Multilayer dielectric radome design method and device based on heterogeneous numerical optimization algorithm

Abstract

The invention provides a multilayer dielectric radome design method and device based on a heterogeneous numerical optimization algorithm, and belongs to the technical field of vehicle-mounted millimeter wave radars. Through combination of multi-layer medium structure optimization and a confidence domain-gene evolution algorithm, the performance and cost problems of the existing single-layer and lens radome are solved. The design method gives consideration to the wave transmission rate, the mechanical strength, the frequency band coverage and the calculation efficiency, is suitable for the high-performance and low-cost design of the vehicle-mounted millimeter wave radar, and has a remarkable engineering application value.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle-mounted millimeter-wave radar technology, and in particular to a method and device for designing a multi-layer dielectric radome based on a heterogeneous numerical optimization algorithm. Background Art

[0002] In existing technologies, to ensure driving safety and achieve accurate detection of millimeter-wave radar, vehicle-mounted radars in the millimeter-wave frequency band are used to detect and identify targets in driving conditions. To ensure high transmittance during millimeter-wave radar signal transmission while taking into account the long-range detection capability of millimeter-wave radars, current millimeter-wave radar antenna covers are mainly designed and molded using engineering plastics, which meet the design requirements of mechanical strength while meeting transmittance requirements.

[0003] However, in existing technologies, most millimeter-wave radars are designed and molded using engineering plastics, and the radome solutions are generally single-layer or lens structures. Single-layer radomes have a more obvious cost advantage, but lens structures have advantages over single-layer structures in long-range detection capabilities and higher antenna gain. Lens-structured radomes are limited by their spherical structure, resulting in a relatively large overall thickness of the radar, and poor conformality during vehicle body installation, making effective stealth installation impossible. Therefore, existing technologies suffer from the following technical problems:

[0004] (1) Single-layer structure: low cost and easy to process, but limited heat dissipation performance, antenna focusing and shaping capabilities, making it difficult to meet the needs of high-performance radar.

[0005] (2) Lens structure: Although it can improve antenna gain and long-distance detection capabilities, its spherical structure results in a large radar thickness and poor vehicle body conformity, making it impossible to achieve stealth installation. It is also costly and has limited promotion.

[0006] The above problems are in urgent need of resolution. Summary of the Invention

[0007] The purpose of the present invention is to overcome at least one technical problem existing in the prior art and to provide a method and device for designing a multi-layer dielectric antenna cover based on a heterogeneous numerical optimization algorithm.

[0008] On the one hand, an embodiment of the present invention provides a multi-layer dielectric radome design method based on a heterogeneous numerical optimization algorithm, the method comprising: S1: constructing a multi-layer dielectric flat plate structure as a basic structural model of the radome; S2: defining the basic structural model parameters of the radome including the thickness d of the i-th dielectric layer; i , relative complex permittivity ε i and relative complex permeability μ i; S3: Based on the basic structural model parameters of the radome, the performance target to be optimized is converted into a quantifiable mathematical target through the normalized transmission matrix, and the quantifiable mathematical target includes the transmission coefficient T and the power transmission coefficient |T| 2 ; S4: Constructing the first objective function and the second objective function based on the transmission coefficient T; S5: Using a heterogeneous numerical optimization algorithm to optimize the dielectric constant gradient distribution based on the first objective function and the second objective function, including: S501: Using a trust region algorithm to perform local optimization based on the first objective function, including dynamically adjusting the trust region radius to solve and obtain the local optimal dielectric constant distribution; S502: Using the local optimal dielectric constant distribution as the initial population starting point of the genetic evolution algorithm; S503: Using the genetic evolution algorithm to start global optimization based on the second objective function at the initial population starting point, including globally searching for the optimal solution through selection, crossover and mutation operations to obtain the global optimal dielectric constant distribution; S6: Constraining the global optimal dielectric constant distribution through a dielectric constant dynamic range model and a porosity model so that the optimization parameters are within the achievable range of the material; S7: Designing the multilayer dielectric antenna cover through the optimization parameters.

[0009] Furthermore, the step S3: based on the basic structural model parameters of the radome, the performance target to be optimized is converted into a quantifiable mathematical target through the normalized transmission matrix, and the quantifiable mathematical target includes the transmission coefficient T and the power transmission coefficient |T| 2 The normalized transmission matrix is used to describe the voltage and current transmission relationship when a plane wave propagates in an N-layer dielectric flat plate structure, including:

[0010]

[0011] Among them, the transmission matrix expression of a single-layer medium is:

[0012]

[0013] in,

[0014]

[0015] Where A and D are used to describe phase delay, B and C are used to describe amplitude modulation caused by wave impedance change, Z0 represents the equivalent wave impedance of air, and Z i represents the equivalent wave impedance of the i-th layer medium, β i represents the propagation constant in the i-th layer of medium, d i represents the thickness of the i-th layer of medium, θ represents the incident angle of the plane wave, ε i represents the relative complex dielectric constant of the i-th layer medium, j is the imaginary unit representing the phase rotation, α iis the refraction angle in the i-th layer of medium, L is the free space wave number, μ i is the relative complex permeability in the i-th layer of medium, T is the transmission coefficient, TE polarization is parallel polarization, and TM polarization is perpendicular polarization.

[0016] Furthermore, the step S4: constructing the first objective function and the second objective function based on the transmission coefficient T includes:

[0017] The first objective function is:

[0018]

[0019] The second objective function is:

[0020]

[0021] Where X is the dielectric constant distribution, M is the total number of frequency points, TE polarization is parallel polarization, TM polarization is perpendicular polarization, is the transmission loss, Where pol is polarization, i is the i-th frequency point, and T is the transmission coefficient.

