CN117195640A - Integrated photon device design method based on reverse design idea - Google Patents
Integrated photon device design method based on reverse design idea Download PDFInfo
- Publication number
- CN117195640A CN117195640A CN202311170726.5A CN202311170726A CN117195640A CN 117195640 A CN117195640 A CN 117195640A CN 202311170726 A CN202311170726 A CN 202311170726A CN 117195640 A CN117195640 A CN 117195640A
- Authority
- CN
- China
- Prior art keywords
- fom
- grids
- grid
- area
- optimization
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 58
- 238000013461 design Methods 0.000 title claims abstract description 56
- 230000006870 function Effects 0.000 claims abstract description 64
- 238000005457 optimization Methods 0.000 claims abstract description 50
- 238000004088 simulation Methods 0.000 claims abstract description 11
- 238000005498 polishing Methods 0.000 claims abstract description 3
- 239000000463 material Substances 0.000 claims description 78
- 239000000758 substrate Substances 0.000 claims description 23
- 101100273916 Schizosaccharomyces pombe (strain 972 / ATCC 24843) wip1 gene Proteins 0.000 claims description 20
- 230000005672 electromagnetic field Effects 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 12
- 101100390985 Schizosaccharomyces pombe (strain 972 / ATCC 24843) fmc1 gene Proteins 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 8
- 230000000694 effects Effects 0.000 claims description 7
- 238000003860 storage Methods 0.000 claims description 7
- 101100460203 Schizosaccharomyces pombe (strain 972 / ATCC 24843) new2 gene Proteins 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 4
- 230000004048 modification Effects 0.000 claims description 4
- 238000012986 modification Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 abstract description 12
- 238000004422 calculation algorithm Methods 0.000 description 8
- 238000004364 calculation method Methods 0.000 description 6
- 238000002834 transmittance Methods 0.000 description 6
- 238000004891 communication Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 230000008033 biological extinction Effects 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N Silicium dioxide Chemical compound O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 description 2
- 230000015556 catabolic process Effects 0.000 description 2
- 238000006731 degradation reaction Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- XUIMIQQOPSSXEZ-UHFFFAOYSA-N Silicon Chemical compound [Si] XUIMIQQOPSSXEZ-UHFFFAOYSA-N 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010894 electron beam technology Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000013386 optimize process Methods 0.000 description 1
- 238000001259 photo etching Methods 0.000 description 1
- 229910052710 silicon Inorganic materials 0.000 description 1
- 239000010703 silicon Substances 0.000 description 1
- 235000012239 silicon dioxide Nutrition 0.000 description 1
- 239000000377 silicon dioxide Substances 0.000 description 1
Landscapes
- Optical Integrated Circuits (AREA)
Abstract
The application discloses an integrated photon device design method based on a reverse design idea, which comprises the following steps: setting the size of the optimized space and the input and output ports, and performing three-dimensional polishing on the optimized space; constructing an initial structure of an optimization space; constructing an index function for evaluating the target performance of the device; calculating an index function value of the initial structure, and giving a judgment threshold value; and sequentially executing global search and local search until the index function value meets the optimization requirement, and stopping optimization. The application only needs to calculate the objective function FOM itself, and does not need to additionally construct other functions. In addition, different heuristic strategies can be introduced, and the heuristic strategies can greatly shorten the simulation time, so that the final structure is easier to process.
Description
Technical Field
The application relates to the technical field of integrated photoelectrons, in particular to a method for optimizing an integrated photon device by using a reverse design idea.
