A theory of design reliability using probability and fuzzy sets

All process designs are subject to uncertainties which make it impossible to state with complete certainty that a design will work. This study presents a measure of design confidence which considers the nature of the uncertainty and the operability of the process as a whole. This measure, called design reliability, quantifies the likelihood that a design will work. Detailed analyses are presented for three special cases: designs subject to random uncertainties, designs subject to fuzzy uncertainties, and designs subject to both random and fuzzy uncertainties. Practical procedures for estimating design reliability in these special cases are also presented.