[0022] Furthermore, the step S501: using a trust region algorithm to perform local optimization based on the first objective function, including dynamically adjusting the trust region radius to solve for the local optimal dielectric constant distribution, includes:

[0023] S5011: Initialize parameters to determine the optimization starting point and initial search range, including: setting the starting point X1, k=1, ν1=0.1, ε>0;

[0024] Among them, ν1 is the initial trust region radius, ε is the convergence threshold, K is the number of iterations, initially 1, and X1 is the initial iteration point;

[0025] S5012: Calculate the objective function value and gradient, including:

[0026] Calculate the k-th step objective function φ k =φ(X k );

[0027] Gradient Vector

[0028] Calculate the second-order derivative according to the Hessian matrix formula,

[0029] Where, φ k is the current iteration point X k The objective function value at g k is the objective function in X k The gradient vector at G k is the Hessian matrix;

[0030] S5013: Convergence judgment based on the modulus of the gradient, including: if the modulus |g k |<ε, then the position is a local minimum point and the calculation stops, otherwise it goes to step S5104;

[0031] S5014: Determine whether the matrix is positive, including: Determine whether G k +ν k Whether I is a positive definite matrix, where I is the identity matrix; ν k is the confidence region radius;

[0032] S5015: In response to the matrix being non-positive definite, the radius of the trust region is expanded, including: if G k +ν k If I is non-positive definite, then ν k+1 =2ν k , go to step S5014;

[0033] S5016: In response to the matrix being positive definite, solving the trial step length, including: obtaining the trial step length by solving a linear equation system, wherein the linear equation system is (G k +ν k I)δ k =-g k , go to step S5017;

[0034] Among them, δ k is the trial step length, g k is the gradient vector;

[0035] S5017: Calculate the actual objective function reduction, including:

[0036] Δφ k =φ k -φ k (X k +δ k );

[0037] S5018: Calculate the reduction of the prediction objective function, including

[0038]

[0039] S5019: Calculating an acceptance rate based on the objective function reduction and the predicted objective function reduction, including:

[0040] ρ k =Δφ k / Δq k ;

[0041] Where, ρ k is the acceptance rate;

[0042] S50110: Dynamically adjusting the confidence region radius according to the acceptance rate, including:

[0043]

[0044] S50111: Update the iteration point based on the acceptance rate, and go to step S5012, including:

[0045] In response to the acceptance rate ρ k ≤0, do not update the iteration point, make X k+1 =X k ;

[0046] In response to the acceptance rate ρ k >0, update the iteration point so that X k+1 =X k +δ k .

[0047] Furthermore, the step S5012: calculating the objective function value and gradient includes:

[0048] The gradient and second-order derivative are calculated using the difference method, and the difference method expression is:

[0049]

[0050] in,

[0051]

[0052] Where, is the gradient, which reflects the change of the objective function φ with the variable x i The rate of change of variable x i Expressed as the dielectric constant of the i-th layer medium, φ i+0.5 Represents x i The objective function value after increasing 0.5h, φ i-0.5 Represents x i The objective function value after reducing 0.5h, φ i-1 Represents x i The objective function value after reducing h, φ i+1 Represents x i The objective function value after increasing h, where h is the difference step size.

[0053] Furthermore, the S502: using the local optimal dielectric constant distribution as the initial population starting point of the genetic evolution algorithm includes: using the local optimal dielectric constant distribution obtained by the trust region algorithm as the initial individual of the genetic evolution algorithm; generating an initial population including multiple dielectric constant distribution schemes.

[0054] Furthermore, the step S503: using the genetic evolution algorithm to perform global optimization based on the second objective function at the starting point of the initial population, including globally searching the optimal solution through selection, crossover and mutation operations to obtain the global optimal dielectric constant distribution includes: S5031: based on the local optimal dielectric constant distribution obtained by the trust region algorithm, generating a dielectric constant containing N pop The initial population of individuals, each individual represents a dielectric constant distribution scheme {ε r1 ,ε r2 ,...,ε rN}; S5032: Calculate the objective function for each individual Evaluate the balance of its wave transmission performance under all frequency bands and polarizations; S5033: Select high-quality individuals according to fitness, generate offspring through crossover and mutation, and form a new population; S5034: When the objective function value of consecutive H generations no longer improves significantly, or reaches the preset number of iterations, the optimal individual is output as the global optimal dielectric constant distribution, H∈N + .

[0055] Furthermore, the step S6: constraining the global optimal dielectric constant distribution by using a dielectric constant dynamic range model and a porosity model so that the optimization parameters are within the material achievable range include:

[0056] The dielectric constant dynamic range model expression is:

[0057] ε ri =ε rmax -ε rmin (DRR-1)sin 2 x i ;

[0058] Used to make ε ri In [ε rmin ,ε rmax ] within the scope;

[0059]

[0060] Where, ε ri represents the relative dielectric constant of the i-th layer medium, ε rmax Expressed as the maximum relative dielectric constant allowed in a multilayer dielectric structure, ε rmin Expressed as the minimum relative dielectric constant allowed in a multilayer dielectric structure, x i is an unconstrained real variable,

[0061] The porosity model expression is:

[0062]

[0063] For the porosity Pi Adjust the relative dielectric constant ε of the i-th layer medium ri , to achieve material mapping.

[0064] Furthermore, the method further comprises:

[0065] By calculating the transmission matrix of each layer, the optimized transmission coefficient and phase delay are verified to ensure that the design indicators are met.

[0066] In a second aspect, an embodiment of the present invention provides a multi-layer dielectric radome design device based on a heterogeneous numerical optimization algorithm. The device is implemented using the above-mentioned multi-layer dielectric radome design method based on a heterogeneous numerical optimization algorithm. The device includes: a model construction unit, adapted to construct a multi-layer dielectric flat plate structure as a basic structural model of the radome; a parameter definition unit, adapted to define the basic structural model parameters of the radome, including the thickness d of the i-th layer of dielectric. i , relative complex permittivity ε i and relative complex permeability μ i A performance target conversion unit is adapted to convert the performance target to be optimized into a quantifiable mathematical target based on the basic structural model parameters of the radome through a normalized transmission matrix, wherein the quantifiable mathematical target includes a transmission coefficient T and a power transmission coefficient |T| 2 ; An objective function construction unit, adapted to construct a first objective function and a second objective function based on the transmission coefficient T; a parameter optimization unit, adapted to optimize the dielectric constant gradient distribution based on the first objective function and the second objective function using a heterogeneous numerical optimization algorithm, including: using a trust region algorithm to perform local optimization based on the first objective function, including dynamically adjusting the trust region radius to solve and obtain the local optimal dielectric constant distribution; using the local optimal dielectric constant distribution as the initial population starting point of the genetic evolution algorithm; using the genetic evolution algorithm to perform global optimization based on the second objective function starting from the initial population starting point, including globally searching for the optimal solution through selection, crossover and mutation operations to obtain the global optimal dielectric constant distribution; a parameter constraint unit, adapted to constrain the global optimal dielectric constant distribution through a dielectric constant dynamic range model and a porosity model, so that the optimization parameters are within the achievable range of the material; an optimization parameter output unit, adapted to design the multilayer dielectric antenna cover through the optimization parameters.

