CN108733924A - A kind of intellectualized design method of digital coding metamaterial unit - Google Patents
A kind of intellectualized design method of digital coding metamaterial unit Download PDFInfo
- Publication number
- CN108733924A CN108733924A CN201810486978.1A CN201810486978A CN108733924A CN 108733924 A CN108733924 A CN 108733924A CN 201810486978 A CN201810486978 A CN 201810486978A CN 108733924 A CN108733924 A CN 108733924A
- Authority
- CN
- China
- Prior art keywords
- 1bit
- design
- unit
- units
- phase
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Aerials With Secondary Devices (AREA)
Abstract
The present invention discloses a kind of intellectualized design method of digital coding metamaterial unit, includes the following steps：1) fundamental design of the 1bit units to building Meta Materials carries out digital coding, Element Polarization wave reflection phase when predicting that disparate modules are arranged in 1bit units using deep learning algorithm for design；2) according to Meta Materials functional requirement, the polarized wave phase difference θ of the 1bit cellular constructions of structure Meta Materials is determined, in conjunction with Binary Particle Swarm Optimization module and 1bit cellular constructions that deep learning module design polarized wave phase difference is θ.The design method of the present invention realizes the Automation Design of the ideally-reflecting phase to more bit units based on deep learning, with high efficiency and simplicity, and autgmentability is good, it can replace software emulation, shorten and obtain the corresponding complexity of encoded element information and time, quickly and easily designs multi-beam multipolarization artificial electromagnetic surface.
Description
Technical field
The present invention relates to a kind of design method of metamaterial unit, more particularly to a kind of intelligence of digital coding metamaterial unit
Design method can be changed, belong to machine learning and digital coding Meta Materials field.
Background technology
There is the artificial composite wood of special conduction or radiation characteristic (negative refraction, zero refraction) when electromagnetic wave is propagated wherein
Material is called Novel manual electromagnetic material (Metamaterials), it may also be said to which Novel manual electromagnetic material is that one kind can be with people
Work design, the electromagnetic material for meeting specific effective dielectric constant and magnetic conductivity requirement.In order to realize the different function of Meta Materials, lead to
Overregulate metamaterial modular construction realize phase variation, then adjust the unit of different reflected phases distribution situation can make into
Ejected wave is through pencil beam of the array structure radiation with high directionality or assigns traveling wave beam.
Phase response for coding unit (16 × 16 random point array element) is determined by the coding of unit, for axis
Symmetrical cell has 264Kind coding and corresponding phase information.Generally use full-wave simulation software CST in the prior art
Microwave Studio are from 264Required coding is obtained in kind coding to encode as metamaterial unit, this design method can consume
Take huge time and efforts.
The concept of deep learning was proposed by Hinton et al. in 2006, and was obtaining development at full speed in recent years, was broken through
The application achievements of property emerge one after another.Deep learning is a kind of method based on to data progress representative learning in machine learning, can
To be modeled automatically to high dimensional data data, the trouble of artificial extraction feature is eliminated, what the system that improves was designed and run
Efficiency.
Particle group optimizing (Particle Swarm Optimization, PSO) algorithm is Kennedy and Eberhart rich
A kind of swarm intelligence Evolutionary Computation that scholar proposes in nineteen ninety-five.Particle swarm optimization algorithm initially proposes to be suitable for solving continuous
The optimization problem in space.On the basis of continuity particle cluster algorithm, Binary Particle Swarm Optimization (Binary PSO,
BPSO) it is suggested the practical sex chromosome mosaicism for solving many Combinatorial Optimizations.In Binary Particle Swarm Optimization, the position of particle is compiled
Code uses binary mode, that is, every one-dimensional component of particle to be indicated with " 0 " and " 1 ".
Based on this, inventor combines deep learning method and particle swarm optimization algorithm, forms the technology of the present invention.
Invention content
Goal of the invention：For encoded element information existing for existing Meta Materials coding unit design method obtain it is complicated and
The problem for obtaining time length, provides a kind of intellectualized design method of digital coding metamaterial unit.
Technical solution：A kind of intellectualized design method of digital coding metamaterial unit of the present invention, including it is following
Step：
1) fundamental design of the 1bit units to building Meta Materials carries out digital coding, pre- using deep learning algorithm for design
Survey Element Polarization wave reflection phase when disparate modules are arranged in 1bit units；
2) according to Meta Materials functional requirement, the polarized wave phase difference θ of the 1bit cellular constructions of structure Meta Materials is determined, in conjunction with
Binary Particle Swarm Optimization module and the 1bit cellular constructions that deep learning module design polarized wave phase difference is θ.
