An integrated approach to analyse system behaviour using knowledge-based approximate reasoning methodologies

Recent advances in technology and the growing complexity of systems have made the job of system-reliability engineers more challenging, because they have to analyse the uncertain behaviour of systems with the help of various analytical techniques that require the knowledge of precise numerical probabilities and component functional dependencies; the information for which is rather difficult to obtain. To cope with such situations, this paper presents the application of knowledge-based approximate reasoning methodologies, that is, fuzzy and grey approaches, as one of the most viable and effective tools for behaviour analysis of systems. An illustrative case from the paper industry is presented. Various parameters of importance to managerial decision-making, such as repair time, failure rate, mean time between failures, availability and expected number of failures related to system performance, are computed to quantify the behaviour in objective terms. Further, a risk-ranking approach-based on fuzzy and grey relational analysis is discussed to prioritise various failure causes associated with the components. Compared with the traditional Failure Mode and Effect Analysis (FMEA), the proposed fuzzy and grey approach provides more pragmatic results.