Table 1 Extended Kalman filter application for cultivation processes
Estimator/
application
type
Cultivation
type/
microorganism
Process
model
Objective
Measured state Reference
Extended
ter/experimental
application
Batch cultivation/E. coli
Dissolved
oxygen mass
balance
Noise filtering
from
dissolved
oxygen
measurements
Dissolved
oxygen
Lee et al.
[19]
Extended
ter/experimental
application
Fed-batch cultivation/
S. cerevisiae
Material balance equation
with Monod
growth rate
kinetics
Parameter
estimation and
substrate
prediction
Glucose concentration with
FIA
Hitzman
et al. [32]
ter/experimental
application
Batch cultivation/
S. cerevisiae
Ideal stirred
tank reactor
model with
Monod
growth kinetics (glucose
and ethanol as
limiting
substrates)
Noise filtering
from
predicted
bioprocess
variables
Biomass, glucose, and ethanol (with
ultrasonic
velocity)
Cha and
Hitzmann
[36]
Extended
ter/experimental
application
Fed-batch cultivation/
S. cerevisiae
A model for
an ideal
stirred tank
reactor in
combination
with Monod
growth
kinetics
Noise filtering
from
predicted
glucose
Glucose concentration with
flow injection
analyses (FIA)
Arndt and
Hitzmann
[37]
Extended
Kalman filter/
simulation
Fed-batch cultivation/
S. cerevisiae
Cybernetic
model of
Jones and
Kompala
Filtering out
noise from the
feed stream
Dilution rate
or the gas–liquid mass
transfer coefficient for
oxygen
Patnaik
[39]
Extended
Kalman filter/
simulation
Fed-batch cultivation/E. coli
General
dynamic
model of bioreactors with
Monod
growth
kinetics
Parameter
estimation and
biomass
prediction
Dissolved and
exhaust oxygen and carbon dioxide
Rocha
et al. [40]
Extended
ter/experimental
application
Fed-batch cultivation/
Bordetella
pertussis
A model with
two parameters which are
calculated
using separate
experiments
Estimation of
specific
growth rate,
biomass, and
oxygen mass
transfer
Dissolved
oxygen
Soons
et al. [42]
(continued)
104
A. Yousefi-Darani et al.