用于气水两相流流型识别的ERT信息融合方法  

AN INFORMATION FUSION METHOD OF ERT DATA FOR GAS-WATER TWO-PHASE FLOW REGIME RECOGNITION

作　　者：谭超[1] 董峰[1] 徐遥远[1] 

机构地区：[1]天津市过程检测与控制重点实验室,天津大学电气与自动化工程学院,天津300072

出　　处：《工程热物理学报》2010年第5期785-788,共4页Journal of Engineering Thermophysics

基　　金：国家自然科学基金项目(No.50776063);天津市应用基础及前沿技术研究计划重点项目(No.08JCZDJC17700)

摘　　要：气水两相流在生产过程与科学研究中有十分重要的地位。流型识别作为两相流研究的基础,对于两相流过程参数精确测量起重要作用。信息融合技术可消除多传感器信息间可能存在的冗余和矛盾,降低其不确定性,实现比单一信息源对被测过程更完全、准确、可靠的测量和描述。本研究分别采用数据级融合与特征级融合实现双截面电阻层析成像系统两测量截面间测量信息融合,结合支持向量机实现水平管道气水两相流流型识别率的提高。通过识别结果的对比,研究了这两类信息融合方法提高流型识别率的原因。Gas-water two-phase flow is widely encountered and of great importance in manufacture processes and scientific researches,its flow pattern is significant to the accurate measurement of its process parameters.Information fusion combines data and related information to achieve more complete,specific and reliable inferences than could be achieved by using a single,independent information source.In this work,the information from a dual-plane Electrical Resistance Tomography （ERT） system is respectfully fused at data level and feature level,to improve the flow regime recognition rate by using the Support Vector Machine（SVM） technique.The two fusion methods are compared and investigated to disclose their respective contribution to flow regime recognition.

关 键 词：电阻层析成像 流型识别 特征提取 数据融合 特征融合 

分 类 号：TK391[动力工程及工程热物理—热能工程]