Human-like focusing metasurface system driven based on an evolutionary learning algorithm with a teacher

Abstract

The present invention discloses a human-eye-like focusing metasurface system driven based on a teacher-involved evolutionary learning algorithm, which is applicable to the technical field of intelligent electromagnetic metasurfaces. The system includes a transmissive metasurface, an array probe, a focusing guide module, and an evolutionary learning module. When an external electromagnetic wave signal passes through the transmissive metasurface, the array probe installed behind the transmissive metasurface detects the external electromagnetic wave data, and the focusing guide module and the evolutionary learning module analyze it and output an adjustment strategy for the transmissive metasurface. The state of the transmissive metasurface changes, the array probe collects new data, the focusing guide module and the evolutionary learning module further analyze the intensity and characteristics of the external electromagnetic wave data, output the next adjustment command, and repeat the above process until focusing at the specified position. The present invention can achieve intelligent focusing at any position under multiple electromagnetic environments, does not require artificial adjustment, and can be used flexibly.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent electromagnetic metasurfaces, and more specifically, to a human-eye-like focusing metasurface system driven based on a supervised-evolutionary learning algorithm.

Background Art

[0002] Light focusing is an old topic that has existed for thousands of years and is widely applied in biology, photonics, and physics, and the research enthusiasm of scientists has not waned. Conventional optical focusing lenses are usually manufactured based on various heavy substrates, and by reasonably designing the microstructure, a beam intensity or shape defined by the user is formed. The emergence of electromagnetic metasurfaces provides the possibility of miniaturization and integration of optical lenses. A metasurface is an artificially designed structure composed of a series of sub-wavelength unit structures. By devising sub-wavelength elements and spatial layouts, researchers have developed devices with multiple functions such as beam polarization, focusing, and imaging. Among many functional devices, metalenses are the most widely applied. Compared with conventional heavy lenses, metalenses can focus incident light with a more compact size.

[0003] In the past decade, scholars have designed many meta-surface focusing systems (metalenses) to achieve functions such as broadband, achromatic, and high efficiency. Achieving adaptive focusing of electromagnetic waves (light) is of great significance in aspects such as electromagnetic spectrum imaging and enhancement of communication channels. However, conventional focusing devices can only operate under a given environment. When the incident environment changes, it is necessary to redesign the structure or unit array, and they do not have adaptive functions. Although many intelligent optical devices have been created by combining artificial intelligence (deep learning) and meta-surfaces, their success largely depends on the quantity and quality of available training data, and information about the environment is required in advance. A single deep learning method may fail when facing the focusing task in a rapidly changing environment. The human eye is a perfect focusing system, highly adaptable to environmental changes, and we can perceive more than 80% of the environmental information with our eyes. Designing a natural intelligent focusing system that can achieve automatic convergence of electromagnetic waves (light) in different environments like the human eye can greatly simplify the device design and contribute to facilitating applications. Therefore, how to provide a meta-surface system adaptable to intelligent focusing is an issue that those skilled in the art should urgently solve.

Summary of the Invention

Means for Solving the Problems

[0004] In view of this, the present invention provides a focusing meta-surface system similar to the human eye driven based on a supervised-evolutionary learning algorithm. Using supervised-evolutionary learning as the basic algorithm framework, when electromagnetic waves randomly incident on the meta-surface, through the learning and evolution of the algorithm, the state of the meta-surface is adaptively adjusted to control the trajectory of the transmitted beam, and the system can achieve focusing at any specified position under a complex electromagnetic environment and can be applied to various applications including electromagnetic space imaging, communication signal enhancement, wireless charging, etc.

[0005] To achieve the above object, the present invention provides the following technical solutions.

[0006] A human-eye-like focusing metasurface system driven based on a supervised-evolutionary learning algorithm includes a transmissive metasurface, an array probe, a focusing guide module, and an evolutionary learning module. When an external electromagnetic wave signal passes through the transmissive metasurface, the array probe installed behind the transmissive metasurface detects the external electromagnetic wave data, and the focusing guide module and the evolutionary learning module analyze the external electromagnetic wave, output an adjustment strategy for the transmissive metasurface, the state of the transmissive metasurface changes, the array probe collects new external electromagnetic wave data, the focusing guide module and the evolutionary learning module further analyze the intensity and characteristics of the external electromagnetic wave data, output the next adjustment command, and repeat the above process until focusing on the specified position.

[0007] Optionally, to stop the iteration after focusing on the specified position, the following judgment conditions need to be satisfied simultaneously.

[0008]

Number

[0009]

Number

[0010]

Number

[0011]

Number

[0012]

Number

[0013] Optionally, the transmissive metasurface is composed of an array of a plurality of unit structures with different switching states. Each unit structure is composed of three layers of metal, and the three layers of metal are separated by two layers of F4B medium. Two PIN switching diodes are welded to the surface of the first layer of metal, and the switching state of the PIN switching diode is controlled by an externally applied voltage.