Above-mentioned steps 1) in, Random Discrete lattice may be selected and do the fundamental designs of 1bit unit particles, the fundamental design by
Air module and metal module composition, are labeled as " 0 ", metal module is labeled as " 1 ", obtains 1bit cell encodings by air module
Pattern.Specifically, special neural network of the structure from cell encoding to reflected phase, 1bit cell encodings on the basis of Resnet
" 0 " and " 1 " is corresponding with " 0 " and " 1 " in deep learning algorithm for design in pattern, the 1bit cell encodings that disparate modules are arranged
Pattern transformation at 0-1 matrixes as network inputs, reflected phase every once discretization to 360 states, various 1bit are mono-
The corresponding reflected phase of primitive encoding pattern is exported as network, to predict unit when disparate modules are arranged in 1bit units
Polarized wave reflected phase.
Above-mentioned steps 2) in, 1bit units are made of " 0 " unit and two units of " 1 " unit, and polarized wave phase difference is θ's
1bit units are：Form the phase of the reflected phase of two units " 0 " of 1bit units and the TM polarized waves and TE polarized waves of " 1 "
Difference is θ, and the phase difference value of the reflected phase of the identical polarized wave of two units is 180 °.
Wherein, it is in conjunction with the design method of Binary Particle Swarm Optimization module and deep learning module：Using two into
Subgroup optimization algorithm module of pelletizing generates primary group, the speed of more new particle and position, and calculates in each iteration suitable
Response；Using the quick calculating of deep learning algorithm for design and the reflected phase of output unit.
Specifically design procedure is：
1) withI=1,2 are used as fitness function, the coding pattern of optimization unit particle to obtain
fitness1Minimum value, i.e. i=1, obtain 1bit units in " 0 " unit, whereinWithThe TM of unit is indicated respectively
With the reflected phase of TE polarized waves；
(2) reflected phase of " 0 " unit is exported, and enters the second wheel and optimizes；
(3) reflected phase of " 0 " unit optimized based on the first round, in second of optimization design, withI=1,2 andAs fitness function, optimize unit particle
Pattern obtains fitness1And fitness2Minimum value, that is, obtain 1bit units in " 1 " unit；Wherein, Respectively
Indicate the reflected phase of the TE polarized waves of unit " 0 " and unit " 1 ".
Advantageous effect：Compared with the prior art, the advantages of the present invention are as follows：The design method of the present invention is based on deep learning
The Automation Design of the ideally-reflecting phase to more bit units is realized, there is high efficiency and simplicity, and autgmentability is good,
Software emulation can be replaced, shortens and obtains the corresponding complexity of encoded element information and time；Meanwhile it also providing a kind of to people
The significant trial of work electromagnetic material intelligent algorithm.
Description of the drawings
Fig. 1 is the axially symmetric structure schematic diagram of metamaterial unit in the present invention；
Fig. 2 is the variation and feature extraction that 1bit cell encoding patterns input characteristic dimension after deep learning neural network
Flow chart；
Fig. 3 is that Binary Particle Swarm Optimization module and deep learning module design 1bit unit knots are combined in the present invention
The method flow diagram of structure；
Fig. 4 is the structural representation for two groups of 1bit units that the polarized wave reflected phase difference that the present invention designs is 0 ° and 180 °
Figure；
Fig. 5 is the structural representation for two groups of 1bit units that the polarized wave reflected phase difference that the present invention designs is -45 ° and 135 °
Figure；
Fig. 6 is the structural representation for two groups of 1bit units that the polarized wave reflected phase difference that the present invention designs is -90 ° and 90 °
Figure；
Fig. 7 is the structural representation for two groups of 1bit units that the polarized wave reflected phase difference that the present invention designs is -135 ° and 45 °
Figure.
Specific implementation mode
Technical scheme of the present invention is described further below in conjunction with the accompanying drawings.
Left-hand circular polarization, the three wave beam multipolarization people of three kinds of wave beams of right-handed circular polarization and linear polarization can be radiated with design
For the metamaterial unit of work resistance electromagnetic surface, the intellectualized design method of the digital coding metamaterial unit of the present invention is said
It is bright.
