Bin Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Bin Li has a total of 63 co-authors. Here are the top co-authors by number of shared publications:
Quan Hu
Univ Elect Sci & Technol China, Sch Elect Sci & Engn, Vacuum Elect Natl Lab, Chengdu 610054, Peoples R China
5 shared publicationsLi Liao
The Vacuum Electronics National Laboratory, School of Electronic Science and Engineering, University of Electronic Science and Technology of China 1 , Chengdu 611731, China
4 shared publications- Yulu Hu
University of Electronic Science and Technology of China
4 shared publications - Jinqing Yu
Hunan Provincial Key Laboratory of High-Energy Scale Physics and Applications, School of Physics and Electronics, Hunan University, Changsha 410082, China
3 shared publications - Quan Hu
University of Electronic Science and Technology of China
3 shared publications Tao Huang
3 shared publications
Latest Publications
- Open Access
Spectral-Prior Guided Multistage Physics-Informed Neural Networks for Highly Accurate PDE Solutions
Physics-Informed Neural Networks (PINNs) are becoming a popular method for solving PDEs, due to their mesh-free nature and their ability to handle high-dimensional problems where traditional numerical solvers often struggle. Despite their promise, the practical application of PINNs is still constrained by several fac- tors, a primary one being their often-limited accuracy. This paper is dedicated to enhancing the accuracy of PINNs by introducing spectral-prior guided multistage strategy. We propose two methods: Spectrum- Informed Multistage Physics-Informed Neural Networks (SI-MSPINNs) and Multistage Physics-Informed Neural Networks with Spectrum Weighted Random Fourier Features (RFF-MSPINNs). The SI-MSPINNs integrate the core mechanism of Spectrum-Informed Multistage Neural Network (SI-MSNNs) and PINNs, in which we extract the Dominant Spectral Pattern (DSP) of residuals by the discrete Fourier transform. This DSP guides the network initialization to alleviate spectral bias, and gradually optimizes the resolution accuracy using a multistage strategy. The RFF-MSPINNs combines random Fourier features with spectral weighting methods, dynamically adjusting the frequency sampling distribution based on the residual power spectral density, allowing the network to prioritize learning high-energy physical modes. Through experimental verification of the Burgers equation and the Helmholtz equation, we show that both models significantly improve the accuracy of the original PINNs.
Open AccessSpectral-Prior Guided Multistage Physics-Informed Neural Networks for Highly Accurate PDE Solutions
Physics-Informed Neural Networks (PINNs) are becoming a popular method for solving PDEs, due to their mesh-free nature and their ability to handle high-dimensional problems where traditional numerical solvers often struggle. Despite their promise, the practical application of PINNs is still constrained by several fac- tors, a primary one being their often-limited accuracy. This paper is dedicated to enhancing the accuracy of PINNs by introducing spectral-prior guided multistage strategy. We propose two methods: Spectrum- Informed Multistage Physics-Informed Neural Networks (SI-MSPINNs) and Multistage Physics-Informed Neural Networks with Spectrum Weighted Random Fourier Features (RFF-MSPINNs). The SI-MSPINNs integrate the core mechanism of Spectrum-Informed Multistage Neural Network (SI-MSNNs) and PINNs, in which we extract the Dominant Spectral Pattern (DSP) of residuals by the discrete Fourier transform. This DSP guides the network initialization to alleviate spectral bias, and gradually optimizes the resolution accuracy using a multistage strategy. The RFF-MSPINNs combines random Fourier features with spectral weighting methods, dynamically adjusting the frequency sampling distribution based on the residual power spectral density, allowing the network to prioritize learning high-energy physical modes. Through experimental verification of the Burgers equation and the Helmholtz equation, we show that both models significantly improve the accuracy of the original PINNs.
- Open Access
- 1 March 2024
- Published by AIP Publishing in AIP Advances
- Vol. 14 (3)
Recent developments on electron optics simulator
Recent developments on the electron optics simulator (EOS) are reported in this paper. The EOS is a finite element two-dimensional (2D) and three-dimensional (3D) electrostatic particle steady-state trajectory code that has been widely used for beam optics modeling of vacuum electronic devices. To extend the application scope and improve the computation efficiency, four significant advances have been made and are presented in this paper. First, the code is reprogrammed to consider physical models that contain dielectrics and to support charge excitation. Second, more primary particle emission models are added in EOS to meet the needs for simulating electron and ion optics devices for diverse applications. For the precise simulation of the secondary electron emission, the Vaughan model is improved, and the Furman model is newly added in EOS. Finally, the algebraic multigrid algorithm has been embedded in EOS, which can significantly improve the computation efficiency for problems with large-scale matrices. For these developments, the necessary principles and corresponding performance are shown in detail in this paper.