[0067] In a third aspect, an embodiment of the present invention further provides an electronic device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the above-mentioned multilayer dielectric antenna cover design method based on heterogeneous numerical optimization algorithm.

[0068] In a fourth aspect, an embodiment of the present invention further provides a readable storage medium, which, when the instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to execute the above-mentioned multi-layer dielectric antenna cover design method based on the heterogeneous numerical optimization algorithm.

[0069] The beneficial effects of the present invention are:

[0070] (1) Conventional millimeter-wave radars are implemented using single-layer engineering plastics or lenses. The present invention innovatively adopts a high-transmittance multi-layer antenna cover structure design, which helps dissipate heat from the radar module and improves overall reliability.

[0071] (2) The innovative use of high-transmittance multi-layer antenna cover structure design effectively enhances the mechanical strength and is suitable for applications in high-stress scenarios such as automotive and industrial environments. It also takes into account characteristics such as high temperature resistance and impact resistance to adapt to harsh environments.

[0072] (3) The innovative use of a high-transmittance multi-layer radome structure design optimizes the radome's structural design to cover a wider frequency band and adapt to the current multi-band communication of vehicle-mounted millimeter-wave radars.

[0073] (4) An innovative multi-layer structure radome design method is adopted, which can flexibly adjust the number of radome layers, materials and thickness according to the requirements of the millimeter wave radar system and optimize the radome's RF performance.

[0074] (5) Compared with the traditional single-layer millimeter-wave radar radome, the innovative radome design can ensure that the manufacturing and processing cost of the radome has only a small increase. The price advantage is obvious due to the lens radome, but the RF performance is significantly improved.

[0075] (6) The innovative use of trust region results to initialize the starting point of the genetic evolution algorithm reduces the search time by more than 70% compared with the traditional genetic evolution algorithm.

[0076] (7) The innovative use of genetic evolution algorithm to optimize the dielectric constant value and the heterogeneous numerical optimization algorithm to achieve the optimization of the final transmission coefficient makes the dielectric constant distribution of the antenna cover smoother, which is conducive to actual engineering implementation. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] The present invention will be further described below with reference to the accompanying drawings and examples.

[0078] Figure 1 This is a flow chart of a multi-layer dielectric antenna cover design method based on a heterogeneous numerical optimization algorithm provided in Example 1 of the present invention.

[0079] Figure 2 This is a schematic diagram of a multi-layer dielectric plate structure provided by Example 1 of the present invention.

[0080] Figure 3This is a flow chart of a trust region algorithm provided in Example 1 of the present invention.

[0081] Figure 4 This is a flow chart of a genetic evolution algorithm provided in Example 1 of the present invention.

[0082] Figure 5 This is a structural schematic diagram of a multi-layer dielectric antenna cover design device based on a heterogeneous numerical optimization algorithm provided by Example 2 of the present invention.

[0083] Figure 6 This is a partial block diagram of an electronic device provided in Example 3 of the present invention. DETAILED DESCRIPTION

[0084] Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe the various operations as sequential processes, many of the operations therein can be implemented in parallel, concurrently, or simultaneously. In addition, the order of the various operations can be rearranged. The process can be terminated when its operation is completed, but can also have additional steps not included in the accompanying drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0085] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0086] The present invention will now be described in detail with reference to the accompanying drawings. This figure is a simplified schematic diagram, which only illustrates the basic structure of the present invention in a schematic manner, and therefore only shows the components related to the present invention.

[0087] To facilitate subsequent understanding, the following professional terms are explained here:

[0088] Trust Region Optimization (TR) is a numerical optimization algorithm used to solve nonlinear programming (NLP) problems. The basic idea behind TR is to construct a local quadratic model to approximate the objective function at each iteration and then search for the optimal solution within a region called the trust region. The size of the trust region automatically adjusts based on the approximation of the model to the true function during the iterations. This algorithm is particularly suitable for problems with complex objective functions that are difficult to solve directly.

[0089] Genetic Algorithm (GA) is an important component of heterogeneous numerical optimization algorithms. It is mainly used for global optimization and solving the global optimal problem of dielectric constant gradient distribution of multi-layer dielectric radome.

[0090] DRR (Dielectric Constant Dynamic Range) is a core engineering constraint parameter used to define the range of variation of the dielectric constant in a multilayer dielectric radome, ensuring the feasibility of the design and material performance.

[0091] Example 1

[0092] For ease of understanding, the following is an overall description of the inventive concept before describing the embodiments of the present invention in detail:

[0093] The present application provides a method and device for designing a multi-layer dielectric radome based on a heterogeneous numerical optimization algorithm, and innovatively proposes a heterogeneous numerical optimization algorithm for optimizing the dielectric constant gradient of the material used in the design of the radome. The dynamic range of the dielectric constant is taken into account during the optimization process. The dynamic range of the dielectric constant is added to the optimization range of the heterogeneous algorithm, and while the dielectric constant is limited to the minimum value, the engineering feasibility and the mechanical properties of the radome structure are guaranteed. The heterogeneous numerical optimization algorithm involved in the embodiment of the present invention adopts a trust region algorithm combined with a genetic evolution algorithm to realize the optimization of a multi-layer dielectric radome with a gradient gradient of bandwidth. First, the gradient gradient multi-layer dielectric radome is optimized using a trust region optimization algorithm, and then the optimization result is used to initialize the genetic evolution algorithm. Finally, the minimum transmission coefficient within the required bandwidth is maximized.

[0094] The specific implementation is as follows:

[0095] like Figure 1 As shown, it is a flow chart of a multi-layer dielectric antenna cover design method based on a heterogeneous numerical optimization algorithm provided by the present invention.

