Simultaneous real-time estimation of maximum substrate uptake capacity and yield coefficient in induced microbial cultures

Abstract

In this work we present a soft sensor to accurately estimate the yield coefficient 𝑌𝑋∕𝑆 and the substrate uptake capacity 𝑞𝑆𝑚𝑎𝑥 in a cultivation process using offgas measurements and a nonlinear state observer with an underlying mechanistic model. The structural observability analysis of the mechanistic model showed that both parameters are observable given the available measurement information. In different simulation scenarios we analyzed under which conditions an accurate estimation is possible when measurements are uncertain. Testing the proposed soft sensor in-silico showed that 𝑞𝑆𝑚𝑎𝑥 and 𝑌𝑋∕𝑆 can be estimated with reasonable accuracy depending on the parameter sensitivity. Verification of the developed state and parameter estimation was carried out in Escherichia coli cultivations. Besides accurate prediction of living biomass, and substrate accumulation, decreasing 𝑌𝑋∕𝑆 and 𝑞𝑆𝑚𝑎𝑥 could be detected. Therefore, the developed soft sensor can be used to control induced cultures and to prevent overfeeding situations.