Accession Number:
ADP003937
Title:
Model-Based Probabilistic Reasoning for Electronics Troubleshooting,
Descriptive Note:
Corporate Author:
NAVY CENTER FOR APPLIED RESEARCH IN ARTIFICIAL INTELLIGENCE WASHINGTON DC
Personal Author(s):
Report Date:
1984-06-01
Pagination or Media Count:
12.0
Abstract:
The Navy A.I. Center is currently developing a series of increasingly sophisticated expert consultant systems for guiding a novice technician through each step of an electronics troubleshooting session. One of the goals is to automatically produce, given a set of initial symptoms, a binary passfail decision tree of testpoints to be checked by the technician. This paper discusses our initial approach using a modified game tree search technique, the gamma miniaverage method. One of the parameters which guides this search technique - the cost of each test - the conditional probability of test outcomes and the proximity to a solution - are provided by a dynamic model of an expert troubleshooters beliefs about what in the device is good and what is bad. This model of beliefs is updated using probabilistic test-results yields plausible-consequences rules. These rules are either provided by an expert technician, or approximately by a model-guided Rule Generator. The model that guides the generation of rules is simple block diagram of the Unit Under Test augmented with component failure rates.