| 精馏塔产品组分的非线性主元回归估计器 |
| |
| 引用本文： | 文忠,俞金寿,华向明. 精馏塔产品组分的非线性主元回归估计器[J]. 石油化工自动化, 1999, 0(1): 23-26 |
| |
| 作者姓名： | 文忠 俞金寿 华向明 |
| |
| 作者单位： | 华东理工大学自动化研究所,上海,200237 |
| |
| 摘 要： | 以一个理想二元精馏塔为对象，讨论基于塔板温度的产品组分估计问题。并指出，相对于线性部分最小二乘估计器，基于神经网络的非线性主元回归估计器具有更好的稳态精度和动态性能。
|
| 关 键 词： | 非线性主元回归 精馏塔 估计器 产品质量 |
A Nonlinear Principal Regression Estimator for Product Composition of a Distillation Tower |
| |
| Wen Zhong,Yu Jinshou,Hua Xiangming. A Nonlinear Principal Regression Estimator for Product Composition of a Distillation Tower[J]. Automation in Petro-chemical Industry, 1999, 0(1): 23-26 |
| |
| Authors: | Wen Zhong Yu Jinshou Hua Xiangming |
| |
| Affiliation: | Wen Zhong,Yu Jinshou Hua Xiangming. Research Institute of Automation,ECUST,Shanghai,200237 |
| |
| Abstract: | To estimate the product composition for an ideal binary distillation tower based on tray temperatures is discussed.After comparing linear principal regress and partial least square estimator,it is found that the nonlinear principal regression estimator based on the neural networks is more accurate in steady state and has better dynamic properties. |
| |
| Keywords: | Partial least squares Nonlinear principal component regression Experiment design Inferential control |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
