach is to employ intelligent optimization methods to find an acceptable optimal solution. However, the performance of existing optimization methods still leaves much room for improvement. Summary of the Invention

[0006] The present invention solves the technical problem that it is difficult to find the optimal solution for the existing multi-objective optimization problem of film system parameters.

[0007] To solve the above technical problems, in a first aspect, the present invention provides a bandpass filter design method, comprising the following steps: S101, randomly generating a plurality of film parameter matrices, and using the plurality of film parameter matrices as an initial population, wherein the film parameter matrices are constructed by material types and thicknesses corresponding to a plurality of film layers of a bandpass filter; S102, iteratively optimizing the initial population using an unbalanced multi-modal multi-objective optimization algorithm based on a two-stage evolution, wherein a bandpass curve corresponding to the film parameter matrix is ​​obtained, and a bandpass performance corresponding to the bandpass curve is obtained based on the bandpass performance. The population is iteratively optimized based on the bandpass performance until a preset number of iterations is met, and multiple optimal solutions of the film parameter matrices are output; S103 , randomly selecting one from the plurality of optimal solutions of the film parameter matrix, determining the material type and thickness corresponding to the plurality of film layers, and designing the bandpass filter.

[0008] Furthermore, in step S102, the unbalanced multi-modal multi-objective optimization algorithm based on two-stage evolution includes the following steps: S201, initialize the number of iterations to 0; S202: When the number of iterations is less than the product of the preset stage demarcation point and the preset number of iterations, the process enters the first iteration stage and executes step S203; otherwise, the process enters the second iteration stage and executes step S206: S203, generating a mating pool according to the Pareto dominance ranks corresponding to different individuals in the current population, and generating an offspring population by simulating binary crossover and polynomial mutation, wherein the Pareto dominance rank is obtained by comparing the bandpass performance corresponding to the current individual with the bandpass performance corresponding to other individuals according to the Pareto dominance relationship; S204. Obtain the fitness of each individual in the first set of parent and child populations in the current iteration phase, and process the convergence of the individuals in the first set according to a preset convergence strategy. The fitness indicates the degree to which an individual is likely to be retained during the population iteration process, and the convergence indicates the degree to which an individual is close to the optimal solution. S205, selecting individuals from the first set according to the fitness to form the next generation population, increasing the number of iterations by 1, and returning to step S202; S206, clustering the current population using a clustering algorithm to obtain multiple subpopulations, each of which corresponds to an optimal solution region; S207. In each of the optimal solution regions, generating a mating pool according to the Pareto dominance levels corresponding to different individuals, and generating an offspring population by simulating binary crossover and polynomial mutation; S208, merging each of the subpopulations and its corresponding offspring population to form a second set, and counting the global non-dominated individuals and dominated individuals with respect to the Pareto dominance level in different second sets to obtain a global non-dominated individual set and a global dominated individual set; S209, if the size of the global non-dominated individual set is smaller than the size of the initial population, move the current non-dominated individuals in the dominated individual set into the non-dominated individual set until its size becomes the same as the initial population; S210: Determine whether the number of the globally non-dominated individuals in the subpopulation is less than a preset convergence threshold: If so, individuals are selected from the non-dominated individual set based on population diversity to form the next generation population; If not, individuals are selected from the non-dominated individual set based on the convergence and the fitness to form the next generation population; S211, adding 1 to the number of iterations. If the number of iterations meets the preset number of iterations, outputting the optimal individual in each of the current optimal solution regions to obtain multiple optimal solutions of the film system parameter matrix; otherwise, returning to step S206.

[0009] Furthermore, the population of the current iteration stage is defined as , where the individuals are In step S204, the preset convergence strategy is used for individual In the population The process of processing the convergence in the above method satisfies the following relationship: ; in, Represents an individual In the population The convergence of , its minimum value is 0, Represents an individual population The number of times other individuals dominate, Represents population The average number of times each individual in is dominated by other individuals, and Respectively represent the current number of iterations and the preset number of iterations, Indicates the preset stage demarcation point, which is a parameter with a value between 0 and 1, and: ; ; To judge individuals Dominate the individual , When represents an individual Pareto dominated individuals .

[0010] Furthermore, in step S210, the preset convergence threshold is specifically: The product of the subpopulation convergence state judgment threshold τ and the average number of the global non-dominated individuals in each of the subpopulations is preset.

