基于气体传感器阵列和独立成分分析的易燃气体检测

本文提出一种基于气体传感器阵列和独立成分分析(ICA)的方法来检测常见易燃性气体,例如一氧化碳(CO)和甲烷(CH4).气体传感器阵列由6个商用金属氧化物半导体传感器组成,分别对这两个独立的气源产生响应.通过分析ICA散点图中CO和CH4的信息成分,实现了这两种易燃气体的定性和定量组分分析.仿真结果表明,用ICA对气体传感器阵列测量数据进行处理,可以提高气体识别的准确率和浓度测量的精度.

Combustible gas detection with gas sensor array and independent component analysis

In this paper, we propose a promising approach to detect ordinary combustible gases,e.g.carbon monoxide (CO) and methane (CH4),using gas sensor array and independent component analysis (ICA). Gas sensor array comprising six commercia metal-oxide semiconductor sensors reponds to two independent gas sources CO and CH4 respectively. Analyzing the information components corresponding to CO and CH4 from the ICA scatter plots,we achieved the qualitative and quantitative gas component analysis of the two combustible gases. Simulation results demonstrate that ICA applied on gas sensor array measurement data is capable of improving the accuracy rate of gas recognition and precision of concentration measurement.