Neural network classification of metal surface properties using a dynamic touch sensor

Author

Brenner, Dean ; Principe, Jose C. ; Doty, Keith L.

Author_Institution

Dept. of Electr. Eng., Florida Univ., Gainesville, FL, USA

Volume

i

fYear

1991

fDate

8-14 Jul 1991

Firstpage

189

Abstract

Discusses an application of neural networks to classify signals produced by a dynamic touch sensor, and achieve automated characterization of metal surfaces during machining. Data are first preprocessed and the spectral features are input to a feedforward one-hidden-layer neural network, trained with backpropagation. The classification accuracy was over 90% for most of the surfaces. The authors discuss the experimental setup, the preprocessing, and a critical view of the classification results

Keywords

classification; computerised pattern recognition; electric sensing devices; learning systems; machining; mechanical engineering computing; metals; neural nets; surface topography measurement; accuracy; automated characterization; backpropagation; dynamic touch sensor; feedforward one-hidden-layer neural network; machining; metal surface properties; preprocessing; signal classification; spectral features; Backpropagation; Force sensors; Machining; Needles; Neural networks; Rough surfaces; Sensor phenomena and characterization; Surface roughness; Surface texture; Tactile sensors;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks, 1991., IJCNN-91-Seattle International Joint Conference on

Conference_Location

Seattle, WA

Print_ISBN

0-7803-0164-1

Type

conf

DOI

10.1109/IJCNN.1991.155174

Filename

155174