Accession Number : ADA319499

Title :   Using a Neural Network and a Statistical Classifier for Aircraft Fault Diagnostics.

Descriptive Note : Final technical rept. Mar 95-Aug 96,

Corporate Author : ARMSTRONG LAB WRIGHT-PATTERSON AFB OH HUMAN RESOURCES DIRECTORATE

Personal Author(s) : Chu, Shing P.

PDF Url : ADA319499

Report Date : AUG 1996

Pagination or Media Count : 22

Abstract : In recent years, several techniques have been developed to create intelligent diagnostic aiding systems. Most of these systems, including the current Integrated Maintenance Information System (IMIS) diagnostics module, involve modeling the systems to be maintained. These systems have the disadvantage of requiring extensive efforts to develop them. A developing technology, neural networks, provides a promising alternative. Neural nets develop diagnostics strategies by learning from past experience with the system, and do not require extensive modeling. Neural networks are well suited to diagnostics applications. This paper presents: A detailed description of a neural network based diagnostic system; An explanation of a radial basis function (RBF) neural network architecture and its construction; An explanation of the construction of a statistical classifier; A description of data representation and method of system optimization; and Performance and experimental result.

Descriptors :   *NEURAL NETS, *AIRCRAFT MAINTENANCE, *COMPUTER AIDED DIAGNOSIS, FIGHTER AIRCRAFT, MILITARY REQUIREMENTS, AIR FORCE RESEARCH, COMBAT READINESS, OPERATIONAL READINESS, COMPUTER ARCHITECTURE, ARTIFICIAL INTELLIGENCE, DIAGNOSTIC EQUIPMENT, MAINTENANCE MANAGEMENT, BAYES THEOREM, AVIATION SAFETY, MAINTENANCE EQUIPMENT.

Subject Categories : Cybernetics
      Attack and Fighter Aircraft

Distribution Statement : APPROVED FOR PUBLIC RELEASE