An inverse design method for realizing an arbitrary radiation pattern of an antenna

Abstract

The present invention discloses an inverse design method for realizing an arbitrary radiation pattern of an antenna, belonging to the technical field of antenna communication. The method of the present invention includes the following steps: Initialization: Set the positions of the scatterers and determine the corresponding far-field expected scattered field according to the set performance indicators; Forward process: Update the total field in the scatterer and the Green's function from the source point to the far-field point based on the background dielectric constant distribution function; Inverse process: Substitute the far-field expected scattered field, the total field in the scatterer, and the Green's function from the source point to the far-field point into the integral equation of the electromagnetic inverse scattering method to obtain the dielectric constant distribution function in the scatterer; Evaluation: If the current performance indicators match the set performance indicators, it is determined as the desired solution, otherwise, this method is executed again until it meets the requirements. The method of the present invention does not rely on existing prior knowledge, has higher efficiency, and can make up for the deficiencies in the forward design method; it is more efficient and flexible compared to the existing inverse methods.

Description

Technical Field

[0001] The present invention belongs to the technical field of antenna communication, and particularly relates to an inverse design method for realizing an arbitrary radiation pattern of an antenna. Background Art

[0002] Radiation patterns with different beam shapes, such as multi-beam, pencil beam, flat-top beam, etc., play important roles in communication fields such as satellites, radars, base stations, etc., and are one of the most important objectives in antenna design.

[0003] Existing shaping technologies can use array synthesis methods or intelligent optimization algorithms to achieve radiation patterns with different beam shapes by controlling the excitation, spacing, arrangement, element structure, etc. of the antenna array. In addition, in the past few decades, the Fabry-Perot resonator principle, the characteristic mode theory, the leaky wave principle, the generalized reflection and refraction laws, etc. have also been used to achieve radiation patterns with different beam shapes. However, generally speaking, the above methods all belong to the forward design method from the structure (including the amplitude and phase of port excitation) to the radiation target design, and the radiation target is used as a judgment condition to judge the design quality in the forward design process.

[0004] The disadvantages of the forward design method are also very obvious. Specifically, there are two points as follows: one is that the forward design method depends on the existing knowledge base and requires a high accumulation of engineers' experience and the reserve of relevant prior knowledge; the other is that the design time cost of the forward design method is high, especially the time cost for full-wave simulation in the process of adjusting and optimizing the antenna structure parameters is huge.

[0005] In addition, in recent years, in the field of nanophotonics, an inverse design concept different from the forward design method has been proposed by introducing deep learning methods. The prior art "Nanophotonic particle simulation and inverse design using artificial neural networks" uses the backpropagation method to apply the trained neural network to the inverse design of nanophotonic problems. The prior art "Generative model for the inverse design of metasurfaces" applies an unsupervised learning system to the inverse design problem of nanophotonic structures. This system has no clear guiding target when processing data and can obtain new results in new data.

[0006] The disadvantages of the inverse design method based on deep learning are also very obvious. Specifically, there are the following two points: First, since the results obtained by the inverse design method based on deep learning are highly sensitive to the quality of the dataset, a high-quality and large dataset needs to be generated to train a good neural network, which makes the computational cost high. Second, this method needs to specifically generate the corresponding dataset for a specific inverse design problem and train the corresponding neural network, which makes this method lack sufficient generalization. Summary of the Invention

[0007] The object of the present invention is to overcome the disadvantages of the existing forward design and reverse design methods, and provide an inverse design method for realizing an arbitrary radiation pattern of an antenna. The electromagnetic inverse scattering method is introduced for the first time and applied to the inverse design process of the antenna cover board.

[0008] The technical problem proposed by the present invention is solved as follows:

[0009] An inverse design method for realizing an arbitrary radiation pattern of an antenna includes the following steps:

[0010] S1. Initialization

[0011] Manually set the scatterer position S′, and the scatterer position S′ is located in the near-field region of the radiating antenna; determine the corresponding far-field expected scattered field according to the set performance index.

