OptoGPT: A foundation model for inverse design in optical multilayer thin film structures

Optical multilayer thin film structures have been widely used in numerous photonic applications.However,existing inverse design methods have many drawbacks because they either fail to quickly adapt to different design targets,or are difficult to suit for different types of structures,e.g.,designing for different materials at each layer.These methods also cannot accommodate versatile design situations under different angles and polarizations.In addition,how to benefit practical fabrications and manufacturing has not been extensively considered yet.In this work,we introduce OptoGPT(Opto Generative Pretrained Transformer),a decoder-only transformer,to solve all these drawbacks and issues simultaneously.

Inverse design of mechanical metamaterial achieving a prescribedconstitutive curve

Besides exhibiting excellent capabilities such as energy absorption,phase-transforming metamaterials offer a vast design space for achieving nonlinear constitutive relations.This is facilitated by switching between different patterns under deformation.However,the related inverse design problem is quite challenging,due to the lack of appropriate mathematical formulation and the convergence issue in the post-buckling analysis of intermediate designs.In this work,periodic unit cells are explicitly described by the moving morphable voids method and effectively analyzed by eliminating the degrees of freedom in void regions.Furthermore,by exploring the Pareto frontiers between error and cost,an inverse design formulation is proposed for unit cells.This formulation aims to achieve a prescribed constitutive curve and is validated through numerical examples and experimental results.The design approach presented here can be extended to the inverse design of other types of mechanical metamaterials with prescribed nonlinear effective properties.

A graph neural network approach to the inverse design for thermal transparency with periodic interparticle system

Recent years have witnessed significant advances in utilizing machine learning-based techniques for thermal metamaterial-based structures and devices to attain favorable thermal transport behaviors.Among the various thermal transport behaviors,achieving thermal transparency stands out as particularly desirable and intriguing.Our earlier work demonstrated the use of a thermal metamaterial-based periodic interparticle system as the underlying structure for manipulating thermal transport behavior and achieving thermal transparency.In this paper,we introduce an approach based on graph neural network to address the complex inverse design problem of determining the design parameters for a thermal metamaterial-based periodic interparticle system with the desired thermal transport behavior.Our work demonstrates that combining graph neural network modeling and inference is an effective approach for solving inverse design problems associated with attaining desirable thermal transport behaviors using thermal metamaterials.

Inverse design of nonlinear phononic crystal configurations based on multi-label classification learning neural networks

Phononic crystals,as artificial composite materials,have sparked significant interest due to their novel characteristics that emerge upon the introduction of nonlinearity.Among these properties,second-harmonic features exhibit potential applications in acoustic frequency conversion,non-reciprocal wave propagation,and non-destructive testing.Precisely manipulating the harmonic band structure presents a major challenge in the design of nonlinear phononic crystals.Traditional design approaches based on parameter adjustments to meet specific application requirements are inefficient and often yield suboptimal performance.Therefore,this paper develops a design methodology using Softmax logistic regression and multi-label classification learning to inversely design the material distribution of nonlinear phononic crystals by exploiting information from harmonic transmission spectra.The results demonstrate that the neural network-based inverse design method can effectively tailor nonlinear phononic crystals with desired functionalities.This work establishes a mapping relationship between the band structure and the material distribution within phononic crystals,providing valuable insights into the inverse design of metamaterials.

High performance integrated photonic circuit based on inverse design method

The basic indexes of all-optical integrated photonic circuits include high-density integration,ultrafast response and ultralow energy consumption.Traditional methods mainly adopt conventional micro/nano-structures.The overall size of the circuit is large,usually reaches hundreds of microns.Besides,it is difficult to balance the ultrafast response and ultra-low energy consumption problem,and the crosstalk between two traditional devices is difficult to overcome.Here,we propose and experimentally demonstrate an approach based on inverse design method to realize a high-density,ultrafast and ultra-low energy consumption integrated photonic circuit with two all-optical switches controlling the input states of an all-optical XOR logic gate.The feature size of the whole circuit is only 2.5μm×7μm,and that of a single device is 2μm×2μm.The distance between two adjacent devices is as small as 1.5μm,within wavelength magnitude scale.Theoretical response time of the circuit is 150 fs,and the threshold energy is within 10 fJ/bit.We have also considered the crosstalk problem.The circuit also realizes a function of identifying two-digit logic signal results.Our work provides a new idea for the design of ultrafast,ultra-low energy consumption all-optical devices and the implementation of high-density photonic integrated circuits.

