All Roads Lead to Fault Diagnosis: Model-Based Reasoning with LYDIA

All Roads Lead to Fault Diagnosis: Model-Based Reasoning with LYDIA

Feldman, A.B.
Pietersma, J.
Van Gemund, A.J.C.

Electrical Engineering, Mathematics and Computer Science

Software Computer Technology

2006-10-05

Model-Based Reasoning (MBR) over qualitative models of complex, real-world systems has proven succesful for automated fault diagnosis, control, and repair. Expressing a system under diagnosis in a formal model and infering a diagnosis given observations are both challenging problems. In this paper we address these challenges. By building a fault model of a realworld artifact (the fuel-system of a light aircraft), we introduce the software package for MBR LYDIA and show its applicability in practice. We demonstrate how structure exploitation and compilation can be used to attack the main challenge to MBR - its high computational cost. Last, we compare our approach to other state-of-the art techniques for MBR and analyze its performance.

http://resolver.tudelft.nl/uuid:762f86fe-aa05-4afd-a51b-aaeacef27016

BNAIC 2006: 18th BeNeLux Conference on Artificial Intelligence, Namur, Belgium, 5-6 October 2006

Institutional Repository

conference paper

(c) 2006 The Author(s)