| 一种基于混合建模的模块化神经网络控制方法 |
| |
| 引用本文： | 杜庆东,刘杰,赵海. 一种基于混合建模的模块化神经网络控制方法[J]. 计算机工程与应用, 2006, 42(11): 215-217 |
| |
| 作者姓名： | 杜庆东 刘杰 赵海 |
| |
| 作者单位： | 沈阳师范大学软件学院,沈阳,110034;东北大学信息学院,沈阳,110006 |
| |
| 基金项目： | 辽宁省教育厅资助项目;吉林省沈阳市重点项目 |
| |
| 摘 要： | 针对传统神经网络用于复杂过程系统的控制时难于收敛的问题,文章提出了基于混合建模的模块化的神经网络模型。采取运行机理建模和神经网络建模相结合的方式,把输入样本空间进行划分,实现基于混合专家网络的建模。试验结果表明,对大型燃煤锅炉供热系统,文章提出的方法可以较好地提高供热系统的稳定性和供热质量。
|
| 关 键 词： | 神经网络 信息融合 供热控制 |
| 文章编号： | 1002-8331-（2006）11-0215-03 |
| 收稿时间： | 2005-09-01 |
| 修稿时间： | 2005-09-01 |
A Modular Neural Networks Control Method Based on Hybrid Modeling |
| |
| Du Qingdong,Liu Jie,Zhao Hai. A Modular Neural Networks Control Method Based on Hybrid Modeling[J]. Computer Engineering and Applications, 2006, 42(11): 215-217 |
| |
| Authors: | Du Qingdong Liu Jie Zhao Hai |
| |
| Affiliation: | School of Software,Shenyang Normal University,Shenyang 110034;School of Information, Northeastern University, Shenyang 110006 |
| |
| Abstract: | Against difficult convergence problem when the traditional Neural Networks are applied to complex process system,a hybrid expert neural networks model is brought forward.In the new method,input samples are divided into sub-space and hybrid expert neural networks model is realized.The experiment indicates that better stability and quality can be obtained through the new neural networks model in the huge coal-burning boiler heating system. |
| |
| Keywords: | neural networks information fusion heating control |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
