On the choice of hyper-parameters of artificial neural networks for stabilized finite element schemes

2021 ◽  
Author(s):  
Subodh M. Joshi ◽  
Thivin Anandh ◽  
Bhanu Teja ◽  
Sashikumaar Ganesan
Keyword(s):  
Neural Networks ◽  
Finite Element ◽  
Artificial Neural Networks ◽  
Stabilized Finite Element ◽  
Artificial Neural ◽  
Finite Element Schemes

scholarly journals A Machine Learning Approach as a Surrogate for a Finite Element Analysis: Status of Research and Application to One Dimensional Systems

Sensors ◽  
2021 ◽  
Vol 21 (5) ◽  
pp. 1654
Author(s):  
Poojitha Vurtur Badarinath ◽  
Maria Chierichetti ◽  
Fatemeh Davoudi Kakhki
Keyword(s):  
Machine Learning ◽  
Neural Networks ◽  
Finite Element Analysis ◽  
Finite Element ◽  
Artificial Neural Networks ◽  
Stress Distribution ◽  
Maintenance Scheduling ◽  
Element Analysis ◽  
One Dimensional ◽  
Artificial Neural

Current maintenance intervals of mechanical systems are scheduled a priori based on the life of the system, resulting in expensive maintenance scheduling, and often undermining the safety of passengers. Going forward, the actual usage of a vehicle will be used to predict stresses in its structure, and therefore, to define a specific maintenance scheduling. Machine learning (ML) algorithms can be used to map a reduced set of data coming from real-time measurements of a structure into a detailed/high-fidelity finite element analysis (FEA) model of the same system. As a result, the FEA-based ML approach will directly estimate the stress distribution over the entire system during operations, thus improving the ability to define ad-hoc, safe, and efficient maintenance procedures. The paper initially presents a review of the current state-of-the-art of ML methods applied to finite elements. A surrogate finite element approach based on ML algorithms is also proposed to estimate the time-varying response of a one-dimensional beam. Several ML regression models, such as decision trees and artificial neural networks, have been developed, and their performance is compared for direct estimation of the stress distribution over a beam structure. The surrogate finite element models based on ML algorithms are able to estimate the response of the beam accurately, with artificial neural networks providing more accurate results.

scholarly journals Optimization of a thin-walled element geometry using a system integrating neural networks and finite element method

2017 ◽  
Vol 62 (1) ◽  
pp. 435-442 ◽  
Author(s):  
P. Golewski ◽  
J. Gajewski ◽  
T. Sadowski
Keyword(s):  
Neural Networks ◽  
Finite Element Method ◽  
Finite Element ◽  
Artificial Neural Networks ◽  
Integrated System ◽  
Multilayer Perceptrons ◽  
Optimization Of Geometry ◽  
Thin Walled ◽  
Artificial Neural ◽  
Element Method

Abstract Artificial neural networks [ANNs] are an effective method for predicting and classifying variables. This article presents the application of an integrated system based on artificial neural networks and calculations by the finite element method [FEM] for the optimization of geometry of a thin-walled element of an air structure. To ensure optimal structure, the structure’s geometry was modified by creating side holes and ribs, also with holes. The main criterion of optimization was to reduce the structure’s weight at the lowest possible deformation of the tested object. The numerical tests concerned a fragment of an elevator used in the “Bryza” aircraft. The tests were conducted for networks with radial basis functions [RBF] and multilayer perceptrons [MLP]. The calculations described in the paper are an attempt at testing the FEM - ANN system with respect to design optimization.

Damage quantification in foam core sandwich composites via finite element model updating and artificial neural networks

2020 ◽  
Vol 234 (21) ◽  
pp. 4288-4304
Author(s):  
Ali Mardanshahi ◽  
Masoud Mardanshahi ◽  
Ahmad Izadi
Keyword(s):  
Neural Networks ◽  
Finite Element ◽  
Artificial Neural Networks ◽  
Finite Element Model ◽  
Model Updating ◽  
Sandwich Structures ◽  
Element Model ◽  
Foam Core ◽  
Damage Quantification ◽  
Artificial Neural

