摘要
作为中国制造重要支柱的铸造业目前面临合格率低、成本高的严峻形势,原因在于过分依赖经验、忽视数据的分析利用。通过影响铸件质量的多因素数据积累,利用主元分析排除次要因素,并借助多元回归将关键因素和质量特性指标相关联,建立起一个预测铸件产品质量的数学模型。通过生产验证其有效性,有助于优化生产条件,提高铸件质量。
Being the important part of ＂Made in China＂, the casting enterprises are facing the serious problem such as low eligibility and high cost because of relying too much on experiences and neglecting the meaning of data. This paper is to construct a casting-product quality model by the data with help of PCA and regression. The effectiveness has been proved by actual data and lent itself to improving product quality.
出处
《工业工程与管理》
2006年第3期32-36,共5页
Industrial Engineering and Management
关键词
质量管理
主元分析
回归分析
quality management
PCA
regression analysis