Artificial neural network based inverse design method for circular sliding slopes

Current design method for circular sliding slopes is not so reasonable that it often results in slope (sliding.) As a result, artificial neural network (ANN) is used to establish an artificial neural network based inverse design method for circular sliding slopes. A sample set containing 21 successful circular sliding slopes excavated in the past is used to train the network. A test sample of 3 successful circular sliding slopes excavated in the past is used to test the trained network. The test results show that the ANN based inverse design method is valid and can be applied to the design of circular sliding slopes.

Benchmarking deep learning-based models on nanophotonic inverse design problems

Photonic inverse design concerns the problem of finding photonic structures with target optical properties.However,traditional methods based on optimization algorithms are time-consuming and computationally expensive.Recently,deep learning-based approaches have been developed to tackle the problem of inverse design efficiently.Although most of these neural network models have demonstrated high accuracy in different inverse design problems,no previous study has examined the potential effects under given constraints in nanomanufacturing.Additionally,the relative strength of different deep learning-based inverse design approaches has not been fully investigated.Here,we benchmark three commonly used deep learning models in inverse design:Tandem networks,Variational Auto-Encoders,and Generative Adversarial Networks.We provide detailed comparisons in terms of their accuracy,diversity,and robustness.We find that tandem networks and Variational Auto-Encoders give the best accuracy,while Generative Adversarial Networks lead to the most diverse predictions.Our findings could serve as a guideline for researchers to select the model that can best suit their design criteria and fabrication considerations.In addition,our code and data are publicly available,which could be used for future inverse design model development and benchmarking.

Review on Applications of 3D Inverse Design Method for Pump

The 3D inverse design method, which methodology is far superior to the conventional design method that based on geometrical description, is gradually applied in pump blade design. However, no complete description about the method is outlined. Also, there are no general rules available to set the two important input parameters, blade loading distribution and stacking condition. In this sense, the basic theory and the mechanism why the design method can suppress the formation of secondary flow are summarized. And also, several typical pump design cases with different specific speeds ranging from centrifugal pump to axial pump are surveyed. The results indicates that, for centrifugal pump and mixed pump or turbine, the ratio of blade loading on the hub to that on the shroud is more than unit in the fore part of the blade, whereas in the aft part, the ratio is decreased to satisfy the same wrap angle for hub and shroud. And the choice of blade loading type depends on the balancing of efficiency and cavitation. If the cavitation is more weighted, the better choice is aft-loaded, otherwise, the fore-loaded or mid-loaded is preferable to improve the efficiency. The stacking condition, which is an auxiliary to suppress the secondary flow, can have great effect on the jet-wake outflow and the operation range for pump. Ultimately, how to link the design method to modem optimization techniques is illustrated. With the know-how design methodology and the know-how systematic optimization approach, the application of optimization design is promising for engineering. This paper summarizes the 3D inverse design method systematically.

Inverse design for material anisotropy and its application for a compact X-cut TFLN on-chip wavelength demultiplexer

Inverse design focuses on identifying photonic structures to optimize the performance of photonic devices.Conventional scalar-based inverse design approaches are insufficient to design photonic devices of anisotropic materials such as lithium niobate(LN).To the best of our knowledge,this work proposes for the first time the inverse design method for anisotropic materials to optimize the structure of anisotropic-material based photonics devices.Specifically,the orientation dependent properties of anisotropic materials are included in the adjoint method,which provides a more precise prediction of light propagation within such materials.The proposed method is used to design ultra-compact wavelength division demultiplexers in the X-cut thin-film lithium niobate(TFLN)platform.By benchmarking the device performances of our method with those of classical scalar-based inverse design,we demonstrate that this method properly addresses the critical issue of material anisotropy in the X-cut TFLN platform.This proposed method fills the gap of inverse design of anisotropic materials based photonic devices,which finds prominent applications in TFLN platforms and other anisotropicmaterial based photonic integration platforms.

