摘要
针对当前传统的化工厂危险气体泄漏检测系统中,在多层爆炸性气体和弱爆炸性气体的条件下,传统的智能检测方法存在较大误差的问题。设计了一套基于虚拟仪器的气味挥发类中药品质测量系统,利用气敏传感器组成阵列进行前端数据采集,采用USB-DATA AD816数据采集器实时采集信号,以LabVIEW2009为开发平台,通过改进K均值聚类对多层爆炸气体聚类分析,通过弱爆炸信号加强,对混合爆炸气体信号进行有效的分类测量,实现化工厂挥发类爆炸性气体泄漏检测,实验结果表明,该方法能够较好的实现检测功能,克服传统方法存在的弊端。
In view of the current traditional chemical plant dangerous gas leakage detection system, explosive gas and weak in multilayer explosive gas, under the condition of the traditional intelligent detection method is larger error problem. Designed a set of virtual instrument based on the smell of the traditional Chinese medicine quality of volatile measurement system, the use of gas sensor array front-end DATA acquisition, the USB-DATA AD816 DATA collector collect real-time signal to LabVIEW2009 for development platform, through improving k-means clustering of multilayer explosive gas clustering analysis, through the weak signal strengthen explosion, hybrid gas explosion signal effectively classification of measurement, and realize the chemical plant volatile explosive gas of leak detection, the experimental results show that the method can better detection function, overcome the disadvantages of traditional methods.
出处
《科技通报》
北大核心
2012年第7期80-83,共4页
Bulletin of Science and Technology
基金
国家自然科学基金项目(61111130183)
关键词
爆炸性气体
模式识别
改进均值聚类
explosive gas
pattern recognition
improved mean clustering
