Autonomy in space will put an increasing demand on the development of sophisticated online fault management systems, an important aspect of the sophistication required being the ability to react appropriately to unforeseen anomalies. Generally, this ability is expected of knowledge-based systems (KBS), since principally they need not depend on a premeditation of all possible anomalies, but on knowledge about the functional structure of the monitored system. However, this ability can still not be realized if the encoded knowledge is based on two-valued logic, as is the case with many KBS. Instead, it is shown that a knowledge representation based on multi-valued logic is required. As a practical illustration, the behaviour of an expert system for fault diagnosis in the cooling loop of EURECA is described. In this analysis the expression 'treatment' of unforeseen situations will be understood in the sense of direct response, i.e. assessment and control of unforeseen situations immediately when they occur, on the basis of the knowledge available. The aspect of learning, which could be considered to be another aspect of the treatment of unforeseen situations, will not be addressed in this treatise.
Treatment of unforeseen situations by online knowledge-based diagnostic systems
Behandlung unvorhergesehener Situationen durch ein wissensbasiertes Online-Diagnosesystem
1987
10 Seiten, 5 Bilder, 5 Quellen
Aufsatz (Konferenz)
Englisch
A Plant Control Expert System to Cope with Unforeseen Situations
British Library Online Contents | 1993
|Wiley | 2013
|IET | 2013
|Wiley | 2015