摘要
近年来,葡萄籽油因其丰富的营养成分和健康益处而受到广泛关注,但市场上掺假现象日益严重,迫切需要一种快速、准确的检测技术。针对以上应用需求,本文旨在开发一种基于Sammon映射结合极限学习机(ELM)的方法,以实现可见差分光谱对葡萄籽油掺假浓度的高效检测。首先,利用可见差分光谱技术获取不同掺假浓度的葡萄籽油样本的光谱数据。通过Sammon映射方法,对获得的光谱数据进行特征提取和降维处理,有效消除噪声和冗余信息,提取出与掺假浓度相关的主要特征。Sammon映射通过保留数据的局部结构,增强了样本间的可分性,为后续的建模提供了更加清晰的特征空间。接着,采用极限学习机进行模型训练和预测。ELM具有训练速度快、泛化能力强的优点,能够在较短时间内建立高效的分类模型。通过对训练集的学习,ELM能够准确地识别不同掺假浓度的葡萄籽油,并实现对未知样本的快速预测,模型对测试集预测的RMSE为0.0051,R 2为0.94。实验结果表明,结合Sammon映射的极限学习机在检测葡萄籽油掺假浓度方面展现了优异的性能。该方法不仅提高了检测的可靠性和准确性,还为葡萄籽油的质量控制提供了一种新的技术手段。
In recent years,grape seed oil has attracted extensive attention because of its rich nutritional components and health benefits.However,the adulteration phenomenon in the market is becoming increasingly serious,and a rapid and accurate detection technology is urgently needed.Given the above application requirements,this paper aims to develop a method based on Sammon Mapping and extreme learning machine(ELM)to realize the efficient detection of grape seed oil adulteration concentration by visible difference spectroscopy.Firstly,the spectral data of grape seed oil samples with different adulterated concentrations were obtained using visible difference spectroscopy.Through the Sammon Mapping method,feature extraction and dimension reduction processing were carried out on the obtained spectral data to eliminate noise and redundant information effectively,and the main features related to adulteration concentration were extracted.By preserving the local structure of the data,the Sammon Mapping enhances the separability between samples and provides a clearer feature space for subsequent modeling.Then,an extreme learning machine is used for model training and prediction.Elm has the advantages of fast training speed and strong generalization ability and can establish an efficient classification model quickly.By learning from the training set,elm can accurately identify different adulterated concentrations of grape seed oil,and realize the rapid prediction of unknown samples.The RMSE and R 2 predicted by the model for the test set are 0.0051 and 0.94,respectively.The experimental results show that the extreme learning machine combined with the Sammon Mapping has excellent performance in detecting the adulterated concentration of grape seed oil.This method not only improves the reliability and accuracy of detection but also provides a new technical means for the quality control of grape seed oil.
作者
王旭
任晓鸽
WANG Xu;REN Xiaoge(College of Information Engineering,Xuchang Electrical Vocational College,Xuchang,Henan 461002)
出处
《光散射学报》
2026年第1期299-306,共8页
The Journal of Light Scattering
基金
中国高校产学研创新基金(2021ITA09018)
河南省重点研发与推广专项(232400410357)
河南省普通高校就业创业课题(JYB2025120)。
关键词
Sammon映射
极限学习机
可见透射光谱
大豆油
葡萄籽油
Sammon Mapping
Extreme learning machine
Visible transmission spectroscopy
Soybean oil
Grape seed oil