1) fundamental design of the 1bit units to building Meta Materials carries out digital coding, pre- using deep learning algorithm for design
Survey Element Polarization wave reflection phase when disparate modules are arranged in 1bit units；
First, we select Random Discrete lattice to be the fundamental design of unit particle, such as Fig. 1.The lattice pattern includes 16
× 16 blockages, wherein grey blockage is metal derby, and white blockage is air block, that is, does not cover any material.
Grey blockage is labeled as " 1 ", and white blockage is labeled as " 0 ", and opposite with " 1 " and " 0 " in deep learning algorithm for design
It answers, is labeled as 1bit cell encodings.The two dimension arrangement of different " 0 " and " 1 ", that is, different cell encodings, determine list
The reflected phase of elementary particle.In our design, cell encoding is encoded using axisymmetric, thus has 264Kind coding
Pattern has also corresponded to various reflected phase.
In deep learning module, we devise dedicated 101 layer network based on Resnet, 1bit cell encoding patterns
0-1 matrixes are converted to as network inputs；Reflected phase is every once discretization to 360 states, by 1bit cell encodings pair
The reflected phase answered is exported as network.It is trained by a large amount of data, we make e-learning to from being encoded to phase
Relationship, and reached required precision.Fig. 2 shows the variation of characteristic dimension and feature after cell encoding input neural network
Extraction process finally extracts 2048 characteristic points, and is converted to output phase feature to by a fully-connected network
Prediction.Table 1 shows the specific convolution operation involved in the constructed each block structure of neural network, not by small convolution kernel
It is disconnected to stack, it increases the acceptance region of feature extraction and so that the feature extracted is more and more advanced.
Involved specific convolution operation in each block structure of neural network constructed by table 1
2) according to Meta Materials functional requirement, the polarized wave phase difference that the 1bit cellular constructions of structure Meta Materials need to meet is determined
θ, in conjunction with the 1bit unit knots that Binary Particle Swarm Optimization module and deep learning module design polarized wave phase difference are θ
Structure.
The metamaterial unit that the present invention designs is for building three wave beam multipolarization artificial electromagnetic surfaces, three wave beam multipolarization people
Work resistance electromagnetic surface must be encoded to obtain by the 1bit cell arrays of 3bit groups.The composition of 1bit codings is defined as two reflected phase phases
Poor 180 ° of unit, such as θ=0 ° and the two of 180 ° groups of units or θ=45 ° and -135 ° of two groups of units etc.；2bit is encoded
Composition be defined as the unit that four reflected phases differ 90 °, such as θ=- 90 °, 0 °, 90 ° and 180 ° of four groups of units；3bit
The composition of coding is defined as the unit that eight reflected phases differ 45 °, such as θ=- 135 °, -90 °, -45 °, 0 °, -45 °, 90 °,
135 ° and 180 ° of eight groups of units, the i.e. present invention need to design polarized wave phase difference be θ be -135 °, -90 °, -45 °, 0 °, -45 °,
90 °, 135 ° and 180 ° of eight groups of 1bit cellular constructions.
Constitute two units " 0 " in 1 group of 1bit unit and " 1 " x polarized waves (transverse-magnetic, TM,
Polarized wave) the phase difference of reflected phase of reflected phase and y polarized waves (transverse-electric, TE, polarized wave) be
θ, and the phase difference value of the identical polarized radiation phase of two units is 180 °.In 1 group of 1bit unit " 0 " and " 1 " unit with
Binary system particle coding " 0 " and " 1 " in binary particle swarm algorithm is corresponding.
Fig. 3 is the design cycle based on Binary Particle Swarm Optimization module and deep learning module, is obtained with this
There are the 1bit anisotropy coding units of fixation reflex phase difference on orthogonal direction.This design program includes two modules, from
Particle swarm optimization algorithm (BPSO) module and deep learning module are dissipated, Binary Particle Swarm Optimization module is initial for generating
Population, the speed of more new particle and position, and fitness is calculated in each iteration；Deep learning for quickly calculate and
The reflected phase of output unit.Specific design procedure is as follows：
(1) withI=1,2 are used as fitness function, the pattern of optimization unit particle to obtain
fitness1Minimum value, i.e. i=1, obtain 1bit units in " 0 ", that is, its TM and TE polarized waves reflected phase
Difference is closest to θ；WithThe reflected phase of the TM and TE polarized waves of unit is indicated respectively.