Background
Reverse design is a common design concept in current industrial production. In the field of integrated photonics, a traditional integrated photonic device design is based on theoretical analysis of a specific structure to obtain processing characteristics of a device under the specific structure for an optical signal, and then tuning functions of the device by adjusting a small number of parameters, so that the finally obtained design is generally a combination of a plurality of highly symmetrical structures. The basic idea of reverse engineering is to first determine the required performance of the optical device and then construct the device structure by computer-aided analysis to enable performance. The design has extremely high degree of freedom of optimization, can break through the limitation of the traditional photon device and realize the function which cannot be achieved by the traditional photon device. The reverse design of the integrated photon device mainly adopts an accompanying simulation (Adjoint simulation) algorithm, the algorithm firstly assumes that the refractive index of the material in the optimization area can be continuously changed, and after the optimization result meets the requirement, the continuously changed refractive index is binarized (0 and 1 respectively correspond to the substrate and the waveguide area). The index binarization process inevitably leads to poor optimization results, the effect of which is not controllable. If the deterioration effect is severe, the optimization with the accompanying simulation needs to be continued, and even further initial structures need to be selected for re-optimization, which results in that the efficiency of the accompanying simulation optimization algorithm is not very high. Furthermore, the companion calculation function employed by the companion simulation is calculated separately for each function-specific device, and if the optimization objective function is difficult to represent by a simple analytical function, the selection of the companion calculation function becomes very difficult. Therefore, in order to increase the efficiency of integrated photonic reverse design, a new approach that is more versatile and practical is needed.
Disclosure of Invention
The application aims to overcome the defects in the prior art and provides a novel integrated photonics reverse design optimization method adopting intelligent search and greedy threshold mechanism. The method is suitable for any linear and nonlinear photon device optimization problem, and a calculation function does not need to be modified according to a specific problem; the output of each iteration is a structure capable of being processed in a delivering way, the performance of the objective function of the result of each iteration is improved compared with that of the last iteration, and the problem of performance degradation caused by binarization does not exist. Therefore, the method provided by the application has the advantages of controllability, universality and high efficiency compared with the accompanying simulation algorithm.
The application aims at realizing the following technical scheme:
an integrated photonic device design method based on a reverse design concept includes:
s1, setting the size of an optimized space and input and output ports, designating the material types of a substrate and a waveguide, and performing three-dimensional polishing on the optimized space to form a plurality of grids, wherein each grid is formed by the substrate or the waveguide material;
s2, constructing an initial structure of an optimization space, wherein all grids in the initial structure are single structures composed of substrate materials or waveguide materials or known integrated photon device structures close to target performance; for example, to design a 90 degree turn waveguide, it is known to use a 90 degree turn waveguide structure as the initial structure in integrated photonics devices.
S3, constructing an index function FOM for evaluating the target performance of the integrated photonic device based on an electromagnetic field simulation method, wherein the smaller the FOM is, the better the target performance index is;
s4, calculating index function value FOM of the initial integrated photon structure by using index function FOM 0 Given two decision thresholds G th And L th Let iteration number t=0;
s5, executing global search, randomly selecting an area B in the optimization space, inverting the material properties of all grids in the area B, and calculating to obtain the index function value FOM of the integrated photonic device new1 The method comprises the steps of carrying out a first treatment on the surface of the If (FOM) 0 －FOM new1 )/FOM 0 ＞G th Executing step S6, otherwise, recovering the material properties of all grids in the region B to a state before the material properties are reversed, and re-executing step S5;
s6, reducing the area B into a smaller area S by taking the geometric center of the area B as a reference point; restoring the material properties of all grids in the region B to a state before the material properties are reversed, and then reversing the material properties of all grids in the region S, and calculating to obtain the index function value FOM of the integrated photonic device new2 Make FOM 0 ＝FOM new2 T=t+1, step S7 is performed;
s7, marking all grids which are not in the area S in the optimization space as an unchecked state;
s8, if all grids adjacent to the area S are checked, executing a step S10, otherwise executing a step S9;
s9, executing local search, and arbitrarily selecting one of the areas which are adjacent to the area S and not yet in the optimized spaceThe checked area X is used for inverting the material attribute of the area X and calculating to obtain the index function value FOM of the current device new3 The method comprises the steps of carrying out a first treatment on the surface of the If (FOM) 0 －FOM new3 )/FOM 0 ＞L th FOM is made to 0 ＝FOM new3 S=s & -X, t=t+1, execute step S8; otherwise, restoring the material properties of all grids in the region X to the state before modification, marking that all grids in the region X are inspected, and executing step S8;
s10, if FOM 0 If the value of (2) meets the optimization requirement, the optimization is stopped to obtain a final design structure, otherwise, the step S5 is executed.
