Yinpeng Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Yinpeng Wang has a total of 31 co-authors. Here are the top co-authors by number of shared publications:
- Qiang Ren
School of Electronic and Information Engineering, Beihang University, Beijing, China
14 shared publications Yongzhong Li
7 shared publicationsShutong Qi
6 shared publicationsHongyu Gao
School of Electronics and Information Engineering Beihang University 37th Xueyuan Road, Haidian District Beijing 100191 China
2 shared publicationsNianru Wang
Beihang University,School of Electronic and Information Engineering,China
2 shared publicationsShihong Zhang
Beihang University
2 shared publications
Latest Publications
- Open Access
- 9 March 2026
- Published by Wiley in Advanced Materials Technologies
- Vol. 11 (13)
Transducers Across Scales and Frequencies: A System‐Level Framework for Multiphysics Integration and Co‐Design
Transducers underpin modern sensing, actuation, and modulation by converting physical signals into electrical or optical representations. Despite rapid advances across materials, fabrication, and device architectures, individual transduction mechanisms remain constrained by intrinsic trade‐offs among bandwidth, sensitivity, speed, energy consumption, and integrability. This review examines transducer technologies across mechanical, acoustic, electromagnetic, and optical domains, and shows that performance evolution is not only increasingly governed by the discovery of new mechanisms, but also by the system‐level coordination of established ones. By organizing representative platforms according to physical scale, operating frequency, and accessible degrees of freedom, we reveal how distinct mechanisms occupy complementary performance envelopes across Hertz‐to‐THz regimes. We highlight how heterogeneous integration and multiphysics co‐design enable these envelopes to be traversed through coordinated architectures that combine flexible interfaces, electromechanical systems, metasurfaces, and photonic circuits. This perspective reframes transducers from isolated interfaces into programmable system nodes that jointly support sensing, modulation, and information processing. The resulting framework provides a foundation for designing reconfigurable and scalable transducer systems for sustainable applications, precision imaging, adaptive communication, edge intelligence, and emerging quantum technologies.
- 9 March 2026
- Published by Wiley in Advanced Materials Technologies
- Vol. 11 (13)
Open AccessTransducers Across Scales and Frequencies: A System‐Level Framework for Multiphysics Integration and Co‐Design
Transducers underpin modern sensing, actuation, and modulation by converting physical signals into electrical or optical representations. Despite rapid advances across materials, fabrication, and device architectures, individual transduction mechanisms remain constrained by intrinsic trade‐offs among bandwidth, sensitivity, speed, energy consumption, and integrability. This review examines transducer technologies across mechanical, acoustic, electromagnetic, and optical domains, and shows that performance evolution is not only increasingly governed by the discovery of new mechanisms, but also by the system‐level coordination of established ones. By organizing representative platforms according to physical scale, operating frequency, and accessible degrees of freedom, we reveal how distinct mechanisms occupy complementary performance envelopes across Hertz‐to‐THz regimes. We highlight how heterogeneous integration and multiphysics co‐design enable these envelopes to be traversed through coordinated architectures that combine flexible interfaces, electromechanical systems, metasurfaces, and photonic circuits. This perspective reframes transducers from isolated interfaces into programmable system nodes that jointly support sensing, modulation, and information processing. The resulting framework provides a foundation for designing reconfigurable and scalable transducer systems for sustainable applications, precision imaging, adaptive communication, edge intelligence, and emerging quantum technologies.
