Abstract
A deterministic parameter design method is applied to system reliability analysis. The method is used with a simulated annealing algorithm to determine the optimum nominal values of component failure probabilities which correspond to minimum variability in system reliability, when individual component failure probabilities are subject to specified tolerances or variability limits. The optimisation is carried out subject to constraints on component failure probabilities, nominal system reliability and component costs.
Additional information
Notes on contributors
D.B. Parkinson
David Parkinson has degrees of B.Sc. (Physics) and Ph.D. (Engineering) and is a Fellow of the UK Institute of Diagnostic Engineers. Originally employed in Aerospace and Glass manufacturing, since 1975 he has taught both undergraduate and postgraduate courses in the Engineering Department of the University of Liverpool, UK. His research interests have been mainly in the fields of CAD, Reliability and Robust Design. He is currently working on projects concerning the wider application of parameter design.