基于极限学习机的氧化镁熔池尺寸软测量研究  

Research on soft sensing of size of MgO melt pool applied with extreme learning machine

作　　者：李铁[1] 王振[1] 王宁会[1] 

机构地区：[1]大连理工大学电气工程学院,辽宁大连116024

出　　处：《大连理工大学学报》2013年第1期127-131,共5页Journal of Dalian University of Technology

基　　金："八六三"国家高技术研究发展计划资助项目(2008AA03A325)

摘　　要：在电弧炉中建立大小合适的氧化镁熔池是制备氧化镁单晶的基础.为了实现对氧化镁熔池大小的控制,通过对电弧炉传热理论的分析,找出影响氧化镁熔池尺寸的主要因素,采用极限学习机对氧化镁熔池进行软测量研究,并通过与使用支持向量机的软测量模型进行比较检测了该模型的学习能力和泛化性能.实验结果表明,应用极限学习机极大地提高了前向神经网络的学习速度,同时具有较好的预测结果,有助于提高氧化镁熔池的控制精度.The size of the MgO melt pool in the electric arc furnace is important to grow high-purity MgO single crystals. In order to control the size of the MgO melt pool precisely, a soft-sensor based on extreme learning machine （ELM） is proposed according to the analysis of the heat transfer theory of the electric arc furnace. The learning capability and generalization performance of the model are examined by comparison to the model based on the support vector machine （SVM）. The comparative results show that ELM has similar control accuracy compared with SVM, but it has obvious advantages in learning speed for feedforward neural network.

关 键 词：氧化镁单晶 氧化镁熔池 极限学习机 软测量 

分 类 号：TF806.6[冶金工程—有色金属冶金]