Fuzzy Decision Support System (FDSS) for Conducting FMEA

摘要

Traditional Failure Mode Effects Analysis(FMEA), the system safety and reliability analysts evaluate the risk or criticality associated with the item failure modes in FMEA using descriptive terms, such as, low, important, high and very high. Owing to the subjective and qualitative nature of information, interdependences among various failure modes are often misinterpreted and decisions are made accordingly, which is indeed the serious limitation of the traditional FMEA based upon Risk Priority Number (RPN) calculations. To help the analyst's inefficient and effective analysis the paper presents a fuzzy logic based approach for conducting FMEA. In the proposed approach, the parameters used in FMEA analysis, ie, frequency of occurrence of failure (S{sub}f), severity (S) and non-detectability(S{sub}d) are represented as members of a fuzzy set, combined by matching them against the rules in a rule base, evaluated by an inference system and then defuzzified to assess the riskiness level of the failure by computing respective RPN. To illustrate the proposed fuzzy logic approach a case study from a process industry has been discussed. The traditional RPN is calculated and compared with fuzzy RPN obtained by Fuzzy Decision Support System (FDSS). The results demonstrated the inherent potential of fuzzy inference modelling in generating a knowledge base to deal with the problems of handling vague and qualitative information in a consistent and logical manner.