Deep Neural Network Machine Learning for Oscillatory Navier-Stokes Flows and Nonlinear Operators, and High Dimensional Fokker-Planck Equations

用于振荡纳维-斯托克斯流和非线性算子以及高维福克-普朗克方程的深度神经网络机器学习

基本信息

  • 批准号:
    2207449
  • 负责人:
  • 金额:
    $ 34.88万
  • 依托单位:
  • 依托单位国家:
    美国
  • 项目类别:
    Standard Grant
  • 财政年份:
    2022
  • 资助国家:
    美国
  • 起止时间:
    2022-08-01 至 2025-07-31
  • 项目状态:
    未结题

项目摘要

The goal of this project is to develop machine learning (ML) technology with deep neural networks (DNNs) to study a wide range of physical and chemical phenomena. DNN has demonstrated its tremendous power in artificial intelligence (AI) applications such as speech recognition, automatic self-driving, image classification, etc. This research will attempt to harness the power of the DNN for scientific discoveries by increasing computational and simulation capabilities in learning and understanding complex phenomena such as fluid flows for ship building applications, seismic wave predictions, chemical reactions in drug designs. In this project, to address some of the key computational challenges in scientific computing, new classes of deep neural network (DNN) ML algorithms will be developed with capabilities for better frequency resolution for simulating oscillatory fluid flows in complex geometries and nonlinear operators in oscillatory function spaces, and increased powers for addressing the curse of dimensionality (CoD) in solving high dimensional Fokker-Planck equations (FPE) from transition path theory (TPT) of complex biochemical systems. Specifically, research will be carried out in three major computational issues relevant to scientific and engineering computing: (a) To develop multiscale DNNs as a viable meshless method for solving time dependent highly oscillatory Navier-Stokes flows in complex domains, and a practical alternative method to traditional numerical methods with no costly mesh generation or linear system solvers. (b) To develop multiscale DNN learning algorithms for nonlinear operators in highly oscillatory function spaces for seismic wave prediction, and forward and inverse scattering applications. (c) To develop forward and backward stochastic differential equation (FBSDE) based multiscale DNNs for boundary value problems of high dimensional PDEs such as Fokker-Planck equations arising from statistical description of physical systems with application in computing committor functions and transition rates in transition path sampling theory of complex chemical and biological systems. This research will have a broad impact in improving the capability of ML for scientific computing in many ways. Efficient meshless DNN algorithms for oscillatory flows in complex geometry and representation of nonlinear operator for physical qualities of highly oscillatory nature can reduce the cost of many optimization, and forward and inverse problems in scientific and engineering research. Overcoming the CoD in solving high dimensional Fokker-Planck equations with the multi-scale FBSDE DNN algorithms will have a wide range impact on both the ML research and many scientific areas such as material sciences and optimal control, financial engineering, as well as to biological system dynamics. Moreover, the project will place strong emphasis on data science education with new ML course development and activities in the Data Science Institute at SMU.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
该项目的目标是开发具有深度神经网络(DNN)的机器学习(ML)技术来研究广泛的物理和化学现象。 DNN 已在语音识别、自动驾驶、图像分类等人工智能 (AI) 应用中展示了其巨大的力量。本研究将尝试通过提高学习和理解复杂现象(例如造船应用中的流体流动、地震波预测、药物设计中的化学反应)的计算和模拟能力,利用 DNN 的力量进行科学发现。在该项目中,为了解决科学计算中的一些关键计算挑战,将开发新型深度神经网络 (DNN) ML 算法,该算法具有更好的频率分辨率,可以模拟复杂几何形状中的振荡流体流动和振荡函数空间中的非线性算子,并增强解决高维 Fokker-Planck 方程中维数灾难 (CoD) 的能力 (FPE)来自复杂生化系统的过渡路径理论(TPT)。具体来说,研究将针对与科学和工程计算相关的三个主要计算问题进行:(a)开发多尺度DNN作为一种可行的无网格方法,用于解决复杂领域中时间依赖性高振荡纳维-斯托克斯流,以及传统数值方法的实用替代方法,无需昂贵的网格生成或线性系统求解器。 (b) 开发用于地震波预测以及前向和逆向散射应用的高振荡函数空间中的非线性算子的多尺度 DNN 学习算法。 (c) 开发基于前向和后向随机微分方程(FBSDE)的多尺度 DNN,用于解决高维偏微分方程的边值问题,例如由物理系统统计描述产生的 Fokker-Planck 方程,并应用于计算复杂化学和生物系统转移路径采样理论中的提交函数和转移率。这项研究将以多种方式对提高机器学习科学计算能力产生广泛影响。针对复杂几何中的振荡流的高效无网格 DNN 算法以及针对高振荡性质的物理性质的非线性算子表示可以降低科学和工程研究中许多优化以及正向和逆向问题的成本。使用多尺度 FBSDE DNN 算法克服求解高维 Fokker-Planck 方程时的 CoD 将对机器学习研究和材料科学和最优控制、金融工程以及生物系统动力学等许多科学领域产生广泛的影响。此外,该项目将重点强调数据科学教育,以及 SMU 数据科学研究所的新 ML 课程开发和活动。该奖项反映了 NSF 的法定使命,并通过使用基金会的智力价值和更广泛的影响审查标准进行评估,被认为值得支持。

