Multivariate-based evaluation methods for system performance

It is hot research topic to evaluate the equipment performance quickly and accurately. The performance of complex product with high reliability and long-life is indicated frequently by more than single feature variable. And it is possible that multiple degradation modes coexist due to diverse number of performance degradation modes and their correlations. Considering the effects of the different degradation processes on the system performance and the dependences between the multivariate, the system performance status can be described accurately, which then will improve the safety and maintenance efficiency of the system and reduce its life cycle cost. Based on the State-of-the-art at home and abroad, this paper presents two kinds of performance evaluation methods, one is based on leading variables and the other is hybrid variables. Especially the typical methods and several performance degradation models used in the above two evaluation processes are introduced, in order to provide academic and engineering reference for the related research.