(2) reflected phase of " 0 " unit is exported, and passes in and out the second wheel optimization；
(3) reflected phase of " 0 " unit optimized based on the first round, in second of optimization design, withI=1,2 andAs fitness function, optimize unit particle
Pattern obtain fitness1And fitness2Minimum value, that is, obtain " 1 " in 1bit units, that is, its poles TM and TE
Change the nearly θ of reflected phase differential of wave, and two units " 0 " of 1bit and the phase difference of " 1 " are close to 180 °；
By this design method, it is -135 ° that can obtain polarized wave phase difference θ, -90 °, -45 °, 0 °, -45 °, 90 °,
135 ° and 180 ° of 8 groups of 1bit units, such as Fig. 4~7.Realization one can be combined using the 1bit units of full income 3bit groups
Three wave beam multipolarizations (left-hand circular polarization, right-handed circular polarization and linear polarization) artificial electromagnetic surface.
Claims (6)
1. a kind of intellectualized design method of digital coding metamaterial unit, which is characterized in that include the following steps：
1) fundamental design of the 1bit units to building Meta Materials carries out digital coding, is predicted using deep learning algorithm for design
Element Polarization wave reflection phase when disparate modules are arranged in 1bit units；
2) according to Meta Materials functional requirement, determine the polarized wave phase difference θ of the 1bit cellular constructions of structure Meta Materials, in conjunction with two into
Subgroup optimization algorithm module of pelletizing and the 1bit cellular constructions that deep learning module design polarized wave phase difference is θ.
2. the intellectualized design method of digital coding metamaterial unit according to claim 1, which is characterized in that step 1)
In, select Random Discrete lattice to do the fundamental design of 1bit unit particles, the fundamental design is by air module and metal module group
At by air module labeled as " 0 ", metal module is labeled as " 1 ", obtains 1bit cell encoding patterns.
3. the intellectualized design method of digital coding metamaterial unit according to claim 2, which is characterized in that step 1)
In, special neural network of the structure from cell encoding to reflected phase on the basis of Resnet, in 1bit cell encoding patterns " 0 "
" 1 " is corresponding with " 0 " and " 1 " in deep learning algorithm for design, the 1bit cell encoding pattern transformations that disparate modules are arranged
At 0-1 matrixes as network inputs, reflected phase is every once discretization to 360 states, by various 1bit cell encodings figures
The corresponding reflected phase of case is exported as network, to which Element Polarization wave when predicting disparate modules arrangement in 1bit units is anti-
Penetrate phase.
4. the intellectualized design method of digital coding metamaterial unit according to claim 2, which is characterized in that step 2)
In, the 1bit units that the polarized wave phase difference is θ are：Constitute the poles TM of two units " 0 " and " 1 " in 1 group of 1bit unit
The phase difference for changing the reflected phase of wave and TE polarized waves is θ, and the phase of the reflected phase of the identical polarized wave of two units
Difference is 180 °.
5. the intellectualized design method of digital coding metamaterial unit according to claim 4, which is characterized in that step 2)
In, the design method of the combination Binary Particle Swarm Optimization module and deep learning module is：Using binary system particle
Colony optimization algorithm module generates primary group, the speed of more new particle and position, and calculates fitness in each iteration；It adopts
With the quick calculating of deep learning algorithm for design and the reflected phase of output unit.
