Title :
Improved diagnosis by combining structural and process knowledge
Author :
Christiansen, Lars ; Fay, Alexander ; Opgenoorth, Bernd ; Neidig, Jörg
Author_Institution :
Inst. of Autom. Technol., Helmut-Schmidt-Univ., Hamburg, Germany
Abstract :
This paper presents an approach for using process knowledge in addition to structural knowledge for plant-wide diagnosis. System diagnosis based on fault-detection and fault-diagnosis often uses structural knowledge (e.g. hierarchies of sub-systems and components) to identify the root cause of faults and malfunctions. The structural knowledge describes the internal relationships of a plant and is particularly suitable for diagnosis if an assembly or component breaks down, which affects other parts of the system. However, the root cause of faults cannot always be identified based on structural knowledge alone, but information about the processes (e.g. production processes) is required in addition. Based on an integration of the VDI/VDE-guideline 3682 “Formalized process description” the authors present a method to consider process knowledge for an improved system diagnosis.
Keywords :
fault diagnosis; knowledge management; manufacturing systems; VDI/VDE-guideline 3682; fault-detection; fault-diagnosis; formalized process description; plant-wide diagnosis; process knowledge; structural knowledge; system diagnosis; Assembly; Belts; Fault diagnosis; Guidelines; Object oriented modeling;
Conference_Titel :
Emerging Technologies & Factory Automation (ETFA), 2011 IEEE 16th Conference on
Conference_Location :
Toulouse
Print_ISBN :
978-1-4577-0017-0
Electronic_ISBN :
1946-0740
DOI :
10.1109/ETFA.2011.6059056