[0014] Optionally, the unit structure of the transmissive metasurface can conduct current in the forward and reverse directions when positive and negative voltages are externally connected, can realize a binary phase under the incidence of electromagnetic waves from -50° to 50°, and has a transmittance greater than 95% under the operating frequency.

[0015] Optionally, the focusing guide module is a focusing directional convolutional neural network, and the focusing directional convolutional neural network is composed of a mapping from the electric field e t to the compensation phase Δψ generated by all unit structures in the transmissive metasurface t and is configured,

[0016]

Number

[0017]

Number

[0018] Optionally, the evolutionary learning module adjusts the voltage of each unit state of the transmissive metasurface based on the compensation phase Δψ t and the collected electric field e t at the output time t + 1, and the voltage update scheme is as follows:

[0019]

Number

[0020]

Number

[0021]

Number

[0022] Optionally, the calculation methods of m t and v t are as follows:

[0023] [Number]

[0024] [Number] where g t is the gradient, m t-1 is the first moment of the gradient at time t - 1, and v t-1 is the second moment of the gradient at time t - 1.

[0025] As can be seen from the above technical solutions, compared with the prior art, the present invention discloses and provides a human - eye - like focusing metasurface system driven based on a supervised - evolutionary learning algorithm, and has the following beneficial effects.

[0026] 1. The adaptive focusing system designed in the present invention can achieve intelligent focusing at any position under multiple electromagnetic environments, without the need for artificial adjustment, and can be used flexibly.

[0027] 2. The present invention constructs a framework of a supervised - evolutionary learning algorithm, adaptively and intelligently adjusts the output scheme under different environments, and can make up for the defect that the prior machine learning cannot adapt to the environment.

[0028] 3. The sub - wavelength metasurface structure designed in the present invention is small in volume, easy to manufacture, and easy to integrate and implement.

[0029] 4. The metasurface unit with high transmittance designed in the present invention can reach a transmittance of more than 95% under different incident angles, has a high transmittance, and stable performance.

[0030] To more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings that need to be used in the following description of the embodiments or the prior art will be briefly described. Obviously, the drawings described below are only embodiments of the present invention, and those skilled in the art can also obtain other drawings based on the provided drawings without the need for creative labor.

Brief Description of the Drawings

[0031]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Best Mode for Carrying Out the Invention

[0032] Hereinafter, with reference to the drawings of the embodiments of the present invention, the technical solution of the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative labor belong to the protection scope of the present invention.

[0033] The embodiments of the present invention disclose a human-eye-like focusing metasurface system driven based on a teacher-involved evolutionary learning algorithm. As shown in FIG. 1, it includes a transmissive metasurface, an array probe, a focusing guide module, and an evolutionary learning module. When an external electromagnetic wave signal passes through the transmissive metasurface, the array probe installed behind the transmissive metasurface detects the external electromagnetic wave data. The focusing guide module and the evolutionary learning module analyze the external electromagnetic wave, output an adjustment strategy for the transmissive metasurface, the state of the transmissive metasurface changes, the array probe collects new external electromagnetic wave data, the focusing guide module and the evolutionary learning module further analyze the intensity and characteristics of the external electromagnetic wave data, output the next adjustment command, and repeat the above process until focusing on the specified position.

[0034] Specifically, as the principle of the human-eye-like adaptive focusing system, incident light passes through the lens and reaches the photoreceptor cells on the retina. At this time, the optical signal is converted into an electrical signal and perceived by the human brain. The brain analyzes the signal and then outputs an adjustment strategy for the lens through the nervous system. At this time, the ciliary muscle / lens of the eye expands and contracts based on the command of the system, the intensity of the signal received by the photoreceptor cells changes, the human brain further analyzes the signal characteristics at this time, outputs an adjustment command, and repeats the above process until the light converges completely on the retina. In the present invention, the transmissive metasurface functions as the lens, the array probe functions as the photoreceptor cells, and the focusing guide module and the evolutionary learning module perform analysis as the human brain.

[0035] Specifically, the framework of the supervised-evolutionary learning algorithm composed of the focusing guide module and the evolutionary learning module, the transmissive metasurface structure, and the external electromagnetic environment constitute a closed-loop adaptive iteration system of "environmental data collection - algorithm prediction - metasurface adjustment - environmental data collection". The external electromagnetic environment includes a single-source electromagnetic environment incident from any direction, a multi-source electromagnetic environment, and an electromagnetic environment with unknown scatterers. The supervised-evolutionary learning algorithm is composed of a supervised learning process and an evolutionary learning process. The core of the supervised learning is the convolutional neural network, and the core of the evolutionary learning is the adaptive moment estimation (Adam) gradient descent algorithm.

