论文部分内容阅读
提出了利用近红外光谱技术对重金属污染泥蚶的快速检测方法。以冷冻干燥磨粉的泥蚶肌肉为试验对象,方案设计分为两种:对照泥蚶和单一重金属(Cd,Cu,Pb或Zn)污染泥蚶的分类(设计Ⅰ);所有样本包括对照泥蚶和4种重金属污染泥蚶的分类(设计Ⅱ)。采用两种识别算法,即最小二乘支持向量机和随机森林,对设计I和设计II分别建立分类模型并进行预测。预测结果表明:设计Ⅰ:最小二乘支持向量机和随机森林的平均预测正确率分别为100%和95%;设计Ⅱ:最小二乘支持向量机和随机森林的平均预测正确率分别为96%和92%。利用近红外光谱技术快速检测重金属污染泥蚶具有可行性,可为泥蚶重金属污染提供一种快速检测方法。
A rapid detection method of heavy metal contaminated sludge using near infrared spectroscopy was proposed. Freeze-dried milled muscle was used as test object. The scheme was divided into two types: control (control Ⅰ), control Ⅰ (control Ⅰ) and contamination control Ⅱ (control Ⅰ); all samples included control mud蚶 and four kinds of heavy metal pollution mud 蚶 classification (Design Ⅱ). Using two recognition algorithms, namely, least-squares support vector machine and random forest, classification models are designed and predicted for design I and design II, respectively. The results show that: Design Ⅰ: the average prediction accuracy of least squares support vector machines and random forests are 100% and 95% respectively; Design Ⅱ: the average prediction accuracy of least squares support vector machines and random forests are 96% And 92%. The rapid detection of heavy metal contaminated sludge by near-infrared spectroscopy is feasible, which can provide a rapid detection method for heavy metal pollution of sludge.