文献类型：期刊文献

英文题名：A genetic neural fuzzy system-based quality prediction model for injection process

作者：Li, Erguo[1]; Jia, Li[1]; Yu, Jinshou[1]

机构：[1]E China Univ Sci & Technol, Res Inst Automat, Shanghai 200237, Peoples R China

年份：2002

卷号：26

期号：9

起止页码：1253

外文期刊名：COMPUTERS & CHEMICAL ENGINEERING

收录：;EI(收录号：2002417135330);WOS:【SCI-EXPANDED(收录号:WOS:000178688800008)】；

语种：英文

外文关键词：neural fuzzy system; genetic algorithm; B-P algorithm; quality prediction; injection process

摘要：In this paper, a genetic neural fuzzy system (GNFS) is presented and a hybrid learning algorithm divided into two stages is proposed to train GNFS. During first learning stage, Genetic algorithm is used to optimize the structure of GNFS and the membership function of each fuzzy term because of its capability of parallel and global search. On the basis of optimized training stage, the back-propagation algorithm (B-P algorithm) is chosen to update the parameters of GNFS to improve the system precision. The proposed GNFS is used to predict the weight of modeled part in injection process. The process of constructing quality prediction model for injection process based on GNFS is introduced. The results predicted by the constructed model show it can perform very well. The comparison between the presented GNFS and the other model based on regression and the neural network is made. The comparison verifies the proposed GNFS has superior performance and good generalization capability and also can apply to other industrial process. (C) 2002 Elsevier Science Ltd. All rights reserved.