TY - JOUR
T1 - A statistical comparison of neuroclassifiers and feature selection methods for gearbox fault diagnosis under realistic conditions
AU - Pacheco, Fannia
AU - Valente de Oliveira, José
AU - Sánchez, René Vinicio
AU - Cerrada, Mariela
AU - Cabrera, Diego
AU - Li, Chuan
AU - Zurita, Grover
AU - Artés, Mariano
N1 - Publisher Copyright:
© 2016 Elsevier B.V.
PY - 2016/6/19
Y1 - 2016/6/19
N2 - Gearboxes are crucial devices in rotating power transmission systems with applications in a variety of industries. Gearbox faults can cause catastrophic physical consequences, long equipment downtimes, and severe production costs. Several artificial neural networks, learning algorithms, and feature selection methods have been used in the diagnosis of the gearbox healthy state. Given a specific gearbox, this study investigates how these approaches compare with each other in terms of the typical fault classification accuracy but also in terms of the area under curve (AUC), where the curve refers to the precision-recall curve otherwise known as receiver operating characteristic (ROC) curve. In particular, the comparison aims at identifying whether there are statistically significant (dis)similarities among six feature selection methods, and seven pairs of neural nets with different learning rules. Genetic algorithm based, entropy based, linear discriminants, principal components, most neighbors first, and non-negative matrix factorization are the studied feature selection methods. Feed forward perceptrons, cascade forward, probabilistic nets, and radial basis function neural nets are evaluated. Six supervised and one unsupervised learning rules are considered. Both parametric and nonparametric statistical tests are employed. A ranking process is defined to elect the best approach, when available. An experimental setup was especially prepared to ensure operating conditions as realistic as possible.
AB - Gearboxes are crucial devices in rotating power transmission systems with applications in a variety of industries. Gearbox faults can cause catastrophic physical consequences, long equipment downtimes, and severe production costs. Several artificial neural networks, learning algorithms, and feature selection methods have been used in the diagnosis of the gearbox healthy state. Given a specific gearbox, this study investigates how these approaches compare with each other in terms of the typical fault classification accuracy but also in terms of the area under curve (AUC), where the curve refers to the precision-recall curve otherwise known as receiver operating characteristic (ROC) curve. In particular, the comparison aims at identifying whether there are statistically significant (dis)similarities among six feature selection methods, and seven pairs of neural nets with different learning rules. Genetic algorithm based, entropy based, linear discriminants, principal components, most neighbors first, and non-negative matrix factorization are the studied feature selection methods. Feed forward perceptrons, cascade forward, probabilistic nets, and radial basis function neural nets are evaluated. Six supervised and one unsupervised learning rules are considered. Both parametric and nonparametric statistical tests are employed. A ranking process is defined to elect the best approach, when available. An experimental setup was especially prepared to ensure operating conditions as realistic as possible.
KW - Classification
KW - Fault diagnosis
KW - Feature selection
KW - Gearbox
KW - Neural networks
KW - Statistic tests
UR - http://www.scopus.com/inward/record.url?scp=84959544325&partnerID=8YFLogxK
U2 - 10.1016/j.neucom.2016.02.028
DO - 10.1016/j.neucom.2016.02.028
M3 - Article
AN - SCOPUS:84959544325
SN - 0925-2312
VL - 194
SP - 192
EP - 206
JO - Neurocomputing
JF - Neurocomputing
ER -