[0011] In a second aspect, the present invention further provides a bandpass filter design system, comprising: an initial generation module, configured to randomly generate a plurality of film parameter matrices and use the plurality of film parameter matrices as an initial population, wherein the film parameter matrices are constructed by material types and thicknesses corresponding to a plurality of film layers of the bandpass filter; an optimization module for iteratively optimizing the initial population based on an unbalanced multi-modal multi-objective optimization algorithm using a two-stage evolutionary method, wherein a bandpass curve corresponding to the film parameter matrix is ​​obtained, and a bandpass performance corresponding to the bandpass curve is obtained based on the bandpass curve, and the population is iteratively optimized based on the bandpass performance until a preset number of iterations is met, and multiple optimal solutions of the film parameter matrices are output; The design module is used to select any one of the optimal solutions of the film parameter matrix, determine the material type and thickness corresponding to the multiple film layers, and design the bandpass filter.

[0012] Furthermore, the unbalanced multi-modal multi-objective optimization algorithm based on two-stage evolution executed by the optimization module includes the following steps: S201, initialize the number of iterations to 0; S202: When the number of iterations is less than the product of the preset stage demarcation point and the preset number of iterations, the process enters the first iteration stage and executes step S203; otherwise, the process enters the second iteration stage and executes step S206: S203, generating a mating pool according to the Pareto dominance ranks corresponding to different individuals in the current population, and generating an offspring population by simulating binary crossover and polynomial mutation, wherein the Pareto dominance rank is obtained by comparing the bandpass performance corresponding to the current individual with the bandpass performance corresponding to other individuals according to the Pareto dominance relationship; S204. Obtain the fitness of each individual in the first set of parent and child populations in the current iteration phase, and process the convergence of the individuals in the first set according to a preset convergence strategy. The fitness indicates the degree to which an individual is likely to be retained during the population iteration process, and the convergence indicates the degree to which an individual is close to the optimal solution. S205, selecting individuals from the first set according to the fitness to form the next generation population, increasing the number of iterations by 1, and returning to step S202; S206, clustering the current population using a clustering algorithm to obtain multiple subpopulations, each of which corresponds to an optimal solution region; S207. In each of the optimal solution regions, generating a mating pool according to the Pareto dominance levels corresponding to different individuals, and generating an offspring population by simulating binary crossover and polynomial mutation; S208, merging each of the subpopulations and its corresponding offspring population to form a second set, and counting the global non-dominated individuals and dominated individuals with respect to the Pareto dominance level in different second sets to obtain a global non-dominated individual set and a global dominated individual set; S209, if the size of the global non-dominated individual set is smaller than the size of the initial population, move the current non-dominated individuals in the dominated individual set into the non-dominated individual set until its size becomes the same as the initial population; S210: Determine whether the number of the globally non-dominated individuals in the subpopulation is less than a preset convergence threshold: If so, individuals are selected from the non-dominated individual set based on population diversity to form the next generation population; If not, individuals are selected from the non-dominated individual set based on the convergence and the fitness to form the next generation population; S211, adding 1 to the number of iterations. If the number of iterations meets the preset number of iterations, outputting the optimal individual in each of the current optimal solution regions to obtain multiple optimal solutions of the film system parameter matrix; otherwise, returning to step S206.

[0013] Furthermore, the population of the current iteration stage is defined as , where the individuals are In step S204, the preset convergence strategy is used for individual In the population The process of processing the convergence in the above method satisfies the following relationship: ; in, Represents an individual In the population The convergence of , its minimum value is 0, Represents an individual population The number of times other individuals dominate, Represents population The average number of times each individual in is dominated by other individuals, and Respectively represent the current number of iterations and the preset number of iterations, Indicates the preset stage demarcation point, which is a parameter with a value between 0 and 1, and: ; ; To judge individuals Dominate the individual , When represents an individual Pareto dominated individuals .

[0014] Furthermore, in step S210, the preset convergence threshold is specifically: The product of the subpopulation convergence state judgment threshold τ and the average number of the global non-dominated individuals in each of the subpopulations is preset.

[0015] In a third aspect, the present invention further provides a computer device comprising: a memory, a processor, and a bandpass filter design program stored in the memory and executable on the processor, wherein when the processor executes the bandpass filter design program, the steps of the bandpass filter design method as described in any one of the above embodiments are implemented.

[0016] In a fourth aspect, the present invention further provides a storage medium storing a bandpass filter design program, which, when executed by a processor, implements the steps of the bandpass filter design method as described in any one of the above embodiments.

