Neural networks(NNs),especially electronic-based NNs,have been rapidly developed in the past few decades.However,the electronic-based NNs rely more on highly advanced and heavy power-consuming hardware,facing its bott...Neural networks(NNs),especially electronic-based NNs,have been rapidly developed in the past few decades.However,the electronic-based NNs rely more on highly advanced and heavy power-consuming hardware,facing its bottleneck due to the slowdown of Moore's law.Optical neural networks(ONNs),in which NNs are realized via optical components with information carried by photons at the speed of light,are drawing more attention nowadays.Despite the advantages of higher processing speed and lower system power consumption,one major challenge is to realize reliable and reusable algorithms in physical approaches,particularly nonlinear functions,for higher accuracy.In this paper,a versatile parametric-process-based ONN is demonstrated with its adaptable nonlinear computation realized using the highly nonlinear fiber(HNLF).With the specially designed modelocked laser(MLL)and dispersive Fourier transform(DFT)algorithm,the overall computation frame rate can reach up to 40 MHz.Compared to ONNs using only linear computations,this system is able to improve the classification accuracies from 81.8%to 88.8%for the MNIST-digit dataset,and from 80.3%to 97.6%for the Vowel spoken audio dataset,without any hardware modifications.展开更多
Nonlinear physical systems hold great promise for energy-efficient and lowhardware-cost information processing.However,their computational capabilities remain constrained by the complexity and tunability of system non...Nonlinear physical systems hold great promise for energy-efficient and lowhardware-cost information processing.However,their computational capabilities remain constrained by the complexity and tunability of system nonlinearity.Here we report a dual-ferroelectric gate-tunable memristor with a dipole coupling effect,achieving enlarged hysteresis,rich temporal dynamics,and nonvolatile heterosynaptic plasticity.By harnessing the dynamic nonlinearity of the dual-ferroelectric memristor,multimodal reservoir computing with an in-material fusion strategy has been achieved,which is demonstrated with a multimodal object recognition task.By exploring the static nonlinearity of the dual-ferroelectric memristor,nonlinear in-memory computing is realized with gate-tunable nonlinear functions,which successfully accelerates the Euclidean distance computation in the K-means clustering task.This work achieves strong coupling between the intrinsic physical dynamics and computational functionalities,offering new opportunities for more efficient hardware-accelerated systems.展开更多
Nonlinear computation is essential for a wide range of information processing tasks,yet implementing nonlinear functions using optical systems remains a challenge due to the weak and power-intensive nature of optical ...Nonlinear computation is essential for a wide range of information processing tasks,yet implementing nonlinear functions using optical systems remains a challenge due to the weak and power-intensive nature of optical nonlinearities.Overcoming this limitation without relying on nonlinear optical materials could unlock unprecedented opportunities for ultrafast and parallel optical computing systems.Here,we demonstrate that large-scale nonlinear computation can be performed using linear optics through optimized diffractive processors composed of passive phase-only surfaces.In this framework,the input variables of nonlinear functions are encoded into the phase of an optical wavefront—e.g.,via a spatial light modulator(SLM)—and transformed by an optimized diffractive structure with spatially varying point-spread functions to yield output intensities that approximate a large set of unique nonlinear functions–all in parallel.We provide proof establishing that this architecture serves as a universal function approximator for an arbitrary set of bandlimited nonlinear functions,also covering wavelength-multiplexed nonlinear functions as well as multi-variate and complex-valued functions that are all-optically cascadable.Our analysis also indicates the successful approximation of typical nonlinear activation functions commonly used in neural networks,including the sigmoid,tanh,ReLU(rectified linear unit),and softplus.We numerically demonstrate the parallel computation of one million distinct nonlinear functions,accurately executed at wavelength-scale spatial density at the output of a diffractive optical processor.Furthermore,we experimentally validated this framework using in situ optical learning and approximated 35 unique nonlinear functions in a single shot using a compact setup consisting of an SLM and an image sensor.These results establish diffractive optical processors as a scalable platform for massively parallel universal nonlinear function approximation,paving the way for new capabilities in analog optical computing based on linear materials.展开更多
The stochastic resonance （SR） in a time-delayed mono-stable system driven by multiplicative white noise, additive white noise, additive dichotomous noise as well as a periodic square-wave signal is considered from t...The stochastic resonance （SR） in a time-delayed mono-stable system driven by multiplicative white noise, additive white noise, additive dichotomous noise as well as a periodic square-wave signal is considered from the view of the signal-to-noise ratio （SNR）. It is found that the SNR increases monotonically with the increase of the delay time. The SNR exhibits the SR behavior when it is plotted as a function of intensities of the noises, displaying the asymmetry of the dichotomous noise. The SNR varies non-monotonically with the increase of the system parameter and the amplitude of the input square-wave signal.展开更多
