Resumo

MARCANO, Cosme R. Use of hidden markov models for modeling the anode effect process in aluminum electrolytic reduction potcells. uct [online]. 2010, vol.14, n.55, pp.103-108. ISSN 2542-3401.

Hidden Markov Models (HMM) are used in the current work to model the anode effect process in electrolytic potcells for aluminum reduction of the P-19s type, which operates at an average current of 160KA and a nominal voltage of 4.3VDC. Based on a sequence of observations over the voltage behavior, and the fluctuation of the alumina concentration (Al₂O₃) in the electrolytic bath, the HMM designed here is used to determine the most likely sequence of states of the superstructure that supports the anodes and of the breaking/feeding subsystem which could cause that observations sequence. Furthermore, the probability of occurrence of that states sequence and the most likely next state the system will take are calculated by using the Viterbi’s Algorithm

Palavras-chave : Anode Effects; Hidden Markov Model; State Prediction; Viterbi’s Algorithm.