-
无线电子鼻系统的软件采用主成分分析方法(PCA)与反向传播(BP)神经网络相结合,主流程图如图4所示。
从软件设计可以看出,该系统主要包括3个过程:数据采集、特征提取以及网络学习。
(1)数据采集
设传感器阵列的第i个传感器的测量电阻值为
${R_i}$ ,参考电阻为${R_0}$ ,则有:$${R_i} = {R_0}\left(\dfrac{{{V_{cc}}}}{{{V_0}}} - 1\right)$$ (1) (2)数据处理
由于环境因素的影响,传感器的测量存在误差,为此需要对数据进行如下处理:
$$\overline {{R_i}} = \dfrac{{\displaystyle\sum\limits_{j = 1}^n {{R_{ij}} - \max {R_{ij}} - \min {R_{ij}}} }}{{n - 2}}$$ (2) 式中:n为传感器每秒钟的采用次数;
$\max {R_{ij}}$ 为采样的最大值;$\min {R_{ij}}$ 为采样的最小值。利用PCA对得到的3个传感器的数据进行归一化处理,如公式(3)所示:
$${y_i} = \frac{{\overline {{X_i}} - \min \overline {{X_i}} }}{{\max \overline {{X_i}} - \min \overline {{X_i}} }}$$ (3) 式中:
$\overline {{X_i}} $ 为样本数据,其值为$\overline {{R_i}} $ 经过DA转换后获得;$\max \overline {{X_i}} $ 表示3个传感器的平均输出最大值;$\min \overline {{X_i}} $ 表示3个传感器的平均输出最小值,i=1,2,3。(3)网络学习
BP神经网络具有很强的自组织、自适应和自学习能力,可以将网络学习与训练的结果输出与数据库中的特征值进行比对,从而确定样本的类别。图5给出了电子鼻系统的用户界面。
Design of wireless electronic nose based on near infrared spectral absorption technology
-
摘要: 基于近红外光谱吸收技术,设计并开发出了一套能够高效、准确地对目标气体进行检测的电子鼻系统。电子鼻系统主要包括近红外激光发射单元、气室单元、系统控制单元、人机界面单元。主成分分析(PCA)算法和反向传播(BP)神经网络通过LabVIEW所提供的MATLAB Script节点,集成到上位机软件中,并用来对采集到的数据进行分析。结果表明,该电子鼻在网络训练次数达到1 000次以上时达到稳定,且精度达到0.000 1。对白醋、米醋和苹果醋进行食品分辨,识别准确率达到100%,实现了高精度、高稳定度和高分辨率的设计目标,具有较好的应用前景。Abstract: An e-nose system based on near infrared spectral absorption technology which can be used to detect target gases accurately and efficiently was designed and implemented in this paper. The e-nose was composed of near infrared laser emission unit, gas cell unit, system controller unit, and human-machine interface. Principal Component Analysis (PCA) algorithm and Back Propagation (BP) neural network for analyzing the collected date were integrated into the upper computer software by the MATLAB Script node supplied by LabVIEW. The results indicate that the e-nose is stable, and its accuracy is 0.0001, when the time of the networks training reach more than 1 000. The recognition accuracy rate in recognizing the white vinegar, rice vinegar and apple vinegar is 100%, which can achieves the design goal of high-precision, high-stability, high resolution, and behaves good application prospects.
-
Key words:
- near infrared /
- spectral absorption technology /
- principal component analysis /
- back propagation /
- LabVIEW /
- e-nose
-
[1] Mahmoudi E. Electronic nose technology and its applications [J]. Sensors & Transducers, 2009, 107(8): 17-25. [2] Liu Shuhai, Zheng Chuantao, Xu Lin, et al. Detection system of resistive gas sensor based on constant current source [J]. Journal of Jilin University(Information Science Edition), 2012, 30(4): 376-380. (in Chinese) doi: 10.3969/j.issn.1671-5896.2012.04.008 [3] Song Haisheng, Su Xiaoyun, Zhao Xiaolin. Intelligent electronic nose system based on LabVIEW platform [J]. Transducer and Microsystem Technologies, 2012, 31(4): 86-88. (in Chinese) doi: 10.3969/j.issn.1000-9787.2012.04.027 [4] Zhang Qinyi, Sun Wei, Xie Changsheng, et al. Quantitative analysis of volatile organic compounds in air based on electronic nose [J]. Transducer and Microsystem Technologies, 2006, 25(11): 21-24. (in Chinese) doi: 10.3969/j.issn.1000-9787.2006.11.007 [5] Zhao Wan, Mei Xiaodong, Zhu Zhe, et al. Design of electronic nose based on LabVIEW and PCA analysis method [J]. Journal of Jilin University(Information Science Edition), 2013, 31(2): 147-151. (in Chinese) [6] Zhang S, Xie C, Bai Z, et al. Spoiling and formalde-hyde-containing detections in octopus with an E-nose [J]. Food Chemistry, 2009, 113(4): 1346-1350. [7] Wang Zhuo, Min Kunlong, Bai Xuebing, et al. Temperature control of semiconductor lasers in gas sensors [J]. Transducer and Microsystem Technologies, 2013, 32(6): 43-46. (in Chinese) doi: 10.3969/j.issn.1000-9787.2013.06.013 [8] Shi Zhibiao, Huang Shengquan, Fan Xuebing, et al. Research on electronic nose based on biological olfaction [J]. China Mechanical Engineering, 2007, 18(23): 2810-2813. (in Chinese) doi: 10.3321/j.issn:1004-132x.2007.23.010 [9] Sun Yonghai, Sun Zhonglei, Li Yu. Meat artificial olfactory system based on composite GA-RBF network [J]. Journal of Jilin University(Engineering and Technology Edition), 2007, 37(5): 1209-1213. (in Chinese) doi: 10.3969/j.issn.1671-5497.2007.05.045 [10] Carmel L, Levy S, Lancet D, et al. A feature extraction method for chemical sensors in electronic noses [J]. Sensors and Actuators B, 2003(93): 67-76. [11] Gao Mingxing, Wang Xiaoning, Gong Yan, et al. Study on application of electronic nose in cotton fabric odor detection [J]. Transducer and Microsystem Technologies, 2011, 30(6): 33-35. (in Chinese) doi: 10.3969/j.issn.1000-9787.2011.06.011 [12] Li Jiangyong, Luo Dehan, Chen Yimin. Realization of embedded electronic nose system based on ARM-Linx [J]. Computer Technology and Development, 2012, 22(11): 209-211, 215. (in Chinese) [13] Liu Kuixue, Chen Lihua, Lv Qinghua, et al. Wireless toxic gas concentration acquisition system [J]. Journal of Instrumentation, 2006, 27(S2): 284-286. (in Chinese)