G. Fiol-Roig (Spain)
Intelligent supervision of dynamic processes, intelligent control, evolutionary databases, adaptive control Systems, real time supervision.
Dynamic processes are characterized by their evolutionary behaviour over time, defining a sequence of operation states of the system. Disturbance Control constitutes one of the main handicaps when trying to lead the system behaviour to an optimal operation state. Knowledge-Based Real Time Supervision arises as an attempt to guarantee and justify the quality of the system response, facing any situation which the system may fall into. However, retrieving and processing useful knowledge from the control system is a complex task, since a lot of raw data coming from sensors must be analized and updated constantly. Three aspects are closely connected with a Knowledge Based Real Time Supervisor: time constraints imposed to the supervision process in developping its tasks, temporal updating of knowledge about the dynamic system and generation of qualitative knowledge from quantitative data coming from the sensors of the system. The main features of the functional components characterizing a Knowledge-Based Architecture for Real Time Supervision are described in this work, particularly those to do with the storage, updating and generation of qualitative knowledge about the system, the diagnosis of disturbances and the decision-making process.
Important Links:
Go Back