[0017] The beneficial effect achieved by the present invention lies in proposing a bandpass filter design method based on a two-stage evolutionary unbalanced multimodal multi-objective optimization method. This method takes the film parameter matrix of the bandpass filter as the optimization object, and through a phased population convergence adjustment strategy and an environment selection method, effectively solves the problem of difficult-to-search areas being ignored or underdeveloped in the unbalanced multimodal multi-objective optimization, facilitates the rapid acquisition of optimal solutions for multiple areas, and improves the efficiency of bandpass filter design. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The present invention will be described in detail below with reference to the accompanying drawings. The above and other aspects of the present invention will become clearer and easier to understand through the detailed description made with reference to the following drawings. In the accompanying drawings: Figure 1 1. It is a schematic flow chart of the steps of the bandpass filter design method provided by the present invention; Figure 2 1 is a schematic structural diagram of a bandpass filter design system provided by an embodiment of the present invention; Figure 3 It is a structural diagram of a computer device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0020] Example 1 Please refer to Figure 1 , Figure 1 1 is a flow chart of the steps of the bandpass filter design method provided by the present invention, wherein the bandpass filter design method comprises the following steps: S101 , randomly generating a plurality of film parameter matrices, and using the plurality of film parameter matrices as an initial population, wherein the film parameter matrices are constructed by material types and thicknesses corresponding to a plurality of film layers of a bandpass filter.

[0021] In the embodiment of the present invention, the film parameter matrix is ​​constructed by taking the film material type and thickness as matrix elements. For example, a film parameter matrix is ​​defined as C: ; Among them, each row in C corresponds to a film layer, Indicates the refractive index of the material used in the first film layer. Indicates the extinction coefficient of the material used in the first film layer, Indicates the thickness of the first film layer; Represents the refractive index of the material used in the mth film layer, Represents the extinction coefficient of the material used in the mth film layer, Represents the thickness of the mth film layer.

[0022] The method of the embodiment of the present invention is to solve the optimal solution that can be achieved for the possible combination of film layer materials and thicknesses, so the matrix planning and design can be carried out for different types of film layer materials, wherein the film layer materials include but are not limited to various metal materials, such as aluminum (Al), silver (Ag), gold (Au), germanium (Ge), various oxide materials, such as titanium dioxide , silicon dioxide , alumina , zirconium oxide , yttrium trioxide , hafnium oxide , various fluoride materials, such as magnesium fluoride , calcium fluoride , various sulfide / selenide materials, such as zinc sulfide , zinc selenide , and various special functional materials such as silicon nitride wait.

[0023] S102. An unbalanced multi-modal multi-objective optimization algorithm based on two-stage evolution is used to iteratively optimize the initial population, wherein a bandpass curve corresponding to the film system parameter matrix is ​​obtained, and the bandpass performance corresponding to the bandpass curve is obtained according to the bandpass curve. The population is iteratively optimized according to the bandpass performance until a preset number of iterations is met, and multiple optimal solutions of the film system parameter matrices are output.

[0024] The bandpass curve can be obtained based on the finite-difference time-domain method described in the background art. An unbalanced multi-modal multi-objective optimization algorithm based on two-stage evolution is a multi-objective optimization solution method proposed in an embodiment of the present invention.

[0025] Specifically, in step S102, the unbalanced multi-modal multi-objective optimization algorithm based on two-stage evolution includes the following steps: S201, initialize the number of iterations to 0; S202: When the number of iterations is less than the product of the preset stage demarcation point and the preset number of iterations, the process enters the first iteration stage and executes step S203; otherwise, the process enters the second iteration stage and executes step S206: S203, generating a mating pool according to the Pareto dominance ranks corresponding to different individuals in the current population, and generating an offspring population by simulating binary crossover and polynomial mutation, wherein the Pareto dominance rank is obtained by comparing the bandpass performance corresponding to the current individual with the bandpass performance corresponding to other individuals according to the Pareto dominance relationship; S204. Obtain the fitness of each individual in the first set of parent and child populations in the current iteration phase, and process the convergence of the individuals in the first set according to a preset convergence strategy. The fitness indicates the degree to which an individual is likely to be retained during the population iteration process, and the convergence indicates the degree to which an individual is close to the optimal solution. S205, selecting individuals from the first set according to the fitness to form the next generation population, increasing the number of iterations by 1, and returning to step S202; S206, clustering the current population using a clustering algorithm to obtain multiple subpopulations, each of which corresponds to an optimal solution region; S207. In each of the optimal solution regions, generating a mating pool according to the Pareto dominance levels corresponding to different individuals, and generating an offspring population by simulating binary crossover and polynomial mutation; S208, merging each of the subpopulations and its corresponding offspring population to form a second set, and counting the global non-dominated individuals and dominated individuals with respect to the Pareto dominance level in different second sets to obtain a global non-dominated individual set and a global dominated individual set; S209, if the size of the global non-dominated individual set is smaller than the size of the initial population, move the current non-dominated individuals in the dominated individual set into the non-dominated individual set until its size becomes the same as the initial population; S210: Determine whether the number of the globally non-dominated individuals in the subpopulation is less than a preset convergence threshold: If so, individuals are selected from the non-dominated individual set based on population diversity to form the next generation population; If not, individuals are selected from the non-dominated individual set based on the convergence and the fitness to form the next generation population; S211, adding 1 to the number of iterations. If the number of iterations meets the preset number of iterations, outputting the optimal individual in each of the current optimal solution regions to obtain multiple optimal solutions of the film system parameter matrix; otherwise, returning to step S206.

