基于三维荧光光谱和ISSA-SVM的食用植物油鉴别

Classification of edible vegetable oils based on three-dimensional fluorescence spectroscopy and ISSA-SVM

[目的]提高食用植物油的分类精度,建立基于三维荧光光谱和ISSA-SVM的食用植物油鉴别模型。[方法]结合三维荧光光谱特征信息,运用改进的麻雀搜索算法优化SVM模型参数,构建一个融合三维荧光光谱信息特征和ISSA-SVM模型的食用植物油鉴别方法。[结果]与SVM模型、GA-SVM模型、PSO-GA模型和SSA-SVM模型相比,ISSA-SVM模型的食用植物油分类精度最高,为100%。[结论]ISSA-SVM模型具有更高的收敛效率、系统稳定性以及避免局部最优解的能力,可以有效应对复杂多变的样本分类任务。

[Objective]To improve the classification accuracy of edible vegetable oils,an identification model based on three-dimensional fluorescence spectroscopy and ISSA-SVM was established.[Methods]Combining the feature information of three-dimensional fluorescence spectroscopy,an improved sparrow search algorithm was used to optimize the parameters of the SVM model,constructing an edible vegetable oil identification method that integrates the characteristics of three-dimensional fluorescence spectroscopy information and the ISSA-SVM model.[Results]Compared with the SVM model,GA-SVM model,PSO-SVM model,and SSA-SVM model,the classification accuracy of the ISSA-SVM model for edible vegetable oils reached 100%.[Conclusion]The ISSA-SVM model has higher convergence efficiency,system stability,and the ability to avoid local optimal solutions,which can effectively cope with complex and variable sample classification tasks.