Home // ADAPTIVE 2010, The Second International Conference on Adaptive and Self-Adaptive Systems and Applications // View article

Hidden State Observation for Adaptive Process Controls

Authors:
Melanie Senn
Norbert Link

Keywords: statistical process model; hidden state prediction; regression analysis; dimension reduction; deep drawing.

Abstract:
In many manufacturing processes it is not possible to measure on-line the state variable values that describe the system state and are essential for process control. Instead, only quantities related to the state variables can be observed. Machine learning approaches are applied to model the relation between observed quantities and state variables. The characterization of a process by its state variables at any point in time can then be used to adequately adjust the process parameters to obtain a desired final state. Also, multiple process controls of a process chain can be linked using standardized transfer state variables between the single processes. This allows the optimization of the entire process chain with respect to the desired properties of the final workpiece. This paper proposes a general method to extract state variables from observable quantities by modeling their relation from experimental data with data mining methods. After transforming the data to a space of de-correlated variables, the relation is estimated via regression methods. Using Principal Component Analysis and Artificial Neural Networks we obtain a system capable of estimating the process state in real time. The feasibility of our approach is shown with data from numerical simulation of a deep drawing process.

Pages: 52 to 57

Copyright: Copyright (c) IARIA, 2010

Publication date: November 21, 2010

Published in: conference

ISSN: 2308-4146

ISBN: 978-1-61208-109-0

Location: Lisbon, Portugal

Dates: from November 21, 2010 to November 26, 2010