[0026] The unbalanced multi-modal multi-objective optimization algorithm based on two-stage evolution proposed in the embodiment of the present invention is a multi-objective optimization method, which involves the concepts of Pareto dominance level and Pareto dominance relationship. For ease of understanding, the embodiment of the present invention explains the above concepts: Pareto dominance is the core rule for judging the performance of two individuals in multi-objective optimization. In simple terms, for two individuals p and q: If q performs as well as p on all objectives (i.e., the objective values ​​are equal or better), and its performance on at least one objective is significantly better than p, then it is said that "q Pareto dominates p", denoted as ; On the contrary, if the two do not satisfy the above conditions (that is, each has its own advantages and disadvantages), then they are in a "non-dominant" relationship.

[0027] The Pareto dominance hierarchy is a classification of the superiority and inferiority of all individuals in a population based on the above dominance relationship. The core is: Level 1: Individuals in the population that are not dominated by any other individuals (i.e., the first tier of global optimality).

[0028] Level 2: Individuals that are dominated only by individuals in Level 1 and not by individuals in other levels (second tier). Similarly, the smaller the level number, the better the individual's overall performance in the population.

[0029] In the embodiment of the present invention, the dominance rank is used to generate the mating pool (individuals with higher rank are preferentially selected as parents), which is an important basis for guiding the population to evolve in a more optimal direction.

[0030] In the embodiment of the present invention, the population in the current iteration phase is defined as , where the individuals are In step S204, the preset convergence strategy is used for individual In the population The process of processing the convergence in the above method satisfies the following relationship: ; Represents an individual In the population The convergence of , its minimum value is 0, Represents an individual population The number of times other individuals dominate, Represents population The average number of times each individual in is dominated by other individuals, and Respectively represent the current number of iterations and the preset number of iterations, Indicates the preset stage demarcation point, which is a parameter with a value between 0 and 1, and: ; ; To judge individuals Dominate the individual , When represents an individual Pareto dominated individuals .

[0031] Step S205 selects individuals from the first set according to fitness to form the next generation population. Individuals in In terms of The larger the value is, the worse the current convergence level is. When other indicators such as diversity are the same, individual The greater the possibility of being deleted during the environmental selection process, the less likely it is that the area where it is located will be further explored in the subsequent evolution. However, the difficult search area of ​​multi-objective optimization is precisely the area where the individual convergence degree is relatively poor. Measuring the convergence of individuals is not conducive to fully exploring difficult-to-search areas.

[0032] Therefore, based on the convergence calculation method in the above embodiment, the embodiment of the present invention is Based on the current population average dominance number, , and adopts truncation mechanism By limiting the lower limit of the convergence value, the individual convergence value in the easy-to-search area with good convergence degree can be reduced to a limited extent, and the individual convergence value in the difficult-to-search area can be reduced to a greater extent, thereby reducing the overall difference in the individual convergence values ​​in different areas and reducing the impact of convergence on the optimization process; Furthermore, as the optimization process progresses, the convergence of individuals in each region will generally get better and better, and the number of times each individual is dominated by other individuals will become less and less. In order to avoid the influence of excessive force on the convergence value, which will cause the convergence to lose its necessary influence on the optimization process, adaptive weights are used in the calculation process of convergence. Gradually reduce the average number of dominant species in the current population that is reduced in each round of evolution in the first iteration Specifically, in the convergence calculation of the initial iteration, the maximum reduction for each individual is twice , and then based on the number of iterations The growth is gradually decreasing When entering the last iteration of the first stage, the reduction factor is reduced to one. In addition, In the expression The reason for lowering the individual convergence value for the benchmark is that This effectively characterizes the overall convergence state of the current population, providing a reasonable basis for the reduction range of convergence values ​​in each iteration. This design avoids excessive reductions that cause the convergence of most individuals to reach zero, thus losing the significance of individual convergence differences in the optimization process. It also avoids insufficient reductions that cause the convergence of most individuals to fall below zero, resulting in only the same degree of reduction, which fails to differentiate the individual convergence values ​​in different regions.