The main idea of this paper is to propose a nondestructive evaluation (NDE) system for two types of damages, core cracking and skin/core debonding, in fiberglass/foam core sandwich structures based on the inverse eigensensitivity-based finite element model updating using the modal test results, and the artificial neural networks. First, the modal testing was conducted on the fabricated fiberglass/foam core sandwich specimens, in the intact and damaged states, and the natural frequencies were extracted. Finite element modeling and inverse eigensensitivity-based model updating of the intact and damaged sandwich structures were conducted and the parameters of the models were identified. Afterward, the updated finite element models were employed to generate a large dataset of the first five harmonic frequencies of the damaged sandwich structures with different damage sizes and locations. This dataset was adopted to train the machine learning models for detection, localization, and size estimation of the core cracking and skin/core debonding damages. A multilayer perceptron neural network classification model was used for detection of types of damages and also a multilayer perceptron neural network regression model was fitted to the dataset for automatically estimation of the locations and dimensions of damages. This intelligent system of damage quantification was also used to make predictions on real damaged specimens not seen by the system. The results indicated that the extracted natural frequencies from the accurate finite element model, in coordination with the experimental data, and using the artificial neural networks can provide an effective system for nondestructive evaluation of foam core sandwich structures.

scholarly journals Prediction of uniaxial tensile flow using finite element-based indentation and optimized artificial neural networks

2020 ◽  
Vol 196 ◽  
pp. 109104 ◽  
Author(s):  
Kyeongjae Jeong ◽  
Hyukjae Lee ◽  
Oh Min Kwon ◽  
Jinwook Jung ◽  
Dongil Kwon ◽  
...  
Keyword(s):  
Neural Networks ◽  
Finite Element ◽  
Artificial Neural Networks ◽  
Uniaxial Tensile ◽  
Artificial Neural

Model Updating of a Steel Truss Based on Artificial Neural Networks

2011 ◽  
Vol 121-126 ◽  
pp. 1363-1366
Author(s):  
Shi Lei Zhang ◽  
Shao Feng Chen ◽  
Huan Ding Wang ◽  
Wei Wang
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Finite Element ◽  
Artificial Neural Networks ◽  
Model Updating ◽  
Back Propagation ◽  
Element Model ◽  
Steel Truss ◽  
Artificial Neural ◽  
The Finite Element Model

Based on the artificial neural network, the parameters of a steel truss are identified. And the finite element model of truss is corrected. In order to improve the efficiency of model updating by artificial neural networks, the momentum is introduced into the back propagation algorithm. Based on the theory of probability and mathematical statistics, the expectation confidence interval of the measured deflections and strains is obtained. In this way, the samples to train the neural network are optimized. The numerical results show that the back propagation neural network proposed on this paper is able to correct the finite element model of the truss effectively.

Determination of Internal Forces Using Artificial Neural Networks

2017 ◽  
Vol 21 ◽  
pp. 151-157
Author(s):  
Alexandrina Elena Pandelea ◽  
Mihai Budescu ◽  
Lucian Soveja ◽  
Maria Solonaru
Keyword(s):  
Neural Networks ◽  
Finite Element ◽  
Artificial Neural Networks ◽  
Computer Program ◽  
Shear Force ◽  
Engineering Structures ◽  
Internal Forces ◽  
Artificial Neural ◽  
Construction Works

Design and verification of engineering structures require knowing the numerical values ​​of sectional internal forces as close to reality, considering that the intervention construction works are correlated with these values.Most of the computer programs are working with finite element method, which was designed by engineers and founded by mathematicians. After running the computer program, stresses and deformations maps are generated as results.Considering these results, using artificial neural networks, a computer program has been designed, which is able to determine internal forces of a section, namely axial force, shear force and bending moment.Neural network input parameters consist of color maps resulted from numerical modeling, numerical values ​​of the normal and tangential tensions and dimensions of the structural element.This procedure is particularly useful when using finite element programs that do not have the ability to determine sectional internal forces.

scholarly journals A Review of Finite Element Analysis and Artificial Neural Networks as Failure Pressure Prediction Tools for Corroded Pipelines

Materials ◽  
2021 ◽  
Vol 14 (20) ◽  
pp. 6135
Author(s):  
Suria Devi Vijaya Kumar ◽  
Michael Lo Yin Kai ◽  
Thibankumar Arumugam ◽  
Saravanan Karuppanan
Keyword(s):  
Neural Networks ◽  
Finite Element ◽  
Artificial Neural Networks ◽  
Structural Integrity ◽  
The Finite Element Method ◽  
Element Analysis ◽  
Strength Assessment ◽  
Failure Pressure ◽  
Prediction Tools ◽  
Artificial Neural

This paper discusses the capabilities of artificial neural networks (ANNs) when integrated with the finite element method (FEM) and utilized as prediction tools to predict the failure pressure of corroded pipelines. The use of conventional residual strength assessment methods has proven to produce predictions that are conservative, and this, in turn, costs companies by leading to premature maintenance and replacement. ANNs and FEM have proven to be strong failure pressure prediction tools, and they are being utilized to replace the time-consuming methods and conventional codes. FEM is widely used to evaluate the structural integrity of corroded pipelines, and the integration of ANNs into this process greatly reduces the time taken to obtain accurate results.

Sign in / Sign up

Export Citation Format

Share Document