文献类型：期刊文献

英文题名：Process modeling method based on an improved elman neural network

作者：Qi, Zhongji[1]; Liu, Mandan[1]; Wang, Honggang[2]

机构：[1] College of Information Science and Technology, East China University of Science and Technology, Shanghai 200237, China; [2] State-Key Laboratory of Chemical Engineering, East China University of Science and Technology, Shanghai 200237, China

年份：2008

起止页码：8184

外文期刊名：Proceedings of the World Congress on Intelligent Control and Automation (WCICA)

收录：EI(收录号：20083911600629)

语种：英文

外文关键词：State feedback - Elman neural networks - Ethylene - Cracks

摘要：A new improved Elman Neural Network, OAIF-Elman(Output-Add-Input Feedback Elman) network is presented based on the Elman neural network in this paper. The new Elman network is endowed with state-feedback and output-feedback by an additional output-context layer, which is added to the former input to get a new input of network. It is applied for setting up ethylene cracking severity soft sensor model, and the parameters of network are optimized by genetic algorithm. The simulation results indicate that the OAIF-Elman network presented in this paper have better fitting and predictive result than basic Elman and other 4 kinds of improved Elman network under the same conditions, meanwhile the satisfying model of cracking severity is obtained. ? 2008 IEEE.