The purpose of this work is the development of a trained artificial neural network for surrogate modeling of the mechanical response of elasto-viscoplastic grain *** this end,a U-Net-based convolutional neural network...
详细信息
The purpose of this work is the development of a trained artificial neural network for surrogate modeling of the mechanical response of elasto-viscoplastic grain *** this end,a U-Net-based convolutional neural network(CNN)is trained using results for the von Mises stress field from the numerical solution of initial-boundary-value problems(IBVPs)for mechanical equilibrium in such microstructures subject to quasi-static uniaxial *** resulting trained CNN(tCNN)accurately reproduces the von Mises stress field about 500 times faster than numerical solutions of the corresponding IBVP based on spectral *** of the tCNN to test cases based on microstructure morphologies and boundary conditions not contained in the training dataset is also investigated and discussed.
MEMS devices suffer from errors in manufacturing processes such as lithography, etching, thinning, etc. These errors cause non-negligible deviations between the actual device performance and the ideal performance obta...
详细信息
ISBN:
(数字)9798350359831
ISBN:
(纸本)9798350359848
MEMS devices suffer from errors in manufacturing processes such as lithography, etching, thinning, etc. These errors cause non-negligible deviations between the actual device performance and the ideal performance obtained from finite element method (FEM) simulation in the design stage. Often, parametric sweeps via FEM simulations at heuristically selected geometrical and material parameter samples are implemented, in order to analyze the corresponding device performance deviations. This analysis is empirical and often fails due to large number of possible design parameter variations caused by manufacturing. Performing parametric sweeps for all possible parameter instances would result in an exponential increase in FEM simulations. In this paper, we propose a novel active learning enhanced deep-learning surrogate model, which replaces the FEM simulation and faithfully predicts the device performance deviations under numerous parameter variances. This provides a powerful tool for predicting device yield at the design stage. Furthermore, the proposed active learning technique could be applied to efficient training of deep learning models for other MEMS design problems.
Machine tools are permanently exposed to complex static, dynamic and thermic loads. This often results in an undesired dis¬placement of the tool center point (TCP), causing errors in the production process and th...
详细信息
Mathematical models of the human heart increasingly play a vital role in understanding the working mechanisms of the heart, both under healthy functioning and during disease. The ultimate aim is to aid medical practit...
详细信息
An increasing amount of the collected data are high-dimensional multi-way arrays (tensors), and it is crucial for efficient learning algorithms to exploit this tensorial structure as much as possible. The ever present...
详细信息
An increasing amount of the collected data are high-dimensional multi-way arrays (tensors), and it is crucial for efficient learning algorithms to exploit this tensorial structure as much as possible. The ever present curse of dimensionality for high dimensional data and the loss of structure when vectorizing the data motivates the use of tailored low-rank tensor classification methods. In the presence of small amounts of training data, kernel methods offer an attractive choice as they provide the possibility for a nonlinear decision boundary. We develop the Tensor Train Multi-way Multi-level Kernel (TT-MMK), which combines the simplicity of the Canonical Polyadic decomposition, the classification power of the Dual Structure-preserving Support Vector Machine, and the reliability of the Tensor Train (TT) approximation. We show by experiments that the TT-MMK method is usually more reliable computationally, less sensitive to tuning parameters, and gives higher prediction accuracy in the SVM classification when benchmarked against other state-of-the-art techniques.
With the recent realization of exascale performace by Oak Ridge National Laboratory’s Frontier supercomputer, reducing communication in kernels like QR factorization has become even more imperative. Low-synchronizati...
详细信息
We construct a neural network model of S-parameters, from which the S-parameters can be quickly pre-dicted. Numerical tests on a filter model show that the proposed method accurately predicts the S-parameters with mul...
详细信息
ISBN:
(数字)9798350351231
ISBN:
(纸本)9798350351248
We construct a neural network model of S-parameters, from which the S-parameters can be quickly pre-dicted. Numerical tests on a filter model show that the proposed method accurately predicts the S-parameters with multiple sharp resonances.
The purpose of the current work is the development of a so-called physics-encoded Fourier neural operator (PeFNO) for surrogate modeling of the quasi-static equilibrium stress field in solids. Rather than accounting f...
详细信息
Interpolatory necessary optimality conditions for H2-optimal reduced-order modeling of unstructured linear time-invariant (LTI) systems are well-known. Based on previous work on L2-optimal reduced-order modeling of st...
详细信息
The tensor t-function, a formalism that generalizes the well-known concept of matrix functions to third-order tensors, is introduced in [K. Lund, The tensor t-function: a definition for functions of third-order tensor...
详细信息
暂无评论