The phenomenon of entropic stochastic resonance （ESR） in a two-dimensional confined system driven by a transverse periodic force is investigated when the colored fluctuation is included in the system. Applying the m...The phenomenon of entropic stochastic resonance （ESR） in a two-dimensional confined system driven by a transverse periodic force is investigated when the colored fluctuation is included in the system. Applying the method of unified colored noise approximation, the approximate Fokker-Planck equation can be derived in the absence of the periodic force. Through the escaping rate of the Brownian particle from one well to the other, the power spectral amplification can be obtained. It is found that increasing the values of the noise correlation time and the signal frequency can suppress the ESR of the system.展开更多
We study the dynamics of tumor cell growth with time-delayed feedback driven by multiplicative noise in an asymmetrical bistable potential well. For a small delay time, the analytical solutions of the probability dist...We study the dynamics of tumor cell growth with time-delayed feedback driven by multiplicative noise in an asymmetrical bistable potential well. For a small delay time, the analytical solutions of the probability distribution and the first passage time show that, with the increasing delay time, the peak of the probability distribution in a lower population state would increase, but in a higher population state it decreases. It is shown that the multiplicative noise and the time delay play opposite roles in the tumor cell growth.展开更多
We report the modulational instability （MI） analysis for the modulation equations governing the propagation of coherent polarized light through a nematic liquid crystal （NLC） slab, in the limit of low light intens...We report the modulational instability （MI） analysis for the modulation equations governing the propagation of coherent polarized light through a nematic liquid crystal （NLC） slab, in the limit of low light intensity and local material response. The linear stability analysis of the nonlinear plane wave solutions is performed by considering both the wave vectors （k,l） of the basic states and wave vectors （K,L） of the perturbations as free parameters. We compute the MI gain, and the MI gain peak is reduced and the stable bandwidth is widened with the increase of the strength of the applied electric field. Further, we invoke the extended homogeneous balance method and Exp-function method aided with symbolic computation and obtain a series of periodic solitonic humps of nematicon profiles admitting the propagation of laser light in the NLC medium.展开更多
We study the features of a single q-breather （SQB） in a Fermi-Pasta-Ulam lattice by the numerical method, and obtain that the stability of SQB correlates to coupling constant K and nonlinear parameter β. No matter ...We study the features of a single q-breather （SQB） in a Fermi-Pasta-Ulam lattice by the numerical method, and obtain that the stability of SQB correlates to coupling constant K and nonlinear parameter β. No matter whether K or β increases, the periodic SQB can be transformed into a quasiperiodic SQB or a chaotic SQB. We also obtain the conditions of excitation of periodic, quasiperiodic and chaotic SQBs.展开更多
A new simple model of self-driven particles in scale-free networks is introduced to understand the emergence of condensation in the natural world. In the model, at each time step, particles are driven to choose their ...A new simple model of self-driven particles in scale-free networks is introduced to understand the emergence of condensation in the natural world. In the model, at each time step, particles are driven to choose their next habitats according to the particle numbers at both the present habitats and neighbors. It is found that the hub effect results in the condensation. The present numerical results as well as the theoretical analysis of condensation transition show the criterion of condensation. Both simulations and theoretical analysis display that there are three phases for different hopping probabilities δ： non-condensation （δ〈δc）, partial condensation （δc≤ δ〈1）, and complete condensation （δ≥δc＇= 1）, and the mean occupation particle number at the stationary state is also obtained. Moreover, the noise effect on self-driven particles is studied, and it makes particle numbers at nodes tend towards being identical.展开更多
We demonstrate that the projective synchronization can be observed in coupled fractional-order chaotic systems. A new systematic and powerful coupling scheme is developed to investigate the projective synchronization ...We demonstrate that the projective synchronization can be observed in coupled fractional-order chaotic systems. A new systematic and powerful coupling scheme is developed to investigate the projective synchronization via the open-plus-closed-loop control, which allows us to arbitrarily manipulate the scaling factor of projective synchronization. The proposed scheme is proved analytically on the basis of the stability theorem of the fractional differential equations. Numerical simulations on the fraction-order chaotic Chen system are presented to justify the theoretical analysis.展开更多
