Dissertation/Thesis Abstract

Use of Data Fusion and SPC in Vibrations Detection
by Kazadi, Marcel M., M.S., Southern Illinois University at Edwardsville, 2013, 80; 1545819
Abstract (Summary)

There are several techniques used to detect the natural frequencies of rotating bars used in machining processes as vibration causes serious quality issues. These vibrations are most of time measured with a single sensor and often corrupted by noise interfering with useful signal. Also, the features that contain the informative signatures about vibration signals are extracted using low spectrum analysis methods such as Fast Fourier transform that are not good at detecting transients in a machine vibration signals. With the recent advances in sensor and wireless technology, new methods of more direct measurement and control of vibration are now possible. This study focuses on the use of data fusion techniques to more accurately measure and control the oscillation of the bar stock in fore mentioned process. The proposed framework uses an array of sensors such as accelerometers and acoustic sensors in various locations in the process and investigates application of sensor fusion techniques to meaningfully consolidate signals from different sensors. The resulting signal is monitored using data fusion and statistical process control techniques adopted from Quality Control to identify small changes in the vibration. The results that the SPC is able to detect transients in bar feeder vibration signals at low frequencies, and severe vibrations at different natural frequencies for high frequencies. The results show also that SPCs obtained from the data fusion have tight control limits indicating that the noise is not incorporated in the fused signal.

Supplemental Files

Some files may require a special program or browser plug-in. More Information

Indexing (document details)
Advisor: Karacal, Cem S.
Commitee: Ko, Hoo S., Lee, Felix H.
School: Southern Illinois University at Edwardsville
Department: Mechanical and Industrial Engineering
School Location: United States -- Illinois
Source: MAI 52/02M(E), Masters Abstracts International
Source Type: DISSERTATION
Subjects: Statistics, Mechanical engineering
Keywords: Bar feeder, Data fusion, Parametric models, Vibration
Publication Number: 1545819
ISBN: 9781303425943
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest