|
|||||
|
|
| 基于Bootstrap多神经网络的软测量方法 | |
| 引用本文： | 周长,张杰,吕文祥,刘先广,黄德先.基于Bootstrap多神经网络的软测量方法[J].控制工程,2009,16(4). |
| 作者姓名： | 周长 张杰 吕文祥 刘先广 黄德先 |
| 作者单位： | 1. 清华大学自动化系,北京,100084 2. 纽卡斯尔大学,化工与先进材料学院,英国,纽卡斯尔,NEI 3. 北京清大腾飞公司,北京,100084 |
| 基金项目： | 国家高技术研究发展计划(863计划) |
| 摘 要： | 针对原油蒸馏过程常规软测量模型难以适应原油进料性质变化的问题,提出Bootstrap多神经网络的非线性软测量处理策略.通过Bootstrap算法复制出训练集样本空间上的多个样本子空间,训练出多神经网络模型,避免了单个神经网络易于陷入局部最优及过度训练的弱点,具有较高的准确率和泛化能力.本处理策略用于建立常压塔一线干点的软测量模型,仿真结果表明模型预测准确率和鲁棒性较好,对原油性质变化具有较好的适应性.该方法将会改进实际蒸馏过程在进料性质变化情况下的产品质量指标的软测量精度. |
| 关 键 词： | 原油蒸馏 软测量 多神经网络 |
Soft-sensor Based on Bootstrap Aggregated Neural Network |
|
| ZHOU Chang,ZHANG Jie,LV Wen-xiang,LIU Xian-guang,HUANG De-xian.Soft-sensor Based on Bootstrap Aggregated Neural Network[J].Control Engineering of China,2009,16(4). | |
| Authors: | ZHOU Chang ZHANG Jie LV Wen-xiang LIU Xian-guang HUANG De-xian |
| Affiliation: | 1.Department of Automation;Tsinghua University;Beijing 100084;China;2.School of Chemical Engineering and Advanced Materials;Newcastle University;Newcastle upon Tyne NE1 7RU;UK;3.Beijing Techfly Company;China |
| Abstract: | A nonlinear soft-sensing strategy with bootstrap aggregated neural network is proposed to solve the poor adaptability of conventional soft-sensor methods when feedstock varies in crude oil distillation.A bootstrap aggregated neural network shows better accuracy and generalization capability than a single neural network which can be trapped in a local minimum or over-fitted the training data.The proposed strategy is used for developing a soft-sensor for the end point of kerosene product of a simulated atmosp... |
| Keywords: | Bootstrap crude oil distillation soft-sensor Bootstrap multiple neural network |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |
