Possibilistic approaches to reliability based inspection planning

摘要

Structural reliability before and after inspections is dependent on many factors best expressed in linguistic format and can hardly be transformed into probabilistic information. Subjectivity and fuzziness are inherent in the evaluation of the system reliability. The alternative method of expressing cognitive uncertainty is through fuzzy set theories. It can be envisaged that both Cognitive and non-cognitive uncertainties exist in one and the same variable. Using fuzzy set theory, one can express failure possibility as a subjective measure of hardware and human reliability. The uncertainties in the loads and behaviour offshore structural systems and need to reduce maintenance costs have called for investigations into reliability based in-service inspections planning. The optimised inspection plan is considered as a function of fatigue life. The parameters in the fatigue life estimation equations for fatigue crack growth are considered fuzzy/random. The membership functions of the parameters are assumed as triangular. The possibility transforms are used to transform the triangular fuzzy numbers to the Gaussian fuzzy numbers. The failure possibility/fuzzy reliability index is the evaluated using fuzzy arithmetic/calculus. Optimal inspection interval can be obtained by estimating updated failure possibility as a function of time. The time between the two inspections would be when predicted failure possibility or fuzzy reliability index approaches the permissible failure possibility or fuzzy reliability index. The inspection plans can thus be based on possibility of failure, which takes into account subjective uncertainties in the analysis and inspections in fuzzy format.