Citation

BibTex format

@article{Yue:2016:10.1142/S1756973716400059,
author = {Yue, N and Sharif, Khodaei Z},
doi = {10.1142/S1756973716400059},
journal = {Journal of Multiscale Modeling},
title = {Assessment of impact detection tchniques for aeronautical application: ANN vs. LSSVM},
url = {http://dx.doi.org/10.1142/S1756973716400059},
volume = {07},
year = {2016}
}

RIS format (EndNote, RefMan)

TY  - JOUR
AB - The Impact localisation in composite panels is assessed using two machinelearning techniques: least square support vector machines (LSSVM) and artificialneural networks (ANN) with local strain signals from piezoelectric sensors. Sensorsignals from impact experiments on a composite plate as well as signals simulated by afinite element model are used to train and test models. A comparative study shows thatLSSVM achieves better accuracy than ANN on identifying location of impacts for acombination of large mass impact and small mass impact, in particular when less datais available for training which is more appropriate for real aeronautical application.Additionally, LSSVM is more capable of identifying new impact events which have notbeen considered in the training process.
AU - Yue,N
AU - Sharif,Khodaei Z
DO - 10.1142/S1756973716400059
PY - 2016///
SN - 1756-9745
TI - Assessment of impact detection tchniques for aeronautical application: ANN vs. LSSVM
T2 - Journal of Multiscale Modeling
UR - http://dx.doi.org/10.1142/S1756973716400059
UR - http://hdl.handle.net/10044/1/39436
VL - 07
ER -

logo