Bioprocess monitoring and control via adaptive sensor calibration

Authors

  • Daniel Krause,

    Corresponding author
    1. (Bio-)Process Technology and Process Analysis, Life Science Engineering, Technische Universität München, Freising, Germany
    • (Bio-) Process Technology and Process Analysis, Life Science Engineering, Technische Universität München, Weihenstephaner Steig 20, 85354 Freising, Germany
    Search for more papers by this author
  • Stephan Birle,

    1. (Bio-)Process Technology and Process Analysis, Life Science Engineering, Technische Universität München, Freising, Germany
    Search for more papers by this author
  • Mohamed A. Hussein,

    1. (Bio-)Process Technology and Process Analysis, Life Science Engineering, Technische Universität München, Freising, Germany
    Search for more papers by this author
  • Thomas Becker

    1. (Bio-)Process Technology and Process Analysis, Life Science Engineering, Technische Universität München, Freising, Germany
    Search for more papers by this author

Abstract

To ensure optimal product quality of bioprocesses, it is necessary to develop intelligent control systems with integrated monitoring of key parameters. Having optimal yeast propagation in brewing technology is important to increase the efficiency of subsequent processes. Major drawbacks are: lacks in online detection of yeast attributes and temporal control schemes. One solution is to accurately detect essential process parameters combined with expert knowledge of linguistic control mechanisms. Those needs can be fulfilled by fuzzy logic or state observers including process dynamics associated with accurate multivariate calibration of sensing devices. Ultrasonic-based devices could monitor key parameter but their inline implementation is limited due to influences of the temperature and gas bubbles. Thus, incipient stages for calibration of the device including temperature dependencies using time and frequency properties of ultrasonic waves are presented. A multivariate model using offline measurements with a maximum prediction error of 0.48 g/100 g is reported in this study. Additionally, we show preliminary results of a mechanistic model for the temperature dependency of yeast growth adapted from the literature (biomass and ethanol production, substrate consumption). The results will lead to flexible control of temperature and aeration resulting in vital yeast and enhanced transparency of propagation progress according to the demands.

Ancillary