Kalman Filter for Indirect Measurement of Electrolytic Bath State Variables: Tuning Design and Practical Aspects

• Main Authors: Carlos A. Braga, Joâo V. da Fonseca Neto, Nilton F. Nagem, Jorge A. Farid, Fábio Nogueira da Silva
• Format: Article
• Language: English
• Published: IFSA Publishing, S.L. 2008-04-01
• Series: Sensors & Transducers
• Subjects: Kalman Filter Tuning; Reduction; Indirect Measurement; Electrolytic Bath and QR duality principle
• Online Access: http://www.sensorsportal.com/HTML/DIGEST/march_08/Special_Issue_Vol_90/P_SI_32.pdf

The development Kalman filter tuning model based on QR duality principle of the gain is the main issue of this article. The filter design is oriented to measure the most important state variable of the electrolytic bath, the percentual of alumina. The technical solution encloses on line evaluation of the Kalman filter working with a real production pot. The main goal is to compute a set of filter gains that represents the behavior of the alumina inside the cell. The design and analysis of the Q and R covariances matrices are exercised in order to find a pattern of the reduction cell resistance variations that may be associated with the Al2O3 concentration. The filter bandwidth tuning is performed by increasing or decreasing the filter bandpass from the Q and R variations. This research goes in the direction of practical aspects limits of the indirect measurement system implementations. The robustness of the filter is evaluated in terms of observability, roundoff and modeling errors.