支持向量机在丁二酸发酵过程建模中的应用

Application of SVM in modeling of succinic acid fermentation

支持向量机是一种新的机器学习算法,它采用结构风险最小化准则,能有效提高模型的泛化能力。本文针对生物转化法生产丁二酸发酵过程机理复杂、高度非线性、生物参数难以实时在线测量等特点,介绍了支持向量机回归建模算法在Matlab软件中的实现过程,对产物丁二酸浓度建立了预测模型,研究了SVM的小样本学习、泛化能力。仿真结果表明,与神经网络相比,SVM算法具有更好的推广能力,使得在未来工业化丁二酸发酵生产过程中针对丁二酸浓度的在线预估与优化控制成为可能。

Support vector machines （SVM） is a new machine learning algorithm, employing the criteria of structural risk minimization, which can improve the generalization of model. In accordance with the features of complicated mechanism, non-linear and hardship to get real-time and on-line biology parameters in succinic acid fermentation process, this paper introduces SVM algorithm and its application based on Matlab in dentils. The characteristics of SVM, such as the learning capability based on small samples and the good characteristic of generalization are presented by its applications to building prediction model of succinic acid concentration. The simulation shows that the SVM has a better generalization than others, and it is an effective method for succinic acid fermentation process modeling, predicting and control real-time on-line in the future industrialization.