[0033] In general, in the first iterative stage, the number of individuals in the easy-to-search area of ​​the final population will still be relatively large, and the convergence will be relatively better, but the diversity will be relatively poor; the number of individuals in the difficult-to-search area will be relatively small, and the convergence will be relatively poor, but the diversity will be relatively good.

[0034] In the second iterative phase, a clustering algorithm is first used to identify multiple optimal solution regions. Different environmental selection operators (S210) are then employed based on the current number of globally non-dominated individuals in each subpopulation corresponding to each optimal solution region. This is because if the current number of globally non-dominated individuals in one or more subpopulations is less than a preset convergence threshold, this indicates that the current convergence level of the regions corresponding to these subpopulations is significantly lower than that of other regions, and additional development efforts in these regions are required to achieve balanced development of the optimal solutions. Conversely, if the current number of globally non-dominated individuals in all subpopulations is greater than the preset convergence threshold, this indicates that the current convergence levels of the regions corresponding to the subpopulations are relatively close. Given the inherent randomness of the evolutionary algorithm, balanced development of the optimal solutions can be achieved without additional attention to the development of certain regions.

[0035] Specifically, in step S210, the preset convergence threshold is: The product of the subpopulation convergence state judgment threshold τ and the average number of the global non-dominated individuals in each of the subpopulations is preset.

[0036] In step S210, when selecting individuals from the set of non-dominated individuals obtained based on the dominance relationship to generate the next generation population, the selection is made only from the perspective of diversity without considering convergence, thereby tending to retain relevant individuals in the difficult-to-search area with relatively good diversity but relatively poor convergence. In other words, the development of the difficult-to-search area is strengthened, which is conducive to the balanced development of the optimal solution area. When no regions require additional development during the optimization process, a fitness-based environmental selection operator is employed. Fitness is defined conventionally as the sum of convergence and diversity, with convergence directly expressed as the number of times the current individual is dominated by others in the population. In this case, when selecting individuals from the set of non-dominated individuals to form the next generation population, both convergence and diversity are considered, ensuring balanced development of the optimal solution region.

[0037] S103 , randomly selecting one from the plurality of optimal solutions of the film parameter matrix, determining the material type and thickness corresponding to the plurality of film layers, and designing the bandpass filter.

[0038] It's understandable that because multiple optimal solution regions are searched and optimized simultaneously, the number of optimal solutions ultimately obtained is the same as the number of regions identified by clustering. For multi-objective optimization, these optimal solutions are all usable and meet the originally set optimization objective (bandpass performance). For bandpass filter design, simply select one of the multiple optimal solutions for the film parameter matrix and design and manufacture the bandpass filter based on the corresponding film layer, material, and thickness.

[0039] The beneficial effect achieved by the present invention lies in proposing a bandpass filter design method based on a two-stage evolutionary unbalanced multimodal multi-objective optimization method. This method takes the film parameter matrix of the bandpass filter as the optimization object, and through a phased population convergence adjustment strategy and an environment selection method, effectively solves the problem of difficult-to-search areas being ignored or underdeveloped in the unbalanced multimodal multi-objective optimization, facilitates the rapid acquisition of optimal solutions for multiple areas, and improves the efficiency of bandpass filter design.

[0040] Example 2 The embodiment of the present invention also provides a bandpass filter design system 300, please refer to Figure 2 , Figure 2 : is a schematic diagram of the structure of a bandpass filter design system provided by an embodiment of the present invention, which includes: An initial generation module 301 is used to randomly generate a plurality of film parameter matrices and use the plurality of film parameter matrices as an initial population, wherein the film parameter matrices are constructed by material types and thicknesses corresponding to a plurality of film layers of the bandpass filter; an optimization module 302 for iteratively optimizing the initial population based on an unbalanced multi-modal multi-objective optimization algorithm using a two-stage evolutionary method, wherein a bandpass curve corresponding to the film parameter matrix is ​​obtained, and a bandpass performance corresponding to the bandpass curve is obtained based on the bandpass curve. The population is iteratively optimized based on the bandpass performance until a preset number of iterations is met, and multiple optimal solutions of the film parameter matrices are output; The design module 303 is used to select any one of the optimal solutions of the film parameter matrix, determine the material type and thickness corresponding to the multiple film layers, and design the bandpass filter.

[0041] The bandpass filter design system 300 can implement the steps in the bandpass filter design method in the above embodiment and can achieve the same technical effects. Please refer to the description in the above embodiment and will not be repeated here.

[0042] Example 3 The embodiment of the present invention also provides a computer device, please refer to Figure 3 , Figure 3 4 is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. The computer device 400 includes: a memory 402, a processor 401, and a program stored in the memory 402 and executable on the processor 401.

