化学计量学法处理可见光谱定量测定纸浆卡伯值

Quantitative Determination of Pulp Kappa Number by Visible Light With Stoichiometric Methods

本文将一元线性回归、多元线性回归和主成分回归分析的化学计量学方法应用于纸浆卡伯值的在线测量中，通过测量蒸煮液的吸光度来预测纸浆的卡伯值。结果显示：三种方法所确立的测量模型均具有较好的学习精度和预测精度，其中尤以多元线性回归模型同时具有较高的学习精度和预测鲁棒性。

The pulp Kappa number is determined quantitatively by measurement of aborbance of cooking liquor.Three stoichiometric methods,monovariable linear regression,multivariable linear regression (MLR) and principal component regression (PCR) are applied for data processing.The results of data processing show that the predicting models determined by three methods all have good learning accuracy and predicting accuracy,and the best model is determined by multivariable linear regression.