项目成果

期刊论文数量(2)
专著数量(0)
科研奖励数量(0)
会议论文数量(0)
专利数量(0)
FBSDE based neural network algorithms for high-dimensional quasilinear parabolic PDEs
基于 FBSDE 的高维拟线性抛物线偏微分方程神经网络算法
  • DOI:
    10.1016/j.jcp.2022.111557
  • 发表时间:
    2022
  • 期刊:
  • 影响因子:
    4.1
  • 作者:
    Zhang, Wenzhong;Cai, Wei
  • 通讯作者:
    Cai, Wei
Linearized Learning with Multiscale Deep Neural Networks for Stationary Navier-Stokes Equations with Oscillatory Solutions
使用多尺度深度神经网络进行线性学习,用于具有振荡解的稳态纳维-斯托克斯方程
{{ item.title }}
{{ item.translation_title }}
  • DOI:
    {{ item.doi }}
  • 发表时间:
    {{ item.publish_year }}
  • 期刊:
  • 影响因子:
    {{ item.factor }}
  • 作者:
    {{ item.authors }}
  • 通讯作者:
    {{ item.author }}

数据更新时间:{{ journalArticles.updateTime }}

{{ item.title }}
  • 作者:
    {{ item.author }}

数据更新时间:{{ monograph.updateTime }}

{{ item.title }}
  • 作者:
    {{ item.author }}

数据更新时间:{{ sciAawards.updateTime }}

{{ item.title }}
  • 作者:
    {{ item.author }}

数据更新时间:{{ conferencePapers.updateTime }}

{{ item.title }}
  • 作者:
    {{ item.author }}

数据更新时间:{{ patent.updateTime }}

Wei Cai其他文献

GW29-e1902 Branch Ostial Optimization Treatment and Optimized Provisional T-Stenting with Polymeric Bioresorbable Scaffolds: Ex Vivo Morphologic and Hemodynamic Examination
GW29-e1902 分支口优化治疗和使用聚合物生物可吸收支架优化临时 T 形支架:离体形态和血流动力学检查
Transcriptome profiling analysis of sex-based differentially expressed mRNAs and lncRNAs in the brains of mature zebrafish (Danio rerio)
成熟斑马鱼 (Danio rerio) 大脑中基于性别的差异表达 mRNA 和 lncRNA 的转录组分析
  • DOI:
    10.1186/s12864-019-6197-9
  • 发表时间:
    2019-07
  • 期刊:
  • 影响因子:
    4.4
  • 作者:
    Yuan Wenliang;Jiang Shouwen;Sun Dan;Wu Zhichao;Wei Cai;Dai Chaoxu;Jiang Linhua;Peng Sihua
  • 通讯作者:
    Peng Sihua
Understanding and manipulating the intrinsic point defect in α-MgAgSb for higher thermoelectric performance
了解和操纵α-MgAgSb 中的固有点缺陷以获得更高的热电性能
Influence of Particle Size on the Spin Pinning Effect in the fcc-FePt Nanoparticles
粒径对 fcc-FePt 纳米粒子自旋钉扎效应的影响
  • DOI:
    10.1007/s10948-019-5091-7
  • 发表时间:
    2019-04
  • 期刊:
  • 影响因子:
    1.8
  • 作者:
    Jing Yu;Dong Han;Yao Ying;Liang Qiao;Jingwu Zheng;Wangchang Li;Juan Li;Wei Cai;Shenglei Che;Naoki Wakiya;Hisao Suzuki
  • 通讯作者:
    Hisao Suzuki
Soundprint Feature Analysis of Main Transformers in a 500kV Substation
500kV变电站主变压器声纹特征分析

Wei Cai的其他文献

{{ item.title }}
{{ item.translation_title }}
  • DOI:
    {{ item.doi }}
  • 发表时间:
    {{ item.publish_year }}
  • 期刊:
  • 影响因子:
    {{ item.factor }}
  • 作者:
    {{ item.authors }}
  • 通讯作者:
    {{ item.author }}

