Study on Soft Sensing Technology of Penicillin Fermentation Based on PLS and SVR

摘要

Penicillin fermentation process is highly nonlinear, time-varying and uncertain. Parameters such as key biological substrate concentration, cell concentration, product concentration of penicillin fermentation process are difficult to real-time measured on-line. Offline measuring method has time delay, and it is difficult to realize real-time control.in the fermentation process. In order to solve the problem, a soft measuring method combined with partial least squares (PLS) and support vector regression (SVR) is put forward. First of all, the principal components were extracted from input data space by PLS, and then the SVR algorithm was used to establish the soft sensor model which can accurately predict the important parameters of penicillin fermentation process. The experimental results shown that compared with the traditional modeling methods, the measuring accuracy of PLS-SVR soft measurement model has good tracking performance and strong generalization ability. The method proposed in the paper can meet online measurement of penicillin mycelium concentration in the fermentation process, the method is an effective for soft sensing.