Automating end-of-line quality control tests of small brushed DC motors

Francisco Rodriguez Verdugo, Enrico Concettoni, Alessandro Bastari, Cristina Cristalli, Maurizio de Palma

Abstract

Robust end-of-line tests are important in manufacturing to monitor production quality and to guarantee that only fault-free products are delivered to customers. By automating quality control tests, subjectivity can be removed from this important process and traceability of defects can be improved. This paper describes a prototype station that automatically tests and classifies small brushed DC motors. In order to perform fault diagnosis, four different quantities are monitored: acoustic noise, mechanical vibration, voltage and current. A three-step approach consisting in a feature extraction phase, a feature space dimensional reduction through multivariate analysis and a final neural network classifier is proposed and preliminary results are in good agreement with experts’ assessments.

Download

IMEKO-TC10-2013-025.pdf

DOI

-

IMEKO TC

TC10 - Measurement for Diagnostics, Optimization and Control

Back to the proceedings