基于UV-Vis检测养殖水体中化学需氧量含量研究

发布时间：2018-05-23 08:50

本文选题：极限学习机 + 化学需氧量　；参考：《光谱学与光谱分析》2016年01期

【摘要】：采用紫外可见光谱(UV-Vis)与极限学习机算法检测水体化学需氧量(chemical oxygen demand,COD)含量研究。采集135份水样进行紫外可见波段全光谱扫描,结合变量标准化(standard normal variate,SNV),多元散射校正(MSC)和一阶微分(1st D)对原始数据进行预处理,然后采用竞争性自适应重加权算法(competitive adaptive reweighted sampling,CARS)、随机青蛙(Random frog)算法和遗传算法进行特征波长选择。基于全光谱建立了偏最小二乘回归(partial least squares,PLS)和基于特征波长建立了极限学习机算法(extreme learning machine,ELM)模型。结果表明:使用CARS提取的9个特征波长建立的ELM模型的预测效果最优,决定系数R~2为0.82,预测均方根误差RMSEP为14.48mg·L~(-1),RPD值为2.34。说明使用CARS变量选择算法获取UV-Vis光谱特征波长,应用极限学习机建模,可以准确、快速的检测养殖水体中COD含量,为实现养殖水体COD的动态快速检测以及水体其他微量物质含量参数检测打下基础。
[Abstract]:The content of chemical oxygen and COD in water was detected by UV-Visand extreme learning machine algorithm. A total of 135water samples were collected for ultraviolet and visible band full-spectral scanning, and the original data were preprocessed by combining with standard normal variates, multivariate scattering correction (MSC) and first-order differential 1st D). Then competitive adaptive reweighted sampling algorithm, Random Frog algorithm and genetic algorithm are used to select the characteristic wavelength. Based on full spectrum, partial least square regression (partial least squares) and extreme learning machine (LLM) algorithm based on characteristic wavelength are established. The results show that the prediction effect of the ELM model based on 9 characteristic wavelengths extracted by CARS is the best, the determination coefficient Rn2 is 0.82and the root mean square error (RMSEP) of the prediction is 14.48mg Leng ~ (-1) RPD (2.34). It is shown that using CARS variable selection algorithm to obtain the characteristic wavelength of UV-Vis spectrum and modeling with extreme learning machine can accurately and quickly detect the content of COD in aquiculture water. It lays a foundation for the rapid and dynamic detection of COD in aquaculture water and the detection of other trace material content parameters.
【作者单位】：华东交通大学轨道交通学院;
【基金】：国家自然科学基金项目(61565005) 江西省科技支撑项目(20142BDH80021,20151BAB207009)资助
【分类号】：S912;O657.32

【相似文献】