- 1 March 2024
- Published by AIP Publishing in AIP Advances
- Vol. 14 (3)
Open AccessRecent developments on electron optics simulator
Recent developments on the electron optics simulator (EOS) are reported in this paper. The EOS is a finite element two-dimensional (2D) and three-dimensional (3D) electrostatic particle steady-state trajectory code that has been widely used for beam optics modeling of vacuum electronic devices. To extend the application scope and improve the computation efficiency, four significant advances have been made and are presented in this paper. First, the code is reprogrammed to consider physical models that contain dielectrics and to support charge excitation. Second, more primary particle emission models are added in EOS to meet the needs for simulating electron and ion optics devices for diverse applications. For the precise simulation of the secondary electron emission, the Vaughan model is improved, and the Furman model is newly added in EOS. Finally, the algebraic multigrid algorithm has been embedded in EOS, which can significantly improve the computation efficiency for problems with large-scale matrices. For these developments, the necessary principles and corresponding performance are shown in detail in this paper.
- Open Access
Research of the Algebraic Multigrid Method for Electron Optical Simulator
At present, electron optical simulator (EOS) takes a long time to solve linear FEM systems. The algebraic multigrid preconditioned conjugate gradient (AMGPCG) method can improve the efficiency of solving systems. This paper is focused on the implementation of the AMGPCG method in EOS. The aggregation-based scheme, which uses two passes of a pairwise matching algorithm and the K-cyle scheme, is adopted in the aggregation-based algebraic multigrid method. Numerical experiments show the advantages and disadvantages of the AMG algorithm in peak memory and solving efficiency. The AMGPCG is more efficient than the iterative methods used in the past and only needs one coarsening when EOS computes the particle motion trajectory.
Open AccessResearch of the Algebraic Multigrid Method for Electron Optical Simulator
At present, electron optical simulator (EOS) takes a long time to solve linear FEM systems. The algebraic multigrid preconditioned conjugate gradient (AMGPCG) method can improve the efficiency of solving systems. This paper is focused on the implementation of the AMGPCG method in EOS. The aggregation-based scheme, which uses two passes of a pairwise matching algorithm and the K-cyle scheme, is adopted in the aggregation-based algebraic multigrid method. Numerical experiments show the advantages and disadvantages of the AMG algorithm in peak memory and solving efficiency. The AMGPCG is more efficient than the iterative methods used in the past and only needs one coarsening when EOS computes the particle motion trajectory.
- 25 January 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Components, Packaging and Manufacturing Technology
- Vol. 11 (3) , 395-406
An Interior Penalty Domain Decomposition Method for Thermal Analysis of 3-D Integrated Systems
This article presents a finite-element-based domain decomposition method (DDM) for thermal analysis of 3-D large-scale and geometrically complicated integrated systems. The presented DDM decomposes the entire problem domain into many nonoverlapping subdomains, and a novel transmission condition (TC) based on a fictitious thermal contact resistance is introduced and then imposed through an interior penalty (IP) Galerkin procedure to enforce continuity at the interfaces between subdomains. The major advantage of the novel TC combined with the IP Galerkin procedure is that no extra unknowns are introduced and the system matrix arising from the finite element method will be symmetric and positive definite. Furthermore, the solving procedure of matrix equation adopts the preconditioned conjugate gradient method with a two-level preconditioner. To verify the convergence and accuracy, this article calculates the transient temperature profile of a brick and compares the computed numerical approximation with the exact solution. Meanwhile, the solving advantage of the proposed DDM is shown by comparing with Robin-TC DDM. Finally, numerical results for a 3-D integration package are shown to demonstrate the accuracy and superior performance of the proposed DDM.