Realization of advanced passive silicon photonic devices with subwavelength grating structures developed by efficient inverse design

Compact passive silicon photonic devices with high performance are always desired for future largescale photonic integration.Inverse design provides a promising approach to realize new-generation photonic devices,while it is still very challenging to realize complex photonic devices for most inverse designs reported previously due to the limits of computational resources.Here,we present the realization of several representative advanced passive silicon photonic devices with complex optimization,including a sixchannel mode(de)multiplexer,a broadband 90 deg hybrid,and a flat-top wavelength demultiplexer.These devices are designed inversely by optimizing a subwavelength grating(SWG)region and the multimode excitation and the multimode interference are manipulated.Particularly,such SWG structures are more fabrication-friendly than those random nanostructures introduced in previous inverse designs.The realized photonic devices have decent performances in a broad bandwidth with a low excess loss of<1 dB,which is much lower than that of previous inverse-designed devices.The present inverse design strategy shows great effectiveness for designing advanced photonic devices with complex requirements(which is beyond the capability of previous inverse designs)by using affordable computational resources.

Inverse design-based metamaterial transparent device and its multilayer realization

We propose an inverse method to determine the material parameters of a transparent device without any knowledge of the corresponding transformation function. The required parameters are independently obtained and expressed as functions of the introduced generator. Moreover, to remove the inhomogeneity and anisotropy of material parameters, a layered transparent device composed of only homogeneous and isotropic materials is presented based on the effective medium theory. The feasibility of using the layered device in antenna protection is also investigated. Full-wave simulation is carried out for verification. This work paves a new way toward designing metamaterial devices without specifying the underlying coordinate transformation, and has great guiding significance for the practical fabrication of transparent devices.

Inverse design and realization of an optical cavity-based displacement transducer with arbitrary responses

Optical cavity has long been critical for a variety of applications ranging from precise measurement to spectral analysis.A number of theories and methods have been successful in describing the optical response of a stratified optical cavity,while the inverse problem,especially the inverse design of a displacement sensitive cavity,remains a significant challenge due to the cost of computation and comprehensive performance requirements.This paper reports a novel inverse design methodology combining the characteristic matrix method,mixed-discrete variables optimization algorithm,and Monte Carlo method-based tolerance analysis.The material characteristics are indexed to enable the mixed-discrete variables optimization,which yields considerable speed and efficiency improvements.This method allows arbitrary response adjustment with technical feasibility and gives a glimpse into the analytical characterization of the optical response.Two entirely different light-displacement responses,including an asymmetric sawtooth-like response and a highly symmetric response,are dug out and experimentally achieved,which fully confirms the validity of the method.The compact Fabry-Perot cavities have a good balance between performance and feasibility,making them promising candidates for displacement transducers.More importantly,the proposed inverse design paves the way for a universal design of optical cavities,or even nanophotonic devices.