Further, the optimization space refers to the space occupied by the integrated photonic device to be optimized.
Further, the material property of the grid refers to a material type filled in a corresponding grid position, and the material type comprises a substrate or waveguide material.
Further, when the global search is executed in step S5, the following several or all heuristic strategies are adopted according to the actual optimization effect:
(1) If the materials of all the grids in the area B are substrates or waveguides and the materials of all the grids in the given area around the area B are substrates or waveguides, directly skipping the area B, and re-executing the step S5;
(2) Electromagnetic field energy at all meshes within the examination region B, if the sum of the electromagnetic field energies is smaller than a given threshold E th Directly skipping over the step B, and re-executing the step S5;
(3) After inverting the material properties of all grids in the region B, calculating the index function of the modified integrated photonic device based on two-dimensional FDTD to obtain an estimated value FOM est If (FOM) 0 －FOM est )/FOM 0 ＜αG th α is a given coefficient, directly skipping the region B, and re-executing step S5; otherwise, calculating index function value FOM by using three-dimensional FDTD new1 If (FOM) 0 －FOM new1 )/FOM 0 ＞G th FOM is made to 0 ＝FOM new1 Step S6 is executed, otherwise, the area B is covered withRestoring the material property with the grid to a state before the material property is reversed, and re-executing the step S5;
(4) Defining a gradient function for all grids of the optimization space: let the iteration number at a certain moment be t, and the index function value of the integrated photon device be FOM 0 For any grid r in the optimization space, the material property of the grid r is inverted, the index function value of the device is changed into FOM (r), and then the gradient function G (r, t) is defined as:
for any grid r, if the current iteration number t c G (r, t) c ) Unknown value, then one-dimensional interpolation is performed on G (r, t) along the time dimension t, according to t c G (r, t) is predicted by gradient values of the grid r under different previous iteration times c ) The method comprises the steps of carrying out a first treatment on the surface of the Then the iteration times t are fixed c Three dimensional pairs G (r, t c ) Performing three-dimensional spatial interpolation, predicting gradient values of each grid under the current iteration number according to the spatial interpolation result, obtaining grid positions with the fastest gradient drop according to the formula (1) and material properties of each grid r, selecting a region B near the grid, and executing step S5 to ensure that the maximum probability meets a threshold condition (FOM 0 －FOM new1 )/FOM 0 ＞G th 。
Further, the design structure finally obtained in step S10 is spatially filtered according to the processing precision, and the fine structure which cannot be processed is smoothed by convoluting the design structure with the spatial filter.
Further, the material property inversion refers to changing the current material of the grid into a waveguide when the current material of the grid is a substrate; when the grid current material is a waveguide, the grid current material is changed to a substrate.
Further, the electromagnetic field simulation method comprises a three-dimensional time domain finite difference method FDTD and a finite element analysis method.
The application also provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the integrated photonics device design method based on the reverse design concept when executing the program.
The present application also provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the integrated photonic device design method based on the reverse design concept.
Compared with the prior art, the technical scheme of the application has the following beneficial effects:
1. the target performance of the device is continuously improved in the execution process, and the problem of performance degradation caused by binarization exists in the traditional accompanying simulation algorithm.
2. The output of each iteration of the application is a deliverable structure, while the traditional structure with continuous change of refractive index before binarization by a simulation algorithm cannot be actually processed.
3. The traditional adjoint simulation algorithm needs to construct different adjoint calculation functions aiming at different objective functions FOM, when the objective functions FOM are very complex, the construction of the adjoint calculation functions is very difficult, and the application only needs to calculate the objective functions FOM and does not need to additionally construct other functions.
4. Aiming at different optimization tasks, different heuristic strategies can be introduced, the heuristic strategies can greatly shorten the simulation time, the final structure is easier to process, and the traditional accompanying simulation algorithm does not support the heuristic strategies.
Drawings
FIG. 1 shows the optimized space size and input/output ports in an embodiment of the present application.
FIG. 2 is an initial structure of an optimization space in an embodiment of the application.
FIG. 3 is a diagram illustrating a global search process according to an embodiment of the present application.
