Statistical Process Monitoring of Nonlinear Profiles Using Wavelets

Journal of Quality Technology vol. 41 issue 2 - April 2009

Abstract:[This abstract is based on the authors' abstract.] The complex data records collected by many modern industrial processes do not readily allow the use of traditional statistical process-control techniques. An observation from a process might consist of n pairs of x and y data. Such data structures or relationships between x and y are called profiles. This study presents a semiparametric wavelet method for monitoring changes in sequence of nonlinear profiles. The method uses the spatial-adaptive properties of wavelets to detect profile changes. The method is used to differentiate between different radar profiles. Results indicate the method can quickly detect a variety of changes from a given in-control profile.

Keywords: Average run length (ARL); Control charts; Likelihood methods; Nonlinear models; Parametric models; Statistical process control (SPC); Thresholds