USING MULTIVARIATE LATENT MODELS TO MONITOR A PRINTING MACHINE AND PREDICT MACHINE FAILURE | ||||
The International Conference on Applied Mechanics and Mechanical Engineering | ||||
Article 19, Volume 13, 13th International Conference on Applied Mechanics and Mechanical Engineering., May 2008, Page 49-62 PDF (551.41 K) | ||||
Document Type: Original Article | ||||
DOI: 10.21608/amme.2008.38997 | ||||
View on SCiNiTO | ||||
Authors | ||||
SAID E. M.; HUSSEIN W. M.; SALEM A. M.; MOHAMED K. I. | ||||
Egyptian Armed Forces. | ||||
Abstract | ||||
ABSTRACT Condition-Based Monitoring involves continuous collection and interpretation of data relating to the operating conditions of critical components of the machine. To ensure high printing quality while maximizing productivity, an on-line process monitoring system is required to take the place of an expert’s judgment. This paper outlines the use of a statistical multivariate technique called Principal Component Analysis (PCA), as a health monitoring technique, and applies it to monitor a GTO type of printing machine. This approach is used to integrate vibration data taken at different positions and directions on the printing machine. Experiments were conducted on the machine for different operating speeds under two conditions, new and worn drive belt. The results showed that the proposed technique can be used for printing machine monitoring and can successfully differentiate between new process conditions. | ||||
Keywords | ||||
Printing monitoring; Multivariate Analysis; Principal component analysis; Vibration monitoring | ||||
Statistics Article View: 114 PDF Download: 185 |
||||