由于蜂蜜蜜种多,成分复杂,加之蜂蜜掺假方式繁多,采用传统的方式很难对蜂蜜进行快速准确的鉴别.通过对国内多个地区的蜂蜜进行调研,采集来自全国20个省份多个蜜种的蜂蜜,利用中红外光谱仪对样品进行光谱扫描,采用主成分分析和聚类分析的方法,利用化学计量软件进行模型的建立.该识别模型不仅能较准确地判别蜂蜜是否掺假(准确率为95.36%),还能对添加量在10%以上的掺假方式进行预判,判别准确率为97.78%,符合判别模型的建立要求.利用中红外光谱技术对蜂蜜掺假进行鉴别的方法有效、可行.%Due to the honey composition is complicated and the honey adulteration ways are multifarious, the traditional way of honey identification is difficult to identify accurately. Through researching honey in several parts of the country, several kinds of honey were collected from 20 provinces all over the country, then the mid infra-red spectrum was used to scan the samples. Chemical metrology software was used to build a model by using the method of principal component analysis and cluster analysis. The recognition model could not only carry on the discrimination about whether honey adulteration (the accuracy rate was 95.36%), but also anticipate the adulteration way of additive amount more than 10%, the accuracy rate was 97.78%, which met the requirements of the discriminant model. The method for identifying honey adulteration by using mid infra-red spectrum is effective and feasible.
展开▼
