Study on steam turbine fault diagnosis and maintenance service grid system

Abstract

It is important to ensure the safe operation, quick fault diagnosis, and maintenance of steam turbines. This is a challenging problem since steam turbines tend to be geographically dispersed and it is difficult to offer a timely response to a facility that can be a considerable distance from the maintenance division. The manufacturing grid is resource sharing technology that could provide a solution to this problem. A framework to apply this concept to steam turbine fault diagnosis and maintenance operations is proposed in this paper. In the proposed manufacturing grid the fault type is identified by an artificial neural network algorithm and the maintenance solution is given on the basis case-based reasoning with the framework of rough set theory. A prototype system is developed on the basis of proposed methodologies to demonstrate their feasibility.