[0096] As an example, the method includes:

[0097] S1: Construct a multi-layer dielectric flat plate structure as the basic structural model of the radome; S2: Define the basic structural model parameters of the radome including the thickness d of the i-th dielectric layer i , relative complex permittivity ε i and relative complex permeability μ i ; S3: Based on the basic structural model parameters of the radome, the performance target to be optimized is converted into a quantifiable mathematical target through the normalized transmission matrix, and the quantifiable mathematical target includes the transmission coefficient T and the power transmission coefficient |T| 2S4: Constructing a first objective function and a second objective function based on the transmission coefficient T; S5: Using a heterogeneous numerical optimization algorithm to optimize the dielectric constant gradient distribution based on the first objective function and the second objective function, including: S501: Using a trust region algorithm to perform local optimization based on the first objective function, including dynamically adjusting the trust region radius to solve for the local optimal dielectric constant distribution; S502: Using the local optimal dielectric constant distribution as the initial population starting point of the genetic evolution algorithm; S503: Using the genetic evolution algorithm to perform global optimization based on the second objective function starting from the initial population starting point, including globally searching for the optimal solution through selection, crossover, and mutation operations to obtain the global optimal dielectric constant distribution; S6: Constraining the global optimal dielectric constant distribution through a dielectric constant dynamic range model and a porosity model to ensure that the optimization parameters are within the material's achievable range; S7: Designing the multilayer dielectric radome using the optimization parameters. Performance objectives include high wave transmittance, low reflection loss, broadband coverage, and other performance objectives. In this embodiment, high wave transmittance and low reflection loss are emphasized.

[0098] In some possible implementations, combined Figure 2 As shown, the incident angle of the plane wave is θ, the thickness, relative complex permittivity, and relative complex permeability of the first layer are d1, ε1, and μ1 respectively, and the thickness, relative complex permittivity, and relative complex permeability of the i-th layer are d i , ε i =ε ri (1-jtanδ i ), μ i .

[0099] In some feasible embodiments, step S3: based on the basic structural model parameters of the radome, the performance target to be optimized is converted into a quantifiable mathematical target through a normalized transmission matrix, and the quantifiable mathematical target includes a transmission coefficient T and a power transmission coefficient |T| 2 include:

[0100] The normalized transmission matrix is used to describe the voltage and current transmission relationship when a plane wave propagates in an N-layer dielectric slab structure, and includes:

[0101]

[0102] Among them, the transmission matrix expression of a single-layer medium is:

[0103]

[0104] in,

[0105]

[0106] Where A and D are used to describe phase delay, B and C are used to describe amplitude modulation caused by wave impedance change, Z0 represents the equivalent wave impedance of air, and Z i represents the equivalent wave impedance of the i-th layer medium, β i represents the propagation constant in the i-th layer of medium, d i represents the thickness of the i-th layer of medium, θ represents the incident angle of the plane wave, ε i represents the relative complex dielectric constant of the i-th layer medium, j is the imaginary unit representing the phase rotation, α i is the refraction angle in the i-th layer of medium, L is the free space wave number, μ i is the relative complex permeability in the i-th layer of medium, T is the transmission coefficient, TE polarization is parallel polarization, and TM polarization is perpendicular polarization.

[0107] Specifically, each layer of the matrix represents the propagation of the electromagnetic wave within the layer (phase delay) and the wave impedance matching at the interface (amplitude modulation). The multiplication of multiple layers of matrices is equivalent to the cascade transmission process from the first layer to the Nth layer. For example: the first layer of the matrix processes the propagation of the incident wave in the first layer and the interface reflection / transmission; the second layer of the matrix processes the propagation of the wave in the second layer, and so on. The final output is the total voltage / current relationship after passing through all layers.

[0108] Specifically, if magnetic materials are involved, the above β i The calculation formula needs to be rewritten as:

[0109]

[0110] In some possible implementations, the power transfer coefficient is defined as |T| 2 In general, when the incident angle θ≠0, the Z of two different polarizations TE (parallel) and TM (vertical) i Different, so the transmission coefficients of the two are different, that is, T TE ≠T TM The transmission coefficient is a complex number. The concept of inserted phase delay used in the design of this high-transmittance radome is:

[0111] Θ = arg(T) - arg(T0);

[0112] The physical meaning of the inserted phase delay Θ is: the additional phase difference between the phase arg(T) of a plane wave in a multilayer medium and the phase arg(T0) of the plane wave in air of the same thickness. This parameter reflects the phase effect of the dielectric plate on the plane wave. Since the radome dielectric changes the equivalent phase distribution of the antenna array unit, this parameter is the largest source of the radome's pointing error. At the same time, Z i / Z0 expression analysis shows that when θ=0,μ i =ε i, i∈[1,N], then α=0, Z i / Z0≡1, for the multi-layer dielectric antenna cover involved in the present invention, only the imaginary part of the electromagnetic parameters of the material will produce attenuation during the transmission process, and no reflection will be generated. The high wave transmittance characteristics of the medium are used to design a wave-transmitting antenna cover. In other words, the inserted phase delay is a key indicator to measure the phase modulation of electromagnetic waves by multi-layer media. Its core function is to quantify the phase deviation caused by the medium, and it is directly related to the aiming accuracy of the radar system. Therefore, by optimizing parameters such as the dielectric constant distribution and thickness (combined with the trust region algorithm and the genetic evolution algorithm), the phase delay can be effectively controlled to achieve "low phase error" wave transmission performance. This provides a theoretical basis for the optimization target targeted by the heterogeneous numerical optimization algorithm proposed below.

[0113] In some feasible implementations, the step S4: constructing the first objective function and the second objective function based on the transmission coefficient T includes:

[0114] The first objective function is:

[0115]

[0116] The second objective function is:

[0117]

[0118] Where X is the dielectric constant distribution, M is the total number of frequency points, TE polarization is parallel polarization, TM polarization is perpendicular polarization, is the transmission loss, Where pol is polarization, i is the i-th frequency point, and T is the transmission coefficient.

