Automatic diagnosis based on reliability analysis (ADORA).

摘要

In this dissertation I worked under the supervision of Professor Joanne Bechta Dugan in the Charles L. Brown Department of Electrical and Computer Engineering at the University of Virginia. We designed a methodology for Automated Diagnostics based On Reliability Analysis (ADORA).; ADORA methodology combines diagnosis approaches based on entropy decision making with quantitative and qualitative information produced by reliability analysis. ADORA produces a diagnostic decision tree (DDT) that designates the best test to apply next, based on the evidence gathered thus far, cost and information gained by each test, and the quantitative, unreliabilities, and qualitative, structural, analysis provided by the reliability analysis model.; Within this dissertation we introduce a new importance factor, the cost and diagnosis importance factor. Which allows us to optimize the cost of diagnosis. We also design a new modeling method that extends the dynamic fault tree to facilitate capturing information from evidence sources such as sensors.; ADORA has been designed as a stand alone tool and also has been implemented and incorporated into Galileo which was funded by NASA. Several invited journal papers, proceeding papers, and conference papers have been published in regards to ADORA. Also a provisional patent application for ADORA was submitted on April 1st 2004.