[0012] S2. Forward process

[0013] The scatterer is equivalent to a secondary source under the irradiation of the radiating antenna. The scatterer is divided into several grids to obtain several scatterer source points r′; based on the current background dielectric constant distribution function Update the total field within the scatterer position S′ Update the Green's function from the scatterer source point r′ to the far-field point where (x′, y′) is the two-dimensional coordinate of the grid divided by the scatterer position S′;

[0014] S3. Inverse process

[0015] Substitute the far-field expected scattered field The total field within the scatterer position S′ and the Green's function from the scatterer source point r′ to the far-field point into the electromagnetic inverse scattering integral equation to obtain the dielectric constant distribution function ε r (x′, y′);

[0016] S4. Evaluation

[0017] Based on the radiating antenna, the scatterer position S′ and the corresponding dielectric constant distribution function εr (x′, y′) obtains the current performance index and the total far - field Determine whether the current performance index converges. If so, determine the current relative permittivity distribution function ε r (x′, y′) as the desired solution, and the structure corresponding to the current relative permittivity distribution function ε r (x′, y′) as the structure of the scatterer position S′; otherwise, set the relative permittivity distribution function ε r (x′, y′) as the background relative permittivity distribution function Execute S1 - S4 until convergence.

[0018] Furthermore, in step S1, the corresponding far - field desired scattered field is determined according to the set performance index The specific process is:

[0019] S1.1. Set the directivity coefficient D des (θ), where the angle θ ∈ [0, 2π), and the electric field strength of the desired total far - field is:

[0020]

[0021] where P r is the radiation power of the antenna, η 0 is the wave impedance in free space, r far is the distance between the antenna radiation source point and the far - field point;

[0022] S1.2. Arbitrarily set the phase The desired total far - field is:

[0023]

[0024] S1.3. Apply an excitation to the antenna and use the near - far field transformation method to obtain the far - field incident field in the direction of the angle θ based on the current background relative permittivity distribution function Then the far - field desired scattered field in the direction of the angle θ is:

[0025]

[0026] Furthermore, the formula for updating the Green's function in step S2 is:

[0027]

[0028]

[0028] where is the Laplace operator, k0 is the free space wave number, and δ(r far - r′) is the impulse excitation function from the source point r′ of the scatterer to the far-field point.

[0029] Furthermore, in step S3, the electromagnetic inverse scattering integral equation is:

[0030]

[0031] where 1 ≤ i ≤ m, m is the number of discrete points of the far-field expected scattered field, and θ i is the i-th angular value in the angle θ ∈ [0, 2π), and δε r (x′, y′) is the contrast of the relative permittivity distribution function;

[0032] After solving δε r (x′, y′) from the above formula, calculate the relative permittivity distribution function ε r (x′, y′):

[0033]

[0034] Furthermore, in step S4, the specific process of determining whether the current performance index converges is as follows:

[0035] Calculate the iteration convergence value:

[0036]

[0037] If the iteration convergence value ITV ≤ the set threshold, it is determined to converge.

[0038] The beneficial effects of the present invention are:

[0039] The method of the present invention first introduces the electromagnetic inverse scattering method and applies it to the inverse design of arbitrary radiation patterns of antennas, and can reasonably set the required radiation patterns with specific shapes according to different application scenarios and different operating frequency bands. Compared with the forward design method, the inverse design method of the present invention uses the design target as a known condition to participate in the inverse design of arbitrary radiation patterns of antennas, does not rely on existing prior knowledge, has higher efficiency, and can make up for the deficiencies in the forward design method.

[0040] Compared with the inverse design method based on deep learning, the method of the present invention does not require the time cost of obtaining a large training data set and training a neural network, and has higher efficiency. At the same time, the method of the present invention can update the expected scattered field according to different design targets to realize the inverse design process of different design targets, and does not need to specially generate corresponding data sets and train corresponding neural networks for each specific inverse design problem, and the method is more flexible. Description of the Drawings

[0041] Figure 1 Schematic diagram of the process of the inverse design method according to the present invention;

[0042] Figure 2 Structural diagram of a one-dimensional antenna with a rectangular flat-top beam having a beam width of 90° at 2.5 GHz in the desired radiation field pattern and the corresponding scatterer position S';

[0043] Figure 3 Structural diagram of a one-dimensional antenna with a single beam having a peak gain pointing in the 25° direction at 2.5 GHz in the desired radiation field pattern and the corresponding scatterer position S';

[0044] Figure 4 Structural diagram of a one-dimensional antenna with a dual beam having a peak gain pointing at 35° at 2.5 GHz in the desired radiation field pattern and the corresponding scatterer position S';

[0045] Figure 5 Structural diagram of a two-dimensional antenna with a rectangular flat-top beam having a beam width of 40° at 2.5 GHz in the desired radiation field pattern and the corresponding scatterer position S';

