采用智能神经元网络技术的新型气体泄漏检测技术  

New Generation Technology of Gas Leak Detection Based on Artificial Neural Network

作　　者：刘秀玲 马云鹂 

机构地区：[1]中石化集团公司安庆分公司,安徽安庆246001 [2]梅思安(中国)安全设备有限公司,江苏苏州215000

出　　处：《仪器仪表用户》2014年第4期98-100,共3页Instrumentation

摘　　要：简述了常规气体泄漏检测面临的潜在风险,引入超声波气体泄漏检测的概念。复杂多变的环境背景噪声的影响,以及其常规超声波探测技术手段的局限性,对超声波气体泄漏检测造成较大的应用限制。第三代超声波气体泄漏检测技术采用了先进的智能神经元网络(ANN)技术,采用模糊模式识别,而不是常规的阀值判断技术,不但检测能力大大提高,同时有效避免了误报警现象,提高了安全防护等级。The potential risk of traditional gas leak detection is briefly described,then the ultrasonic technology is introduced.However due to impact from changeful and complex background noise and the limitation of old-fashion acoustic technology as well,the ultrasonic gas leak detection applications are quite limited.The third generation gas leak technology applies start-ofthe-art Artificial Neural Network（ANN）,instead of traditional threshold based judgement or simple fingerprint reading.ANN technology provides wider range of detection,more importantly,it＇s much more easier to install and operate than ever.

关 键 词：气体泄漏 超声波 神经元网络 误报警 气体检测 

分 类 号：U467[机械工程—车辆工程]