Release Time:2019-03-10  Hits:

Indexed by: Journal Article

Date of Publication: 2009-12-01

Journal: INVERSE PROBLEMS IN SCIENCE AND ENGINEERING

Included Journals: Scopus、EI、SCIE

Volume: 17

Issue: 8

Page Number: 999-1017

ISSN: 1741-5977

Key Words: bimodularity; inverse problem; artificial neural network

Abstract: This article suggests the utilization of artificial neural network to estimate bimodular constitutive parameters, including extensional/compressive moduli, and extensional/compressive Poisson's ratios. By combining a smoothing function with an initial stress scheme, solutions of the direct bimodular problems are provided by finite element (FE) techniques, and are employed as input to train the networks. One- and two-dimensional numerical examples are presented to illustrate the performance of the network, and good results are achieved. Several factors affecting network performance are discussed.