摘要
以COD标准液为研究对象,基于特定激发波长下的荧光发射光谱数据,采用化学计量学算法对水质COD进行了检测,分析了水的温度、浊度和pH的变化对发射光谱的影响,并对相关参数的影响进行了补偿校正。在此基础上建立了多参量共同作用时对水质COD预测模型的补偿校正方法。首先采用荧光光谱法对浓度范围为1~55 mg·L^(-1)水质化学需氧量(COD)标准溶液进行三维荧光光谱的采集,去除散射峰以后采用基于蚁群算法优化的偏最小二乘法(ACO-iPLS)对不同激发波长(Ex=255~285 nm,间隔为5 nm)下的荧光发射光谱(Em=275~450nm)数据进行特征提取并采用基于粒子群优化的最小二乘支持向量机算法(PSO-LSSVM)进行预测模型的建立。结果表明,不同激发波长下的荧光发射光谱数据模型的检验集决定系数R■在0.961 8~0.998 1范围内,当采用波长为Ex=270 nm的激发光作用时所激发出的荧光发射光谱数据所建模型的效果最优,其检验集决定系数R■=0.998 1,预测均方根误差RMSEP=0.348 3 mg·L^(-1)。其次,对温度、浊度、 pH对荧光光谱法检测水质COD的影响进行了分析,并给出了相应的补偿模型。结果表明,温度和浊度在检测水质COD时对荧光光谱的影响不可忽略,但通过建立补偿模型可以对其影响进行有效的补偿校正,温度补偿后荧光数据模型的整体平均偏差Bias=0.130 6 mg·L^(-1),经浊度补偿后可以很好的校正浊度变化对荧光光谱法检测水质COD的影响,而pH范围在4~12.3内变化时对荧光光谱的影响相对较小,因此可忽略。最后,结合单一影响因素的分析结果,对荧光光谱法检测水质COD时水体的多种环境因素(温度、浊度、 pH)共同作用的影响进行了分析。实验结果表明,忽略pH影响后,可以采用对温度和浊度同时补偿的方法对二者的影响进行有效的校正。该结果可为水质参数光学传感器在调试过程中抑制环境因素的影响提供参考。
In this paper, the COD standard liquid is used as the research object, and the water COD is detected by the chemometrics algorithm based on the fluorescence emission spectrum data of specific excitation wavelength. During the detection process, the influences of temperature, turbidity and pH on the fluorescence spectrum are analyzed, and the compensation correction is performed on the influence of the related parameters. Firstly, excitation-emission matrix(EEM) spectra of the COD standard solution whose concentration ranges between 1 and 55 mg·L^-1 are collected by fluorescence spectrophotometer, after the scattering peaks are removed, the partial least squares based on the ant colony(ACO-iPLS) algorithm is used for extracting feature for the fluorescence emission spectra(Em=275~450 nm) at different excitation wavelengths(Ex=255~285 nm, with the interval 5 nm) and the least squares support vector machine algorithm with particle swarm optimization(PSO-LSSVM) is used to establish the prediction model. The results show that the determination coefficient of the validation set(R2^p) of the fluorescence emission spectrum data model at different excitation wavelengths is within the range of 0.961 8~0.998 1, of which the effect of the fluorescence emission spectrum data model at Ex=270 nm is the optimal, and the determination coefficient(R2^p) and the root mean square error of prediction(RMSEP) are R2^p=0.998 1, RMSEP=0.348 3 mg·L^-1, respectively. Secondly, the influences of temperature, turbidity and pH on the water COD detection by fluorescence spectrometry are analyzed, and the corresponding compensation model is obtained. The results demonstrate that the effect of temperature and turbidity on the fluorescence spectrum cannot be ignored, but the compensation model can be established to correct the interference effectively. The mean deviation(Bias) of fluorescence model after temperature compensation is 0.130 6 mg·L^-1, and the influence of turbidity change on COD detection by fluorescence spectrometry can be well corrected after turbidity compensation, while the effect of pH range in 4~12.3 on the fluorescence spectrum is relatively small, so it can be ignored. Finally, combined with the analysis results of single influence factors, the effects of various environmental factors(temperature, turbidity and pH) on the detection of water quality COD by fluorescence spectrometry are analyzed. The result shows that after neglecting the influence of pH, the influences of temperature and turbidity on the fluorescence spectrum can be corrected effectively. The results of the paper can serve as reference for water quality parameter optical sensors in suppressing environmental factors during commissioning.
作者
周昆鹏
白旭芳
毕卫红
ZHOU Kun-peng;BAI Xu-fang;BI Wei-hong(School of Physics and Electronic Information, Inner Mongolia University for Nationalities, Tongliao 028000, China;The Key Laboratory for Special Fiber and Fiber Sensor of Hebei Province, Yanshan University, Qinhuangdao 066004, China)
出处
《光谱学与光谱分析》
SCIE
EI
CAS
CSCD
北大核心
2019年第4期1097-1102,共6页
Spectroscopy and Spectral Analysis
基金
国家重点研发计划项目(2017YFC1403800)
河北省重点研发计划项目(18273302D)
内蒙古民族大学博士科研启动基金项目(BS432)
内蒙古自治区高等学校科学研究项目(NJZY17202)资助
关键词
水质化学需氧量(COD)
荧光光谱法
预测模型
环境因素
影响分析
Water chemical oxygen demand (COD)
Fluorescent spectroscopy
Prediction model
Environmental factors
Impact analysis