- 15 November 2025
- Published by Elsevier in Water Research
- Vol. 289 (Pt B) , 124962
Diffusion for Diffusion: A versatile multiphysics fields refinement framework in pollutants transportation
- 15 November 2025
- Published by Elsevier in Water Research
- Vol. 289 (Pt B) , 124962
Diffusion for Diffusion: A versatile multiphysics fields refinement framework in pollutants transportation
- 28 October 2024
- Published by Optica Publishing Group in Optical Materials Express
- Vol. 14 (11) , 2740-2754
Multi-receptive-field physics-informed neural network for complex electromagnetic media
Acquiring the electromagnetic response of intricate media at the nanoscale constitutes a pivotal phase in the design intricacies of nanophotonic apparatuses. Conventional numerical algorithms often necessitate intricate and specialized treatments to accommodate the unique properties of the medium, coupled with substantial computational time and resource demands. In recent years, the advent of deep learning technology has heralded numerous advancements in the domain of computational electromagnetics, albeit with a scarcity of solvers tailored for versatile complex media. Consequently, this study introduces an innovative multi-receptive-field physics-informed neural network (MRF-PINN) designed to tackle nano optical scattering predicaments inherent in media exhibiting dispersion, inhomogeneity, anisotropy, nonlinearity, and chirality. This framework adeptly captures electromagnetic perturbations surrounding scatterers via variable-scale receptive fields, thereby enhancing prediction precision. Within the training regimen, a scale balancing algorithm is proposed to expedite network convergence. Empirical findings demonstrate that a fully trained MRF-PINN proficiently reconstructs electromagnetic field distributions within complex nanomaterials within a mere tens of milliseconds of inference time. Such quasi real-time capabilities herald a novel approach to supplant the arduous forward calculation processes inherent in nanomaterial design workflows.
- 28 October 2024
- Published by Optica Publishing Group in Optical Materials Express
- Vol. 14 (11) , 2740-2754
Multi-receptive-field physics-informed neural network for complex electromagnetic media
Acquiring the electromagnetic response of intricate media at the nanoscale constitutes a pivotal phase in the design intricacies of nanophotonic apparatuses. Conventional numerical algorithms often necessitate intricate and specialized treatments to accommodate the unique properties of the medium, coupled with substantial computational time and resource demands. In recent years, the advent of deep learning technology has heralded numerous advancements in the domain of computational electromagnetics, albeit with a scarcity of solvers tailored for versatile complex media. Consequently, this study introduces an innovative multi-receptive-field physics-informed neural network (MRF-PINN) designed to tackle nano optical scattering predicaments inherent in media exhibiting dispersion, inhomogeneity, anisotropy, nonlinearity, and chirality. This framework adeptly captures electromagnetic perturbations surrounding scatterers via variable-scale receptive fields, thereby enhancing prediction precision. Within the training regimen, a scale balancing algorithm is proposed to expedite network convergence. Empirical findings demonstrate that a fully trained MRF-PINN proficiently reconstructs electromagnetic field distributions within complex nanomaterials within a mere tens of milliseconds of inference time. Such quasi real-time capabilities herald a novel approach to supplant the arduous forward calculation processes inherent in nanomaterial design workflows.
- 19 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Multiscale and Multiphysics Computational Techniques
- Vol. 9, 341-352
Deep Multiphysics Fields Solver Established on Operator Learning Transformer and Finite Element Method
The accurate acquisition of unknown multiphysics fields in specified regions is vital for industrial production. Traditional computational approaches often require dense mesh generation to achieve precise numerical results, leading to substantial computational resource consumption and extended processing times. However, recent advancements in deep learning (DL) have introduced alternative solutions to computational physics problems. This paper presents a novel multiphysics field solver that integrates operator learning with classical finite element methods (FEM). The overall structure of the framework is a Transformer based on the attention mechanism, with a loss function incorporating physical constraints. The network takes the result of a coarse grid finite element calculation as input, while the output target is the value of a dense grid computation. Compared to traditional DL frameworks, the proposed architecture consistently maintains low error rates across a range of input resolutions. Additionally, the high efficiency of graphics processing units (GPUs) enables fully trained networks to generate solutions in quasi-real time, demonstrating significant potential for practical applications.
- 19 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Multiscale and Multiphysics Computational Techniques
- Vol. 9, 341-352
Deep Multiphysics Fields Solver Established on Operator Learning Transformer and Finite Element Method
The accurate acquisition of unknown multiphysics fields in specified regions is vital for industrial production. Traditional computational approaches often require dense mesh generation to achieve precise numerical results, leading to substantial computational resource consumption and extended processing times. However, recent advancements in deep learning (DL) have introduced alternative solutions to computational physics problems. This paper presents a novel multiphysics field solver that integrates operator learning with classical finite element methods (FEM). The overall structure of the framework is a Transformer based on the attention mechanism, with a loss function incorporating physical constraints. The network takes the result of a coarse grid finite element calculation as input, while the output target is the value of a dense grid computation. Compared to traditional DL frameworks, the proposed architecture consistently maintains low error rates across a range of input resolutions. Additionally, the high efficiency of graphics processing units (GPUs) enables fully trained networks to generate solutions in quasi-real time, demonstrating significant potential for practical applications.
