Approximate Evaluation of Reliability and Related Quantities via Perturbation Techniques

Reliability evaluation of fault tolerant control systems (FTCS) that include sequential tests for failure detection and identification involve the transient analysis of finite-state semi-Markov chains of very large dimension. Such models for the time horizons of interest are intractable even for sim...

Full description

Saved in:
Bibliographic Details
Main Authors: Walker, Bruce K, Srichander, Ramaswamy
Format: Report
Language:English
Subjects:
ARCHITECTURE
CHAINS
CONTROL SYSTEMS
DETECTION
FAILURE
FAULT TOLERANT COMPUTING
FAULTS
MARKOV PROCESSES
PE61102F
PERTURBATIONS
RELIABILITY
SEQUENTIAL ANALYSIS
SIZES(DIMENSIONS)
STATISTICAL TESTS
Statistics and Probability
SYSTEMS ANALYSIS
TEST AND EVALUATION
TOLERANCE
TRANSIENTS
TRANSITIONS
WUAFOSR2304A5
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Reliability evaluation of fault tolerant control systems (FTCS) that include sequential tests for failure detection and identification involve the transient analysis of finite-state semi-Markov chains of very large dimension. Such models for the time horizons of interest are intractable even for simple architectures. This has motivated the work summarized in this report as well as the work that was accomplished under a previous AFOSR grant. The basis for the work is the idea of asymptotic aggregation of semi-Markov chains that include slow and fast transitions. The extension of earlier asymptotic aggregation results (primarily due to Korolyuk) to models that include decomposed classes that are non-ergodic and the subsequent application of these results to FTCS models was the subject of studies conducted under the previous grant. The research efforts reported here concentrate on the application of the results from the previous study to more general FTCS architectures.