FIG. 4 is a diagram illustrating a local search implementation in accordance with an embodiment of the present application.
FIG. 5 is a final optimization result of an embodiment of the present application.
FIG. 6 shows the frequency response of two output ports in the design of FIG. 5 according to an embodiment of the present application.
FIG. 7 shows the final optimization result after the heuristic strategy (2) is adopted in the embodiment of the application.
FIG. 8 is a spatially filtered result of the design of FIG. 5 in accordance with an embodiment of the present application.
Fig. 9 is an effect diagram of the structure of fig. 6 after spatial filtering.
Detailed Description
The application is described in further detail below with reference to the drawings and the specific examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.
The specific embodiment provides a method for optimizing an integrated photon device based on a reverse design idea, the specific flow is shown in fig. 1, and the method is verified by adopting a Lumerical+MATLAB mixed development environment, wherein Lumerical is used for executing FDTD calculation to obtain FOM function values, and MATLAB is used for completing global search and local search; the method comprises the following steps:
first, as shown in FIG. 2, the length, width and height dimensions of the optimized section were set to 3.5. Mu.m.times.3.5. Mu.m.times.220 nm in Lumerical. The substrate material is silicon dioxide (white) and the waveguide material is silicon (black). The length and width dimensions of the throwing grid are 20nm×20nm×220nm, namely, the length and width directions respectively throw the optimized space 175 parts, and the height direction does not throw the space (so that the manufacture of a design structure can be completed only by one photoetching or electron beam exposure process when a chip is processed), the thrown grid can be represented by a simple 0-1 matrix of 175×175, 0 represents that the grid material is a substrate, 1 represents that the grid material is a waveguide, and the general three-dimensional space optimization problem proposed by the application is simplified into a two-dimensional plane optimization problem in the specific embodiment. In addition, fig. 2 also shows that the input port and the output port of the device are waveguides with the width of 600nm, and the center-to-center distance between the output ports 1 and 2 is 2500nm. The final objective of the integrated photonic device is to realize the wavelength division multiplexing function, and lasers with the wavelengths of 1550nm and 1950nm enter an optimized space from an input port, wherein the output port 1 only outputs the laser with the wavelength of 1950nm, and the output port 2 only outputs the laser with the wavelength of 1550 nm.
Fig. 3 shows an initial structure of the optimization space in this embodiment. The initial structure is a simple Y-beam splitter. With this configuration, the laser energy at wavelengths 1550nm and 1950nm are equally divided into two output ports, which obviously do not meet the target function and need to be improved by an optimization method. According to the target performance to be realized by the integrated photonic device, defining an index function as follows:
wherein P is in_1550 、P in_1950 The power of the laser with the wavelength of 1550nm and 1950nm entering the optimized space from the input port is respectively P out1_1950 Is the laser power of 1950nm wavelength, P, output from output port 1 out2_1550 Is the laser power of 1550nm wavelength outputted from the output port 2. As can be seen from the formula, fom=0 reaches the minimum value only when the laser power of 1950nm is all output from the output port 1 and the laser power of 1550nm is all output from the output port 2. In the specific embodiment, the FOM is less than 0.3, and the optimization target is achieved. For the initial configuration of the Y-beam splitter given in FIG. 3, the initial value of the index function is FOM 0 =1-0.5×0.5=0.75. To complete the subsequent optimization search, two decision thresholds G are set in MATLAB th ＝0.0001，L th =0, the iteration number t=0 is set.
Fig. 4 shows the execution of the global search in this embodiment. The randomly selected small region B in fig. 4 (a) is a circle with a radius of 80nm, and the material properties of all the grids in the small region need to be inverted, and since the original materials in the selected circular region in fig. 4 (a) are all substrates, all the materials in the circular region become black waveguide materials after the material properties are inverted. Calculating the index function value FOM of the given structure of FIG. 4 (a) new1 The method comprises the steps of carrying out a first treatment on the surface of the Because (FOM 0 －FOM new1 )/FOM 0 < 0.0001, so it is necessary to mix all the mesh materials in the circleThe attribute is restored to the substrate, and the global searching process is re-executed. After several global searches, the present embodiment finds a threshold condition (FOM) 0 －FOM new1 )/FOM 0 Circular region B > 0.0001, as shown in fig. 4 (B), so that the local search process can be entered.
