Application of multiple support vector machines model based on FCM for temperature prediction in blast furnace

王义康,刘祥官

DOI: https://doi.org/10.3969/j.issn.1000-7059.2012.03.005

2012-01-01

Abstract:A kind of multiple support vector machines model based on fuzzy C-mean clustering （FCM） is proposed in order to improve the temperature prediction model with single support vector machine （SVM） in intelligent control expert system for blast furnace. At first, the training set is clas- sified into several groups by using of fuzzy C-mean clustering method. Then each group is trained through support vector machine and the corresponding sub-models are established. After that, the same sample in testing set is predicted separately and the memberships between each sample and each group are set as the weights. Finally, the predicted result is obtained by weighted sum. The analysis of on-line data shows that the multiple support vector machines model based on fuzzy C-mean clustering has. better prediction performance in many aspects compared with the single SVM model. The hit rate reaches 86% in consecutive 100 furnaces based on the proposed model.