Direct laser-written aperiodic photonic volume elements for complex light shaping with high efficiency:inverse design and fabrication

Light plays a central role in many applications.The key to unlocking its versatility lies in shaping it into the most appropriate form for the task at hand.Specifically tailored refractive index modifications,directly manufactured inside glass using a short pulsed laser,enable an almost arbitrary control of the light flow.However,the stringent requirements for quantitative knowledge of these modifications,as well as for fabrication precision,have so far prevented the fabrication of light-efficient aperiodic photonic volume elements(APVEs).Here,we present a powerful approach to the design and manufacturing of light-efficient APVEs.We optimize application-specific three-dimensional arrangements of hundreds of thousands of microscopic voxels and manufacture them using femtosecond direct laser writing inside millimeter-sized glass volumes.We experimentally achieve unprecedented diffraction efficiencies up to 80%,which is enabled by precise voxel characterization and adaptive optics during fabrication.We demonstrate APVEs with various functionalities,including a spatial mode converter and combined intensity shaping and wavelength multiplexing.Our elements can be freely designed and are efficient,compact,and robust.Our approach is not limited to borosilicate glass but is potentially extendable to other substrates,including birefringent and nonlinear materials,giving a preview of even broader functionalities,including polarization modulation and dynamic elements.

Inverse design of highly efficient and broadband mode splitter on SOI platform

Mode splitters that directly separate modes without changing their orders are highly promising to improve the flexibility of the mode-division multiplexing systems.In this paper,we design a high-performance mode splitter on the silicon-oninsulator platform with a compact footprint of 14μm×2.5μm using an inverse design method based on shape optimization.The fabrication of this mode splitter requires only a single lithography step and exhibits good fabrication tolerances.The experimental results show that the proposed device exhibits state-of-the-art insertion loss(<0.9 dB)and cross talk(<-16 dB)over a broad bandwidth(1500-1600 nm).Furthermore,the shape optimization method used is implemented to design a dual-mode(de)multiplexer,and the experimental results fulfill the design objective,demonstrating the excellent generality of the design method in this paper.

A general alternate loading technique and its applications in the inverse designs of centrifugal and mixed-flow pump impellers

For the inverse designs of centrifugal and mixed-flow pump impellers,clarifying the generation process of secondary flows and putting forward corresponding suppression measures is an important approach to improve the impeller performance.In this paper,to provide a better qualitative insight into the generation mechanism of secondary flows in the impeller,a simple kinematic equation is derived based on the ideal assumptions,which indicates that the potential rothalpy gradient(PRG)is the most important dynamic source that actively induces secondary vortical flows.Induced by the natural adverse PRG on the S1 and S2 stream surfaces,two typical secondary flows,H-S and P-S secondary flows,are clearly presented.To specially suppress these typical secondary flows,a general alternate loading technique(GALT)is proposed,aiming to adjust the real blade loadingδp to control the PRG features.At the blade fore part,theδp on the hub streamline should be slowly increased to avoid breakneck growth of the potential rothalpy to reduce adverse streamwise PRG on the S2 streamsurface.At the blade middle part,theδp should be moderately decreased to reduce adverse streamwise PRG on the S1 streamsurface.At the blade aft part,the difference in theδp between the shroud and hub streamlines should be decreased faster to control the exit uniformity.By applying the GALT to the impeller designs of three typical pump types in hydraulic engineering,the organizational effect of the PRG on fundamental flow structures is proven.The GALT can effectively control the PRG distributions and suppress the secondary flows,thereby widening the pump’s high-efficiency zone,improving flow uniformity and suppressing pressure fluctuations.Compared with the current Z-G method and the ALT,the GALT can meet the requirements of"de-experience"better,thereby enabling the designers to obtain good products explicitly and quickly.

Inverse design of an integrated-nanophotonics optical neural network

Artificial neural networks have dramatically improved the performance of many machine-learning applications such as image recognition and natural language processing. However, the electronic hardware implementations of the above-mentioned tasks are facing performance ceiling because Moore’s Law is slowing down. In this article, we propose an optical neural network architecture based on optical scattering units to implement deep learning tasks with fast speed, low power consumption and small footprint.The optical scattering units allow light to scatter back and forward within a small region and can be optimized through an inverse design method. The optical scattering units can implement high-precision stochastic matrix multiplication with mean squared error < 10-4 and a mere 4*4 um2 footprint.Furthermore, an optical neural network framework based on optical scattering units is constructed by introducing "Kernel Matrix", which can achieve 97.1% accuracy on the classic image classification dataset MNIST.