In FIG. 5 (a), a circular region B with a radius of 80nm is first reduced to a circular region S with a radius of 40nm, the material properties of all grids in B are restored to the state before the reverse, and then the material properties of all grids in S are reversed, and the index function value FOM of the device is calculated new2 Make FOM 0 ＝FOM new2 T=t+1. All grids in the optimization space that are not in S are marked as unchecked.
Then, in FIG. 5 (b), a circular region X with a radius of 40nm, which is adjacent to S and has not yet been inspected, is arbitrarily selected, the material properties of X are inverted, and the index function value FOM of the current device is calculated new3 The method comprises the steps of carrying out a first treatment on the surface of the If (FOM) 0 －FOM new3 )/FOM 0 > 0, let FOM 0 ＝FOM new3 S=s &. X, t=t+1, otherwise, the material properties of all grids in X are restored to the pre-modification state, and all grids marked X are inspected. The above local search process is repeatedly performed, and after 107 times of local optimization, all grids adjacent to S have been inspected, and as shown in fig. 5 (c), the local search is stopped.
FOM at this time 0 = 0.7436 > 0.3, the optimization objective is not satisfied, so the global search and local search processes are re-performed. Finally, through 1931 iterations, FOM 0 = 0.2662 meets the requirements, and the optimized design is shown in fig. 6. The frequency response of the two output ports is shown in fig. 7. The transmittance of the output port 1 to 1550nm and 1950nm lasers is 0.0113 and 0.8522 respectively, the transmittance of the output port 2 to 1550nm and 1950nm lasers is 0.8611 and 0.0221 respectively, and the extinction ratio of the two ports is more than 15dB, so that the expected wavelength division multiplexing function is realized.
The optimization method often needs to be tried several times to find the condition (FOM) meeting the threshold value when global searching is performed 0 －FOM new1 )/FOM 0 Region B > 0.0001. In order to speed up the global search, this embodiment employs several heuristic strategies:
(1) If all the current materials of the grids in the area B are substrates, and the materials in the circular area with the circle center of the area B as the circle center and the radius of 400nm are substrates, the area B is far away from the waveguide area, the global search can directly skip the area B, and the FOM function value after material inversion does not need to be calculated and whether the threshold condition is met or not is indicated.
(2) Checking the electromagnetic field power at all grids in region B if the sum of the electromagnetic field powers P total ＜P max 30, wherein P max The maximum value of the electromagnetic field power in the current optimization space shows that the electromagnetic field passes through almost no area B, and the influence of changing the material property of the area B on the FOM function value is small, so that the global search can directly skip the area B and search other areas again.
(3) After inverting the material properties of all grids in the B, adopting the two-dimensional FDTD of Lumerical to rapidly estimate the index function value FOM of the device est If (FOM) 0 －FOM est )/FOM 0 < 0.5x0.0001, then B can be skipped directly and the other areas can be searched again. Otherwise, adopting three-dimensional FDTD of Lumerical to accurately calculate index function value FOM new1 If (FOM) 0 －FOM new1 )/FOM 0 > 0.0001, FOM is prepared 0 ＝FOM new1 And executing local search, otherwise, restoring the material properties of all grids in the B to a state before inversion, and observing other areas by using global search.
