基于SVM的多传感器信息融合  

Information fusion algorithm based on SVM

作　　者：孙颖[1] 刘玉满[1] 龚稳[1] 

机构地区：[1]长春工业大学 电气与电子工程学院,吉林长春130012

出　　处：《长春工业大学学报》2013年第3期300-304,共5页Journal of Changchun University of Technology

基　　金：吉林省教育厅基金资助项目(吉教科合字2010300)

摘　　要：针对现代工业生产采用大量不同传感器来监测和控制生产过程,被测量之间存在较大模糊关系,提出了一种基于SVM理论的多传感器信息融合方法。通过"对电弧炉炼钢供电曲线进行优化"表明,该方法对解决小样本、高维特征空间和不确定条件下的多传感器信息融合问题,具有较好的融合效果。Based on modern industrial production line with many kinds of sensors and monitor environments,a multi-sensor fusion algorithm based on SVM theory is proposed considering the fuzzy relationship between the measured value.The application in the electric arc furnace steelmaking supply curve optimization shows that the algorithm can solve the multi-sensor fusion problem under the conditions of high dimensional space and short samples.

关 键 词：SVM(支持向量机) 多传感器 信息融合 电弧炉 供电曲线 

分 类 号：TM92[电气工程—电力电子与电力传动]