[0036] Furthermore, to stop the iteration after focusing on the specified position, the following judgment conditions need to be satisfied simultaneously.

[0037]

Number

[0038]

Number

[0039]

Number

[0040]

Number

[0041]

Number

[0042] Furthermore, the transmissive metasurface is composed of an array of a plurality of unit structures with different switching states. The unit structure is shown in FIG. 2. Each unit structure is composed of three layers of metal, and the three layers of metal are separated by two layers of F4B media. Two PIN switching diodes are welded to the surface of the first layer of metal, and the switching state of the PIN switching diode is controlled by an externally applied voltage.

[0043] Specifically, the adjustable transmissive unit structure is composed of three layers of metal layers (Cu) and two layers of dielectric layers (relative permittivity 2.65), and the middle is an adhesive layer with a relative permittivity of 4.4. Two PIN diodes are welded to the surface of the uppermost metal layer, and the diodes are commonly grounded through the middle holes. As shown in FIG. 3, the dashed line and the solid line represent different states of the two diodes respectively. When the diode is in the open state, its resistance value is 2.1 Ω and its capacitance value is 0 fF. When the diode is in the closed state, the resistance value is 0 Ω and the capacitance is 50 fF. The abscissa is the frequency, the left ordinate is the phase, and the right ordinate is the amplitude. It can be seen that when the diode is in different states, its phase is inverted by 180°, and below the operating frequency (5.85 GHz to 5.95 GHz), the amplitudes are all -1 dB or more.

[0044] Furthermore, the unit structure of the transmissive metasurface can allow current to flow in both forward and reverse directions when positive and negative voltages are externally connected, can realize a binary phase under the incidence of electromagnetic waves from -50° to 50°, and has a transmittance greater than 95% at the operating frequency.

[0045] Furthermore, the focusing guide module is a focusing directional convolutional neural network, and the focusing directional convolutional neural network is the mapping from the electric field e t to the compensation phase Δψ generated by all unit structures in the transmissive metasurface t and is composed of

[0046]

Number

[0047]

Number

[0048] Furthermore, the evolutionary learning module adjusts the voltage of each unit state of the transmissive metasurface based on the compensation phase Δψ t and the output time t + 1 of the collected electric field e t , and the voltage update scheme is as follows:

[0049]

Number

[0050]

Number

[0051]

Number

[0052] Furthermore, the calculation methods of m t and v t are as follows:

[0053]

Number

[0054]

Number

[0055]

Number

[0056] Furthermore, as shown in FIG. 4, the supervised-evolutionary learning algorithm generally includes two modules: an operation cycle and an evolving cycle. When an electromagnetic wave is incident on the tunable metasurface, (a) the detection array collects electric field data, (b) the data is transmitted to the supervised guidance network, (c) the network outputs an adjusted phase and transfers it to the control side, and (d) at this time, the algorithm outputs an adjustment method for each unit on the metasurface, and (e) the state of the metasurface changes. This process is repeated until the task of the focusing point is completed. During the supervised-evolutionary process, the algorithm simultaneously collects the traversed data (f) to facilitate the acceleration of subsequent iterations.

[0057] Furthermore, at different positions of the tunable focusing metasurface under different electromagnetic environments, the focusing results are as follows. FIGS. 5, 6, and 7 are respectively three-dimensional schematic diagrams of the electric field at the first iteration and at the end of the iteration in a single-source electromagnetic environment, a comparison diagram of one-dimensional electric field data (theory, first iteration, end of iteration), and a focusing result diagram at different focusing positions. FIGS. 8, 9, and 10 are respectively three-dimensional schematic diagrams of the electric field at the first iteration and at the end of the iteration in a double-source electromagnetic environment, a comparison diagram of one-dimensional electric field data (theory, first iteration, end of iteration), and a focusing result diagram at different focusing positions. FIGS. 11, 12, and 13 are respectively three-dimensional schematic diagrams of the electric field at the first iteration and at the end of the iteration in an electromagnetic environment with a scattering obstacle, a comparison diagram of one-dimensional electric field data (theory, first iteration, end of iteration), and a focusing result diagram at different focusing positions. It can be seen that the supervised-evolutionary learning algorithm proposed in the present invention has strong adaptability and learning ability under different electromagnetic environments and can achieve focusing at any specified position.

[0058] Each embodiment of this specification is described step by step. Each embodiment focuses on the differences from other embodiments, and the same or similar parts between each embodiment may be referred to each other.

[0059] Based on the above description of the disclosed embodiments, those skilled in the art can implement or use the present invention. Various corrections to these embodiments are obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to these embodiments shown herein, but rather conforms to the broadest scope consistent with the principles and novel features disclosed herein.