(4) According to the formulaThe gradient function of this particular embodiment can be obtained. Since the height direction of the optimization space is not divided, the grid r has only two dimensions in the optimization space, the iteration times are used as the time dimension, and all the values of the gradient function G form a three-dimensional matrix GM. The global search and the local search can update the corresponding elements in the GM matrix each time the index function FOM is calculated. For any one gridr, if the current iteration number t c G (r, t) c ) Unknown value, then one-dimensional interpolation is performed on G (r, t) along the time dimension t, according to t c G (r, t) is predicted by gradient values of the grid r under different previous iteration times c ) The method comprises the steps of carrying out a first treatment on the surface of the Then the iteration times t are fixed c Along two dimensional pairs of space G (r, t c ) The two-dimensional interpolation is carried out, the gradient value of each grid under the current iteration number can be predicted according to the interpolation result, then the grid position with the fastest gradient drop can be obtained according to the formula 1 and the material attribute of each grid r, the grid is taken as the center of a circle, the center area B with the radius of 80nm has high probability of meeting the threshold value condition (FOM) 0 －FOM new1 )/FOM 0 ＞0.0001。
In this embodiment, a function switch option is added to the MATLAB program, and the four heuristic strategies can be selectively turned on or off according to the optimized process and effect, or other heuristic strategies can be expanded and added. Fig. 8 shows the final optimization result after the heuristic strategy (2) is started, the transmittance of the laser with the wavelength of 1550nm at the output ports 1 and 2 is 0.025 and 0.789 respectively, the transmittance of the laser with the wavelength of 1950nm at the output ports 1 and 2 is 0.765 and 0.091 respectively, and the extinction ratio of the two ports is more than 9dB. Compared with fig. 6, the area through which the electromagnetic field energy passes is selected as much as possible for each optimization, so that the design structure is more compact, the micro-structure is fewer, and the micro-nano processing and manufacturing are more convenient.
Finally, in order to ensure that the final design structure is processable, the embodiment performs spatial filtering on the structure of fig. 6, the effect after filtering is as shown in fig. 9, compared with fig. 6, the micro structure is smoothed, after smoothing, the transmittance of the laser with the wavelength of 1550nm at the output ports 1 and 2 is respectively 0.027 and 0.733, the transmittance of the laser with the wavelength of 1950nm at the output ports 1 and 2 is respectively 0.702 and 0.009, and the extinction ratio of the two ports is more than 14dB.
Preferably, the embodiment of the present application further provides a specific implementation manner of an electronic device capable of implementing all the steps in the integrated photonic device design method based on the reverse design concept in the foregoing embodiment, where the electronic device specifically includes the following contents:
a processor (processor), a memory (memory), a communication interface (Communications Interface), and a bus;
the processor, the memory and the communication interface complete communication with each other through buses; the communication interface is used for realizing information transmission among relevant equipment such as server-side equipment, metering equipment and user-side equipment.
The processor is configured to invoke the computer program in the memory, and when the processor executes the computer program, the processor implements all the steps in the integrated photonic device design method based on the reverse design concept in the above embodiment.
The embodiment of the present application also provides a computer-readable storage medium capable of implementing all the steps in the integrated photonic device design method based on the reverse design concept in the above embodiment, and the computer-readable storage medium has a computer program stored thereon, which when executed by a processor, implements all the steps in the integrated photonic device design method based on the reverse design concept in the above embodiment.
The foregoing describes specific embodiments of the present disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
Although the application provides method operational steps as an example or a flowchart, more or fewer operational steps may be included based on conventional or non-inventive labor. The order of steps recited in the embodiments is merely one way of performing the order of steps and does not represent a unique order of execution. When implemented by an actual device or client product, the instructions may be executed sequentially or in parallel (e.g., in a parallel processor or multi-threaded processing environment) as shown in the embodiments or figures.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The application is not limited to the embodiments described above. The above description of specific embodiments is intended to describe and illustrate the technical aspects of the present application, and is intended to be illustrative only and not limiting. Numerous specific modifications can be made by those skilled in the art without departing from the spirit of the application and scope of the claims, which are within the scope of the application.
