DocumentCode
1858335
Title
State reduction for semi-Markov reliability models
Author
White, Allan L. ; Palumbo, Daniel L.
Author_Institution
NASA Langley Res. Center, Hampton, VA, USA
fYear
1990
fDate
23-25 Jan 1990
Firstpage
280
Lastpage
285
Abstract
Trimming, a method for reducing the number of states in a semi-Markov reliability model, is described, and an error bound is derived. The error bound uses only three parameters from the semi-Markov model-the maximum sum of rates for failure transitions leaving any state, the maximum average holding time for a recovery-mode state, and the operating time for the system. The error bound can be computed before any model generation takes places which means the modeler can decide immediately whether the model can be trimmed. The trimming has a precise and simple description and thus can be easily included in a program that generates reliability models. The simplest version of the error bound for trimming is presented. More accurate versions can be obtained by requesting more information about the system being modeled
Keywords
Markov processes; failure analysis; reliability; error bound; failure; holding time; operating time; recovery-mode; semi-Markov reliability models; state reduction; trimming; Aerospace control; Computer architecture; Concrete; Digital control; Failure analysis; Mathematical model; NASA; Reduced order systems; Redundancy; System recovery;
fLanguage
English
Publisher
ieee
Conference_Titel
Reliability and Maintainability Symposium, 1990. Proceedings., Annual
Conference_Location
Los Angeles, CA
Type
conf
DOI
10.1109/ARMS.1990.67970
Filename
67970
Link To Document