Process-based diagnosis: An approach to understanding novel failures
Collins, John William
This item is only available for download by members of the University of Illinois community. Students, faculty, and staff at the U of I may log in with your NetID and password to view the item. If you are trying to access an Illinois-restricted dissertation or thesis, you can request a copy through your library's Inter-Library Loan office or purchase a copy directly from ProQuest.
Permalink
https://hdl.handle.net/2142/21460
Description
Title
Process-based diagnosis: An approach to understanding novel failures
Author(s)
Collins, John William
Issue Date
1994
Doctoral Committee Chair(s)
Winslett, Marianne
Department of Study
Computer Science
Discipline
Computer Science
Degree Granting Institution
University of Illinois at Urbana-Champaign
Degree Name
Ph.D.
Degree Level
Dissertation
Keyword(s)
Computer Science
Language
eng
Abstract
This thesis describes a diagnostic technique for explaining unanticipated modes of failure in continuous-variable systems. Previous approaches in model-based diagnosis have traditionally suffered from either a dependence on explicit fault models or a tendency to produce unintuitive results. This research aims at achieving the explanatory power of explicit fault models, without sacrificing the robustness of consistency-based diagnosis. The unique compositional nature of the process-centered models of Qualitative Process Theory makes the application of model-based diagnostic techniques both non-trivial and rewarding. Rather than relying on explicit fault models, this approach utilizes a general domain theory to model the broken device. Given a sufficiently broad domain theory, symptoms are explained in terms of a transformed physical structure. Generative fault models replace explicit, pre-enumerated fault models, thereby increasing robustness for identifying novel faults. This approach combines the efficiency of the consistency-based approach with the explanatory power of abductive backchaining. Candidates generated using a consistency-based approach are used to focus the abductive search for a structural model of the failed system. An implementation built on a modified ATMS and an incremental qualitative envisioner is tested on a number of examples. The systems examined are taken primarily from the domain of thermodynamics, but also include some simple circuits.
Use this login method if you
don't
have an
@illinois.edu
email address.
(Oops, I do have one)
IDEALS migrated to a new platform on June 23, 2022. If you created
your account prior to this date, you will have to reset your password
using the forgot-password link below.
