Résumé: With rapid innovations in information technology and sensing technology, increasingly less expensive and smart devices with multiple heterogeneous on-board sensors, networked through wireless links and deployable in large numbers, are distributed throughout complex Multistage Manufacturing Systems (MMSs). These technologies provide unprecedented opportunities for quality improvement. If product-sensing data are obtained via certain distributed sensing and information system, the problem of whether the faults of a manufacturing system are diagnosable is of great interest to both academia and industry. In this study, the diagnosability of the process faults in a MMS is defined in a general way using a linear input-output model, which does not depend on specific diagnosis algorithms. The condition of faults diagnosability, the diagnosability matrix and indices are defined and derived. Finally, the methodology is illustrated by a machining process and a hot deformation process.

Langue: Anglais