- Open Access
- 13 June 2024
- Published by Elsevier in International Journal of Thermofluids
- Vol. 23
Coupled multiphysics solver for irregular regions based on graph neural network
- 13 June 2024
- Published by Elsevier in International Journal of Thermofluids
- Vol. 23
Open AccessCoupled multiphysics solver for irregular regions based on graph neural network
- Open Access
Wave transformation -- Simulation and Application
This paper focus on the theoretical analysis and simulation of electromagnetic wave transforms, which is widely encountered in teaching physics. When the electromagnetic wave is not consistent with the shape of the object, it is often necessary to introduce electromagnetic wave transforms to analyze the interaction between the two. In this paper, we discuss three common waves, namely plane wave, cylindrical wave and spherical wave. We first analyze the mathematical transformation equations between them, and then use computer programs to draw a simulation image to intuitively show the derivation results. During the derivation, we corrected an error in a classic electromagnetic textbook. Finally, several specific electromagnetic examples are solved by electromagnetic wave transformation. These examples include the scattering of plane waves by ideal conductor cylinder, sphere and ideal dielectric cylinder. The simulation results are displayed intuitively in the form of images.
Open AccessWave transformation -- Simulation and Application
This paper focus on the theoretical analysis and simulation of electromagnetic wave transforms, which is widely encountered in teaching physics. When the electromagnetic wave is not consistent with the shape of the object, it is often necessary to introduce electromagnetic wave transforms to analyze the interaction between the two. In this paper, we discuss three common waves, namely plane wave, cylindrical wave and spherical wave. We first analyze the mathematical transformation equations between them, and then use computer programs to draw a simulation image to intuitively show the derivation results. During the derivation, we corrected an error in a classic electromagnetic textbook. Finally, several specific electromagnetic examples are solved by electromagnetic wave transformation. These examples include the scattering of plane waves by ideal conductor cylinder, sphere and ideal dielectric cylinder. The simulation results are displayed intuitively in the form of images.
- 4 August 2022
- Published by Wiley in Advanced Theory and Simulations
- Vol. 5 (11)
Differential Operator Approximation Based Tightly Coupled Multiphysics Solver Using Cascaded Fourier Network
Calculating the unknown coupled multiphysics fields from the known boundary condition is of great significance in computational physics. Existing classical algorithms are usually time consuming and resource demanding. Explosive growth in deep learning (DL) has provided a substitutive way to accelerate the calculation process by fully making use of the parallel computing capability of the graphics processing unit. In this work, a cascaded DL framework is presented to compute the coupled multiphysics fields related to electrical potential, temperature, velocity, etc. The whole framework is comprised of an input module and several Fourier modules, which achieves to acquire the coupled fields from the explicit boundary conditions. Compared with conventional networks, the Fourier network emerges consistent average error among different resolutions which saves training costs. As a result, a fully trained network can obtain the expected physics fields in real time, offering rosy prospects for practical scenarios.
- 4 August 2022
- Published by Wiley in Advanced Theory and Simulations
- Vol. 5 (11)
Differential Operator Approximation Based Tightly Coupled Multiphysics Solver Using Cascaded Fourier Network
Calculating the unknown coupled multiphysics fields from the known boundary condition is of great significance in computational physics. Existing classical algorithms are usually time consuming and resource demanding. Explosive growth in deep learning (DL) has provided a substitutive way to accelerate the calculation process by fully making use of the parallel computing capability of the graphics processing unit. In this work, a cascaded DL framework is presented to compute the coupled multiphysics fields related to electrical potential, temperature, velocity, etc. The whole framework is comprised of an input module and several Fourier modules, which achieves to acquire the coupled fields from the explicit boundary conditions. Compared with conventional networks, the Fourier network emerges consistent average error among different resolutions which saves training costs. As a result, a fully trained network can obtain the expected physics fields in real time, offering rosy prospects for practical scenarios.