6. the intellectualized design method of digital coding metamaterial unit according to claim 5, which is characterized in that step 2)
In, the design procedure of the combination Binary Particle Swarm Optimization module and deep learning module includes：
(1) withAs fitness function, the coding pattern of optimization unit particle obtains
fitness1Minimum value, i.e. i=1, obtain 1bit units in " 0 " unit, whereinWithRespectively indicate unit TM and
The reflected phase of TE polarized waves；
(2) reflected phase of " 0 " unit is exported, and enters the second wheel and optimizes；
(3) reflected phase of " 0 " unit optimized based on the first round, in second of optimization design, withWithAs fitness function, optimize unit grain
The pattern of son obtains fitness1And fitness2Minimum value, that is, obtain 1bit units in " 1 " unit；Wherein, Point
Not Biao Shi unit " 0 " and unit " 1 " TE polarized waves reflected phase.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810486978.1A CN108733924B (en) | 2018-05-21 | 2018-05-21 | Intelligent design method of digital coding metamaterial unit |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810486978.1A CN108733924B (en) | 2018-05-21 | 2018-05-21 | Intelligent design method of digital coding metamaterial unit |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108733924A true CN108733924A (en) | 2018-11-02 |
| CN108733924B CN108733924B (en) | 2022-05-10 |
Family
ID=63937666
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810486978.1A Active CN108733924B (en) | 2018-05-21 | 2018-05-21 | Intelligent design method of digital coding metamaterial unit |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108733924B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110110398A (en) * | 2019-04-19 | 2019-08-09 | 西安工程大学 | A kind of super surface automatic design method based on convolution self-encoding encoder |
| CN110245413A (en) * | 2019-06-11 | 2019-09-17 | 西北工业大学 | A kind of design method of Meta Materials absorbent structure neural network based |
| CN110518361A (en) * | 2019-08-22 | 2019-11-29 | 西安电子科技大学 | A kind of adjustable super surface cell of reflection of 2 bit phases |
| CN110647024A (en) * | 2019-10-30 | 2020-01-03 | 武汉大学 | Method for realizing circuitous phase coding multiplexing based on super-surface array structure |
| CN110718755A (en) * | 2019-10-16 | 2020-01-21 | 东南大学 | Integrated low-profile digital coding antenna |
| CN113270157A (en) * | 2021-05-20 | 2021-08-17 | 兰州大学 | Mechanical metamaterial design and optimization method based on machine learning |
| CN117610317A (en) * | 2024-01-19 | 2024-02-27 | 湖北工业大学 | Multi-bit super-surface phase arrangement optimization method based on deep learning |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1650688A1 (en) * | 2004-10-20 | 2006-04-26 | International Business Machines Corporation | Automatic addition of power connections to chip power supply network |
| WO2006108768A1 (en) * | 2005-04-15 | 2006-10-19 | France Telecom | Method and device for decoding a variable length code taking into account an a priori probability information |
| US20150213374A1 (en) * | 2014-01-24 | 2015-07-30 | International Business Machines Corporation | Detecting hotspots using machine learning on diffraction patterns |
| CN104868252A (en) * | 2015-06-05 | 2015-08-26 | 东南大学 | 1-bit microwave anisotropic electromagnetic coding meta-material |
| WO2015169109A1 (en) * | 2014-05-06 | 2015-11-12 | 国家电网公司 | A transformer noise suppression method |
| CN106410418A (en) * | 2016-08-11 | 2017-02-15 | 东南大学 | Dual-functional anisotropic electromagnetic encoding metamaterial applied to microwave band, and basic unit structure and design method |
| CN107704673A (en) * | 2017-09-26 | 2018-02-16 | 中国人民解放军空军工程大学 | Wideband encoding Meta Materials Fast design method |
-
2018
- 2018-05-21 CN CN201810486978.1A patent/CN108733924B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1650688A1 (en) * | 2004-10-20 | 2006-04-26 | International Business Machines Corporation | Automatic addition of power connections to chip power supply network |
| WO2006108768A1 (en) * | 2005-04-15 | 2006-10-19 | France Telecom | Method and device for decoding a variable length code taking into account an a priori probability information |
| US20150213374A1 (en) * | 2014-01-24 | 2015-07-30 | International Business Machines Corporation | Detecting hotspots using machine learning on diffraction patterns |
| WO2015169109A1 (en) * | 2014-05-06 | 2015-11-12 | 国家电网公司 | A transformer noise suppression method |
| CN104868252A (en) * | 2015-06-05 | 2015-08-26 | 东南大学 | 1-bit microwave anisotropic electromagnetic coding meta-material |
