Fuzzy set‐valued and grey filtering statistical inferences on a system operating data

Abstract

Purpose

Intends to address a fundamental problem in maintenance engineering: how should the shutdown of a production system be scheduled? In this regard, intends to investigate a way to predict the next system failure time based on the system historical performances.

Design/methodology/approach

GM(1,1) model from the grey system theory and the fuzzy set statistics methodologies are used.

Findings

It was found out that the system next unexpected failure time can be predicted by grey system theory model as well as fuzzy set statistics methodology. Particularly, the grey modelling is more direct and less complicated in mathematical treatments.

Research implications

Many maintenance models have developed but most of them are seeking optimality from the viewpoint of probabilistic theory. A new filtering theory based on grey system theory is introduced so that any actual system functioning (failure) time can be effectively partitioned into system characteristic functioning times and repair improvement (damage) times.

Practical implications

In today's highly competitive business world, the effectively address the production system's next failure time can guarantee the quality of the product and safely secure the delivery of product in schedule under contract. The grey filters have effectively addressed the next system failure time which is a function of chronological time of the production system, the system behaviour of near future is clearly shown so that management could utilize this state information for production and maintenance planning.

Originality/value

Provides a viewpoint on system failure‐repair predictions.

Citation

Guo, R. and Love, E. (2005), "Fuzzy set‐valued and grey filtering statistical inferences on a system operating data", Journal of Quality in Maintenance Engineering, Vol. 11 No. 3, pp. 267-278. https://doi.org/10.1108/13552510510616478