- 25 April 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 81-84
Inversion of Sophisticated Thermal Conductivity via Deep Learning
Identifying the thermophysical properties of unknown material is of great significance in computational heat transfer. In this paper, a deep learning (DL) framework is proposed to reconstruct the temperature or position-dependent thermal conductivity in the 2D region. The framework consists of a denoising network and an inversion network, aiming at inhibiting the noise and reconstruct the thermal conductivity, respectively. The architecture of the denoising network is mainly based on the convolutional neural network (CNN) while the inversion network on the fully connected neural network (FCNN). In order to gain sufficient training data, a forward solver based on the finite element method (FEM), combined with a random thermal conductivity generator, is utilized to compute the boundary temperature. To simulate the actual scenarios, Gaussian noises are added to the calculation results. During the training process, the temperature with noise serves as the input of the denoising network while the generated thermal conductivity is regarded as the target of the inversion network. After 500 epochs of training on Keras, the network eventually achieves convergence. To quantitatively depict the performance of the framework, we define the average relative error rate. As a result, the final error rate is 1.03% for the temperature-dependent situation and 2.65% for the position-dependent one. A great merit for DL techniques is that a well-trained framework can give a forecast almost simultaneously with the given input information. In our network, it only spends 1.732ms to predict the unknown thermal conductivity preciously. Since real-time thermal inversion is more and more widely used in industry, it is anticipated that the proposed framework can be extensively adopted in a variety of scenarios.
- 25 April 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 81-84
Inversion of Sophisticated Thermal Conductivity via Deep Learning
Identifying the thermophysical properties of unknown material is of great significance in computational heat transfer. In this paper, a deep learning (DL) framework is proposed to reconstruct the temperature or position-dependent thermal conductivity in the 2D region. The framework consists of a denoising network and an inversion network, aiming at inhibiting the noise and reconstruct the thermal conductivity, respectively. The architecture of the denoising network is mainly based on the convolutional neural network (CNN) while the inversion network on the fully connected neural network (FCNN). In order to gain sufficient training data, a forward solver based on the finite element method (FEM), combined with a random thermal conductivity generator, is utilized to compute the boundary temperature. To simulate the actual scenarios, Gaussian noises are added to the calculation results. During the training process, the temperature with noise serves as the input of the denoising network while the generated thermal conductivity is regarded as the target of the inversion network. After 500 epochs of training on Keras, the network eventually achieves convergence. To quantitatively depict the performance of the framework, we define the average relative error rate. As a result, the final error rate is 1.03% for the temperature-dependent situation and 2.65% for the position-dependent one. A great merit for DL techniques is that a well-trained framework can give a forecast almost simultaneously with the given input information. In our network, it only spends 1.732ms to predict the unknown thermal conductivity preciously. Since real-time thermal inversion is more and more widely used in industry, it is anticipated that the proposed framework can be extensively adopted in a variety of scenarios.
- 1 January 2022
- Published by Springer Nature
Sophisticated Electromagnetic Forward Scattering Solver via Deep Learning
- 1 January 2022
- Published by Springer Nature
Sophisticated Electromagnetic Forward Scattering Solver via Deep Learning
- 1 December 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. ume1, 01-03
Cascaded Network for Inversion of Electrical Conductivity in High Noise Environment
In this paper, a cascade network is proposed to reconstruct the electrical conductivity through the electrical fields, presenting great computing efficiency and high accuracy. Benefit from the unsupervised U-net, the framework is applicable to high noise environment. Compared to traditional algorithms, the proposed physics informed neural networks (PINN) avoids meshing and iterative calculation. Therefore, it can be used to predict the conductivity distribution of objects with complex geometric shapes simultaneously. We believe this work will be benefit to fields related to electrical conductivity inversion problems.
- 1 December 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. ume1, 01-03
Cascaded Network for Inversion of Electrical Conductivity in High Noise Environment
In this paper, a cascade network is proposed to reconstruct the electrical conductivity through the electrical fields, presenting great computing efficiency and high accuracy. Benefit from the unsupervised U-net, the framework is applicable to high noise environment. Compared to traditional algorithms, the proposed physics informed neural networks (PINN) avoids meshing and iterative calculation. Therefore, it can be used to predict the conductivity distribution of objects with complex geometric shapes simultaneously. We believe this work will be benefit to fields related to electrical conductivity inversion problems.