Abstract
In recent years, a series of multivariate statistical methods have been developed for batch process modeling and online monitoring, which take important batch process characteristics into consideration, such as, multiphase, batch dynamics and so on. As a typical batch process, injection molding is a suitable object to employ these methods. In this paper, the potential applications of these methods in industry are shown by monitoring and fault diagnosis results of the injection molding process.