[210] A self-supervised LSTM network for cell temperature prediction in aluminium electrolysis reduction
Y Lei and H R Karimi
Politecnico di Milano, Italy
This paper presents a novel LSTM method for cell temperature prediction in the aluminium electrolysis industry. In the practical electrolysis production, it always suffers the limitation of the sample scarcity, which greatly decreases the performance of the supervised-based prediction models. This paper proposes an enhanced LSTM unit that fully utilises the self-supervised loss as the training process. The Kullback-Leibler (KL) divergence is proposed for learning the similarity of the different unlabelled data. The proposed SSLSTM is further applied to the industrial aluminium electrolysis temperature process. The experimental results demonstrate that the proposed method has state-of-the-art accuracy and robustness.
