| 小波包压缩-RBF网络同时测定润滑油中铁、铜、锌 |
| |
| 引用本文： | 程正军,张运陶.小波包压缩-RBF网络同时测定润滑油中铁、铜、锌[J].分析测试学报,2006,25(3):1-5. |
| |
| 作者姓名： | 程正军 张运陶 |
| |
| 作者单位： | 西华师范大学,应用化学研究所,四川,南充,637002 |
| |
| 摘 要： | 将小波包压缩-RBF网络方法用于润滑油中铁铜锌三组分分光光度同时测定该方法采用小波包函数对光谱数据进行压缩处理，用较大的小波色系数构成新的校正集和预测集代替原始的校正集和预测集，然后用RBF网络进行数据解析。研究表明，用bior2．4小波包处理原始测定数据，最佳小波基用logenerge熵标准，选择适当阈值将变量数由46个压缩成25个（压缩比为0．54），整体预测效果最好。将此方法用于合成样预测．预测结果与实际浓度的相对误差绝对值在2．50％～9．40％之间；用于实际润滑油样品中铁、铜和锌的同时测定，解析值与原子吸收法（AAS）的测定值的相对误差绝对值作3．55％～7．86％之间。应用结果令人满意。
|
| 关 键 词： | 小波包压缩 RBF网络 分光光度法 同时测定 润滑油 |
| 文章编号： | 1004-4957（2006）03-0001-05 |
| 收稿时间： | 2005-06-04 |
| 修稿时间： | 2005-09-13 |
Simultaneous Spectrophotometric Determination of Fe, Cu and Zn in Lubricating Oil Using Wavelet Packet Transform-RBF Neural Network |
| |
| CHENG Zheng-jun,ZHANG Yun-tao.Simultaneous Spectrophotometric Determination of Fe, Cu and Zn in Lubricating Oil Using Wavelet Packet Transform-RBF Neural Network[J].Journal of Instrumental Analysis,2006,25(3):1-5. |
| |
| Authors: | CHENG Zheng-jun ZHANG Yun-tao |
| |
| Affiliation: | Institute of Applied Chemistry, China West Normal University, Nanchong 637002, China |
| |
| Abstract: | A novel method for simultaneous speetrophotometric determination of into, copper and zinc in lubricating oil using wavelet packet transform - RBF neural network is presented. Spectral data of calibration and unknown samples were compressed by wavelet packet to obtain wavelet packet coefficient, respectively. Larger wavelet packet coefficient was used to make up the new calibration and prediction set, and the data analyzed by RBF neural network. Results sbowed that by using bior 2. 4 wavelet packet to treat the original determined data and log energe entropy standard for wavelet basis and by selecting appropriate threshold value to compress the variable number from 46 to 25, best prediction result would be obtained. The relative error of the difference of the results obtained by the predicted and actual concentrations ranged from 2. 50% - 9.40% and that obtained by AAS and this method for the determination of iron, copper and zinc in lubricating oil ranged frnm 3.55%-7.86%. The results are satisfactory. |
| |
| Keywords: | Data compression using wavelet packet transform Radial basis function neural networks Speetrophotometry Simultaneous determination Lubricating oil |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
