基于主成分分析的多路阀再制造质量回归模型优化  

Optimization Based on Principal Component Analysis of the Regression Model of Remanufacture Multiple Valve Quality

作　　者：李军[1] 聂春晖[1] 杨建中[1,2] 

机构地区：[1]重庆交通大学机电与汽车工程学院,重庆400074 [2]重庆港务集团有限公司,重庆401120

出　　处：《液压气动与密封》2013年第7期76-79,共4页Hydraulics Pneumatics & Seals

基　　金：重庆市科委自然基金项目(CSCT2006BB3410);重庆市教委自然科学项目(KJ090408)

摘　　要：采用主成分分析法对影响多路阀再制造产品质量的各生产工艺指标数据进行降维处理,提取主元,减少数据间线性相关性的影响,并建立基于主元的内部泄漏量回归模型。分析结果表明:基于主元的回归模型可简化模型结构,预测结果更接近实际实验结果。It is reduced that the dimension of production process indicators which influence the quality of remanufacture multiple valve products, extracts principal component, eliminates linear correlation between datum in this paper. And the prediction model of internal leakage of remanufaeture multiple valve product is established. The analysis and experimental result shows that the prediction model based on principal component is a simplified model structure, and the calculation results are more close to actual results.

关 键 词：再制造 多路阀 主成分分析 质量模型 

分 类 号：TH137[机械工程—机械制造及自动化]