Automatic Control of Bioprocesses - Measurement, Monitoring, Modelling, and Control of Bioprocesses - page 37

Contents

1 Introduction.......................................................................................................................... 37

2 Controller Design ................................................................................................................ 38

2.1 Direct Measurements .................................................................................................. 38

2.2 Soft Sensors ................................................................................................................ 41

2.3 Control Action ............................................................................................................ 42

3 State-of-the-Art Control Algorithm .................................................................................... 43

3.1 Open Loop-Closed Loop Controller.......................................................................... 43

3.2 PID Control Based on Soft Sensors Measurements .................................................. 44

3.3 Model Linearization-Based Control ........................................................................... 49

3.4 Fuzzy Logic-Based Control........................................................................................ 50

3.5 Artificial Neural Network-Based Control.................................................................. 52

3.6 Model Predictive Control ........................................................................................... 54

3.7 Probing Feeding Controller Strategy ......................................................................... 56

3.8 Extremum-Seeking Control ........................................................................................ 57

3.9 Control Based on A Heuristic Procedure .................................................................. 58

4 Conclusions.......................................................................................................................... 58

References.................................................................................................................................. 59

1 Introduction

Due to competition, industry tends to increase the degree of automation in pro-
duction processes. Only an automated system is never tired and always attentive,
will act reliable, and therefore can provide optimal process operation. It can react
quickly to changes in raw material quality as well as changes in environmental
conditions. As a result, energy and material input can be decreased and process
safety and product yield and quality can be increased. This applies, of course, also
for bioprocesses. The operation of these processes is usually carried out in three
successive steps:

Upstreaming (filling, sterilization, and mixing)
Cultivation/enzyme reaction (growth of cells, bioconversion, and production)
Downstreaming (harvesting, separation, concentrating, and crystallization)
Each step demands a high degree of automation. In the first step, standard
automatic sequence control units are available. The quality of raw materials is of
special importance for the subsequent steps. The automation in the second step is
more complicated, since complex transport processes are combined with a mul-
titude of dynamic biochemical reactions during cultivation. Therefore, one has to
deal with a complex, nonlinear, multiparameter, time-variant system. Little
detailed comprehensive knowledge is available. The microorganisms used for the
synthesis of the product have many inherent closed-loop systems of their own,
which can only be manipulated indirectly through environmental conditions
by physical and chemical variables. Frequently, open-loop control systems are