Knowledge Based System for Fault Iso lation and Diagnosis of Aircraft Maintenance in Ethiopian Airlines
No Thumbnail Available
Date
2009-10
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
Fault isolation and diagnosis is one of the key aircraft maintenance activities performed
in Ethiopian Airlines. The main features of fault diagnosis in aircraft maintenance
involve multidisciplinary knowledge in aircraft functional systems and the indefinite
relationship between .symptom and source or failure. In order 10 facilitate these activities
the experience and knowledge of aircraft technicians plays a significant role in quickly
solving aircraft system failures without delaying flight schedules. However the airline is
losing significant amount of skilled and experienced technicians whose contribution to
the fault diagnosis was invaluable.
A knowledge based system is developed to .facilitate the fault isolation technique., in
aircrafi maintenance using the experience and knowledge of skilled technicians along
with the relevant mail1lenance documents and maintenance records. The domain
knowledge modeling is performed using decision trees in such a way that a rule based
inference can be implemented for knowledge representation. The prototype developed
employs the utilization of,/ac(s and come up with conclusions based on the backward
chaining inference. The rules in (he knowledge base are constructed Form analyzing the
records of the aircraft maintenance resume book along with (he available maintenance
documel1ls. A prolog programming language is used to demonstrate the prototype and
(he result of the study has revealed that there is an accuracy of 86%. Adopting the
knowledge based system is supposed to retain the aircraft maintenance knowledge and
makes the airline stronger by itself rather than depending fully on human experts.
The knowledge based .system needs further works in updating its knowledge base when
new facts are introduced besides of creating generalized pattern for identifying system
faults.
Description
Keywords
Information Science