Performance Improvement of a Multi-State Coherent System using Component Importance Measure

Abstract

In system engineering, numerous efforts have been made for achieving improvement in system performance under a binary set up, where each component, as well as the entire system, has any one of two states; namely, perfect functioning and complete failure. However, there are systems which perform their tasks at various performance levels rather than functioning at only the above two performance levels. These systems are multi-state systems. In these systems, there can be some partially working states or performance levels before the system comes to the state of complete failure. Hence, the need has been felt to develop the procedures for improving the performance of multi-state systems consisting of multi-state components. This article resolves such an issue for a multi-state system using a multi-state component importance measure. The measure developed here is used to assess the impact of individual components on the improvement of system performance. Some basic theory to deal with a homogeneous multi-state coherent system has been developed, and finally, a rule has been derived to improve system performance using the importance measure. The applications of the results have been illustrated through a real-life example.