改进神经网络在芝麻油掺伪检测中的应用(朱清妍).docVIP

改进神经网络在芝麻油掺伪检测中的应用 朱清妍1，潘勇军2 （，）Application of improved neural network in sesame oil adulteration detection Zhu Qing-yan1，Pan yong-jun2 (Hunan College of Information , Hunan ,Changsha, 410200) Abstract：Study of sesame oil adulteration detection problems, improve the accuracy of detection. The traditional physical or chemical sesame oil adulteration detection method operation is complex, expensive equipment, defects to varying degrees. Combining near infrared spectroscopy technology and neural network advantages, improve a RBF neural network _ near infrared spectroscopy sesame oil adulteration detection method ( NIR-RBF). The first extraction of sesame oil by near infrared spectroscopy sample spectral information, and then the principal component analysis of main effective components extraction of spectral information, finally the main effective components input to the neural network learning, get sesame oil adulteration detection results. Using the model established on the incorporation of different types of plant oil and sesame oil were detected, the results show that, compared with other sesame oil adulteration detection method, NIR-RBF improves the detection precision and speed, to reduce the error of detection, is a fast and effective method for detection of adulteration in sesame oil. Keywords: near infrared spectroscopy；neural network；sesame oil；adulteration 1 引言 芝麻油又称香油，是一种古老的油脂品种，营养价值高于其它食用植物油，香味浓郁，是深受人们喜爱，市场售价远高于其它植物油[1]。不法商贩将大豆油、菜籽油、葵花籽油等掺入到芝麻油中，谋取更高的利润。为了保护消费者和生产经营者的合法利益，在加强对芝麻油生产和销售监管的同时，迫切需要建立一种快速、准确、科学、有效的检测方法，对芝麻油掺伪进行有效鉴别[2]。 目前主要采用物理或化学的方法对芝麻油掺伪进行检测，这些方法检测精度都不高，误差较大。近些年来，一些学者提出一些新检测方法。朱杏冬利用紫外分光光度法对芝麻油掺假进行检测，该方法简单快速，结果稳定可靠，但是需要对样品进行复杂的前期处理[3]。电子鼻技术是一种生物传感器分析技术，对掺伪油的定量分析结果较差[4]。相色谱是芝麻油掺伪检测中最常用的技术，检测精度相当的高，平均预测误差都小于5%[5]。吴静珠提出了红外光谱法的芝麻油掺伪检测，检测准确率有时达到97%，但是该方法只能进行定性分析，不能进行定量分析[6]。当前检测精度最高的为核磁共振法，，是一种最先进的芝麻油掺伪鉴别手段，总体检测正确率达到95%以上，但是该检测方法要借助于昂贵的设备，目前尚难于在国内推广普及[7]。近些年来随着光谱学发展，近红外光谱分析技术（Near Infrared Spectroscopy，NIR）得到了迅速发展，其具有操作简单、分析速度快、所需样品少等优点，在石油化工、制药、发