We investigate the game theory in a structured population with the assumption that the evolution of network structure is far faster than that of strategy update. We find that the degree distribution for the finM netwo...We investigate the game theory in a structured population with the assumption that the evolution of network structure is far faster than that of strategy update. We find that the degree distribution for the finM network consists of two distinct parts： the low degree part which is contributed to by defectors and a broadband in the regime with high degree which is formed by cooperators. The structure of the final network and the final strategy pattern have also been numerically proved to be independent of the game parameters.展开更多
In accordance with a new compensation principle of discrete computations,the traditional meteo- rological global (pseudo-) spectral schemes of barotropic primitive equation (s) are transformed into perfect energy cons...In accordance with a new compensation principle of discrete computations,the traditional meteo- rological global (pseudo-) spectral schemes of barotropic primitive equation (s) are transformed into perfect energy conservative fidelity schemes,thus resolving the problems of both nonlinear computa- tional instability and incomplete energy conservation,and raising the computational efficiency of the traditional schemes. As the numerical tests of the new schemes demonstrate,in solving the problem of energy conser- vation in operational computations,the new schemes can eliminate the (nonlinear) computational in- stability and,to some extent even the (nonlinear) computational diverging as found in the traditional schemes,Further contrasts between new and traditional schemes also indicate that,in discrete opera- tional computations,the new scheme in the case of nondivergence is capable of prolonging the valid in- tegral time of the corresponding traditional scheme,and eliminating certain kind of systematical com- putational“climate drift”,meanwhile increasing its computational accuracy and reducing its amount of computation.The working principle of this paper is also applicable to the problem concerning baroclin- ic primitive equations.展开更多
基金Research Grants Council of the Hong Kong Special Administrative Region of China(HKU 17212824,HKU 17210522,HKU C7074-21G,HKU R7003-21,HKU 17205321)Innovation and Technology Commission—Hong Kong(MHP/073/20,MHP/057/21,Health@Inno HK programc)。
文摘Neural networks(NNs),especially electronic-based NNs,have been rapidly developed in the past few decades.However,the electronic-based NNs rely more on highly advanced and heavy power-consuming hardware,facing its bottleneck due to the slowdown of Moore's law.Optical neural networks(ONNs),in which NNs are realized via optical components with information carried by photons at the speed of light,are drawing more attention nowadays.Despite the advantages of higher processing speed and lower system power consumption,one major challenge is to realize reliable and reusable algorithms in physical approaches,particularly nonlinear functions,for higher accuracy.In this paper,a versatile parametric-process-based ONN is demonstrated with its adaptable nonlinear computation realized using the highly nonlinear fiber(HNLF).With the specially designed modelocked laser(MLL)and dispersive Fourier transform(DFT)algorithm,the overall computation frame rate can reach up to 40 MHz.Compared to ONNs using only linear computations,this system is able to improve the classification accuracies from 81.8%to 88.8%for the MNIST-digit dataset,and from 80.3%to 97.6%for the Vowel spoken audio dataset,without any hardware modifications.
基金National Natural Science Foundation of China,Grant/Award Numbers:92164302,8206100486,62404007Guangdong Provincial Key Laboratory of In-Memory Computing Chips,Grant/Award Number:2024B1212020002+5 种基金Shenzhen Science and Technology Program,Grant/Award Number:JCYJ20241202125907011Beijing Natural Science Foundation,Grant/Award Numbers:L234026,L257010,25D40029,25FY3314111 Project,Grant/Award Number:B18001Fok Ying-Tong Education FoundationTencent FoundationChina Postdoctoral Science Foundation,Grant/Award Number:2023M740051。
文摘Nonlinear physical systems hold great promise for energy-efficient and lowhardware-cost information processing.However,their computational capabilities remain constrained by the complexity and tunability of system nonlinearity.Here we report a dual-ferroelectric gate-tunable memristor with a dipole coupling effect,achieving enlarged hysteresis,rich temporal dynamics,and nonvolatile heterosynaptic plasticity.By harnessing the dynamic nonlinearity of the dual-ferroelectric memristor,multimodal reservoir computing with an in-material fusion strategy has been achieved,which is demonstrated with a multimodal object recognition task.By exploring the static nonlinearity of the dual-ferroelectric memristor,nonlinear in-memory computing is realized with gate-tunable nonlinear functions,which successfully accelerates the Euclidean distance computation in the K-means clustering task.This work achieves strong coupling between the intrinsic physical dynamics and computational functionalities,offering new opportunities for more efficient hardware-accelerated systems.
