支持向量机在纸浆蒸煮过程Kappa值软测量中的应用研究

Application of Support Vector Machine Method in Soft Measurement of Kappa Number of Kraft Pulping Process

针对纸浆蒸煮过程机理复杂、影响因素众多和数据不完备条件下纸浆Kappa值预报问题,在介绍支持向量机基本原理和实现算法的基础上,探讨了支持向量机方法在纸浆Kappa值预报中的应用,经过与线性回归方法和人工神经网络方法预报结果比较,表明该方法具有精度高、速度快、泛化能力强的特点,取得了较传统建模方法更好的预报效果。

Aiming at the problem of predicting Kappa number of kraft pulping process under circumstances of complicated process kinetics and poor basic information, the support vector machine method is introduced. The basic theory and algorithm of the method are presented, and application of the method to predict Kappa number is conducted. Comparison is made between SVM methods and the traditional methods (linear regression and artificial neural network). The comparative result indicates that SVM method is high in precision, faster in computation and has a better generalization ability. Based on this new method, we get the better result to predict the Kappa number than the traditional methods.