基于嗅觉可视化技术的食用植物油分类识别Classification and recognition of edible vegetable oils based on olfactory visualization technology
2023
杨干， 李大鹏，文韬，蒋涵，龚中良 YANG Gan, LI Dapeng, WEN Tao, JIANG Han, GONG Zhongliang
为实现山茶油与3种常见食用植物油(菜籽油、大豆油和玉米油)的区分，制备可视化传感器阵列，采用嗅觉可视化技术对4种不同种类的食用植物油进行分类识别。采用主成分分析(PCA)对4种油样的特征数据进行降维，然后将降维后的数据导入K近邻(KNN)、极限学习机(ELM)、支持向量机(SVM) 3种分类模型中进行模型参数优化，对比了3种分类模型的分类结果。结果表明：建立的SVM分类模型性能最优，当输入主成分向量数为7、c=1.741 1、g=4.549 8时，SVM分类模型的测试集分类识别准确率为95.8%，五折交叉验证准确率为89.6%。制得的可视化传感器阵列可以实现4种食用植物油的分类识别，嗅觉可视化技术用于食用植物油检测是可行的。In order to distinguish oil-tea camellia seed oil from three common edible vegetable oils (rapeseed oil, soybean oil and corn oil), visual sensor array was prepared, and four different edible vegetable oils were classified and identified by olfactory visualization technology. Principal component analysis (PCA) was used to reduce the dimension of the characteristic data of the four oil samples. The data after PCA dimensionality reduction was imported into three classification models namely K-Nearest Neighbor (KNN), Extreme Learning Machine (ELM), and Support Vector Machine (SVM), and the model parameters were optimized, and the classification results of the three classification models were compared. The results showed that the established SVM classification model had the best performance. When the number of input principal component vectors was 7, c=1.741 1, and g=4.549 8, the classification and recognition accuracy of the test set of the SVM classification model was 95.8%, and the 5-fold validation accuracy was 89.6%. The visual sensor array can achieve the classification and recognition of four edible vegetable oils, and the olfactory visualization technology is feasible for the classification and identification of edible vegetable oils.
Показать больше [+] Меньше [-]Библиографическая информация
Эту запись предоставил Directory of Open Access Journals