文摘Nonlinear computation is essential for a wide range of information processing tasks,yet implementing nonlinear functions using optical systems remains a challenge due to the weak and power-intensive nature of optical nonlinearities.Overcoming this limitation without relying on nonlinear optical materials could unlock unprecedented opportunities for ultrafast and parallel optical computing systems.Here,we demonstrate that large-scale nonlinear computation can be performed using linear optics through optimized diffractive processors composed of passive phase-only surfaces.In this framework,the input variables of nonlinear functions are encoded into the phase of an optical wavefront—e.g.,via a spatial light modulator(SLM)—and transformed by an optimized diffractive structure with spatially varying point-spread functions to yield output intensities that approximate a large set of unique nonlinear functions–all in parallel.We provide proof establishing that this architecture serves as a universal function approximator for an arbitrary set of bandlimited nonlinear functions,also covering wavelength-multiplexed nonlinear functions as well as multi-variate and complex-valued functions that are all-optically cascadable.Our analysis also indicates the successful approximation of typical nonlinear activation functions commonly used in neural networks,including the sigmoid,tanh,ReLU(rectified linear unit),and softplus.We numerically demonstrate the parallel computation of one million distinct nonlinear functions,accurately executed at wavelength-scale spatial density at the output of a diffractive optical processor.Furthermore,we experimentally validated this framework using in situ optical learning and approximated 35 unique nonlinear functions in a single shot using a compact setup consisting of an SLM and an image sensor.These results establish diffractive optical processors as a scalable platform for massively parallel universal nonlinear function approximation,paving the way for new capabilities in analog optical computing based on linear materials.
文摘The stochastic resonance （SR） in a time-delayed mono-stable system driven by multiplicative white noise, additive white noise, additive dichotomous noise as well as a periodic square-wave signal is considered from the view of the signal-to-noise ratio （SNR）. It is found that the SNR increases monotonically with the increase of the delay time. The SNR exhibits the SR behavior when it is plotted as a function of intensities of the noises, displaying the asymmetry of the dichotomous noise. The SNR varies non-monotonically with the increase of the system parameter and the amplitude of the input square-wave signal.
基金Supported by the National Natural Science Foundation of China under Grant No 10847156.
文摘The phenomenon of entropic stochastic resonance （ESR） in a two-dimensional confined system driven by a transverse periodic force is investigated when the colored fluctuation is included in the system. Applying the method of unified colored noise approximation, the approximate Fokker-Planck equation can be derived in the absence of the periodic force. Through the escaping rate of the Brownian particle from one well to the other, the power spectral amplification can be obtained. It is found that increasing the values of the noise correlation time and the signal frequency can suppress the ESR of the system.
基金Supported by the National Natural Science Foundation of China under Grant No 10975063, and the Fundamental Research Pund for Physics and Mathematics of Lanzhou University.
文摘We study the dynamics of tumor cell growth with time-delayed feedback driven by multiplicative noise in an asymmetrical bistable potential well. For a small delay time, the analytical solutions of the probability distribution and the first passage time show that, with the increasing delay time, the peak of the probability distribution in a lower population state would increase, but in a higher population state it decreases. It is shown that the multiplicative noise and the time delay play opposite roles in the tumor cell growth.
基金One of the authors (L. Kavitha) gratefully acknowledges the financial support from NBHM, India in the form of major research project, BRNS, India in the form of Young Scientist Research Award and ICTP, Italy in the form of Junior AssociateshipUGC, India for financial assistance in the form of Research Award+1 种基金M. Venkatesh acknowledges BSR-Research Fellowship under UGC Non-SAP Scheme, IndiaS. Dhamayanthi thanks the University Research Fellowship (URF) given by Periyar Uni- versity, India.
