Title :
Recognition and detection methods in supervision and control of the manufacturing processes
Author_Institution :
Inst. of Math. & Inf., Acad. of Sci., Vilnius, Lithuania
Abstract :
The paper considers a new approach to the supervision and control of the manufacturing processes. It treats supervision and control of the manufacturing systems as a problem of discrete status, operating conditions and events of dynamic objects recognition. The software of computer aided systems for the manufacturing systems supervision includes tools for output signals clustering, dynamic models of discrete status identification, parameters of dynamic models for each status estimation, observable realisation classification and decision making. A method of diagnostics of the electric engine status based on the analysis of process realisation that takes place in the engine is described
Keywords :
fault diagnosis; manufacturing processes; object recognition; parameter estimation; pattern classification; production control; fault detection; fault diagnosis; identification; manufacturing processes; objects recognition; parameter estimation; pattern classification; signals clustering; supervision; Bayesian methods; Control systems; Engines; Equations; Informatics; Manufacturing processes; Manufacturing systems; Mathematics; Object recognition; Random processes;
Conference_Titel :
American Control Conference, 2000. Proceedings of the 2000
Conference_Location :
Chicago, IL
Print_ISBN :
0-7803-5519-9
DOI :
10.1109/ACC.2000.879565
