摘要
本文对钢铁球团竖炉焙烧质量分类预测的建模进行了研究,描述了球团竖炉焙烧质量分类预测问题,建立了支持向撼机和BP神经网络两种预测模型;在工业现场试验数据的基础上,比较了两种模型分类预测准确性,并考察了SVM模型参数对分类准确性的影响,结果表明:支持向量机分类预测模型的预测准确率可达84%,优于BP神经网络分类预测模型且具有更好的泛化性能.
This paper presents research on quality prediction classifier modeling of iron ore pellet shaft furnace. The model was firstly described abstractly and then constructed based on support vector machine classifier and BP neural network classifier separately. Predictions of the two kinds of models were compared with the field data. Besides, the SVM model parameters＇ influence on model accuracy were discussed as well. The results showed that SVM classifier model could offer better performance than the BP classifier model with accuracy of 84%.
出处
《工程热物理学报》
EI
CAS
CSCD
北大核心
2013年第11期2065-2068,共4页
Journal of Engineering Thermophysics
基金
国家自然科学基金项目(No.51076027)
江苏省杰出青年基金项目(No.BK20130022)
关键词
支持向量机(SVM)
BP神经网络
质量分类预报
support vector machine（SVM）
BP neural network
quality prediction classifier model
