节点文献
结合主元分析和系统聚类的丙烯腈反应器优化
Optimization on reactor of acrylonitrile apparatus based on principal component analysis and system cluster
【摘要】 为提高丙烯腈收率,在分析丙烯腈生产工艺的基础上,结合主元分析与系统聚类两种方法对丙烯腈硫化床反应器进行优化。利用主元分析进行变量选取,然后用系统聚类法进行分类并提取优类样本。在分类操作中通过选用不同的判据得到几组最优运行工况。比较发现类平均距离法、离差平方和法要优于其他方法。优化后丙烯腈平均收率提高了近0.7%,试验结果说明了该方法是很有效的。
【Abstract】 In order to improve the acrylonitrile production ratio, PCA (Principal Component Analysis ) and system cluster was combined to optimize the acrylonitrile reactor, on the base of analyzing the techniques of the acrylonitrile apparatus. The variables were selected by PCA, then the classes were separated and the excellent samples were distilled. A group of optimal operation parameters were gotten by using some different criterions. The results showed that the centroid and ward are better than other methods. The acrylonitrile production ratio increased about 0.7 percent after optimization. The experiment results indicated that the optimization method is effective.
【Key words】 pattern recognition; principal component analysis(PCA); system cluster; acrylonitri;
- 【文献出处】 吉林大学学报(信息科学版) ,Journal of Changchun Post and Telecommunication Institute , 编辑部邮箱 ,2004年04期
- 【分类号】TQ226
- 【被引频次】2
- 【下载频次】73