This article attempts to determine the most accurate classification method for different abalone-flavoring liquids. Three common voltammetric detection methods, namely, linear sweep voltammetry (LSV), cyclic voltammetry (CV), and square-wave voltammetry (SWV), were considered. To compare their classification accuracies of abalone-flavoring liquids, three methods were separately adopted to classify five different abalone-flavoring liquids, using a four-electrode (Au, Pt, Pd, and W) sensor array. Then the data acquired by each method were subject to the principal component analysis (PCA): the first three principal components whose eigenvalues were greater than 1 were extracted from each set of data; the cumulative variance contribution rate and the principal component scores of each method were obtained. The PCA results show that the first three principal components obtained by the CV had the highest cumulative variance contribution rate (91.307%), indicating that the CV can more comprehensively characterize the information of abalone-flavoring liquid samples than the other two methods. According to the principal component scores, compared with those of LSV and SWV, the same kind of samples detected by the CV were highly clustered and the different kinds of samples detected by the CV were greatly dispersed. This indicates that the CV can effectively distinguish between the five abalone-flavoring liquids. Finally, the detection data were further verified through probabilistic neural network and a support vector machine algorithm optimized by genetic algorithm. The results further confirm that the CV is more accurate than the other two methods in the classification of abalone-flavoring liquids. Therefore, the CV was recommended for the classification of abalone-flavoring liquids.

中文翻译：

---

鲍鱼味液体伏安法检测方法的比较

本文试图为不同的鲍鱼调味液确定最准确的分类方法。考虑了三种常见的伏安法检测方法，即线性扫描伏安法（LSV），循环伏安法（CV）和方波伏安法（SWV）。为了比较鲍鱼调味液的分类准确性，分别采用三种方法，使用四电极（Au，Pt，Pd和W）传感器阵列对五种不同的鲍鱼调味液进行分类。然后，对每种方法获取的数据进行主成分分析（PCA）：从每组数据中提取特征值大于1的前三个主成分；获得了每种方法的累积方差贡献率和主成分得分。PCA结果表明，通过CV获得的前三个主成分具有最高的累积方差贡献率（91.307％），这表明CV比其他两种方法可以更全面地表征鲍鱼味液体样品的信息。根据主成分评分，与LSV和SWV相比，CV检测到的相同种类的样品高度聚类，CV检测到的不同种类的样品大大分散。这表明CV可以有效地区分五种鲍鱼调味液。最后，通过概率神经网络和遗传算法优化的支持向量机算法对检测数据进行进一步验证。结果进一步证实，在鲍鱼调味液的分类中，CV比其他两种方法更准确。因此，推荐使用CV对鲍鱼调味液进行分类。