| 基于反应基元的非线性系统灰箱建模方法 |
| |
| 引用本文： | 曹柳林,孙娅苹,吴海燕.基于反应基元的非线性系统灰箱建模方法[J].中南大学学报(自然科学版),2011,42(2). |
| |
| 作者姓名： | 曹柳林 孙娅苹 吴海燕 |
| |
| 作者单位： | 北京化工大学自动化研究所,北京,100029 |
| |
| 基金项目： | 国家自然科学基金资助项目 |
| |
| 摘 要： | 提出一种基于反应基元的建立复杂非线性系统模型的灰箱建模方法.首先根据先验知识及系统特性分析引入过程的初始反应基元,并以此为出发点建立结构逼近神经网络模型,实现基元之间的关联,赋予网络节点实际的物理意义;然后,通过提出的最小化预测误差,结合逐步回归分析方法选择最优反应基元,优化网络结构,建立起表示系统变量关系的灰箱模型.以实际橡胶硫化促进剂制备的间歇反应过程作为实验对象,建立以生成物浓度为输出的数学模型,达到较高的输出预测精度.
|
| 关 键 词： | 灰箱建模 非线性系统 系统辨识 间歇反应 结构逼近神经网络 |
Grey-box modeling based on fundamental genes for nonlinear system |
| |
| CAO Liu-lin,SUN Ya-ping,WU Hai-yan.Grey-box modeling based on fundamental genes for nonlinear system[J].Journal of Central South University:Science and Technology,2011,42(2). |
| |
| Authors: | CAO Liu-lin SUN Ya-ping WU Hai-yan |
| |
| Institution: | CAO Liu-lin,SUN Ya-ping,WU Hai-yan (Institute of Automation,Beijing University of Chemical Technology,Beijing 100029,China) |
| |
| Abstract: | An approach of grey-box modeling based on fundamental genes was developed for modeling dynamic processes with non-linear characteristics.By combination the prior knowledge and systematic behaviors,structure approaching neural network(SANN) was established based on fundamental genes,and the nodes of SANN were given actual significance.Then the optimal fundamental genes were chosen through minimizing the proposed predicted error with stepwise regression analysis(SRA) to optimize the structure of SANN,so as to... |
| |
| Keywords: | grey-box modeling nonlinear system system identification batch reaction structure approaching neural network |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
