CORDIS - Wyniki badań wspieranych przez UE
CORDIS
Zawartość zarchiwizowana w dniu 2024-04-16

An Intelligent Real-Time System for Signal Understanding

Cel

The objective was to design and develop AITRAS, a knowledge-based system for signal understanding. AITRAS comprised a system shell and the design tools for building systems for the real-time analysis and interpretation of instrumentation signals.
The objective of the project is to design and develop an intelligent real time system for signal understanding (AITRAS) which is knowledge based. AITRAS comprises a system shell and the design tools for building systems for the real time analysis and interpretation of instrumentation signals.
The intention was to design AITRAS around a real time inference engine (RTIE), tightly coupled to a signal processing subsystem, and to design and develop an architecture for real time environments.

The workplan for each application began with the elicitation of expert knowledge and the creation of the system specification. The demonstration of the AITRAS prototype followed. Static and dynamic validations were then made and the performance refined within an operational environment. Validation of the generic AITRAS system is carried out on 3 industrial applications:
the nondestructive examination of steam generator tubes in nucelar plants by the analysis of eddy current signals;
the use of the same technique for heat exchangers and condensers;
the fault prediction by the analysis of vibration signals for the maintenance of bearings.
Though sophisticated real-time techniques have been applied successfully in this field, for several reasons their success has been limited. The huge amount of data to be processed slows the response; the experts needed to interpret the results are in short supply; and even the use of knowledge-based systems to emulate expert reasoning does not reduce the delays sufficiently.

In attempting to reduce these shortcomings, the intention was to design AITRAS around a Real-Time Inference Engine (RTIE), tightly coupled to a signal processing subsystem, and to design and develop an architecture for real-time environments.

Temat(-y)

Data not available

Zaproszenie do składania wniosków

Data not available

System finansowania

Data not available

Koordynator

Framentec Cognitech SA
Wkład UE
Brak danych
Adres
Tour Fiat 1 place de la Coupole
92084 Paris la Défense
Francja

Zobacz na mapie

Koszt całkowity
Brak danych

Uczestnicy (4)