Fixture fault diagnosis is a critical component of currently evolving techniques aimed at manufacturing variation reduction. The impact of sensor location on the effectiveness of fault-type discrimination in such diagnostic procedures is significant. This paper proposes a methodology for achieving optimal fault-type discrimination through an optimized configuration of defined 'sensor locales'. The optimization is presented in the context of autobody fixturing - a predominant cause of process variability in automobile assembly. The evaluation criteria is the ability to provide a consistency of best match, in a pattern recognition sense, of any existing fixture error to a classified, anticipated error set. The proposed analytical methodology is novel in addressing the design issue at both the fixture design and the sensor location planning levels - constituting a design for fault detectability. Examples of the locale planning and an application to an industrial fixture configuration are presented to illustrate the proposed methodology.
An optimal sensor location methodology for fixture fault diagnosis
Methodologie zur optimalen Sensorpositionierung für die Vorrichtungsfehlerdiagnose
1995
12 Seiten, 10 Bilder, 3 Tabellen, 17 Quellen
Conference paper
English
Optimal sensor placement for fixture fault diagnosis using Bayesian network
Tema Archive | 2011
|Optimal sensor location for robust fault detection filter
Online Contents | 1999
|Optimal sensor location for robust fault detection filter
SAGE Publications | 1999
|