Claims

1. A human-eye-like focusing metasurface system driven based on a teacher-assisted evolutionary learning algorithm, comprising a transmissive metasurface, an array probe, a focusing guide module, and an evolutionary learning module. When an external electromagnetic wave signal passes through the transmissive metasurface, the array probe installed behind the transmissive metasurface detects the external electromagnetic wave data, and the focusing guide module and the evolutionary learning module analyze the external electromagnetic wave, output an adjustment strategy for the transmissive metasurface, the state of the transmissive metasurface changes, the array probe collects new external electromagnetic wave data, the focusing guide module and the evolutionary learning module further analyze the intensity and characteristics of the external electromagnetic wave data, output the next adjustment command, and repeat the above process until focusing on the specified position. A human-eye-like focusing metasurface system driven based on a teacher-assisted evolutionary learning algorithm, characterized in that.

2. To stop the iteration after focusing on the specified position, it is necessary to simultaneously satisfy the following judgment conditions: 【Number 1】 【Number 2】 【Mathematics 3】 where r(E Ｔｈｅｏｒｙ , E Ｔｅｓｔ ) is the effective correlation coefficient between the theoretical electric field E Ｔｈｅｏｒｙ and the test electric field E Ｔｅｓｔ , C(E Ｔｈｅｏｒｙ , E Ｔｅｓｔ ) is the covariance between the theoretical electric field E Ｔｈｅｏｒｙ and the test electric field E Ｔｅｓｔ , Var[E Ｔｈｅｏｒｙ is the variance of the theoretical electric field E Ｔｈｅｏｒｙ , Var[E Ｔｅｓｔ is the variance of the test electric field E Ｔｅｓｔ , ρ ｍ is the main lobe energy ratio of the electromagnetic wave, [Number 4] is the main lobe energy of the electromagnetic wave, 【Number 5】 is the measured total electric field energy, and ρ ｓ is the total side-lobe energy ratio of the electromagnetic wave, and E Ｓｉｄｅ＿ｌｏｂｅ A human-eye-like focusing metasurface system driven based on a teacher-present evolutionary learning algorithm according to claim 1, characterized in that the side-lobe energy of the electromagnetic wave is E

3. The transmissive metasurface is composed of an array of a plurality of unit structures with different switching states. Each unit structure is composed of three layers of metal, and the three layers of metal are separated by two layers of F4B media. Two PIN switching diodes are welded to the surface of the first layer of metal, and the switching state of the PIN switching diode is controlled by an externally applied voltage. The human-eye-like focusing metasurface system driven based on the teacher-assisted evolutionary learning algorithm according to claim 1, characterized in that.

4. The unit structure of the transmissive metasurface can allow current to flow in both forward and reverse directions when positive and negative voltages are externally connected, can realize a binary phase under the incidence of electromagnetic waves from -50° to 50°, and has a transmittance greater than 95% under the operating frequency. The human-eye-like focusing metasurface system driven based on the teacher-assisted evolutionary learning algorithm according to claim 3, characterized in that.

5. The beam-steering guide module is a beam-steering convolutional neural network, and the beam-steering convolutional neural network is the electric field e ｔ from the compensation phase Δψ generated by all unit structures in the transmissive metasurface ｔ to the mapping and is composed of 【Number 6】 Wherein, 【Number 7】 is the focusing compensation phase that the i-th unit structure needs to satisfy, where i ∈ [1, N], and θ is a network parameter when training the focusing directional convolutional neural network. A human-eye-like focusing metasurface system driven based on the supervised-evolutionary learning algorithm according to claim 3, characterized in that.

6. The evolutionary learning module adjusts the voltage of each unit state of the transmissive metasurface based on the compensation phase Δψ ｔ and the collected electric field e ｔ at the output time t + 1, and the voltage update scheme is as follows: 【Number 8】 where u ｔ and u ｔ＋１ represent the adjusted voltages at time t and time t + 1, respectively, α, ∈ are learning parameters, β １ , β ２ are the decay coefficients of the first-order moment and the second-order moment, respectively, 【Number 9】 is β １ to the power of t, and 【Number 10】 is β ２ to the power of t, and m ｔ is the first moment of the gradient at time t, and v ｔ is the second moment of the gradient at time t, characterized by the teacher-present evolutionary learning algorithm according to claim 5, a human-eye-like focusing metasurface system driven thereby.

7. m ｔ and v ｔ The calculation methods are as follows, 【Number 11】 【Number 12】 where g ｔ is the gradient, m ｔ－１ is the first moment of the gradient at time t - 1, and v ｔ－１ is the second moment of the gradient at time t - 1, a human-eye-like focusing metasurface system driven based on the supervised-evolutionary learning algorithm according to claim 6.

Images