{{ truncateString('Wei Cai', 18)}}的其他基金

Collaborative Research: DMREF: Developing Damage Resistant Materials for Hydrogen Storage and Large-scale Transport
合作研究:DMREF:开发用于储氢和大规模运输的抗损伤材料
  • 批准号:
    2118522
  • 财政年份:
    2021
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Continuing Grant
Collaborative Research: Multi-Scale Modeling and Numerical Methods for Charge Transport in Ion Channels
合作研究:离子通道中电荷传输的多尺度建模和数值方法
  • 批准号:
    1950471
  • 财政年份:
    2020
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Continuing Grant
High Order and Efficient Numerical Methods for Simulating Electromagnetic Phenomena
模拟电磁现象的高阶高效数值方法
  • 批准号:
    1802143
  • 财政年份:
    2017
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Standard Grant
Path Integral Monte Carlo Methods for Computing Polarizability Tensors of Nano-materials and Electrical Impedance Tomography
计算纳米材料极化张量和电阻抗断层扫描的路径积分蒙特卡罗方法
  • 批准号:
    1719303
  • 财政年份:
    2017
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Standard Grant
Path Integral Monte Carlo Methods for Computing Polarizability Tensors of Nano-materials and Electrical Impedance Tomography
计算纳米材料极化张量和电阻抗断层扫描的路径积分蒙特卡罗方法
  • 批准号:
    1764187
  • 财政年份:
    2017
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Standard Grant
High Order and Efficient Numerical Methods for Simulating Electromagnetic Phenomena
模拟电磁现象的高阶高效数值方法
  • 批准号:
    1619713
  • 财政年份:
    2016
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Standard Grant
Student Travel: 7th International Conference on Multiscale Materials Modeling; Berkeley, California; 6-10 October 2014
学生旅行:第七届多尺度材料建模国际会议;
  • 批准号:
    1444609
  • 财政年份:
    2014
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Standard Grant
A parallel Poisson/Helmholtz solver using local boundary integral equation and random walk methods
使用局部边界积分方程和随机游走方法的并行泊松/亥姆霍兹求解器
  • 批准号:
    1315128
  • 财政年份:
    2013
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Standard Grant
Structural Transitions during Catalyzed Growth of Semiconductor Nanowires
半导体纳米线催化生长过程中的结构转变
  • 批准号:
    1206511
  • 财政年份:
    2012
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Continuing Grant
Numerical Methods for Wave Propagations in Inhomogeneous Media
非均匀介质中波传播的数值方法
  • 批准号:
    1005441
  • 财政年份:
    2010
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Standard Grant

相似国自然基金

Neural Process模型的多样化高保真技术研究
  • 批准号:
    62306326
  • 批准年份:
    2023
  • 资助金额:
    30 万元
  • 项目类别:
    青年科学基金项目

相似海外基金

CRII: RI: Deep neural network pruning for fast and reliable visual detection in self-driving vehicles
CRII:RI:深度神经网络修剪,用于自动驾驶车辆中快速可靠的视觉检测
  • 批准号:
    2412285
  • 财政年份:
    2024
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Standard Grant
Development of data-driven multiple sound spot synthesis technology based on deep generative neural network models
基于深度生成神经网络模型的数据驱动多声点合成技术开发
  • 批准号:
    23K11177
  • 财政年份:
    2023
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Grant-in-Aid for Scientific Research (C)
CISE-ANR: Small: Evolutional deep neural network for resolution of high-dimensional partial differential equations
CISE-ANR:小型:用于求解高维偏微分方程的进化深度神经网络
  • 批准号:
    2214925
  • 财政年份:
    2023
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Standard Grant
Elements: A Deep Neural Network-based Drone (UAS) Sensing System for 3D Crop Structure Assessment
Elements:用于 3D 作物结构评估的基于深度神经网络的无人机 (UAS) 传感系统
  • 批准号:
    2334690
  • 财政年份:
    2023
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Standard Grant
CAREER: DeepMatter: A Scalable and Programmable Embedded Deep Neural Network
职业:DeepMatter:可扩展且可编程的嵌入式深度神经网络
  • 批准号:
    2348983
  • 财政年份:
    2023
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Continuing Grant
Construction of a deep graph neural network that prevents over-smoothing
构建防止过度平滑的深度图神经网络
  • 批准号:
    23K11241
  • 财政年份:
    2023
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Grant-in-Aid for Scientific Research (C)
Safe Lyapunov-Based Deep Neural Network Adaptive Control of a Rehabilitative Upper Extremity Hybrid Exoskeleton
基于安全李亚普诺夫的深度神经网络自适应控制康复上肢混合外骨骼
  • 批准号:
    2230971
  • 财政年份:
    2023
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Standard Grant
Provable, Explainable, Efficient, Robust Deep Neural Network for Large-scale Multimedia Retrieval
用于大规模多媒体检索的可证明、可解释、高效、鲁棒的深度神经网络
  • 批准号:
    22KF0369
  • 财政年份:
    2023
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Grant-in-Aid for JSPS Fellows
Development of Deep Learning Methods for Generating and Analyzing Neural Network Microscopy Images
用于生成和分析神经网络显微图像的深度学习方法的开发
  • 批准号:
    23KF0296
  • 财政年份:
    2023
  • 资助金额:
    $ 34.88万
  • 项目类别:
    Grant-in-Aid for JSPS Fellows
Cancer Emulation Analysis with Deep Neural Network
使用深度神经网络进行癌症仿真分析
  • 批准号:
    10725293
  • 财政年份:
    2023
  • 资助金额:
    $ 34.88万
  • 项目类别:
{{ showInfoDetail.title }}

作者:{{ showInfoDetail.author }}

知道了