Claims (9)
1. An integrated photonic device design method based on a reverse design concept is characterized by comprising the following steps:
s1, setting the size of an optimized space and input and output ports, designating the material types of a substrate and a waveguide, and performing three-dimensional polishing on the optimized space to form a plurality of grids, wherein each grid is formed by the substrate or the waveguide material;
s2, constructing an initial structure of an optimization space, wherein all grids in the initial structure are single structures composed of substrate materials or waveguide materials or known integrated photon device structures consistent with target performances;
s3, constructing an index function FOM for evaluating the target performance of the integrated photonic device based on an electromagnetic field simulation method, wherein the smaller the FOM is, the better the target performance index is;
s4, calculating index function value FOM of the initial integrated photon structure by using index function FOM 0 Given two decision thresholds G th And L th Let iteration number t=0;
s5, executing global search, randomly selecting an area B in the optimization space, inverting the material properties of all grids in the area B, and calculating to obtain the index function value FOM of the integrated photonic device new1 The method comprises the steps of carrying out a first treatment on the surface of the If (FOM) 0 －FOM new1 )/FOM 0 ＞G th Executing step S6, otherwise, recovering the material properties of all grids in the region B to a state before the material properties are reversed, and re-executing step S5;
s6, reducing the area B into a smaller area S by taking the geometric center of the area B as a reference point; restoring the material properties of all grids in the region B to a state before the material properties are reversed, and then reversing the material properties of all grids in the region S, and calculating to obtain the index function value FOM of the integrated photonic device new2 Make FOM 0 ＝FOM new2 T=t+1, step S7 is performed;
s7, marking all grids which are not in the area S in the optimization space as an unchecked state;
s8, if all grids adjacent to the area S are checked, executing a step S10, otherwise executing a step S9;
s9, executing local search, arbitrarily selecting an area X which is adjacent to the area S and is not inspected in the optimized space, inverting the material attribute of the area X, and calculating to obtain the index function value FOM of the current device new3 The method comprises the steps of carrying out a first treatment on the surface of the If (FOM) 0 －FOM new3 )/FOM 0 ＞L th FOM is made to 0 ＝FOM new3 S=s & -X, t=t+1, execute step S8; otherwise, restoring the material properties of all grids in the region X to the state before modification, marking that all grids in the region X are inspected, and executing step S8;
s10, if FOM 0 If the value of (2) meets the optimization requirement, the optimization is stopped to obtain a final design structure, otherwise, the step S5 is executed.
2. The method for designing an integrated photonic device based on the idea of reverse design according to claim 1, wherein the optimization space is a space occupied by the integrated photonic device to be optimized.
3. The method of claim 1, wherein the material property of the grid is a material type filled in the corresponding grid position, and the material type includes a substrate or a waveguide material.
4. The method for designing an integrated photonic device based on the reverse design concept as claimed in claim 1, wherein when the global search is performed in step S5, several or all of the following heuristic strategies are adopted according to the actual optimization effect:
(1) If the materials of all the grids in the area B are substrates or waveguides and the materials of all the grids in the given area around the area B are substrates or waveguides, directly skipping the area B, and re-executing the step S5;
(2) Electromagnetic field energy at all meshes within the examination region B, if the sum of the electromagnetic field energies is smaller than a given threshold E th Directly skipping over the step B, and re-executing the step S5;
(3) After inverting the material properties of all grids in the region B, calculating the index function of the modified integrated photonic device based on two-dimensional FDTD to obtain an estimated value FOM est If (FOM) 0 －FOM est )/FOM 0 <αG th α is a given coefficient, directly skipping the region B, and re-executing step S5; otherwise, calculating index function value FOM by using three-dimensional FDTD new1 If (FOM) 0 －FOM new1 )/FOM 0 ＞G th FOM is made to 0 ＝FOM new1 Executing step S6, otherwise, recovering the material properties of all grids in the region B to a state before the material properties are reversed, and re-executing step S5;
(4) Defining a gradient function for all grids of the optimization space: let the iteration number at a certain moment be t, and the index function value of the integrated photon device be FOM 0 For any grid r in the optimization space, the material property of the grid r is inverted, the index function value of the device is changed into FOM (r), and then the gradient function G (r, t) is defined as:
for any grid r, if the current iteration number t c G (r, t) c ) Unknown value, then one-dimensional interpolation is performed on G (r, t) along the time dimension t, according to t c G (r, t) is predicted by gradient values of the grid r under different previous iteration times c ) The method comprises the steps of carrying out a first treatment on the surface of the Then the iteration times t are fixed c Three dimensional pairs G (r, t c ) Performing three-dimensional spatial interpolation, predicting gradient values of each grid under the current iteration number according to the spatial interpolation result, obtaining grid positions with the fastest gradient drop according to the formula (1) and material properties of each grid r, selecting a region B near the grid, and executing step S5 to ensure that the maximum probability meets a threshold condition (FOM 0 －FOM new1 )/FOM 0 ＞G th 。
5. The method of designing an integrated photonic device based on the idea of reverse design according to claim 1, wherein the design structure finally obtained in step S10 is spatially filtered according to the accuracy of the processing technique, and the fine structure which cannot be processed is smoothed by convolving the design structure with a spatial filter.
