Flow profile identification with multipath transducers

详细信息    查看全文

• 作者：Longhui Qin ; Liang Hu ; ^{cmeehuli@zju.edu.cn} ; Kai Mao ; Wenyu Chen ; Xin Fu
• 关键词：Ultrasonic gas flowmeter ; Flow profile identification ; Support vector machine ; Multiclass identifier ; Extreme learning machine
• 刊名：Flow Measurement and Instrumentation
• 出版年：2016
• 出版时间：December 2016
• 年：2016
• 卷：52
• 期：Complete
• 页码：148-156
• 全文大小：1449 K
• 卷排序：52

文摘

An intelligent method of flow profile identification for multipath ultrasonic flowmeter is proposed. The flow profile identifier is constructed based on multiclass support vector machine (SVM) with polynomial kernel. 65 flow profiles of different types are used to train and test the SVM identifier. A comparison with extreme learning machine under different levels of uncertainty reveals effect of the proposed identifier.