[0046] Figure 6 Structural diagram of a two-dimensional antenna with a pen-shaped beam having a peak gain pointing in the 0° direction at 2.5 GHz in the desired radiation field pattern and the corresponding scatterer position S'. Detailed implementation manners

[0047] The following describes the detailed implementation manners of the present invention to facilitate those skilled in the art of the present technology to understand the present invention. However, it should be clear that the present invention is not limited to the scope of the detailed implementation manners. For those of ordinary skill in the art of the present technology, as long as various changes are within the spirit and scope of the present invention defined and determined by the appended claims, these changes are obvious, and all inventions and creations using the concept of the present invention are within the scope of protection.

[0048] This embodiment provides an inverse design method for realizing an arbitrary radiation pattern of an antenna, and its schematic diagram of the process is as Figure 1 shown, including the following steps:

[0049] S1. Initialization

[0050] Set the inverse design solution region, that is, the scatterer position S', according to the actual engineering requirements. The scatterer position S is located in the near-field region of the radiation antenna; determine the corresponding far-field desired scattered field according to the set performance indicators The set performance indicators satisfy the law of conservation of energy and the basic laws of antennas.

[0051] In step S1, the specific process of determining the corresponding far-field desired scattered field according to the set performance indicators is as follows:

[0052] S1.1. Set the direction coefficient D des (θ), where the angle θ ∈ [0, 2π), and the electric field strength of the expected total far - field is:

[0053]

[0054] where P r is the radiation power of the antenna, η 0 is the wave impedance in free space, r far is the distance between the antenna radiation source point and the far - field point;

[0055] S1.2. Arbitrarily set the phase The expected total far - field is:

[0056]

[0057] where j is the imaginary part symbol;

[0058] S1.3. Apply an excitation to the antenna. Based on the current background dielectric constant distribution function (set the initial value to 1 when S1 is executed for the first time), use the near - far field transformation method to obtain the far - field incident field in the direction of the angle θ. Then the expected far - field scattered field in the direction of the angle θ is:

[0059]

[0060] S2. Forward process

[0061] Under the irradiation of the radiating antenna, the scatterer can be equivalent to a secondary source. Divide the scatterer into several grids to obtain several scatterer source points r′; based on the current background dielectric constant distribution function update the total field inside the scatterer position S′ and the Green's function from the scatterer source point r′ to the far - field point, where (x′, y′) are the two - dimensional coordinates of the grid divided by the scatterer position S′.

[0062] The formula for updating the Green's function in step S2 is:

[0063]

[0064] where is the Laplace operator, k 0 is the free - space wave number, δ(r far-r′) is the impulse excitation function from the scatterer source point r′ to the far-field point.

[0065] S3. Inverse process

[0066] Substitute the far-field expected scattered field in S1 and the total field within the scatterer position S′ in S2 into the electromagnetic inverse scattering integral equation to obtain the dielectric constant distribution function ε within the scatterer position S′ r (x′, y′).

[0067] In step S3, the electromagnetic inverse scattering integral equation is:

[0068]

[0069] where 1 ≤ i ≤ m, m is the number of discrete points of the far-field expected scattered field, θ i is the i-th angle value in the angle θ ∈ [0, 2π), δε r (x′, y′) is the dielectric constant distribution function contrast;

[0070] Solve for δε r (x′, y′) from the above equation, and then calculate the dielectric constant distribution function ε r (x′, y′):

[0071]

[0072] S4. Evaluation

[0073] Based on the radiation antenna, the scatterer position S′, and the corresponding dielectric constant distribution function ε r (x′, y′), obtain the current performance index and the far-field total field Judge whether the current performance index converges. If so, determine that the current dielectric constant distribution function ε r (x′, y′) is the desired solution, and the structure corresponding to the current dielectric constant distribution function ε r (x′, y′) is used as the structure of the scatterer position S′; otherwise, let the dielectric constant distribution function ε r (x′, y′) be the background dielectric constant distribution function Execute S1 - S4 until convergence.

[0074] The specific process of judging whether the current performance index converges is as follows:

[0075] Calculate the iterative convergence value:

[0076]

[0077] If the iterative convergence value ITV ≤ the set threshold, then determine convergence.

