Research and Application of Genetic Algorithm-based Optimized Radial Basis Neural Network Model Parameter Design

摘要：

In this paper,the structural features of the radial basis network as well as both the center value of the hidden node and the width parameter s influence on the structure are analyzed;the strategy of optimizing the center value and the width parameter by genetiC algorithm is researched. An above-algorithm based yarn quality forecast model is established,and the result shows that the predictive output of the model basically matches with the actually measured sample,and the network trained is capable of fast and accurately predict the quality indexes.

关键词： RBF Neutral Network Genetic Optimiz-ation Yarn Quality Forecast

作者: Xiujuan Fan

作者单位: Beijing Institute of Fashion Technology Beijing China

会议类型: 国际会议

会议名称: 2009 9th International Conference on Electronic Measurement & Instruments(第九届电子测量与仪器国际会议 ICEMI2009)

会议地点: 北京

会议语种:英文

页码: 1464-1467

在线出版日期: 2009-08-16（万方平台首次上网日期，不代表论文的发表时间）