基于LMSE的葡萄劣变监测仪器的设计  

LMSE Based Grape Deterioration Monitoring Device Design

作　　者：王文重[1] 韩峻峰[2] 董大明[3] 

机构地区：[1]桂林电子科技大学电子工程与自动化学院,广西桂林541004 [2]广西机电职业技术学院电气工程系,南宁530007 [3]国家农业信息化工程技术研究中心,北京100097

出　　处：《农机化研究》2014年第9期111-114,120,共5页Journal of Agricultural Mechanization Research

基　　金："十二五"国家科技支撑计划项目(2013BAD19B02);北京市农林科学院创新能力建设专项(KJCX201102001)

摘　　要：葡萄在运输和贮藏过程中容易发生变质,对其进行预警可有效降低大规模腐败的风险。傅里叶变换光谱表明葡萄在劣变过程中,其挥发性物质里乙醇和二氧化碳(CO2)变化最为明显。为此,通过选用CO2气体传感器模块和乙醇电化学传感器,设计了相应的软硬件,完成了一个低成本、小型的葡萄变质监测仪器,可以实时、定量测量葡萄挥发物质中CO2和乙醇含量。最后,通过模式识别中最小误差平方(LMSE)算法设计一个线性分类器,并采集了77组已知类别样本训练分类器,使仪器初步达到了葡萄变质自动报警的目的。Grapes are liable to deteriorate in transit and storage. But large- scale deterioration can be effectively reduced if the beginning deterioration were detected. Fourier transform infrared spectroscopy indicated that CO2and ethanol are the most significant volatile compounds in the grape deterioration process. Thus,a CO2gas sensor module and an ethanol electrochemistry gas sensor were selected for designing a low- cost and small monitoring device,which can measure the concentration of CO2and ethanol real- time and quantificational. At last,a linear classifier was designed by using the least square method,and was trained by 80 known- category samples. The classifier makes it possible making the device able to alarm the beginning deterioration.

关 键 词：葡萄劣变 CO2传感器 乙醇传感器 LMSE 线性分类器 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]