[0119] In some possible implementations, combined Figure 3 As shown, the step S501: using the trust region algorithm to perform local optimization based on the first objective function, including dynamically adjusting the trust region radius to solve for the local optimal dielectric constant distribution includes:

[0120] S5011: Initialize parameters to determine the optimization starting point and initial search range, including: setting the starting point X1, k=1, ν1=0.1, ε>0;

[0121] Among them, ν1 is the initial trust region radius, ε is the convergence threshold, K is the number of iterations, initially 1, and X1 is the initial iteration point;

[0122] S5012: Calculate the objective function value and gradient, including:

[0123] Calculate the k-th step objective function φ k =φ(X k );

[0124] Gradient Vector

[0125] Calculate the second-order derivative according to the Hessian matrix formula,

[0126] Where, φ k is the current iteration point X k The objective function value at g k is the objective function in X k The gradient vector at G k is the Hessian matrix;

[0127] S5013: Convergence judgment based on the modulus of the gradient, including: if the modulus |g k |<ε, then the position is a local minimum point and the calculation stops, otherwise it goes to step S5104;

[0128] S5014: Determine whether the matrix is positive, including: Determine whether G k +ν k Whether I is a positive definite matrix, where I is the identity matrix; ν k is the confidence region radius;

[0129] S5015: In response to the matrix being non-positive definite, the radius of the trust region is expanded, including: if G k +ν k If I is non-positive definite, then ν k+1 =2ν k , go to step S5014;

[0130] S5016: In response to the matrix being positive definite, solving the trial step length, including: obtaining the trial step length by solving a linear equation system, wherein the linear equation system is (G k +ν k I)δ k =-g k , go to step S5017;

[0131] Among them, δ k is the trial step length, g k is the gradient vector;

[0132] S5017: Calculate the actual objective function reduction, including:

[0133] Δφ k =φ k -φ k (X k +δ k );

[0134] S5018: Calculate the reduction of the prediction objective function, including

[0135]

[0136] S5019: Calculating an acceptance rate based on the objective function reduction and the predicted objective function reduction, including:

[0137] ρ k =Δφ k / Δq k ;

[0138] Where, ρ k is the acceptance rate;

[0139] S50110: Dynamically adjusting the confidence region radius according to the acceptance rate, including:

[0140]

[0141] S50111: Update the iteration point based on the acceptance rate, and go to step S5012, including:

[0142] In response to the acceptance rate ρ k ≤0, do not update the iteration point, make X k+1 =X k ;

[0143] In response to the acceptance rate ρ k >0, update the iteration point so that X k+1 =X k +δ k .

[0144] Specifically, the trust region algorithm directly results in a set of locally optimal solutions for the dielectric constant distribution by minimizing the return loss deviation within a given initial parameter and local neighborhood. This result provides a high-quality initial population for the subsequent genetic evolution algorithm and is the core output of the "local optimization" phase in the heterogeneous numerical optimization algorithm, ultimately serving the design goals of high transmittance and low return loss for the multilayer dielectric radome.

[0145] In some feasible implementations, the step S5012: calculating the objective function value and the gradient includes:

[0146] The gradient and second-order derivative are calculated using the difference method, and the difference method expression is:

[0147]

[0148] in,

[0149]

[0150] Where, is the gradient, which reflects the change of the objective function φ with the variable x i The rate of change of variable x iExpressed as the dielectric constant of the i-th layer medium, φ i+0.5 Represents x i The objective function value after increasing 0.5h, φ i-0.5 Represents x i The objective function value after reducing 0.5h, φ i-1 Represents x i The objective function value after reducing h, φ i+1 Represents x i The objective function value after increasing h, where h is the difference step size.

[0151] The purpose of this step is to solve complex objective functions, which are difficult to calculate analytically using gradients and second-order derivatives. Therefore, a differential method can be used. Specifically, the central difference algorithm is a key support for heterogeneous numerical optimization algorithms. Through high-precision numerical gradient calculations, the trust region algorithm can effectively handle complex electromagnetic objective functions, ensuring rapid convergence to a local optimal solution in optimizing the dielectric constant distribution of multilayer dielectrics. This provides high-quality initial values for the subsequent global optimization of the genetic evolution algorithm, ultimately achieving a balance between wave transmittance, computational efficiency, and engineering feasibility.

[0152] In some feasible implementations, the step S502: using the local optimal dielectric constant distribution as the initial population starting point of the genetic evolution algorithm includes: using the local optimal dielectric constant distribution obtained by the trust region algorithm as the initial individual of the genetic evolution algorithm; generating an initial population containing multiple dielectric constant distribution schemes. Figure 4 As shown, the step S503: using the genetic evolution algorithm to start global optimization based on the second objective function at the starting point of the initial population, including globally searching the optimal solution through selection, crossover and mutation operations to obtain the global optimal dielectric constant distribution includes: S5031: based on the local optimal dielectric constant distribution obtained by the trust region algorithm, generating a dielectric constant containing N pop The initial population of individuals, each individual represents a dielectric constant distribution scheme {ε r1 ,ε r2 ,...,ε rN}; S5032: Calculate the objective function for each individual Evaluate the balance of its wave transmission performance under all frequency bands and polarizations; S5033: Select high-quality individuals according to fitness, generate offspring through crossover and mutation, and form a new population; S5034: When the objective function value of consecutive H generations no longer improves significantly, or reaches the preset number of iterations, the optimal individual is output as the global optimal dielectric constant distribution, H∈N + That is, the significance of using the genetic evolution algorithm here is to first calculate the maximum transmission loss of each frequency point in all frequency points and planning methods. Then take the minimum loss of all frequency points. The ultimate goal is to minimize this "worst-case loss", that is, to maximize the minimum transmission coefficient within the entire frequency band to ensure balanced performance within the bandwidth range.

[0153] This genetic evolution algorithm acts as a "global explorer" within heterogeneous numerical optimization algorithms. By simulating biological evolution, it searches for a globally optimal solution that balances multi-frequency and multi-polarization performance within the high-dimensional solution space of the dielectric constant distribution. Combined with the "local refinement" of trust region optimization, this forms a heterogeneous optimization framework that prioritizes global positioning followed by local refinement, ultimately achieving improvements in both transmittance and computational efficiency.

[0154] That is, the core optimization variable of the heterogeneous numerical optimization algorithm is the dielectric constant gradient distribution. By adjusting the dielectric constant of each layer, the equivalent wave impedance and propagation constant are changed, thereby optimizing the transmission matrix elements (A / B / C / D) and ultimately improving the wave transmission performance.

[0155] In some feasible implementations, in order to ensure the feasibility of the project, the range of the dielectric constant must be limited, and the direct mapping relationship between the dielectric constant and the actual engineering material must be ensured through porosity. Step S6: constraining the global optimal dielectric constant distribution through the dielectric constant dynamic range model and the porosity model so that the optimization parameters are within the material achievable range include:

[0156] The dielectric constant dynamic range model expression is:

[0157] ε ri =ε rmax -ε rmin (DRR-1)sin 2 x i ;

[0158] Used to make ε ri In [ε rmin ,ε rmax ] within the scope;

[0159]

[0160] Where, ε ri represents the relative dielectric constant of the i-th layer medium, ε rmax Expressed as the maximum relative dielectric constant allowed in a multilayer dielectric structure, ε rmin Expressed as the minimum relative dielectric constant allowed in a multilayer dielectric structure, x i is an unconstrained real variable,

[0161] The porosity model expression is:

[0162] P i∈[0,1], 0 means no pores, 1 means full pores.