[0043] The processor 401 calls the program stored in the memory 402 to execute the steps of the method provided in the embodiment of the present invention. Figure 1 , specifically including the following steps: S101, randomly generating a plurality of film parameter matrices, and using the plurality of film parameter matrices as an initial population, wherein the film parameter matrices are constructed by material types and thicknesses corresponding to a plurality of film layers of a bandpass filter; S102, iteratively optimizing the initial population using an unbalanced multi-modal multi-objective optimization algorithm based on a two-stage evolution, wherein a bandpass curve corresponding to the film parameter matrix is ​​obtained, and a bandpass performance corresponding to the bandpass curve is obtained based on the bandpass performance. The population is iteratively optimized based on the bandpass performance until a preset number of iterations is met, and multiple optimal solutions of the film parameter matrices are output; S103 , randomly selecting one from the plurality of optimal solutions of the film parameter matrix, determining the material type and thickness corresponding to the plurality of film layers, and designing the bandpass filter.

[0044] Specifically, in step S102, the unbalanced multi-modal multi-objective optimization algorithm based on two-stage evolution includes the following steps: S201, initialize the number of iterations to 0; S202: When the number of iterations is less than the product of the preset stage demarcation point and the preset number of iterations, the process enters the first iteration stage and executes step S203; otherwise, the process enters the second iteration stage and executes step S206: S203, generating a mating pool according to the Pareto dominance ranks corresponding to different individuals in the current population, and generating an offspring population by simulating binary crossover and polynomial mutation, wherein the Pareto dominance rank is obtained by comparing the bandpass performance corresponding to the current individual with the bandpass performance corresponding to other individuals according to the Pareto dominance relationship; S204. Obtain the fitness of each individual in the first set of parent and child populations in the current iteration phase, and process the convergence of the individuals in the first set according to a preset convergence strategy. The fitness indicates the degree to which an individual is likely to be retained during the population iteration process, and the convergence indicates the degree to which an individual is close to the optimal solution. S205, selecting individuals from the first set according to the fitness to form the next generation population, increasing the number of iterations by 1, and returning to step S202; S206, clustering the current population using a clustering algorithm to obtain multiple subpopulations, each of which corresponds to an optimal solution region; S207. In each of the optimal solution regions, generating a mating pool according to the Pareto dominance levels corresponding to different individuals, and generating an offspring population by simulating binary crossover and polynomial mutation; S208, merging each of the subpopulations and its corresponding offspring population to form a second set, and counting the global non-dominated individuals and dominated individuals with respect to the Pareto dominance level in different second sets to obtain a global non-dominated individual set and a global dominated individual set; S209, if the size of the global non-dominated individual set is smaller than the size of the initial population, move the current non-dominated individuals in the dominated individual set into the non-dominated individual set until its size becomes the same as the initial population; S210: Determine whether the number of the globally non-dominated individuals in the subpopulation is less than a preset convergence threshold: If so, individuals are selected from the non-dominated individual set based on population diversity to form the next generation population; If not, individuals are selected from the non-dominated individual set based on the convergence and the fitness to form the next generation population; S211, adding 1 to the number of iterations. If the number of iterations meets the preset number of iterations, outputting the optimal individual in each of the current optimal solution regions to obtain multiple optimal solutions of the film system parameter matrix; otherwise, returning to step S206.

[0045] Among them, the population of the current iteration stage is defined as , where the individuals are In step S204, the preset convergence strategy is used for individual In the population The process of processing the convergence in the above method satisfies the following relationship: ; Represents an individual In the population The convergence of , its minimum value is 0, Represents an individual population The number of times other individuals dominate, Represents population The average number of times each individual in is dominated by other individuals, and Respectively represent the current number of iterations and the preset number of iterations, Indicates the preset stage demarcation point, which is a parameter with a value between 0 and 1, and: ; ; To judge individuals Dominate the individual , When represents an individual Pareto dominated individuals .

[0046] Furthermore, in step S210, the preset convergence threshold is specifically: The product of the subpopulation convergence state judgment threshold τ and the average number of the global non-dominated individuals in each of the subpopulations is preset.

[0047] The computer device 400 provided in the embodiment of the present invention can implement the steps in the bandpass filter design method in the above embodiment and can achieve the same technical effects. Please refer to the description in the above embodiment and will not repeat them here.

[0048] Example 4 An embodiment of the present invention further provides a storage medium, on which a bandpass filter design program is stored. When the bandpass filter design program is executed by a processor, the various processes and steps in the bandpass filter design method provided in the embodiment of the present invention are implemented, and the same technical effects can be achieved. To avoid repetition, they will not be described here.