6. The integrated photonic device design method based on the reverse design concept as claimed in claim 1, wherein the material property inversion refers to changing the current material of the grid into a waveguide when the current material of the grid is a substrate; when the grid current material is a waveguide, the grid current material is changed to a substrate.
7. The integrated photonic device design method based on the reverse design concept as claimed in claim 1, wherein the electromagnetic field simulation method comprises a three-dimensional time domain finite difference method FDTD and a finite element analysis method.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the integrated photonics device design method based on the reverse design concept of any of claims 1 to 7 when the program is executed by the processor.
9. A computer-readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the integrated photonic device design method based on the reverse design concept as claimed in any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311170726.5A CN117195640A (en) | 2023-09-12 | 2023-09-12 | Integrated photon device design method based on reverse design idea |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311170726.5A CN117195640A (en) | 2023-09-12 | 2023-09-12 | Integrated photon device design method based on reverse design idea |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117195640A true CN117195640A (en) | 2023-12-08 |
Family
ID=88988344
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311170726.5A Pending CN117195640A (en) | 2023-09-12 | 2023-09-12 | Integrated photon device design method based on reverse design idea |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117195640A (en) |
-
2023
- 2023-09-12 CN CN202311170726.5A patent/CN117195640A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Minkov et al. | Inverse design of photonic crystals through automatic differentiation | |
| JP2018529142A (en) | Quantum processing apparatus and method | |
| US10871612B2 (en) | Interferometer and method of designing an interferometer | |
| WO2011146643A2 (en) | Computation efficiency by iterative spatial harmonics order truncation | |
| Boettcher et al. | Fixed-point properties of the Ising ferromagnet on the Hanoi networks | |
| JP2020112756A (en) | Prelearned model generation method, structure design method, computer program, and prelearned model | |
| CN117195640A (en) | Integrated photon device design method based on reverse design idea | |
| CN116341634B (en) | Training method and device for neural structure search model and electronic equipment | |
| CN115616762A (en) | Super-surface design method and device and electronic equipment | |
| Majety et al. | Triangular cross-section beam splitters in silicon carbide for quantum information processing | |
| CN113625379A (en) | Design method, manufacturing method and design device of Fresnel zone plate | |
| Li et al. | On the accuracy of the addition theorem for a scalar Green's function used in the FMM | |
| CN116680766B (en) | Method and device for determining achromatic superlens, electronic equipment and storage medium | |
| Sakamoto et al. | Deep learning design of topological bending waveguide | |
| Parandin et al. | Design and Optimization of a Photonic Crystal-Based All-Optical XOR Gate Using Machine Learning | |
| Angelini et al. | On morphs of 1-plane graphs | |
| Kwiecien et al. | Rcwa/arcwa-an efficient and diligent workhorse for nanophotonic/nanoplasmonic simulations-can it work even better? | |
| Tagawa et al. | Many-hard-objective optimization using differential evolution based on two-stage constraint-handling | |
| Shamin et al. | Sidewall roughness model for optical losses calculation in photonic integrated circuits | |
| Algasi | Uninformed and Informed Search Techniques in Artificial Intelligence | |
| Dourado-Sisnando et al. | Artificial immune systems optimization of the absolute bandgap of photonic crystals | |
| Khoram et al. | Stochastic optimization of nonlinear nanophotonic media for artificial neural inference | |
| Parandin et al. | Enhancing the Performance of Photonic Crystal AND Gates with Machine Learning Optimization | |
| Hamdani et al. | Modelling Fabrication Variability in Silicon Photonic Devices. | |
| Xu et al. | An FPGA-Enabled Framework for Rapid Automated Design of Photonic Integrated Circuits |
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 |