[0163] For the porosity P i Adjust the relative dielectric constant ε of the i-th layer medium ri , to achieve material mapping.

[0164] Specifically, due to sin 2 x i ∈[0,1], so it can be ensured that ε ri In [ε rmin ,ε rmax ] range. That is, due to sin 2 x i It naturally satisfies [0,1] and no additional constraints are required, so the convenient constraint on the dielectric constant is transformed into x i unconstrained optimization problem, so that the algorithm can freely adjust x i , automatically satisfying ε ri scope.

[0165] In some possible implementations, due to the loss tangent (tan δ i , Loss Tangent) is the core parameter for measuring the energy loss of dielectric materials in an alternating electric field. Therefore, when selecting dielectric materials, the loss tangent must be optimized simultaneously. Specifically, the objective function is the sum of the squares of the transmission loss in is the transmission coefficient. The larger the loss tangent, the smaller the transmission coefficient (higher energy loss). Therefore, during the optimization process, the influence of the loss tangent must be minimized to maximize the transmittance.

[0166] Among them, the porosity P i and loss tangent tanδ i The relationship is:

[0167]

[0168] When the porosity P i When the dielectric constant decreases, tanδ i It decreases accordingly, thus meeting the design goal of "high porosity in the inner layer (low loss) and low porosity in the outer layer (high mechanical strength)".

[0169] In some feasible implementations, in order to intuitively reflect the advantages of the design method described in this embodiment, experimental verification was carried out, and the experimental data obtained by using the trust region algorithm, the genetic evolution algorithm and the heterogeneous numerical optimization algorithm under the conditions of the same bandwidth, maximum complex dielectric constant, the outermost layer thickness d1, the total thickness, and the number of layers N are shown in Table 1. Among them, the optimized frequency range is 76GHz-81GHz, and the frequency interval in the sweep calculation process is 0.1GHz, that is, the total number of calculated frequency points is M=51. The gradient multilayer dielectric antenna cover is optimized under different incident angles. The complex relative dielectric constant of the base medium used is ε1=ε max =7(1-j0.006), the change of dielectric constant is achieved by changing the porosity. The minimum relative dielectric constant is ε rmin =1.2. In order to ensure the mechanical strength of the radome structure, the outermost layer is made of a uniform material with the highest dielectric constant of a certain thickness. Assume that the total thickness of the radome is 20 mm.

[0170] Table 1:

[0171]

[0172] It can be seen that the minimum power transmission coefficient calculated by the heterogeneous numerical optimization algorithm described in this embodiment is 0.959, the result calculated by the trust region algorithm is 0.753, and the result calculated by the genetic evolution algorithm is 0.795. It can be seen that the result obtained by the heterogeneous numerical optimization algorithm is the largest, but the calculation time is greatly reduced to 1528 seconds compared to the 5235 seconds of the genetic evolution algorithm, which is the optimal choice for both result accuracy and efficiency. Comparison of the calculation time of the three different optimization methods shows that the calculation time required by the genetic evolution algorithm is the longest. This is because the trust region algorithm is a method based on gradient and Hessian matrix, while the genetic evolution algorithm is a random optimization algorithm. The total optimization time of initializing the genetic evolution algorithm by the trust region optimization method saves more than half of the optimization time of the randomly initialized genetic algorithm.

[0173] Example 2

[0174] See also Figure 5 , this embodiment provides a structural schematic diagram of a multi-layer dielectric antenna cover design device based on a heterogeneous numerical optimization algorithm.

[0175] As an example, the device is implemented using the multilayer dielectric radome design method based on the heterogeneous numerical optimization algorithm described in Example 1, and the device includes:

[0176] The model building unit 500 is adapted to build a multi-layer dielectric flat plate structure as a basic structural model of the radome.

[0177] The parameter definition unit 510 is adapted to define the basic structural model parameters of the radome including the thickness d of the i-th layer medium. i , relative complex permittivity ε i and relative complex permeability μ i .

[0178] The performance target conversion unit 520 is adapted to convert the performance target to be optimized into a quantifiable mathematical target based on the basic structural model parameters of the radome through the normalized transmission matrix, wherein the quantifiable mathematical target includes the transmission coefficient T and the power transmission coefficient |T| 2 .

[0179] The objective function construction unit 530 is adapted to construct a first objective function and a second objective function based on the transmission coefficient T.

[0180] The parameter optimization unit 540 is suitable for optimizing the dielectric constant gradient distribution based on the first objective function and the second objective function using a heterogeneous numerical optimization algorithm, including: using a trust region algorithm to perform local optimization based on the first objective function, including dynamically adjusting the trust region radius to solve and obtain the local optimal dielectric constant distribution; using the local optimal dielectric constant distribution as the initial population starting point of the genetic evolution algorithm; using the genetic evolution algorithm to perform global optimization based on the second objective function starting from the initial population starting point, including globally searching for the optimal solution through selection, crossover and mutation operations to obtain the global optimal dielectric constant distribution.

[0181] The parameter constraint unit 550 is adapted to constrain the global optimal dielectric constant distribution through a dielectric constant dynamic range model and a porosity model, so that the optimized parameters are within the material achievable range.

[0182] The optimization parameter output unit 560 is adapted to design the multi-layer dielectric radome according to the optimization parameters.

[0183] It is not difficult to find that this embodiment is a system embodiment corresponding to the first embodiment, and this embodiment can be implemented in conjunction with the first embodiment. The relevant technical details mentioned in the first embodiment are still valid in this embodiment, and to reduce repetition, they are not repeated here. Accordingly, the relevant technical details mentioned in this embodiment can also be applied to the first embodiment.

[0184] It is worth noting that all modules involved in this embodiment are logical units. In actual applications, a logical unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. Furthermore, to highlight the innovations of this invention, this embodiment does not include units that are not closely related to solving the technical problems proposed by this invention. However, this does not mean that other units do not exist in this embodiment.

