基于麻雀搜索算法的土壤重金属X射线荧光光谱重叠峰解析  

Overlapping Peak Analysis of Soil Heavy Metal X-Ray Fluorescence Spectra Based on Sparrow Search Algorithm

作　　者：陈颖[1] 刘峥莹 肖春艳[2] 赵学亮 李康[3] 庞丽丽[3] 史彦新[3] 李少华[4] CHEN Ying;LIU Zheng-ying;XIAO Chun-yan;ZHAO Xue-liang;LI Kang;PANG Li-li;SHI Yan-xin;LI Shao-hua(Hebei Province Key Laboratory of Test/Measurement Technology and Instrument,School of Electrical Engineering,Yanshan University,Qinhuangdao 066004,China;School of Resources and Environment,Henan Polytechnic University,Jiaozuo 454000,China;Center for Hydrogeology and Environmental Geology,China Geological Survey,Geological Environment Monitoring Engineering Technology Innovation Center of The Ministry of Natural Resources,Baoding 071051,China;Hebei Sailhero Environmental Protection Hi-tech Co.,Ltd.,Shijiazhuang 050000,China)

机构地区：[1]燕山大学电气工程学院河北省测试计量技术及仪器重点实验室,河北秦皇岛066004 [2]河南理工大学资源与环境学院,河南焦作454000 [3]中国地质调查局水文地质环境地质调查中心,自然资源部地质环境监测工程技术创新中心,河北保定071051 [4]河北先河环保科技股份有限公司,河北石家庄050000

出　　处：《光谱学与光谱分析》2021年第7期2175-2180,共6页Spectroscopy and Spectral Analysis

基　　金：国家重点研发计划项目(2018YFC1800903,2016YFC1400601-3);河北省重点研发计划项目(19273901D,20373301D);河北省自然科学基金项目(F2020203066);中国博士后基金项目(2018M630279);河北省博士后择优资助项目(D2018003028);河北省高等学校科学技术研究项目(ZD2018243)资助。

摘　　要：近年来随着土壤重金属污染的加剧,和人们环境意识的逐渐提高,科研人员对快速检测土壤重金属含量方法的研究正在不断深化。目前,X射线荧光分析法(XRF)是广泛应用于土壤重金属污染检测的方法。但由于X射线荧光光谱仪的能量分辨率有限,而一些重金属元素的荧光产额较低,一些元素的相邻谱峰出现了重叠现象。针对XRF法中元素相邻谱峰的重叠问题,提出了一种基于麻雀搜索算法(SSA)的光谱重叠峰解析方法。首先,将从河北保定地区采样得到的土壤,制备出不同含水率、不同重金属元素含量的样本并用X射线荧光光谱仪获取原始光谱数据。接着,对光谱数据进行预处理,采用谱聚类算法剔除异常光谱样本,采用Savitzky-Golay五点二次去噪法和线性本底法完成对光谱的去噪和本底扣除,并对光谱净计数用随机数法生成大量模拟光谱数据,以备后续算法使用。然后,用期望最大化法(EM)对重叠峰进行初步解析,首先设置EM算法的初始参数,并将生成的模拟光谱数据代入EM算法,当达到迭代次数时,即可初步得到高斯混合模型(GMM)中各高斯峰的期望、方差和权重参数。但由于EM算法容易受初始参数设置的影响,且易陷入局部最优而导致结果不准确,还需对EM算法进一步优化。本研究采用SSA对GMM的各参数进行全局优化,在设置SSA算法的基本参数后,将100组由EM算法得到的参数作为该算法的初始种群,并设置合适的适应度函数,通过迭代,最终得到全局最优参数,实现了重叠峰的分解。SSA受参数设置的影响较小,相比于一些传统的优化算法,如遗传算法(GA)、蚁群算法(ACO)、粒子群算法(PSO)等,具有收敛速度快、不易陷入局部最优的特点,因此,采用此算法,可以达到较好的优化效果。通过对重叠峰解析结果的分析表明,该算法可在较少的迭代次数下得到较准确的解析结果,可广泛应用于能谱�In recent years,with the aggravation of soil heavy metal pollution and the gradual improvement of people’s environmental awareness,the research on the rapid detection method of soil heavy metal content has been strengthened rapidly.At present,X-ray Fluorescence analysis(XRF)has been widely used to detect heavy metal pollution in soil.However,due to the limited energy resolution of the X-ray fluorescence spectrometer and the low fluorescence yield of some heavy metal elements,overlapping phenomena occurred in adjacent spectral peaks of some elements.In the cause of overlapping phenomenon often appears between adjacent peaks in X-ray Fluorescence analysis(XRF),a new overlapping peak analysis method based on Sparrow Search Algorithm(SSA)was proposed.Firstly,samples with different moisture content and heavy metal element content were prepared,and original spectral data were obtained by X-ray fluorescence spectrometer from the soil sampled of Baoding,Hebei.Then,the spectral data were preprocessed,the spectral clustering algorithm removed the abnormal spectral samples,the spectral denoising and background subtraction were completed by the Savitzky-Golay five-point quadratic denoising method and the linear background method.The random number method is used to generate a large number of simulated spectral data for the use of subsequent algorithms.After that,expectation-maximization(EM)was applied to analyze overlapping peaks preliminarily.Set the initial parameters of the EM algorithm,and put simulation spectra data into the EM algorithm.When it reached the maximum number of iterations,can preliminarily get parameters of the Gaussian Mixture Model(GMM),expectation,variance and weights of each Gaussian peaks.However,the EM algorithm is easily affected by the initial parameter and is prone to fall into the local optimum,leading to inaccurate results.Therefore,further optimization of the EM algorithm is needed.In this study,SSA was used for global optimization of parameters of the GMM.After setting the basic SSA algorithm

关 键 词：X射线荧光分析法 高斯混合模型 期望最大化法 麻雀搜索算法 重叠峰解析 

分 类 号：O433.4[机械工程—光学工程]