Process Operation Performance Assessment Based on Multisource Heterogeneous Information With Semisupervised Transfer Generative Adversarial Network for Electrofused Magnesium Furnace
Kaiqing Bu, Yan Liu, Fuli Wang
Abstract
The process operation performance assessment in the smelting process of the electrofused magnesium furnace (EFMF) plays a very important role in improving the production efficiency and economic benefits. Most of the existing methods mainly rely on a large number of labeled data, which is difficult to satisfy in practical application. To solve the problem of a small number of labeled samples, a semisupervised transfer generative adversarial network based on source domain (SD) data is proposed to assess the operation performance of EFMF in this article. The core idea is to use the labeled data of SD to train a classifier and predict the unlabeled data of the target domain (TD), and use the generate adversarial network (GAN) to make the multisource heterogeneous information comprehensive features of the TD and SD conform to the similar distribution. The experiment shows that the method can obtain satisfactory assessment results than the compared method.