Frontiers in Energy Research,
Journal Year:
2025,
Volume and Issue:
13
Published: Feb. 19, 2025
With
the
continued
development
and
progress
of
industrialisation,
modernisation,
smart
cities,
global
energy
demand
continues
to
increase.
Photovoltaic
systems
are
used
control
CO
2
emissions
manage
demand.
(PV)
system
public
utility,
effective
planning,
control,
operation
compels
accurate
Global
Horizontal
Irradiance
(GHI)
prediction.
This
paper
is
ardent
about
designing
a
novel
hybrid
GHI
prediction
method:
Bayesian
Optimisation
algorithm-based
Optimized
Deep
Bidirectional
Long
Short
Term
Memory
(BOA-D-BiLSTM).
work
attempts
fine-tune
hyperparameters
employing
optimisation.
Globally
ranked
fifth
in
solar
photovoltaic
deployment,
INDIA
Two
Region
Solar
Dataset
from
NOAA-National
Oceanic
Atmospheric
Administration
was
assess
proposed
BOA-D-BiLSTM
approach
for
long-term
horizon.
The
superior
performance
highlighted
with
help
experimental
results
comparative
analysis
grid
search
random
search.
Furthermore,
forecasting
effectiveness
compared
other
models,
namely,
Persistence
Model,
ARIMA,
BPN,
RNN,
SVR,
Boosted
Tree,
LSTM,
BiLSTM.
Compared
models
according
resulting
evaluation
error
metrics,
suggested
model
has
minor
Root
Mean
Squared
Error:
0.0026
0.0030,
Absolute
Error:0.0016
0.0018,
Mean-Squared
6.6852
×
10
−06
8.8628
R-squared:
0.9994
0.9988
on
both
dataset
1
respectively.
outperforms
baseline
models.
Thus,
viable
potential
distributed
generation
planning
control.
Developments in the Built Environment,
Journal Year:
2024,
Volume and Issue:
18, P. 100465 - 100465
Published: April 1, 2024
In
predicting
building
energy
(affected
by
seasons),
there
are
issues
like
inefficient
hyperparameter
optimization
and
inaccurate
predictions,
it
is
unclear
whether
spatial
temporal
attention
improves
performance.
This
study
proposes
a
method
based
on
Bayesian
Optimization
(BO),
Spatial-Temporal
Attention
(STA),
Long
Short-Term
Memory
(LSTM).
Seven
improved
LSTM
models
(BO-LSTM,
SA-LSTM,
TA-LSTM,
STA-LSTM,
BO-SA-LSTM,
BO-TA-LSTM,
BO-STA-LSTM)
compared
with
the
impacts
of
seasonal
variations
BO-STA-LSTM
analysed
using
different
sample
types
time
domain
analysis.
To
further
demonstrate
efficiency
proposed
method,
comparisons
convolutional
neural
network
(CNN)
(TCN)
performed,
followed
validation
new
datasets.
The
findings
indicate
that
adding
STA
BO
to
enhances
average
prediction
performance
0.0885.
alone
contributes
0.0717,
while
0.0560.
achieves
higher
accuracy
for
similar
test
training
samples
or
size
14016,
effectively
capturing
seasonal,
trend,
peak
patterns.
Additionally,
outperforms
CNN
TCN,
demonstrating
superior
accuracy.
Energies,
Journal Year:
2024,
Volume and Issue:
17(14), P. 3435 - 3435
Published: July 12, 2024
Due
to
the
inherent
intermittency,
variability,
and
randomness,
photovoltaic
(PV)
power
generation
faces
significant
challenges
in
energy
grid
integration.
To
address
these
challenges,
current
research
mainly
focuses
on
developing
more
efficient
management
systems
prediction
technologies.
Through
optimizing
scheduling
integration
PV
generation,
stability
reliability
of
can
be
further
improved.
In
this
study,
a
new
model
is
introduced
that
combines
strengths
convolutional
neural
networks
(CNNs),
long
short-term
memory
(LSTM)
networks,
attention
mechanisms,
so
we
call
algorithm
CNN-LSTM-Attention
(CLA).
addition,
Crested
Porcupine
Optimizer
(CPO)
utilized
solve
problem
generation.
This
abbreviated
as
CPO-CLA.
first
time
CPO
has
been
into
LSTM
for
parameter
optimization.
effectively
capture
univariate
multivariate
series
patterns,
multiple
relevant
target
variables
patterns
(MRTPPs)
are
employed
CPO-CLA
model.
The
results
show
superior
traditional
methods
recent
popular
models
terms
accuracy
stability,
especially
13
h
timestep.
mechanisms
enables
adaptively
focus
most
historical
data
future
prediction.
optimizes
network
parameters,
which
ensures
robust
generalization
ability
great
significance
establishing
trust
market.
Ultimately,
it
will
help
integrate
renewable
reliably
efficiently.
