基于拉曼光谱高斯分峰拟合的硫化锑含量检测方法

发布时间：2018-05-25 00:22

本文选题：拉曼光谱 + 高斯分峰拟合　；参考：《光谱学与光谱分析》2017年12期

【摘要】：针对硫化锑含量采用化学分析方法检测存在操作复杂、检测时间长的问题,提出了一种基于拉曼光谱高斯分峰拟合的硫化锑含量快速检测方法。采用拉曼光谱系统测定锑矿样品的拉曼谱图,对原始拉曼谱图进行平滑去噪、背景扣除、谱段选择及归一化等预处理。基于高斯曲线的高斯峰、峰面积、半高宽和峰位置等信息,建立单个高斯峰的数学模型,提出高斯分峰拟合算法对光谱进行拟合,采用状态转移算法对高斯峰模型进行优化求解,得到表征光谱信息的特征参量,结合偏最小二乘回归方法,确定特征参量和硫化锑含量之间的关系,从而建立模型,实现对硫化锑含量的预测。实验中通过训练样本建立校正模型,对测试样本进行预测,同时从训练样本中随机挑选出检验样本,利用已建立的模型,对其硫化锑含量进行预测,以检验模型的正确性和外推性。实验结果表明:与预处理后全光谱建模相比,采用高斯分峰拟合后建立的预测模型的预测效果更好,证明了模型的正确性和良好外推性。因此,拉曼光谱结合高斯分峰拟合算法应用于锑矿中硫化锑含量的检测是可行的,且测量过程更简单,适用于矿物成分的快速分析。
[Abstract]:In order to solve the problem of complex operation and long detection time of antimony sulfide content by chemical analysis, a rapid detection method of antimony sulfide content based on Gao Si peak fitting of Raman spectrum is proposed. Raman spectra of antimony ore samples were measured by Raman spectroscopy system. The original Raman spectra were pretreated by smoothing denoising, background subtraction, spectrum selection and normalization. Based on the information of Gao Si peak, peak area, half maximum width and peak position of Gao Si curve, the mathematical model of single Gao Si peak is established, and the fitting algorithm of Gao Si split peak is proposed to fit the spectrum. The state transfer algorithm is used to optimize the Gao Si peak model, and the characteristic parameters representing the spectral information are obtained. The relationship between the characteristic parameters and the content of antimony sulfide is determined with the partial least square regression method, and the model is established. The prediction of antimony sulfide content was realized. In the experiment, the calibration model of the training sample was established, and the test sample was predicted. At the same time, the test sample was selected randomly from the training sample, and the content of antimony sulfide was predicted by using the established model. To verify the correctness and extrapolation of the model. The experimental results show that the prediction model based on Gao Si peak fitting is more effective than the full-spectrum model after pretreatment, which proves the correctness and extrapolation of the model. Therefore, it is feasible to apply Raman spectroscopy and Gao Si peak fitting algorithm to the determination of antimony sulfide content in antimony ore, and the measurement process is simpler, which is suitable for the rapid analysis of mineral composition.
【作者单位】：中南大学信息科学与工程学院;
【基金】：国家自然科学基金项目(61473319);国家自然科学基金重点项目(61533021);国家自然科学基金创新研究群体项目(61621062) 中南大学创新驱动计划资助
【分类号】：O657.37;P575

【相似文献】