Arabian Journal of Chemistry,
Год журнала:
2022,
Номер
16(3), С. 104509 - 104509
Опубликована: Дек. 15, 2022
In
this
study,
a
novel
singular
third
order
perturbed
delay
differential
model
(STO-PDDM)
is
designed
with
its
two
types
using
the
traditional
Lane-Emden
model.
The
descriptions
of
delay/shape
perturbed,
and
factors
are
also
presented
for
both
STO-PDDM.
artificial
neural
networks
(ANNs)
along
optimization
global/local
performances
based
on
genetic
algorithm
(GA)
interior-point
(IPA)
have
been
used
to
solve
performed
GAIPA
activation
function
through
form
For
solving
STO-PDDM,
system's
accuracy,
substantiation,
authenticity
by
comparison
obtained
exact
solutions.
accessible
approximate
solutions
evaluate
computational
approach's
robustness,
stability,
correctness,
convergence.
reliability
scheme
different
statistical
measures
Sustainability,
Год журнала:
2023,
Номер
15(17), С. 12914 - 12914
Опубликована: Авг. 26, 2023
The
wind
is
a
crucial
factor
in
various
domains
such
as
weather
forecasting,
the
power
industry,
agriculture,
structural
health
monitoring,
and
so
on.
variability
unpredictable
nature
of
challenge
faced
by
most
wind-energy-based
sectors.
Several
atmospheric
geographical
factors
influence
characteristics.
Many
forecasting
methods
tools
have
been
introduced
since
early
times.
Wind
can
be
carried
out
short-,
medium-,
long-term.
uncertainty
accuracy
techniques.
This
article
brings
general
background
physical,
statistical,
intelligent
approaches
their
used
to
predict
characteristics
challenges—this
work’s
objective
improve
effective
data-driven
models
for
wind-power
production.
investigation
listing
effectiveness
improved
machine
learning
estimate
univariate
wind-energy
time-based
data
crucially
prominent
focus
this
work.
performance
ML
predicting
was
examined
using
ensemble
(ES)
models,
boosted
trees
bagged
trees,
Support
Vector
Regression
(SVR)
with
distinctive
kernels
etc.
Numerous
neural
networks
recently
constructed
speed
due
artificial
intelligence
(AI)
advancement.
Based
on
model
summary,
further
directions
research
application
developments
planned.
Renewable Energy,
Год журнала:
2024,
Номер
224, С. 120188 - 120188
Опубликована: Фев. 19, 2024
The
determinantal
noise
pollution
from
wind
turbines
has
constricted
the
acceptance
of
energy,
posing
health
and
environmental
concerns.
Existing
solutions
often
compromise
turbine
efficiency
or
farm
output,
limiting
full
utilization
energy.
To
address
this
challenge,
we
propose
a
novel
method
effectively
employing
acoustic
metamaterials
(AMMs)
inside
that
leverages
phase
cancellation
for
suppression
energy
enhancement.
Through
combined
retrieval,
propagation
model,
genetic
algorithm,
determine
optimal
layout
structural
design
these
AMMs
to
minimize
noise.
We
present
two
AMM
designs,
full-walled
segmented,
demonstrate
their
effectiveness
both
single-turbine
multiple-turbine
scenarios
by
achieving
91%
(in
Pa)
10%–68%
reduction,
respectively,
compared
reference
layouts.
Furthermore,
exhibit
AMM's
impact
in
enhancing
throughput
farms
installing
an
additional
noise-restricted
area
existing
AMM-equipped
while
maintaining
70%
reduction
levels.
This
approach
paves
way
constructing
near
urban-suburban
areas,
complying
with
landscaping
visual
government
policies,
securing
community
towards
sustainable
power
generation
systems.
Wind Energy,
Год журнала:
2024,
Номер
27(7), С. 667 - 694
Опубликована: Май 13, 2024
Abstract
Short‐term
wind
speed
prediction
is
essential
for
economical
power
utilization.
The
real‐world
data
are
typically
intermittent
and
fluctuating,
presenting
great
challenges
to
existing
shallow
models.
In
this
paper,
we
present
a
novel
deep
hybrid
model
multistep
prediction,
namely,
LR‐FFT‐RP‐MLP/LSTM
(linear
fast
Fourier
transform
rank
pooling
multiple‐layer
perceptron/long
short‐term
memory).
Our
processes
the
local
global
input
features
simultaneously.
We
leverage
RP
feature
extraction
capture
temporal
structure
while
maintaining
order.
Besides,
understand
periodic
patterns,
exploit
FFT
extract
relevant
frequency
components
in
data.
resulting
are,
respectively,
integrated
with
original
fed
into
an
MLP/LSTM
layer
initial
predictions.
Finally,
linear
regression
collaborate
these
predictions
produce
final
prediction.
proposed
evaluated
using
real
collected
from
2010
2020,
demonstrating
superior
forecasting
capabilities
when
compared
state‐of‐the‐art
single
Overall,
study
presents
promising
approach
improving
accuracy
of
forecasting.
