Patent
1995-08-17
1998-09-08
Hafiz, Tariq R.
395 50, 395 60, G06F 1518
Patent
active
058057765
DESCRIPTION:
BRIEF SUMMARY
BACKGROUND OF THE INVENTION
Whenever a technical system fails to carry out its standard function, or to do so satisfactorily, a technical diagnosis of this system is required. The technical diagnosis comprises ascertaining the cause of the observed malfunction and then carrying out a remedial measure, which eliminates the cause. For complex technical systems, a diagnosis requires the knowledge of an experienced human expert. Human experts, however, are expensive and often unavailable. That is why a way must be found to make their knowledge available without having to enlist the human experts themselves.
A modern way of attaining this object is to furnish an expert system for technical diagnosis. Efficient expert systems are able to decisively increase the availability of technical systems and to decrease the cost of their use.
Today an expert system is typically constructed as a knowledge-based system, i.e. as a software system in which the knowledge of the human expert is stored structurally in a knowledge base. The remaining components of the expert system are logically separated from the knowledge base and can therefore be reused; all together, they are called the expert system shell. The elements of the expert system shell are the problem solving component, which carries out a diagnosis and for this utilizes the knowledge base, the explanation component, which optionally explains to the user the actions of the expert system; and the user interface. A knowledge-based system is produced by choosing an expert system shell and filling the knowledge base with the structured knowledge of the human expert.
Diagnostic expert systems are classified as symptom-based and model-based systems: experiential knowledge of the human expert about the malfunctions of the technical system and about the interrelationships between these malfunctions and the symptoms which can be determined during a diagnosis. a model, which describes the structure and the functional cooperation of the modules of the technical system to be diagnosed, and on the other hand a description of the standard behavior of all of the modules and of their behavior in the event of a malfunction.
A model-based expert system can only be built if one can build a precise enough model for the structure and the functional cooperation of the modules of the technical system to be diagnosed. In many applications, this is too expensive or totally impossible, since one cannot describe the inner life of the technical system precisely enough. In this case, only the symptom-based approach remains.
The classic and often-used approach for representing the experiential knowledge of an expert in a formalized way is to structure the knowledge in rules in the form of "IF premise, THEN conclusion". The premise of each of these rules is an individual symptom or a logical concatenation of symptoms, which can be observed in the technical system; the conclusion is a statement about which malfunctions can or must exist upon fulfillment of the premise and which have to be ruled out. This structuring is offered for example by "EXSYS" and "NexpertOBJECT", two tools for building expert systems.
A further often-used approach is the error search tree. The knowledge base contains an error tree, that is a directional graph, whose nodes stand for the malfunctions possible in the technical system and whose edges stand for the causal interrelationships among them. The leaves stand for the possible causes. Furthermore, the knowledge base contains interrelationships between malfunctions and symptoms. The expert system "navigates" in this error search tree with the aim of arriving at a leaf. To do so, it analyzes each symptom produced in order to go from one node to another. This kind of process uses the "testbench" expert system shell, for example.
In building symptom-based expert systems, knowledge acquisition has turned out to be a stumbling block; knowledge acquisition is the series of events by means of which the experiential knowledge of a human diagnostic expert is gained by interroga
REFERENCES:
patent: 4766595 (1988-08-01), Gollomp
patent: 4866635 (1989-09-01), Kahn et al.
patent: 4965743 (1990-10-01), Malin et al.
patent: 5099436 (1992-03-01), McCown et al.
patent: 5119318 (1992-06-01), Paradies et al.
patent: 5121496 (1992-06-01), Harper
patent: 5212768 (1993-05-01), Itsuki et al.
patent: 5241623 (1993-08-01), Shah
patent: 5303332 (1994-04-01), Kirk et al.
patent: 5386498 (1995-01-01), Kakefuda
patent: 5598344 (1997-01-01), Dangelo et al.
patent: 5598511 (1997-01-01), Petrinjak et al.
Juengst Ernst-Werner
Meyer-Gramann Klaus Dieter
Daimler - Benz Aktiengesellschaft
Hafiz Tariq R.
Shah Sanjin
LandOfFree
Device for automatic generation of a knowledge base for a diagno does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Device for automatic generation of a knowledge base for a diagno, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Device for automatic generation of a knowledge base for a diagno will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-1292322