期刊文献+
共找到3篇文章
< 1 >
每页显示 20 50 100
集成学习结合波长选取的有机物红外光谱定量回归方法研究 被引量:3
1
作者 鞠薇 鲁昌华 +2 位作者 张玉钧 陈晓静 蒋薇薇 《光谱学与光谱分析》 SCIE EI CAS CSCD 北大核心 2023年第1期239-247,共9页
研究集成学习方法在有机物红外光谱定量分析中的应用及特征波长选取方法对红外光谱集成学习建模效率和预测精度的影响。以柴油红外光谱的十六烷和总芳香烃含量为研究对象,首先采用极端随机森林(ERT)、线性核支持向量机(LinearSVM)、径... 研究集成学习方法在有机物红外光谱定量分析中的应用及特征波长选取方法对红外光谱集成学习建模效率和预测精度的影响。以柴油红外光谱的十六烷和总芳香烃含量为研究对象,首先采用极端随机森林(ERT)、线性核支持向量机(LinearSVM)、径向基核支持向量机(RBFSVM)和多项式核支持向量机(polySVM)作为基学习器,LinearSVM作为元学习器建立两层Stacking集成学习框架,分析比较单个基学习器与集成学习对柴油红外光谱的定量回归预测精度,与偏最小二乘(PLS)定量回归模型相比,Stacking集成学习模型对柴油光谱的两种有机物含量的预测精度均有提升,其中十六烷含量的ERT模型预测结果最优(r=0.848,RMSEP=1.603,RDP=2.627),总芳香烃含量的Stacking模型预测结果最优(r=0.991,RMSEP=0.526,RDP=9.243);进一步利用组合偏最小二乘(SiPLS)和连续投影算法(SPA)对红外光谱进行特征波长选取,利用优选出的红外光谱特征波长建立集成学习定量回归模型,其中十六烷含量的SiPLS-ERT模型预测结果最优(r=0.893,RMSEP=1.013,RDP=3.051),芳香烃含量的SiPLS-Stacking模型预测结果最优(r=0.998,RMSEP=0.354,RDP=11.475),且模型平均训练时间较全光谱训练时间减少50%以上,建模速度明显提高。研究结果表明,特征波长结合集成学习定量回归建模能够用于有机物红外光谱的定量分析中,与传统定量回归方法相比,该方法的建模效率和预测精度均有较大提高,为进一步研究机器学习在光谱定量分析中的应用提供相关方法支持。 展开更多
关键词 集成学习 定量回归 特征波长选取 有机物红外光谱
下载PDF
Enhanced performance of g-C_3N_4/TiO_2 photocatalysts for degradation of organic pollutants under visible light 被引量:5
2
作者 宋改雪 储震宇 +1 位作者 金万勤 孙红旗 《Chinese Journal of Chemical Engineering》 SCIE EI CAS CSCD 2015年第8期1326-1334,共9页
Photocatalytic degradation is one of the most promising remediation technologies in terms of advanced oxidation processes(AOPs) for water treatment. In this study, novel graphitic carbon nitride/titanium dioxide(gC3N4... Photocatalytic degradation is one of the most promising remediation technologies in terms of advanced oxidation processes(AOPs) for water treatment. In this study, novel graphitic carbon nitride/titanium dioxide(gC3N4/Ti O2) composites were synthesized by a facile sonication method. The physicochemical properties of the photocatalyst with different mass ratios of g-C3N4 to Ti O2 were investigated by X-ray diffraction(XRD), scanning electron microscope(SEM), transmission electron microscopy(TEM), N2 sorption, Fourier transform infrared spectroscopy(FT-IR), X-ray photoelectron spectroscopy(XPS), and UV–vis DRS. The photocatalytic performances were evaluated by degradation of methylene blue. It was found that g-C3N4/Ti O2 with a mass ratio of 1.5:1 exhibited the best degradation performance. Under UV, the degradation rate of g-C3N4/Ti O2 was 6.92 and 2.65 times higher than g-C3N4 and Ti O2, respectively. While under visible light, the enhancement factors became 9.27(to g-C3N4) and 7.03(to Ti O2). The improved photocatalytic activity was ascribed to the interfacial charge transfer between g-C3N4 and Ti O2. This work suggests that hybridization can produce promising solar materials for environmental remediation. 展开更多
关键词 Carbon nitride Titanium dioxide Composite Photodegradation Sonication
下载PDF
Structural and Vibrational Study of PCBM
3
作者 Alfredo Tlahuice-Flores Sergio Mejia-Rosales 《Journal of Chemistry and Chemical Engineering》 2011年第11期1034-1040,共7页
The 1-(3-methoxycarbonyl) propyl-1-phenyl-(6,6)C61 (also called PCBM) is a C60 derivative widely used as an electron-acceptor in organic solar cells. To date, all the infrared spectra reported are experimental, ... The 1-(3-methoxycarbonyl) propyl-1-phenyl-(6,6)C61 (also called PCBM) is a C60 derivative widely used as an electron-acceptor in organic solar cells. To date, all the infrared spectra reported are experimental, mainly because of the calculations needed to study these structures are highly time-consuming. In this report we address for the first time the infrared spectrum calculation of PCBM with Cs symmetry by using the PW91/dnp level as implemented in the Dmol3 code. In this calculation we have found two intense peaks in the IR spectrum, that agree fairly with the 1187 and 1787 cm^-1 measured experimentally. 展开更多
关键词 PCBM DFT IR solar cells Cs symmetry
下载PDF
上一页 1 下一页 到第
使用帮助 返回顶部