[0049] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented using a computer program or hardware associated with instructions. The program can be stored in a computer-readable storage medium, and when executed, the program can include the processes in the above-described method embodiments. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0050] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0051] Through the above description of the embodiments, those skilled in the art will clearly understand that the methods of the above embodiments can be implemented using software plus the necessary general-purpose hardware platform. Of course, hardware can also be used, but in many cases the former is the more preferred implementation method. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, or optical disk) and includes a number of instructions for enabling a terminal (such as a mobile phone, computer, server, air conditioner, or network device) to execute the methods described in the various embodiments of the present invention.

[0052] The embodiments of the present invention are described above in conjunction with the accompanying drawings. What is disclosed is only a preferred embodiment of the present invention. However, the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many equivalent changes in form without departing from the scope of protection of the purpose of the present invention and the claims, which are all within the protection of the present invention.

Claims

1. A bandpass filter design method, characterized in that: The following steps are involved: S101, randomly generating a plurality of film parameter matrices, and using the plurality of film parameter matrices as an initial population, wherein the film parameter matrices are constructed by material types and thicknesses corresponding to a plurality of film layers of a bandpass filter; S102, iteratively optimizing the initial population using an unbalanced multi-modal multi-objective optimization algorithm based on a two-stage evolution, wherein a bandpass curve corresponding to the film parameter matrix is ​​obtained, and a bandpass performance corresponding to the bandpass curve is obtained based on the bandpass performance. The population is iteratively optimized based on the bandpass performance until a preset number of iterations is met, and multiple optimal solutions of the film parameter matrices are output; S103 , randomly selecting one from the plurality of optimal solutions of the film parameter matrix, determining the material type and thickness corresponding to the plurality of film layers, and designing the bandpass filter.

2. The bandpass filter design method according to claim 1, wherein: In step S102, the unbalanced multi-modal multi-objective optimization algorithm based on two-stage evolution includes the following steps: S201, initialize the number of iterations to 0; S202: When the number of iterations is less than the product of the preset stage demarcation point and the preset number of iterations, the process enters the first iteration stage and executes step S203; otherwise, the process enters the second iteration stage and executes step S206: S203, generating a mating pool according to the Pareto dominance ranks corresponding to different individuals in the current population, and generating an offspring population by simulating binary crossover and polynomial mutation, wherein the Pareto dominance rank is obtained by comparing the bandpass performance corresponding to the current individual with the bandpass performance corresponding to other individuals according to the Pareto dominance relationship; S204. Obtain the fitness of each individual in the first set of parent and child populations in the current iteration phase, and process the convergence of the individuals in the first set according to a preset convergence strategy. The fitness indicates the degree to which an individual is likely to be retained during the population iteration process, and the convergence indicates the degree to which an individual is close to the optimal solution. S205, selecting individuals from the first set according to the fitness to form the next generation population, increasing the number of iterations by 1, and returning to step S202; S206, clustering the current population using a clustering algorithm to obtain multiple subpopulations, each of which corresponds to an optimal solution region; S207. In each of the optimal solution regions, generating a mating pool according to the Pareto dominance levels corresponding to different individuals, and generating an offspring population by simulating binary crossover and polynomial mutation; S208, merging each of the subpopulations and its corresponding offspring population to form a second set, and counting the global non-dominated individuals and dominated individuals with respect to the Pareto dominance level in different second sets to obtain a global non-dominated individual set and a global dominated individual set; S209, if the size of the global non-dominated individual set is smaller than the size of the initial population, move the current non-dominated individuals in the dominated individual set into the non-dominated individual set until its size becomes the same as the initial population; S210: Determine whether the number of the globally non-dominated individuals in the subpopulation is less than a preset convergence threshold: If so, individuals are selected from the non-dominated individual set based on population diversity to form the next generation population; If not, individuals are selected from the non-dominated individual set based on the convergence and the fitness to form the next generation population; S211, adding 1 to the number of iterations. If the number of iterations meets the preset number of iterations, outputting the optimal individual in each of the current optimal solution regions to obtain multiple optimal solutions of the film system parameter matrix; otherwise, returning to step S206.

3. The bandpass filter design method according to claim 2, wherein: Define the population of the current iteration stage as , where the individuals are In step S204, the preset convergence strategy is used for individual In the population The process of processing the convergence in the above method satisfies the following relationship: ； in, Represents an individual In the population The convergence of , its minimum value is 0, Represents an individual population The number of times other individuals dominate, Represents population The average number of times each individual in is dominated by other individuals, and Respectively represent the current number of iterations and the preset number of iterations, Indicates the preset stage demarcation point, which is a parameter with a value between 0 and 1, and: ； ； To judge an individual Dominate the individual , When represents an individual Pareto dominant individuals .

