摘要
针对差压式流量计密封性检测中存在效率低、可靠性差和人工影响大等问题,提出了一种基于支持向量机的方法实现流量计密封性声发射检测,通过采集和分析声发射信号来判断流量计的密封性能是否满足要求。检测过程中对采集的声发射信号进行多参数提取,基于支持向量机选择多项式核函数进行系统参数样本数据训练,建立分类器,用于识别流量计密封性正常与否。通过采集不同泄漏情况的流量计声发射信号,正确判断了流量计密封性正常和不正常两种状况,说明此方法能自动识别流量计密封性状况,减小人工检测的弊端,提高生产效率和可靠性。
In order to overcome the problems of low efficiency,poor reliability and great artificial influence in sealing detection of differential pressure flowmeter,a method of acoustic emission detection based on support vector machine(SVM) is proposed,which can judge whether the sealing of flowmeter meets the requirements by collecting and analyzing acoustic emission signals.In the process of detection,multi-parameters are extracted from the collected AE signals,and the system parameter sample data is trained based on the polynomial kernel function selected by support vector machine,and a classifier is established to identify whether the flowmeter seal is normal or not.By collecting acoustic emission signals of flowmeters under different leakage conditions,the normal and abnormal sealing conditions of flowmeters are correctly judged,which shows that this method can automatically identify the sealing conditions of flowmeters,reduce the disadvantages of manual detection,and improve production efficiency and reliability.
作者
路萍
LU Ping(Chongqing Research Institute Co.,Ltd.,China Coal Technology Industry Group,Chongqing 400039,China)
出处
《自动化与仪器仪表》
2020年第5期25-27,32,共4页
Automation & Instrumentation
基金
国家自然科学基金(No.61573073)
重庆市技术创新与应用示范(社会民生类)一般项目(No.cstc2018jscx-msybX0148)。
关键词
声发射
支持向量机
密封性
acoustic emission
support vector machine
sealability
