On-demand design of spectrally sensitive multiband absorbers using an artificial neural network
1 Department of Mechanical Engineering, Pohang University of Science and Technology (POSTECH), Pohang 37673, Republic of Korea
2 Department of Chemical Engineering, Pohang University of Science and Technology (POSTECH), Pohang 37673, Republic of Korea
3 National Institute of Nanomaterials Technology (NINT), Pohang 37673, Republic of Korea
Figures & Tables
Fig. 1. Schematic of the designing grating structures for multiband absorbers. (a) A schematic of ANN for designing grating structures. The network is composed of two artificial neural networks of design network and pre-trained spectrum network. The design network both takes the input reflection spectra and resonant wavelengths, and the pre-trained spectrum network takes design parameters to evaluate the optical reflection spectra of the designed structures. (b) A schematic and (c) an example of optical property of a perfect multiband absorber under investigation. Yellow markers indicate resonant wavelengths.
下载图片 查看原文
Fig. 2. (a) Scanning electron microscope image of a designed grating structure with a scale bar of 1 μm. (b) Target reflection spectrum (black solid line) and designed optical properties obtained from the FDTD simulation (red dotted line) and experiment (yellow dotted line). Grating parameters with [P, Gr, h1, h2, hsub] = [245 nm, 120 nm, 42 nm, 113 nm, 195 nm] are designed by the network. (c) Examples of test results are shown. Black solid lines and red dotted lines are the input and target reflection spectra, respectively, and yellow markers are indexed resonant wavelengths.
下载图片 查看原文
Fig. 3. (a) Target (black solid line) and designed reflection spectrum. Magnetic field distribution (color maps) and electric displacement (arrow surfaces) at the resonant wavelengths of (b) 450 nm, (c) 525 nm, and (d) 600 nm.
下载图片 查看原文
Fig. 4. Design of multiband absorbers with (a) single, (b) double, and (c) triple resonances. The first column shows the target input spectra, and the second column shows the designed responses. The red lines indicate target resonant wavelengths. The third column shows the histogram of the MSE for a total of 51 input spectra. The insets show the average MSE of the test input.
下载图片 查看原文
Fig. 5. Comparison between two networks fed with and without spectral resonant wavelengths. The left is the target input spectra; the middle and the right are the predicted response of the networks without and with spectral information, respectively. The red lines are target resonant wavelengths.
下载图片 查看原文
Fig. 6. Analysis on output parameters for gradually changing target resonant wavelengths. (a) Target spectra with gradually changing resonant target wavelengths and (b) corresponding designed responses. For given varying input spectra, the designed parameters of (c) grating height and substrate height and (d) period and grating width.
下载图片 查看原文
Fig. 7. Network pruning results. Visualization of the trained weights in (a) the original network and (b) the pruned network. For each layer (Ln,n=1,2,…,7), the number of neurons is indicated. MSE histogram of the test data for (c) the original network and (d) the pruned network.
下载图片 查看原文
Fig. 8. Design of multiband absorbers with (a) single, (b) double, and (c) triple resonances using the reduced network. The first column shows target input spectra, and the second column shows the designed response. The red lines indicate target resonant wavelengths. The third column shows the histogram of the MSE for a total of 51 input spectra.
下载图片 查看原文
Table1. Hyperparameters Used in the Training of Two Networks
| | Design Network | Spectrum Network |
|---|
| Number of neurons | [202, 400, 1000, 2000, 1000, 500, 200, 5] | [5, 200, 500, 1000, 500, 200, 101] | | Optimizer | Adam, weight decay | Adam, weight decay | | Learning rate | From to | | | Nonlinear activation functions | Leaky ReLU, | Leaky ReLU, |
|
查看原文
Table2. Number of Neurons in Each Layer
| Layer | 1 | 2 | 3 | 4 | 5 | 6 | 7 | Total |
|---|
| Original | 202 | 400 | 1000 | 2000 | 1000 | 500 | 200 | 5055905 | | Reduced | 202 | 400 | 1000 | 499 | 625 | 474 | 199 | 1651642 |
|
查看原文
Sunae So, Younghwan Yang, Taejun Lee, Junsuk Rho. On-demand design of spectrally sensitive multiband absorbers using an artificial neural network[J]. Photonics Research, 2021, 9(4): 0400B153.
PDF全文
