铸造砂型质量预估的灰色关联法与支持向量机法比较

The Comparison of Gray Relative Analysis and Support Vector Machine in Molding Sand Quality Prediction

型砂性能直接影响了最终砂型的质量，而决定型砂陛能的指标较多且具有模糊性、灰色陛，从而导致砂型质量的预测极为复杂、困难。提出了铸造砂型质量预测的支持向量机建模方法，并且与已有的灰色关联分析方法相比较。实例分析证明，支持向量回归估计模型预测型砂性能更加科学、合理。

The properties of molding sand is vital to the quality of sand mold. The indexes of sand property are multiplex, ambiguity and disfigurement. The prediction of sand mold quality is difficult. A support vector machine model is concluded for the molding sand quality prediction. This model is also be compared with gray relative analysis method. The case study has shown that using vector regression model to predict sand quality is more reasonable.