基于支持向量机的X射线荧光光谱重金属检测模型的建立

Establishment of X-ray fluorescence heavy metal detection model based on support vector machine

目的意义:建立土壤中As、Cr、Cu、Pb、Zn等5种重金属的定量检测模型,为土壤重金属的快速检测提供一种新思路。方法:利用X射线荧光光谱技术结合支持向量机对土壤中的5种重金属元素进行定量检测,在对检测数据进行(-1,1)归一化处理以及K折交叉验证寻优的基础上,讨论了基于网格寻优算法、粒子群算法以及遗传算法的支持向量机回归模型,通过比较均方根误差,得出采用遗传算法进行优化后建模效果最佳。结果:将遗传算法参数设为进化代数200次,种群数量20,交叉率0.4,变异率0.1,在此条件下建模并验证模型准确度和精密度,得到5种重金属检测模型预测值与检测值间决定系数r2分别为0.9821、0.958、0.9764、0.9673和0.9684,交叉验证均方根误差与模型训练集、测试集均方根误差数值较低。结论:模型预测精度高,相关性显著,能够很好的预测土壤中的5种重金属含量,对于提高仪器的快速、准确测定有着重要的意义。

An X-ray fluorescence spectrum quantitative detection models for five heavy metals（Cr,Cu,Zn,Pb and As）in soil were established based on support vector machine.The experiment results showed that the model had a significant correlation;the predicted values were close to the reference values,which indicated a good modeling effect for predicting heavy metals in soil.