- 25 January 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Components, Packaging and Manufacturing Technology
- Vol. 11 (3) , 395-406
An Interior Penalty Domain Decomposition Method for Thermal Analysis of 3-D Integrated Systems
This article presents a finite-element-based domain decomposition method (DDM) for thermal analysis of 3-D large-scale and geometrically complicated integrated systems. The presented DDM decomposes the entire problem domain into many nonoverlapping subdomains, and a novel transmission condition (TC) based on a fictitious thermal contact resistance is introduced and then imposed through an interior penalty (IP) Galerkin procedure to enforce continuity at the interfaces between subdomains. The major advantage of the novel TC combined with the IP Galerkin procedure is that no extra unknowns are introduced and the system matrix arising from the finite element method will be symmetric and positive definite. Furthermore, the solving procedure of matrix equation adopts the preconditioned conjugate gradient method with a two-level preconditioner. To verify the convergence and accuracy, this article calculates the transient temperature profile of a brick and compares the computed numerical approximation with the exact solution. Meanwhile, the solving advantage of the proposed DDM is shown by comparing with Robin-TC DDM. Finally, numerical results for a 3-D integration package are shown to demonstrate the accuracy and superior performance of the proposed DDM.
- 4 June 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Plasma Science
- Vol. 46 (7) , 2505-2511
Analytic Exploration of the Accuracy of Pierce’s Three-Wave Beam-Wave Interaction Theory of Traveling-Wave Tubes
As it is well known that the simplified classical Pierce's three-wave theory can provide scaling insights into traveling-wave tube (TWT) operation for its close-form solutions, and thus becomes a useful guide for TWT design. However, the classical Pierce's three-wave theory shows poor agreement with Lagrangian theory, so determining how to achieve a more accurately Pierce's three-wave theory from Pierce's four-wave theory becomes an open problem. In this paper, a more accurately revised Pierce's three-wave dispersion relation is deduced by using an approximate treatment on the nonlinear term of the dispersion relation of Pierce's four-wave beam-wave interaction theory. Meanwhile, the boundary conditions of classical Pierce's three-wave theory are revised by theoretical analysis. Thus, the revised Pierce's three-wave theory is established. The Pierce small-signal theories are compared to each other and Lagrangian theory on a set of TWT parameters which are based on a single pitch section of C- and Ku- bands TWTs. It is found that the revised Pierce's three-wave theory agrees extremely well with Lagrangian theory and gains more accuracy than the classical Pierce's three-wave theory in the small-signal beam-wave interaction region. Finally, the phase difference between Pierce's four-wave theory and revised Pierce's three-wave theory is discussed.
- 4 June 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Plasma Science
- Vol. 46 (7) , 2505-2511
Analytic Exploration of the Accuracy of Pierce’s Three-Wave Beam-Wave Interaction Theory of Traveling-Wave Tubes
As it is well known that the simplified classical Pierce's three-wave theory can provide scaling insights into traveling-wave tube (TWT) operation for its close-form solutions, and thus becomes a useful guide for TWT design. However, the classical Pierce's three-wave theory shows poor agreement with Lagrangian theory, so determining how to achieve a more accurately Pierce's three-wave theory from Pierce's four-wave theory becomes an open problem. In this paper, a more accurately revised Pierce's three-wave dispersion relation is deduced by using an approximate treatment on the nonlinear term of the dispersion relation of Pierce's four-wave beam-wave interaction theory. Meanwhile, the boundary conditions of classical Pierce's three-wave theory are revised by theoretical analysis. Thus, the revised Pierce's three-wave theory is established. The Pierce small-signal theories are compared to each other and Lagrangian theory on a set of TWT parameters which are based on a single pitch section of C- and Ku- bands TWTs. It is found that the revised Pierce's three-wave theory agrees extremely well with Lagrangian theory and gains more accuracy than the classical Pierce's three-wave theory in the small-signal beam-wave interaction region. Finally, the phase difference between Pierce's four-wave theory and revised Pierce's three-wave theory is discussed.
- 24 February 2015
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 62 (4) , 1319-1326
Magnetic Focusing Simulator: A 3-D Finite-Element Permanent-Magnet Focusing System Design Tool
This paper reports the development of a 3-D finite-element design tool for the design of permanent-magnet focusing system called the magnetic focusing simulator (MFS). It is a new module of microwave-tube simulator suite developed by the University of Electronic Science and Technology of China. MFS specifically targets problem classes including periodic permanent magnet, wiggler, permanent-magnet quadrupoles, periodic cusped magnet, and so on. MFS is a finite-element code of magnetic scalar potential and is a solver of magnetostatic field. It can be used for designing, optimizing, testing, and validating the permanent-magnetic focusing systems. The theory of MFS is discussed in detail in this paper. Besides, the accuracy is compared with the commercial software ANSYS Maxwell, and the performance of MFS is also shown.