[0078] Figure 2 Structural diagram of a one-dimensional antenna with an expected radiation field pattern of a rectangular flat-top beam with a 90° beam width at 2.5 GHz and its corresponding scatterer position S'; Figure 3 Structural diagram of a one-dimensional antenna with an expected radiation field pattern of a single beam with a peak gain pointing in the 25° direction at 2.5 GHz and its corresponding scatterer position S'; Figure 4 Structural diagram of a one-dimensional antenna with an expected radiation field pattern of a dual beam with a peak gain pointing at 35° at 2.5 GHz and its corresponding scatterer position S'; A one-dimensional array is applied as the excitation source in S1.3.

[0079] Figure 5 Structural diagram of a two-dimensional antenna with an expected radiation field pattern of a rectangular flat-top beam with a 40° beam width at 2.5 GHz and its corresponding scatterer position S'; Figure 6 Structural diagram of a two-dimensional antenna with an expected radiation field pattern of a pen beam with a peak gain pointing at 0° at 2.5 GHz and its corresponding scatterer position S'. A 2×2 radiation array is applied as the excitation source in S13.

Claims

1. An inverse design method for realizing an arbitrary radiation pattern of an antenna, characterized in that, it includes the following steps: S1. Initialization Manually set the position of the scatterer S', where the scatterer position S' is located in the near-field region of the radiating antenna; determine the corresponding expected far-field scattered field according to the set performance indicators S2. Forward process The scatterer is equivalent to a secondary source under the illumination of the radiation antenna. The scatterer is divided into several grids to obtain several scatterer source points r'; based on the current background dielectric constant distribution function Update the total field within the scatterer position S' Update the Green's function from the scatterer source point r' to the far-field point where (x', y') are the two-dimensional coordinates of the grids divided in the scatterer position S'; S3. Inverse process The far-field desired scattered field The total field within the scatterer position S' And the Green's function from the scatterer source point r′ to the far-field point Substitute into the electromagnetic inverse scattering integral equation to obtain the dielectric constant distribution function ε r (x′, y′); S4. Evaluation Based on the radiation antenna, the position S' of the scatterer, and the corresponding dielectric constant distribution function ε r (x′, y′) to obtain the current performance index and the total far-field Determine whether the current performance index converges. If so, determine that the current dielectric constant distribution function ε r (x′, y′) is the desired solution, and the structure corresponding to the current dielectric constant distribution function ε r (x′, y′) is used as the structure of the scatterer position S'; otherwise, set the dielectric constant distribution function ε r (x′, y′) as the background dielectric constant distribution function Execute S1 - S4 until convergence.

2. The inverse design method for realizing an arbitrary radiation pattern of an antenna according to claim 1, characterized in that, In step S1, the corresponding far-field expected scattering field is determined according to the set performance indicators The specific process is as follows: S1.

1. Set the direction coefficient D des (θ), where the angle θ ∈ [0, 2π), and the electric field strength of the expected total far-field is as follows: Among them, P r is the radiation power of the antenna, η 0 is the wave impedance in free space, r far is the distance between the antenna radiation source point and the far-field point; S1.

2. Arbitrarily set the phase Total desired far-field is as follows: S1.

3. Apply excitation to the antenna and, based on the current background dielectric constant distribution function Use the near - far field transformation method to obtain the far - zone incident field in the direction of angle θ Then, the far - zone desired scattered field in the direction of angle θ is:

3. The inverse design method for realizing an arbitrary radiation pattern of an antenna according to claim 1, characterized in that, the formula for updating the Green's function in step S2 is: Among them, is the Laplace operator, k 0 is the free-space wavenumber, and δ(r far - r′) is the impulse excitation function from the source point r′ of the scatterer to the far-field point.

4. The inverse design method for realizing an arbitrary radiation pattern of an antenna according to claim 1, characterized in that, in step S3, the electromagnetic inverse scattering integral equation is: where \(1\leq i\leq m\), \(m\) is the number of discrete points of the expected scattered field in the far zone, and \(\theta\) i is the \(i\)-th angular value in the angle \(\theta\in[0, 2\pi)\), and \(\delta\epsilon\) r (x′, y′) is the contrast of the dielectric constant distribution function; Solve for δε from the above equation r After (x′, y′), calculate the dielectric constant distribution function ε r (x′, y′): .

5. The inverse design method for realizing an arbitrary radiation pattern of an antenna according to claim 1, characterized in that, in step S4, the specific process for judging whether the current performance index converges is: Calculate the iterative convergence value: If the iterative convergence value ITV ≤ the set threshold, it is determined to converge.

Images