4. The bandpass filter design method according to claim 2, wherein: In step S210, the preset convergence threshold is specifically: The product of the subpopulation convergence state judgment threshold τ and the average number of the global non-dominated individuals in each of the subpopulations is preset.

5. A bandpass filter design system, characterized in that: include: an initial generation module, configured to randomly generate a plurality of film parameter matrices and use the plurality of film parameter matrices as an initial population, wherein the film parameter matrices are constructed by material types and thicknesses corresponding to a plurality of film layers of the bandpass filter; an optimization module for iteratively optimizing the initial population based on an unbalanced multi-modal multi-objective optimization algorithm using a two-stage evolutionary method, wherein a bandpass curve corresponding to the film parameter matrix is ​​obtained, and a bandpass performance corresponding to the bandpass curve is obtained based on the bandpass curve, and the population is iteratively optimized based on the bandpass performance until a preset number of iterations is met, and multiple optimal solutions of the film parameter matrices are output; The design module is used to select any one of the optimal solutions of the film parameter matrix, determine the material type and thickness corresponding to the multiple film layers, and design the bandpass filter.

6. The bandpass filter design system according to claim 5, characterized in that: The unbalanced multi-modal multi-objective optimization algorithm based on two-stage evolution executed by the optimization module includes the following steps: S201, initialize the number of iterations to 0; S202: When the number of iterations is less than the product of the preset stage demarcation point and the preset number of iterations, the process enters the first iteration stage and executes step S203; otherwise, the process enters the second iteration stage and executes step S206: S203, generating a mating pool according to the Pareto dominance ranks corresponding to different individuals in the current population, and generating an offspring population by simulating binary crossover and polynomial mutation, wherein the Pareto dominance rank is obtained by comparing the bandpass performance corresponding to the current individual with the bandpass performance corresponding to other individuals according to the Pareto dominance relationship; S204. Obtain the fitness of each individual in the first set of parent and child populations in the current iteration phase, and process the convergence of the individuals in the first set according to a preset convergence strategy. The fitness indicates the degree to which an individual is likely to be retained during the population iteration process, and the convergence indicates the degree to which an individual is close to the optimal solution. S205, selecting individuals from the first set according to the fitness to form the next generation population, increasing the number of iterations by 1, and returning to step S202; S206, clustering the current population using a clustering algorithm to obtain multiple subpopulations, each of which corresponds to an optimal solution region; S207. In each of the optimal solution regions, generating a mating pool according to the Pareto dominance levels corresponding to different individuals, and generating an offspring population by simulating binary crossover and polynomial mutation; S208, merging each of the subpopulations and its corresponding offspring population to form a second set, and counting the global non-dominated individuals and dominated individuals with respect to the Pareto dominance level in different second sets to obtain a global non-dominated individual set and a global dominated individual set; S209, if the size of the global non-dominated individual set is smaller than the size of the initial population, move the current non-dominated individuals in the dominated individual set into the non-dominated individual set until its size becomes the same as the initial population; S210: Determine whether the number of the globally non-dominated individuals in the subpopulation is less than a preset convergence threshold: If so, individuals are selected from the non-dominated individual set based on population diversity to form the next generation population; If not, individuals are selected from the non-dominated individual set based on the convergence and the fitness to form the next generation population; S211, adding 1 to the number of iterations. If the number of iterations meets the preset number of iterations, outputting the optimal individual in each of the current optimal solution regions to obtain multiple optimal solutions of the film system parameter matrix; otherwise, returning to step S206.

7. The bandpass filter design system according to claim 6, wherein: Define the population of the current iteration stage as , where the individuals are In step S204, the preset convergence strategy is used for individual In the population The process of processing the convergence in the above method satisfies the following relationship: ； in, Represents an individual In the population The convergence of , its minimum value is 0, Represents an individual population The number of times other individuals dominate, Represents population The average number of times each individual in is dominated by other individuals, and Respectively represent the current number of iterations and the preset number of iterations, Indicates the preset stage demarcation point, which is a parameter with a value between 0 and 1, and: ； ； To judge individuals Dominate the individual , When represents an individual Pareto dominant individuals .

8. The bandpass filter design system according to claim 6, wherein: In step S210, the preset convergence threshold is specifically: The product of the subpopulation convergence state judgment threshold τ and the average number of the global non-dominated individuals in each of the subpopulations is preset.

9. A computer device, characterized in that: include: A memory, a processor, and a program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the bandpass filter design method according to any one of claims 1 to 4 are implemented.

10. A storage medium, characterized in that: The storage medium stores a program, and when the program is executed by the processor, the steps in the bandpass filter design method according to any one of claims 1 to 4 are implemented.

Images