- 24 February 2015
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 62 (4) , 1319-1326
Magnetic Focusing Simulator: A 3-D Finite-Element Permanent-Magnet Focusing System Design Tool
This paper reports the development of a 3-D finite-element design tool for the design of permanent-magnet focusing system called the magnetic focusing simulator (MFS). It is a new module of microwave-tube simulator suite developed by the University of Electronic Science and Technology of China. MFS specifically targets problem classes including periodic permanent magnet, wiggler, permanent-magnet quadrupoles, periodic cusped magnet, and so on. MFS is a finite-element code of magnetic scalar potential and is a solver of magnetostatic field. It can be used for designing, optimizing, testing, and validating the permanent-magnetic focusing systems. The theory of MFS is discussed in detail in this paper. Besides, the accuracy is compared with the commercial software ANSYS Maxwell, and the performance of MFS is also shown.
- 12 September 2013
- Published by Elsevier in High Energy Density Physics
- Vol. 9 (4) , 745-749
Influence of proton beam Coulomb explosion in laser proton acceleration
- 12 September 2013
- Published by Elsevier in High Energy Density Physics
- Vol. 9 (4) , 745-749
Influence of proton beam Coulomb explosion in laser proton acceleration
- 28 August 2013
- Published by Hindawi Limited in Laser and Particle Beams
- Vol. 31 (4) , 597-605
Influence of the initial size of the proton layer in sheath field proton acceleration
We investigate the influence of the initial size of the proton layer on proton acceleration in the interaction of high intensity laser pulses with double-layer targets by using two-dimensional particle-in-cell code. We discuss the influence of proton layer initial sizes on the cut-off energy, energy spread, and divergence angle of proton beam. It is found that Coulomb explosion plays an important role on the proton cut-off energy. This causes the cut-off energy to increase for increasing proton layer thickness, at the expense of energy spread. The proton divergence angle reaches a peak value and then falls again with increasing the width. Proton divergence angle grows with target thickness. It is found that there is an optimal thickness to obtain the narrowest energy spread, which may provide an effective method (change the size of proton layer) to obtain high quality proton beams. This work may serve to improve the understanding of sheath field proton acceleration.
- 28 August 2013
- Published by Hindawi Limited in Laser and Particle Beams
- Vol. 31 (4) , 597-605
Influence of the initial size of the proton layer in sheath field proton acceleration
We investigate the influence of the initial size of the proton layer on proton acceleration in the interaction of high intensity laser pulses with double-layer targets by using two-dimensional particle-in-cell code. We discuss the influence of proton layer initial sizes on the cut-off energy, energy spread, and divergence angle of proton beam. It is found that Coulomb explosion plays an important role on the proton cut-off energy. This causes the cut-off energy to increase for increasing proton layer thickness, at the expense of energy spread. The proton divergence angle reaches a peak value and then falls again with increasing the width. Proton divergence angle grows with target thickness. It is found that there is an optimal thickness to obtain the narrowest energy spread, which may provide an effective method (change the size of proton layer) to obtain high quality proton beams. This work may serve to improve the understanding of sheath field proton acceleration.
- Open Access
- 1 January 2013
- Published by Global Science Press in Communications in Computational Physics
- Vol. 13 (4) , 1134-1150
High-Order Interpolation Algorithms for Charge Conservation in Particle-in-Cell Simulations
High-Order Interpolation Algorithms for Charge Conservation in Particle-in-Cell Simulations - Volume 13 Issue 4 - Jinqing Yu, Xiaolin Jin, Weimin Zhou, Bin Li, Yuqiu Gu
- 1 January 2013
- Published by Global Science Press in Communications in Computational Physics
- Vol. 13 (4) , 1134-1150
Open AccessHigh-Order Interpolation Algorithms for Charge Conservation in Particle-in-Cell Simulations
High-Order Interpolation Algorithms for Charge Conservation in Particle-in-Cell Simulations - Volume 13 Issue 4 - Jinqing Yu, Xiaolin Jin, Weimin Zhou, Bin Li, Yuqiu Gu