中国机械工程2012，Vol.23Issue(1)：51-54,4.
基于主成分分析的BP神经网络内螺纹冷挤压成形质量预测
Forming Quality Forecast for Internal Threads Formed by Cold Extrusion Based on Principal Component Analysis and Neural Networks
摘要
Abstract
Forming quality grade of internal threads formed by cold extrusion was rated synthetically by BP neural network based on pitch diameter, thread pitch, half of thread angle and threads height ratio. For eliminating linear relevance in inter-influencing factors while the data pre-processing, major factors that affected the forming quality of internal threads formed by cold extrusion were extracted by principal component analysis. The experimental results show that the neural networks input by the processed data by this method become simple, with improved convergence rate and forecast accuracy. This method realizes the forecast for quality grade of internal threads formed by cold extrusion precisely. Also it provides a new solution for detection of internal thread quality.关键词
内螺纹/成形质量预测/主成分分析/神经网络Key words
internal thread/forming quality forecast/principal component analysis/neural network分类
矿业与冶金引用本文复制引用
张敏,黎向锋,左敦稳,缪宏..基于主成分分析的BP神经网络内螺纹冷挤压成形质量预测[J].中国机械工程,2012,23(1):51-54,4.基金项目
空军装备部"十一五"预研项目 （）
