Prediction of Biogas Production Volumes from Household Organic Waste Based on Machine Learning
Energies,
Год журнала:
2024,
Номер
17(7), С. 1786 - 1786
Опубликована: Апрель 8, 2024
The
article
proposes
to
use
machine
learning
as
one
of
the
areas
artificial
intelligence
forecast
volume
biogas
production
from
household
organic
waste.
five
regression
algorithms
(Linear
Regression,
Ridge
Lasso
Random
Forest
and
Gradient
Boosting
Regression)
create
an
effective
model
for
forecasting
waste
is
considered.
Based
on
comparison
these
by
MSE
MAE
indicators,
quality
training
their
accuracy
during
are
evaluated.
proposed
algorithm
creating
a
volumes
involves
implementation
10
main
3
auxiliary
steps.
Their
advantage
that
they
aid
in
performance
component
data
analysis,
which
carried
out
based
method
reducing
dimensionality
set,
increasing
interpretability,
minimizing
risk
loss.
An
analysis
2433
was
out,
characterizes
formation
food
(FW)
yard
(YW)
according
four
features.
Data
preparation
performed
using
Jupyter
Notebook
environment
Python.
We
select
substantiate
On
basis
conducted
research,
advantages
disadvantages
used
building
models
determined.
It
found
two
models,
“Random
Regressor”
“Gradient
Regressor”,
show
best
indicators.
other
three
inferior
were
not
considered
further.
To
determine
we
choose
Regressor
be
more
accurate
compared
Regressor.
This
confirmed
fact
set
7.14
times
smaller
than
model.
At
same
time,
2.67
both
worse
test
indicates
overtraining
tendencies.
has
sets.
established
most
provides
=
0.088
smallest
absolute
errors
predictions.
Further
systematic
improvement
new
will
ensure
its
maintain
competitive
advantages.
Язык: Английский
Optimizing energy systems of livestock farms with computational intelligence for achieving energy autonomy
Scientific Reports,
Год журнала:
2025,
Номер
15(1)
Опубликована: Март 28, 2025
The
relevance
of
the
study
is
due
to
need
increase
energy
autonomy
livestock
farms
by
introducing
innovative
solutions
based
on
computational
intelligence.
Given
significant
consumption
farms,
as
well
reduced
dependence
traditional
sources,
there
a
optimise
systems
using
renewable
sources.
aim
research
develop
model
for
integrating
intelligence
achieve
their
autonomy.
use
models
will
allow
farmers
manage
more
efficiently,
minimise
carbon
emissions,
and
overall
stability
supply.
object
including
subject
methods
optimisation
used
resource
management.
paper
develops
optimising
genetic
algorithm
that
involves
systematic
implementation
5
steps.
In
contrast
static
models,
proposed
takes
into
account
possibility
dynamic
adaptation
structure
supply
system
real
production
conditions.
This
done
taking
demand
external
factors
such
power
grid
failures
weather
multi-criteria
approach
simultaneously
reduces
CO₂
costs
increases
sustainability
farms.
in
provides
flexible
parameter
settings
search
an
optimal
solution
context
variable
complex
system.
Based
model,
Python
3.10
program
was
created
perform
labour-intensive
calculations
According
results
testing
at
farm
Volyn
Nova
LLC
(Volyn
region,
Ukraine),
it
found
optimised
allows
reducing
emissions
from
1263
kg/day
92.3
increasing
Prospects
further
include
other
types
development
integration
combined
several
Язык: Английский
European Green Deal: Substantiation of the Rational Configuration of the Bioenergy Production System from Organic Waste
Energies,
Год журнала:
2024,
Номер
17(17), С. 4513 - 4513
Опубликована: Сен. 9, 2024
A
review
of
the
current
state
theory
and
practice
bioenergy
production
from
waste
allowed
us
to
identify
scientific
applied
problem
substantiating
rational
configuration
a
modular
anaerobic
system,
taking
into
account
volume
organic
generated
in
settlements.
To
solve
this
problem,
paper
develops
an
approach
algorithm
for
matching
system
with
amount
residential
areas.
Unlike
existing
tools,
takes
peculiarities
areas,
which
is
basis
accurate
forecasting
generation
and,
accordingly,
determining
system.
In
addition,
each
scenarios,
digestion
process
modeled,
allows
determine
functional
indicators
that
underlie
determination
terms
cost
environmental
performance.
Based
on
use
developed
tools
conditions
Golosko
area,
Lviv
(Ukraine),
possible
scenarios
installation
systems
are
substantiated.
It
was
found
greatest
annual
benefits
obtained
processing
mixed
food
yard
waste.
The
payback
period
investments
given
area
largely
depends
their
ranges
3.3
8.4
years,
differ
other
by
2.5
times.
This
indicates
toolkit
practical
value,
as
it
coordination
real
conditions.
future,
recommended
proposed
decision
support
model
biomass
energy
resource
ensures
Язык: Английский