| CN106410418A (en) * | 2016-08-11 | 2017-02-15 | 东南大学 | Dual-functional anisotropic electromagnetic encoding metamaterial applied to microwave band, and basic unit structure and design method |
| CN107704673A (en) * | 2017-09-26 | 2018-02-16 | 中国人民解放军空军工程大学 | Wideband encoding Meta Materials Fast design method |
Non-Patent Citations (2)
| Title |
|---|
| Q. ZHANG: "Coding Metasurface for Shaping Beams through Software-Based Approach", 《11TH INTERNATIONAL CONGRESS ON ENGINEERED MATERIAL PLATFORMS FOR NOVEL WAVE PHENOMENA – METAMATERIALS》 * |
| 刘峻峰: "太赫兹信息超材料与超表面", 《雷达学报》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110110398A (en) * | 2019-04-19 | 2019-08-09 | 西安工程大学 | A kind of super surface automatic design method based on convolution self-encoding encoder |
| CN110245413A (en) * | 2019-06-11 | 2019-09-17 | 西北工业大学 | A kind of design method of Meta Materials absorbent structure neural network based |
| CN110518361A (en) * | 2019-08-22 | 2019-11-29 | 西安电子科技大学 | A kind of adjustable super surface cell of reflection of 2 bit phases |
| CN110518361B (en) * | 2019-08-22 | 2021-01-12 | 西安电子科技大学 | 2-bit phase-adjustable reflection super-surface unit |
| CN110718755A (en) * | 2019-10-16 | 2020-01-21 | 东南大学 | Integrated low-profile digital coding antenna |
| CN110718755B (en) * | 2019-10-16 | 2021-01-05 | 东南大学 | Integrated low-profile digital coding antenna |
| CN110647024A (en) * | 2019-10-30 | 2020-01-03 | 武汉大学 | Method for realizing circuitous phase coding multiplexing based on super-surface array structure |
| CN110647024B (en) * | 2019-10-30 | 2020-10-09 | 武汉大学 | Method for realizing circuitous phase coding multiplexing based on super-surface array structure |
| CN113270157A (en) * | 2021-05-20 | 2021-08-17 | 兰州大学 | Mechanical metamaterial design and optimization method based on machine learning |
| CN117610317A (en) * | 2024-01-19 | 2024-02-27 | 湖北工业大学 | Multi-bit super-surface phase arrangement optimization method based on deep learning |
| CN117610317B (en) * | 2024-01-19 | 2024-04-12 | 湖北工业大学 | Multi-bit super-surface phase arrangement optimization method based on deep learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108733924B (en) | 2022-05-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108733924A (en) | A kind of intellectualized design method of digital coding metamaterial unit | |
| Dong et al. | Fast multi-objective optimization of multi-parameter antenna structures based on improved BPNN surrogate model | |
| Cui et al. | A modified efficient KNN method for antenna optimization and design | |
| Naseri et al. | A generative machine learning-based approach for inverse design of multilayer metasurfaces | |
| Weile et al. | Genetic algorithm optimization applied to electromagnetics: A review | |
| Berthelier et al. | Deep model compression and architecture optimization for embedded systems: A survey | |
| CN108763708A (en) | Method based on deep learning design multi-beam multipolarization artificial electromagnetic surface | |
| CN111611683A (en) | Electromagnetic super-surface design method and device based on deep learning | |
| Wang et al. | Efficient gradient‐based optimisation of pixel antenna with large‐scale connections | |
| CN110083934B (en) | Electromagnetic super-surface design method based on support vector machine algorithm | |
| Nazari et al. | Tot-net: An endeavor toward optimizing ternary neural networks | |
| Naseri et al. | A combined machine-learning/optimization-based approach for inverse design of nonuniform bianisotropic metasurfaces | |
| CN110020721B (en) | Target detection deep learning network optimization method based on parameter compression | |
| CN111199126A (en) | Rapid optimization design method of patch antenna | |
| On et al. | Optimal design of microwave absorber using novel variational autoencoder from a latent space search strategy | |
| Nasr et al. | Hybrid genetic algorithm for constrained nonlinear optimization problems | |
| CN111898316A (en) | Construction method and application of super-surface structure design model | |
| Naik et al. | Survey on comparative study of pruning mechanism on mobilenetv3 model | |
| CN107704673B (en) | Rapid design method for broadband coding metamaterial | |
| Ding et al. | Antenna optimization based on auto-context broad learning system | |
| Chen et al. | Joint sparse neural network compression via multi-application multi-objective optimization | |
| CN114239439B (en) | Automatic filter design method based on tribal algorithm | |
| Tan et al. | Artificial neural network application in prediction of concrete embedded antenna performance | |
| CN115146544A (en) | Array antenna design method adopting knowledge and data hybrid driving | |
| Liu et al. | Adapruner: Adaptive channel pruning and effective weights inheritance |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |