Pattern Identification for Feed Control Strategy Using Fuzzy Neural Algorithm

Smelters have a difficult task in the reduction of the green house gas emission (GHG) by decreasing anode effect. When alumina buck concentration reaches critical levels an anode effects occurs and express itself as a suddenly increase in voltage. Vertical Stub Soderberg (VSS) Side Break pots had no improvements on alumina control in the past decade due the complexity of the problem. The pot is fed every two hours with a fix amount of alumina and the actual feed adjustment is done in a manual daily basis. Based on Prebaked feed control strategy; a model was developed based on a pattern identification algorithm using neuro-fuzzy networks. This algorithm will determine the patterns of the alumina concentration using the pseudo resistance shape curve of the pot. This information provides the amount of alumina that will be fed in the next cycle without mucking the pot and avoiding anode effect

[1]  M. Vermasvuori,et al.  The use of Kohonen self-organizing maps in process monitoring , 2002, Proceedings First International IEEE Symposium Intelligent Systems.

[2]  H. Vogt,et al.  The voltage of alumina reduction cells prior to the anode effect , 2002 .

[3]  G. Tontini.  Pattern identification in statistical process control using fuzzy neural networks , 1996, Proceedings of IEEE 5th International Fuzzy Systems.

[4]  Zeng Shuiping,et al.  Diagnosis System of the Anode Faults for Alumina Reduction Cell , 2006, Sixth International Conference on Intelligent Systems Design and Applications.

[5]  Chuen-Tsai Sun,et al.  Functional equivalence between radial basis function networks and fuzzy inference systems , 1993, IEEE Trans. Neural Networks.

[6]  Kjell Kalgraf,et al.  THEORY OF BUBBLE NOISE , BATH HEIGHT AND ANODE QUALITY , 2006 .