DOE-Based Hybrid Automatic Process Control for Variation Reduction

Abstract

The quality improvement of industrial processes is achieved by maintaining their mean continuously close to the target value while reducing process variability. When there are strong noise factors including observable noises and unobservable ones in the process, use of parameter design, feedforward control or feedback control alone may not be effective. In this paper, an integrated online control strategy which is able to compensate all sources of noise factors is developed. The framework of the proposed hybrid control scheme is given and a two-step approach for the integrated design of feedforward and feedback controllers is also presented.