METAMODELLING USING SIMULATION TRAINED ARTIFICIAL NEURAL NETWORK FOR STRUCTURE APPLICATIONS

K. Mahboob; A. Khan; M.N. Khan; Q. Awais; K.A. Khan; T. Azid

No Detail Found!

METAMODELLING USING SIMULATION TRAINED ARTIFICIAL NEURAL NETWORK FOR STRUCTURE APPLICATIONS

Author(s)

K. Mahboob Department of Mechanical Engineering, University of Engineering and Technology Lahore, Pakistan ; Department of Mechanical Engineering, Chenab College of Engineering and Technology Gujranwala, Pakistan

A. Khan Department of Mechanical Engineering, University of Engineering and Technology Lahore, Pakistan

M.N. Khan Department of Electrical Engineering, The University of Lahore,Lahore,Pakistan

Q. Awais Department of Electrical Engineering, Chenab College of Engineering and Technology Gujranwala,Pakistan

K.A. Khan Department of Electrical Engineering, The University of Lahore,Lahore,Pakistan

T. Azid Department of Mechanical Engineering, University of Engineering and Technology Lahore, Pakistan

Abstract

Usage of neural network for metamodeling is quite common in different fields of engineering. Engineering simulations' cost directed us to replace it with some metamodel. In this paper neural network is used for predictions and an optimal ANN structure is predicted and modeled. A truss structure is modeled and simulations are carried out with different variations. Neural network is trained and verified for data derived from non linear truss analysis. It is observed that neural network results are in conformance with the simulation results and simulation result base metamodel was formed successfully.

Publication Details

Page(s) 145-149

DOI DOI not available

Published Journal: Pakistan Journal of Science, Volume: 71, Issue: 4S, Year: 2019

Request the full text from PASTIC

Keywords

Artificial Neural Network, FEM, Non linear Truss structure