A categorical controller for run-to-run processes
Lists
Only you see this
Abstract
Qualitative characteristics of products and processes are frequently encountered in diverse applications. Low-resolution categorical information is particularly important in a run-to-run process when numerical laboratory measurement is difficult to obtain without significant time delay. In this paper, we develop a run-to-run control scheme based on categorical observations, which features using qualitative information in an effective and formal way for process adjustment where quantitative information is not available. A two-phase modeling and adjustment strategy is introduced: ordinary regression and classification methodologies are used for modeling in Phase I, and a latent-model originated control law, categorical run-to-run (R2R) controller, is presented in Phase II. Simulation results show that the proposed algorithm exhibits competitive control performance, less adjustment efforts, and a larger stable region, comparing with the situation where accurate information is available and the classical Exponentially Weighted Moving Average (EWMA) controller is implemented.
BibTeX
@inproceedings{wang2005categorical,
author = {Kaibo Wang and Fugee Tsung},
title = {{A categorical controller for run-to-run processes}},
booktitle = {Rare},
year = {2005},
}Loading related content...