文摘We report the modulational instability （MI） analysis for the modulation equations governing the propagation of coherent polarized light through a nematic liquid crystal （NLC） slab, in the limit of low light intensity and local material response. The linear stability analysis of the nonlinear plane wave solutions is performed by considering both the wave vectors （k,l） of the basic states and wave vectors （K,L） of the perturbations as free parameters. We compute the MI gain, and the MI gain peak is reduced and the stable bandwidth is widened with the increase of the strength of the applied electric field. Further, we invoke the extended homogeneous balance method and Exp-function method aided with symbolic computation and obtain a series of periodic solitonic humps of nematicon profiles admitting the propagation of laser light in the NLC medium.
基金Supported by the National Natural Science Foundation of China under Grant No 1057400, and the Natural Science Foundation of Heilongjiang Province under Grant No A200506.
文摘We study the features of a single q-breather （SQB） in a Fermi-Pasta-Ulam lattice by the numerical method, and obtain that the stability of SQB correlates to coupling constant K and nonlinear parameter β. No matter whether K or β increases, the periodic SQB can be transformed into a quasiperiodic SQB or a chaotic SQB. We also obtain the conditions of excitation of periodic, quasiperiodic and chaotic SQBs.
文摘A new simple model of self-driven particles in scale-free networks is introduced to understand the emergence of condensation in the natural world. In the model, at each time step, particles are driven to choose their next habitats according to the particle numbers at both the present habitats and neighbors. It is found that the hub effect results in the condensation. The present numerical results as well as the theoretical analysis of condensation transition show the criterion of condensation. Both simulations and theoretical analysis display that there are three phases for different hopping probabilities δ： non-condensation （δ〈δc）, partial condensation （δc≤ δ〈1）, and complete condensation （δ≥δc＇= 1）, and the mean occupation particle number at the stationary state is also obtained. Moreover, the noise effect on self-driven particles is studied, and it makes particle numbers at nodes tend towards being identical.
基金Supported by the National Natural Science Foundation of China under Grant Nos 10871074 and 60704045, Research and for the Doctoral Program of Higher Education of China under Grant No 20070558053, and the Natural Science Foundation of Guangdong Province under Grant No 9451042001004076.
文摘We demonstrate that the projective synchronization can be observed in coupled fractional-order chaotic systems. A new systematic and powerful coupling scheme is developed to investigate the projective synchronization via the open-plus-closed-loop control, which allows us to arbitrarily manipulate the scaling factor of projective synchronization. The proposed scheme is proved analytically on the basis of the stability theorem of the fractional differential equations. Numerical simulations on the fraction-order chaotic Chen system are presented to justify the theoretical analysis.
基金Supported by the New Century Excellent Talent Project of the Ministry of Education of China under Grant No NECT-07-0112, and the National Natural Science Foundation of China under Grant No 10775022.
文摘We investigate the game theory in a structured population with the assumption that the evolution of network structure is far faster than that of strategy update. We find that the degree distribution for the finM network consists of two distinct parts： the low degree part which is contributed to by defectors and a broadband in the regime with high degree which is formed by cooperators. The structure of the final network and the final strategy pattern have also been numerically proved to be independent of the game parameters.
基金Sponsored partly by Priority-Scientific-Projects for China's 7th and 8th Five-Year Plana Priority Project of the Director's Foundation of the Institute of Atmospheric PhysicsChinese Academy of Sciences.
文摘In accordance with a new compensation principle of discrete computations,the traditional meteo- rological global (pseudo-) spectral schemes of barotropic primitive equation (s) are transformed into perfect energy conservative fidelity schemes,thus resolving the problems of both nonlinear computa- tional instability and incomplete energy conservation,and raising the computational efficiency of the traditional schemes. As the numerical tests of the new schemes demonstrate,in solving the problem of energy conser- vation in operational computations,the new schemes can eliminate the (nonlinear) computational in- stability and,to some extent even the (nonlinear) computational diverging as found in the traditional schemes,Further contrasts between new and traditional schemes also indicate that,in discrete opera- tional computations,the new scheme in the case of nondivergence is capable of prolonging the valid in- tegral time of the corresponding traditional scheme,and eliminating certain kind of systematical com- putational“climate drift”,meanwhile increasing its computational accuracy and reducing its amount of computation.The working principle of this paper is also applicable to the problem concerning baroclin- ic primitive equations.
