| Improved Marquardt Algorithm for Training Neural Networks for Chemical Process Modeling |
| |
| 作者姓名： | 吴建昱 何小荣 |
| |
| 作者单位： | WU Jianyu，HE Xiaorong Department of Chemical Engineering，Tsinghua University，Beijing 100084，China |
| |
| 摘 要： | IntroductionResearch on artificial neural networks(ANN) hasmade great progress during the past few years.Neural networks have been widely used in chemicalprocesses.Among all kinds of networks,the back-propagation (BP) network is the most commonchoice forits high capability of nonlinear mapping,study and classification. Through adjustingnetwork weights according to samples,the BPnetwork can simulate systems with complexnonlinear mapping relationships,such as chemicalprocesses. The most com…
|
Improved Marquardt Algorithm for Training Neural Networks for Chemical Process Modeling |
| |
| WU Jianyu,HE Xiaorong.Improved Marquardt Algorithm for Training Neural Networks for Chemical Process Modeling[J].Tsinghua Science and Technology,2002,7(5). |
| |
| Authors: | WU Jianyu HE Xiaorong |
| |
| Institution: | WU Jianyu,HE Xiaorong Department of Chemical Engineering,Tsinghua University,Beijing 100084,China |
| |
| Abstract: | Back-propagation (BP) artificial neural networks have been widely used to model chemical processes. BP networks are often trained using the generalized delta-rule (GDR) algorithm but application of such networks is limited because of the low convergent speed of the algorithm. This paper presents a new algorithm incorporating the Marquardt algorithm into the BP algorithm for training feedforward BP neural networks. The new algorithm was tested with several case studies and used to model the Reid vapor pressure (RVP) of stabilizer gasoline. The new algorithm has faster convergence and is much more efficient than the GDR algorithm. |
| |
| Keywords: | neural network Marquardt algorithm training |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
