室内环境智能多气体传感器系统建模方法研究  

Research on Modeling Method of Intelligent Multi Gas Sensor System for Indoor Environment

作　　者：叶廷东[1] 赖晶亮[1] 高艺康 梁豪[1] 

机构地区：[1]广东轻工职业技术学院计算机工程系,广东广州510260

出　　处：《广东轻工职业技术学院学报》2016年第2期1-5,共5页Journal of Guangdong Industry Polytechnic

基　　金：广东省科技计划项目(2015A020214025;2015A070710030);广州市科技计划项目(2013J4100077);广东轻工职业技术学院创新强校工程项目(2A11105)

摘　　要：针对室内环境多气体监测的需求,提出了一种智能多气体传感器系统模型,该模型采用分布式传感形式,先对各传感器的检测信号进行小波多尺度去噪、多传感信息TEDS解耦自校正,最后用最大后验概率估计各传感器对真实值的支持水平,并由此对多传感器进行融合,获得对真实值的最优估计。基于该模型,使用双CPU开发了室内微环境监测系统,应用研究表明,气体甲醛检测分辨率可达0.01mg/m^3,气体烟雾检测误差为±4%,实现了对室内气体的准确监测。According to the demand of multi gas monitoring in indoor environment, the paper proposed a model of intelligent multi gas sensor system. In the model, the distributed sensing form is adopted, first, the model processes the detection signal denoising of each sensor by wavelet multi-scale, and TEDS decoupling self-calibration of multi-sensing information;last, it uses maximum a posteriori （ MAP） to estimate the support level of each sensor for the true value;by it, the model fuses multi sensor information to obtain the optimal estimate of the true value. Based on the model, an indoor micro environment monitoring system is developed by using the dual CPU. The application research shows that the measurement resolution of the gas formaldehyde is up to 0. 01mg/m3 , the measurement error of gas smoke is ± 4%, and it realized the accurate monitoring of the indoor gases.

关 键 词：气体传感 多尺度 去噪 自校正 信息融合 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]