| 基于BP神经网络在卷烟焦油预测中的应用 |
| |
| 引用本文： | 杨再波,叶冲,韩伟,钟才宁,孙成斌,毛海立.基于BP神经网络在卷烟焦油预测中的应用[J].黔南民族师范学院学报,2006,26(6):5-8. |
| |
| 作者姓名： | 杨再波 叶冲 韩伟 钟才宁 孙成斌 毛海立 |
| |
| 作者单位： | 1. 黔南民族师范学院,化学化工系,贵州,都匀,558000
2. 贵州黄果树烟草集团公司,技术中心,贵州,贵阳,550003 |
| |
| 摘 要： | 提出利用数值优化改进的BP算法建立卷烟焦油的预测模型,以卷烟常规化学成分总糖、总氮、总氯作为神经网络的输入,焦油作为输出,并将这些指标作归一化处理,然后通过教师样本数据对网络进行充分的训练,获得适宜的参数矩阵,最后用训练好的网络对检验样本数据进行预测,预测效果相当显著,相对偏差在-1.81%~1.31%之间。
|
| 关 键 词： | 常规化学成分 神经网络 BP算法 卷烟 焦油 |
| 文章编号： | 1005-6769（2006）06-0005-04 |
| 收稿时间： | 2006-09-04 |
| 修稿时间： | 2006年9月4日 |
Application of BP Artificial Neural Network Forcasting the Cigarette Tar |
| |
| YANG Zai-bo,YE Chong,HAN Wei,ZHONG Cai-ning,SUN Cheng-bin,MAO Hai-li.Application of BP Artificial Neural Network Forcasting the Cigarette Tar[J].Journal of Qiannan Normal College of Nationalities,2006,26(6):5-8. |
| |
| Authors: | YANG Zai-bo YE Chong HAN Wei ZHONG Cai-ning SUN Cheng-bin MAO Hai-li |
| |
| Institution: | 1. Department of Chemistry; Qiannan Normal University for Nationalities; Duyun 558000, China; 2. Technology of the centre GuiZhou huangguoshu tobacco group GuiYang 550003, Chnia |
| |
| Abstract: | The BP neural network, which was improved through the numerical optimization method, was applied to build the model d forecasting the cigarette tar, The general chemical components - total sugar, total nitrogen and total chlorine is as the inputs of the network and the tar is as the outputs. Then standardizing the general chemical index and training the network well though training - samples to obtain the best network parameters vector, Finally, making use of the successfully trained network to forecast the tar content of the testing - sample. The study showed that the method provided better result than other reported ones and has promising in cigarette industry, the error is from- 1,81% to 1.31%. |
| |
| Keywords: | General chemical components Neural Network BP Cigarette Tar |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
