Soft sensor of Lysine fermentation based on fuzzy support vector machines

Abstract

Fermentation process is a complex nonlinear dynamic coupling process. As it is very difficult to measure the primary biolog parametre online. It limits the advanced control in the biochemical processes. The least-squares support vector machine which used to model the Lysine fermentation process by soft sensor is established. In order to overcome the overfitting problem caused by noises and outliers in support vector machines, a fuzzy membership model based on support vector data decscription is proposed to fuzzify all the training data.Then the model is introduced into least square support vector machines. The simulation example shows that the FLS-SVM could measure the key parameters, which could not be measured online during the course of lysine fermentation, with a high precision.