Options
1993
Conference Paper
Titel
Akzeleratorunterstützte Fuzzy-Klassifikation zur Echtzeit-Prozeßüberwachung
Abstract
Monitoring of processes can not be described by a complete and exact mathematic model in any case. These supervision tasks can be simply modelled via fuzzy pattern classification. The approach bases on appropriate features embodying mediate or immediate information on the process status to be monitored. Applying fuzzy sets uncertainty in the model structure as well as in the measuring signals to be analysed can be considered. An accelerator system basing on programable gate arrays is used for real-time monitoring. The system introduced enables a rapid reaction of the on-line decision support system. Ranges of application and examples of utilization are briefly presented.