[0185] Example 3

[0186] See also Figure 6 An embodiment of the present invention also provides an electronic device, comprising: a memory and a processor; the memory stores at least one program instruction; the processor loads and executes the at least one program instruction to implement the multilayer dielectric antenna cover design method based on the heterogeneous numerical optimization algorithm provided in Example 1.

[0187] The memory 702 and processor 701 are connected using a bus. The bus can include any number of interconnected buses and bridges, connecting various circuits of one or more processors 701 and memory 702. The bus can also connect various other circuits such as peripheral devices, voltage regulators, and power management circuits. These are all well known in the art and, therefore, are not described further herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be a single component or multiple components, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. Data processed by the processor 701 is transmitted over a wireless medium via an antenna. Furthermore, the antenna receives data and transmits it to the processor 701.

[0188] The processor 701 is responsible for managing the bus and general processing, and can also provide various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. The memory 702 can be used to store data used by the processor 701 when performing operations.

[0189] Example 4

[0190] An embodiment of the present invention further provides a storage medium storing a method for designing a multilayer dielectric radome based on a heterogeneous numerical optimization algorithm. When executed by a processor, this program for designing a multilayer dielectric radome based on a heterogeneous numerical optimization algorithm implements the steps of the method for designing a multilayer dielectric radome based on a heterogeneous numerical optimization algorithm as described above. Because this storage medium utilizes all the technical solutions of all the aforementioned embodiments, it possesses at least all the beneficial effects brought about by the technical solutions of these embodiments, and therefore will not be further elaborated upon here.

[0191] The above is only an embodiment of the present invention. Common knowledge such as the known specific structures and characteristics in the scheme is not described in detail here. Ordinary technicians in the field are aware of all common technical knowledge in the technical field of the invention before the application date or priority date, can obtain all existing technologies in the field, and have the ability to apply conventional experimental means before that date. Ordinary technicians in the field can improve and implement this scheme in combination with their own abilities under the inspiration given by this application. Some typical known structures or known methods should not become obstacles for ordinary technicians in the field to implement this application. It should be pointed out that for those skilled in the art, without departing from the structure of the present invention, several variations and improvements can be made, which should also be regarded as the scope of protection of the present invention. These will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A multi-layer dielectric radome design method based on a heterogeneous numerical optimization algorithm, characterized in that: The method comprises: S1: constructing a multi-layer dielectric flat plate structure as the basic structural model of the radome; S2: Define the basic structural model parameters of the radome including the thickness d of the i-th layer of medium i , relative complex permittivity ε i and relative complex permeability μ i ; S3: Based on the basic structural model parameters of the radome, the performance target to be optimized is converted into a quantifiable mathematical target through the normalized transmission matrix. The quantifiable mathematical target includes the transmission coefficient T and the power transmission coefficient |T| 2 ; S4: constructing a first objective function and a second objective function based on the transmission coefficient T; S5: Optimizing the dielectric constant gradient distribution based on the first objective function and the second objective function using a heterogeneous numerical optimization algorithm, including: S501: performing local optimization based on the first objective function using a trust region algorithm, including dynamically adjusting the trust region radius to obtain a local optimal dielectric constant distribution; S502: Using the local optimal dielectric constant distribution as the initial population starting point of the genetic evolution algorithm; S503: Using a genetic evolution algorithm to perform global optimization based on the second objective function starting from the initial population, including globally searching for an optimal solution through selection, crossover, and mutation operations to obtain a global optimal dielectric constant distribution; S6: constraining the global optimal dielectric constant distribution through a dielectric constant dynamic range model and a porosity model so that the optimization parameters are within the achievable range of the material; S7: Designing the multi-layer dielectric antenna cover according to the optimization parameters.

2. The multi-layer dielectric radome design method based on heterogeneous numerical optimization algorithm according to claim 1 is characterized in that: The step S3: based on the basic structural model parameters of the radome, the performance target to be optimized is converted into a quantifiable mathematical target through the normalized transmission matrix, and the quantifiable mathematical target includes the transmission coefficient T and the power transmission coefficient |T| 2 include: The normalized transmission matrix is used to describe the voltage and current transmission relationship when a plane wave propagates in an N-layer dielectric slab structure, and includes: Among them, the transmission matrix expression of a single-layer medium is: in, Where A and D are used to describe phase delay, B and C are used to describe amplitude modulation caused by wave impedance change, Z0 represents the equivalent wave impedance of air, and Z i represents the equivalent wave impedance of the i-th layer medium, β i represents the propagation constant in the i-th layer of medium, d i represents the thickness of the i-th layer of medium, θ represents the incident angle of the plane wave, ε i represents the relative complex dielectric constant of the i-th layer medium, j is the imaginary unit representing the phase rotation, α i is the refraction angle in the i-th layer of medium, L is the free space wave number, μ i is the relative complex permeability in the i-th layer of medium, T is the transmission coefficient, TE polarization is parallel polarization, and TM polarization is perpendicular polarization.

3. The multi-layer dielectric radome design method based on heterogeneous numerical optimization algorithm according to claim 2 is characterized in that: The step S4: constructing the first objective function and the second objective function based on the transmission coefficient T includes: The first objective function is: The second objective function is: Where X is the dielectric constant distribution, M is the total number of frequency points, TE polarization is parallel polarization, TM polarization is perpendicular polarization, is the transmission loss, Where pol is polarization, i is the i-th frequency point, and T is the transmission coefficient.

