Multivariate Approach for Equipment Health Monitoring

摘要

This paper deals with a method for equipment health monitoring through a multivariate approach. The approach concentrates on the autonomous classification of different process variable types which are defined by particular variable behaviors and the investigation of specified statistical feature extraction methods for every respective class. Also the computation of the equipment health factor (EHF) is presented. We provide a general method for detection of unknown failures through a multivariate approach to minimize production costs while still maintaining high yield.