LS-SVM Based Software Sensor for Fed-batch Yeast Fermentation and Comparative Studies

Least square support vector machine (LS-SVM) is a very powerful tool for pattern recognition and function estimation. In this paper, LS-SVM has been used to construct software sensors in an application to a fed-batch yeast fermentation process. Comparisons have been made between results from LS-SVM and software sensors using multiway partial least squares (MPLS) and extended Kalman filters (EKF). The LS-SVM algorithm is introduced firstly and then applied to a yeast fed-batch fermentation process to provide soft-sensing facilities. The soft-sensing capabilities of the LS-SVM approach are found to compare favorably with the results using EKF and MPLS