电子鼻系统对危险爆炸物的识别研究

Reseach on recognizing explosives by electronic nose

采用纳米氧化锌厚膜气体传感器组成的阵列对硝铵、矿山炸药、苦味酸、2,4二硝基甲苯(DNT)4种典型爆炸物样品进行了实验。称取4mg爆炸物样品,采用动态实验对18只传感器进行筛选、优化,得到6个最优元件组成的阵列。采用静态实验考察了传感器的检测能力,结果表明:随着待测样品浓度的增加,6个传感器的灵敏度均呈指数上升,且该方法能够检测到4种爆炸物样品的最小浓度均低至3.34μg/L。为进一步与实际应用相接近,采用完全动态小剂量实验,结果表明:在15.4μg/L浓度处提取斜率最大值为特征值对阵列进行判别函数分析,可使4种典型爆炸物在该浓度水平能够完全区分。

Four kinds of representative explosives （NH4NO3, mineral explosives, picric acid and DNT） are measured by the gas sensor arrays, which are composed of eighteen doped nano-ZnO thick film sensors. 4 mg of each sample is weighed out and dynamic sampling method is used to optimize the sensor arrays, then an optimal sensor array composed of six better sensors are gained, static sampling method is used to exam the detection ability of the gas sensors. The results show that with the increase of concentrations of samples, the sensitivities of six sensors all increase, and the detection concentrations of samples are low to 3.34 μg/L under the laboratory conditions. For the sake of application,full dynamic sampling method is used and the results demonstrate that all samples can be identified completely at 15. 4 μg/L when extracting maximum of slope as the feature to discriminant function analysis.