Process signal classification for aluminum cells

Larry E. Banta; Philip L. Biedler; Congxia Dai

doi:10.1117/12.443120

3 October 2001 Process signal classification for aluminum cells

Larry E. Banta, Philip L. Biedler, Congxia Dai

Proceedings Volume 4565, Intelligent Systems in Design and Manufacturing IV; (2001) https://doi.org/10.1117/12.443120
Event: Intelligent Systems and Advanced Manufacturing, 2001, Boston, MA, United States

Abstract

The control of the aluminum reduction process is typically done based on measurements of cell voltage and current. Both signals are very noisy. Voltage fluctuations come from numerous sources such as CO₂ bubble generation, current fluctuations, metal pad motion, anode movement, electrical arcing, and changes in alumina concentration. The latter is an important control variable, but its effects on cell voltage are small and are often swamped by the signals from other sources. To date, control system designers have attempted to deal with this problem by extreme low-pass filtering and by suspension of all anode movements and alumina feed operations during the measurement period. This strategy is fairly effective, but leads in most case to sub- optimal cell operation most of the time. Research at West Virginia University and Century Aluminum seeks to improve upon the traditional strategies by incorporating knowledge about the process and prior cell control actions in the data analysis operation. For example, selective bandpass filtering and model-based feature detection strategies can be used to identify and isolate various portions of the composite signal. This paper discuses the methods being developed and the results of their application to this critical industrial process.

Citation Download Citation

Larry E. Banta, Philip L. Biedler, and Congxia Dai "Process signal classification for aluminum cells", Proc. SPIE 4565, Intelligent Systems in Design and Manufacturing IV, (3 October 2001); https://doi.org/10.1117/12.443120

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available