Inverse design of digital nanophotonic devices using the adjoint method

A high-efficiency inverse design of"digital"subwavelength nanophotonic devices using the adjoint method is proposed.We design a single-mode 3dB power divider and a dual-mode demultiplexer to demonstrate the efficiency of the proposed inverse design approach,called the digitized adjoint method,for single-and dual-object optimization,respectively.The optimization comprises three stages:1)continuous variation for an"analog"pattern;2)forced permittivity biasing for a"quasi-digital"pattern;and 3)a multilevel digital pattern.Compared with the conventional brute-force method,the proposed method can improve design efficiency by about five times,and the performance optimization can reach approximately the same level.The method takes advantages of adjoint sensitivity analysis and digital subwavelength structure and creates a new way for the efficient and high-performance design of compact digital subwavelength nanophotonic devices,which could overcome the efficiency bottleneck of the brute-force method,which is restricted by the number of pixels of a digital pattern,and improve the device performance by extending a conventional binary pattern to a multilevel one.

An inverse design method for supercritical airfoil based on conditional generative models

Inverse design has long been an efficient and powerful design tool in the aircraft industry.In this paper,a novel inverse design method for supercritical airfoils is proposed based on generative models in deep learning.A Conditional Variational Auto Encoder(CVAE)and an integrated generative network CVAE-GAN that combines the CVAE with the Wasserstein Generative Adversarial Networks(WGAN),are conducted as generative models.They are used to generate target wall Mach distributions for the inverse design that matches specified features,such as locations of suction peak,shock and aft loading.Qualitative and quantitative results show that both adopted generative models can generate diverse and realistic wall Mach number distributions satisfying the given features.The CVAE-GAN model outperforms the CVAE model and achieves better reconstruction accuracies for all the samples in the dataset.Furthermore,a deep neural network for nonlinear mapping is adopted to obtain the airfoil shape corresponding to the target wall Mach number distribution.The performances of the designed deep neural network are fully demonstrated and a smoothness measurement is proposed to quantify small oscillations in the airfoil surface,proving the authenticity and accuracy of the generated airfoil shapes.

Inverse design of mission success space for combat aircraft contribution evaluation

This paper presents a novel design method of the Mission Success Space(MSS) for the evaluation on aircraft contribution effectiveness. MSS concept was proposed for giving success criterion of a mission and judging the success by conventional mission effectiveness with regards to the aircraft capabilities. This space is created by the Mission Success Function(MSF) and the original Effectiveness Index Space(EIS) where empirical equations are usually assumed to be MSFs. Based on this MSS concept, this paper firstly defines the MSS-based evaluation, then further summarizes the evaluation process of the Contribution to System-of-Systems(CSS). More importantly, based on the thought of Inverse Design(ID), a new design method of MSF is presented comprehensively analyzing aircraft's CSS in a combat mission without using any empirical MSF. The definition of MSS based ID is given and the design procedure is sequentially introduced. Two different confrontation cases are depicted with many details as the simulation validation: Air-to-ground and Penetration. There are two design variables considered for designing MSS in the latter case while only one for the former. However, in both cases, the best design is given by evaluating the Gaussian fitting performance of CSS.

Entropy based inverse design of aircraft mission success space in system-of-systems confrontation

This paper presents a practical and efficient design method for aircraft Mission Success Space(MSS)based on the entropy measurement(EM).First,fundamentals regarding MSS,Inverse Design(ID)and entropy are discussed.Then,two EM schemes of entropy-based ID and the whole MSS ID procedure are given to demonstrate alternative ways of entropy quantification and MSS design.After that,Genetic Algorithm(GA)is utilized as a search algorithm to find the optimal MSS design with the minimum objective,entropy,in each EM scheme.A simulation case of aircraft penetration mission is given for the method validation where the best aircraft MSS design is obtained by GA according to the minimum entropy.Results from two schemes are compared at the end.