4. The multi-layer dielectric radome design method based on heterogeneous numerical optimization algorithm according to claim 3 is characterized in that: The step S501: using a trust region algorithm to perform local optimization based on the first objective function, including dynamically adjusting the trust region radius to solve for the local optimal dielectric constant distribution, includes: S5011: Initialize parameters to determine the optimization starting point and initial search range, including: setting the starting point X1, k=1, ν1=0.1, ε>0; Among them, ν1 is the initial trust region radius, ε is the convergence threshold, K is the number of iterations, initially 1, and X1 is the initial iteration point; S5012: Calculate the objective function value and gradient, including: Calculate the k-th step objective function φ k =φ(X k ); Gradient Vector Calculate the second-order derivative according to the Hessian matrix formula, Where, φ k is the current iteration point X k The objective function value at g k is the objective function in X k The gradient vector at G k is the Hessian matrix; S5013: Convergence judgment based on the modulus of the gradient, including: if the modulus |g k |<ε, then the position is a local minimum point and the calculation stops, otherwise it goes to step S5104; S5014: Determine whether the matrix is positive, including: Determine whether G k +ν k Whether I is a positive definite matrix, where I is the identity matrix; ν k is the confidence region radius; S5015: In response to the matrix being non-positive definite, the radius of the trust region is expanded, including: if G k +ν k If I is non-positive definite, then ν k+1 =2ν k , go to step S5014; S5016: In response to the matrix being positive definite, solving the trial step length, including: obtaining the trial step length by solving a linear equation system, wherein the linear equation system is (G k +ν k I)δ k =-g k , go to step S5017; Among them, δ k is the trial step length, g k is the gradient vector; S5017: Calculate the actual objective function reduction, including: Df k =φ k -f k (X k +d k ); S5018: Calculate the reduction of the prediction objective function, including S5019: Calculating an acceptance rate based on the objective function reduction and the predicted objective function reduction, including: r k =Df k / Δq k ; Where, ρ k is the acceptance rate; S50110: Dynamically adjusting the confidence region radius according to the acceptance rate, including: S50111: Update the iteration point based on the acceptance rate, and go to step S5012, including: In response to the acceptance rate ρ k ≤0, do not update the iteration point, make X k+1 =X k ; In response to the acceptance rate ρ k >0, update the iteration point so that X k+1 =X k +δ k .

5. The multi-layer dielectric radome design method based on heterogeneous numerical optimization algorithm according to claim 4 is characterized in that: The step S5012: calculating the objective function value and the gradient includes: The gradient and second-order derivative are calculated using the difference method, and the difference method expression is: in, Where, is the gradient, which reflects the change of the objective function φ with the variable x i The rate of change of variable x i Expressed as the dielectric constant of the i-th layer medium, φ i+0.5 Represents x i The objective function value after increasing 0.5h, φ i-0.5 Represents x i The objective function value after reducing 0.5h, φ i-1 Represents x i The objective function value after reducing h, φ i+1 Represents x i The objective function value after increasing h, where h is the difference step size.

6. The multi-layer dielectric radome design method based on heterogeneous numerical optimization algorithm according to claim 4 is characterized in that: The step S502 of using the local optimal dielectric constant distribution as the initial population starting point of the genetic evolution algorithm includes: The local optimal dielectric constant distribution obtained by the trust region algorithm is used as the initial individual of the genetic evolution algorithm; Generate an initial population containing multiple dielectric constant distribution schemes.

7. The method for designing a multilayer dielectric radome based on a heterogeneous numerical optimization algorithm according to claim 3, wherein: The step S503: using a genetic evolution algorithm to perform global optimization based on the second objective function starting from the initial population, including performing global search for the optimal solution through selection, crossover and mutation operations to obtain the global optimal dielectric constant distribution includes: S5031: Generate the local optimal dielectric constant distribution based on the trust region algorithm, including N pop The initial population of individuals, each individual represents a dielectric constant distribution scheme {ε r1 ,ε r2 ,...,ε rN }; S5032: Calculate the objective function for each individual Evaluate the balance of its wave transmission performance across all frequency bands and polarizations; S5033: Select high-quality individuals according to fitness, generate offspring through crossover and mutation, and form a new population; S5034: When the objective function value of consecutive H generations no longer improves significantly, or reaches the preset number of iterations, the optimal individual is output as the global optimal dielectric constant distribution, H∈N + .

8. The method for designing a multilayer dielectric radome based on a heterogeneous numerical optimization algorithm according to claim 1, wherein: The step S6: constraining the global optimal dielectric constant distribution by using the dielectric constant dynamic range model and the porosity model so that the optimization parameters are within the material achievable range include: The dielectric constant dynamic range model expression is: e ri =e rmax -e rmin (DRR-1)sin 2 x i ; To make ε ri In [ε rmin ,ε rmax ] within the scope; Where, ε ri represents the relative dielectric constant of the i-th layer medium, ε rmax Expressed as the maximum relative dielectric constant allowed in a multilayer dielectric structure, ε rmin Expressed as the minimum relative dielectric constant allowed in a multilayer dielectric structure, x i is an unconstrained real variable, The porosity model expression is: For the porosity P i Adjust the relative dielectric constant ε of the i-th layer medium ri , to achieve material mapping.

9. The method for designing a multilayer dielectric radome based on a heterogeneous numerical optimization algorithm according to claim 1, wherein: The method further comprises: By calculating the transmission matrix of each layer, the optimized transmission coefficient and phase delay are verified to ensure that the design indicators are met.

10. A device for designing a multi-layer dielectric radome based on a heterogeneous numerical optimization algorithm, wherein the device is implemented using the method for designing a multi-layer dielectric radome based on a heterogeneous numerical optimization algorithm according to any one of claims 1 to 9, and wherein: The device comprises: A model building unit, adapted to build a multi-layer dielectric flat plate structure as a basic structural model of the radome; The parameter definition unit is adapted to define the basic structural model parameters of the radome including the thickness d of the i-th layer medium. i , relative complex permittivity ε i and relative complex permeability μ i ; A performance target conversion unit is adapted to convert the performance target to be optimized into a quantifiable mathematical target based on the basic structural model parameters of the radome through a normalized transmission matrix, wherein the quantifiable mathematical target includes a transmission coefficient T and a power transmission coefficient |T| 2 ; an objective function construction unit, adapted to construct a first objective function and a second objective function based on the transmission coefficient T; A parameter optimization unit is adapted to optimize the dielectric constant gradient distribution based on the first objective function and the second objective function using a heterogeneous numerical optimization algorithm, including: using a trust region algorithm to perform local optimization based on the first objective function, including dynamically adjusting the trust region radius to solve for a local optimal dielectric constant distribution; using the local optimal dielectric constant distribution as an initial population starting point of a genetic evolution algorithm; using the genetic evolution algorithm to perform global optimization based on the second objective function starting from the initial population starting point, including globally searching for an optimal solution through selection, crossover, and mutation operations to obtain a global optimal dielectric constant distribution; a parameter constraint unit adapted to constrain the global optimal dielectric constant distribution through a dielectric constant dynamic range model and a porosity model so that the optimized parameters are within a material achievable range; The optimization parameter output unit is adapted to design the multi-layer dielectric radome according to the optimization parameters.

Images