As
renewable
energy,
particularly
regional
photovoltaic
(PV),
becomes
more
prevalent
in
the
power
grid,
accurate
forecasting
of
its
output
is
paramount
for
efficient
operation.
However,
challenges
persist,
including
lack
reliable
data,
inappropriate
data
usage,
and
computational
burdens
stemming
from
vast
number
dispersed
nature
PV
installations.
To
address
these
problems,
a
prediction
based
on
transfer
learning
satellite
cloud
imagery
proposed.
Firstly,
an
algorithmic
architecture
composed
gray-level
co-occurrence
matrix
(GLCM)
random
forest
(RF)
established
extracting
texture
features
(TFs)
images
selecting
TFs
with
highest
correlation
to
irradiance.
Furthermore,
attention
mechanism
(AM)
long
short-term
memory
(LSTM)
employed
at
reconstruct
significant
TFs.
These
reconstructed
are
then
integrated
into
training
model,
aiming
enhance
between
outcome.
Finally,
structure
combine
convolutional
neural
network
(CNN)
LSTM
taken
as
maximum
mean
discrepancy
(MMD)
algorithm
utilized
measure
correlations
source
target
stations.
Both
single
located
UK
station
China
analysis
verify
effectiveness,
several
benchmark
methods
have
been
compared,
approach
this
research
demonstrated
superior
performance.
Energy Reports,
Journal Year:
2024,
Volume and Issue:
11, P. 1711 - 1722
Published: Jan. 20, 2024
Power
grid
stability
depends
on
the
ability
to
forecast
solar
power
generation
at
regional
level.
Most
previous
research
probabilistic
forecasting
has
focused
use
of
machine
learning
predict
output
individual
plants
rather
than
generation,
and
few
studies
have
considered
effects
seasonal
weather
patterns
In
this
study,
climate
geographic
data
were
collected
from
83
stations
between
2019
2021
for
in
developing
a
model
by
which
generation.
The
results
pattern
analysis
based
unsupervised
ensemble
voting
used
build
quantile
regression
short-term
prediction
capacity.
efficacy
was
assessed
using
48
PV
plants,
included
four
sub-datasets
pertaining
target
regions.
Highly
accurate
consistently
obtained
across
all
regions
both
winter
summer
seasons.
proposed
outperformed
conventional
deterministic
6.55%
4.03%
terms
total
normalized
mean
absolute
error
(NMAE).
Prediction
intervals
generated
could
be
as
input
parameters
dispatch
decisions.
Advances in Atmospheric Sciences,
Journal Year:
2024,
Volume and Issue:
42(2), P. 269 - 296
Published: Dec. 28, 2024
Abstract
The
fundamental
scientific
and
engineering
knowledge
concerning
the
solar
power
curve,
which
maps
irradiance
other
auxiliary
meteorological
variables
to
photovoltaic
output
power,
has
been
gathered
put
forward
in
preceding
tutorial
review.
Despite
many
pages
of
that
review,
it
was
incomplete
sense
did
not
elaborate
on
applications
this
very
important
tool
energy
meteorology.
Indeed,
curves
are
ubiquitously
needed
a
broad
spectrum
forecasting
resource
assessment
tasks.
Hence,
review
should
continue
from
where
left
off
present
examples
usage
curves.
In
nutshell,
together
with
one,
elucidate
how
surface
shortwave
radiation
data,
be
they
ground-based,
satellite-retrieved,
or
model-output,
bridged
various
system
operations
via
Sensors,
Journal Year:
2024,
Volume and Issue:
24(2), P. 397 - 397
Published: Jan. 9, 2024
Aerosols
play
a
crucial
role
in
the
surface
radiative
budget
by
absorbing
and
scattering
both
shortwave
longwave
radiation.
While
most
aerosol
types
exhibit
relatively
minor
forcing
when
compared
to
their
counterparts,
dust
aerosols
stand
out
for
substantial
forcing.
In
this
study,
radiometers,
sun
photometer,
microwave
radiometer
parameterization
scheme
clear-sky
radiation
estimation
were
integrated
investigate
properties
of
aerosols.
During
an
event
Xianghe,
North
China
Plain,
from
25
April
27
2018,
composition
(anthropogenic
dust)
optical
depth
(AOD,
ranging
0.3
1.5)
changed
considerably.
A
notable
effect
(SARE)
was
revealed
system
(reaching
its
peak
at
-131.27
W·m
Energies,
Journal Year:
2024,
Volume and Issue:
17(12), P. 2913 - 2913
Published: June 13, 2024
This
study
assesses
the
efficacy
of
Heliosat-2
algorithm
for
estimating
solar
radiation,
comparing
its
outputs
against
ground
measurements
across
seven
distinct
countries:
Netherlands,
Spain,
Japan,
Namibia,
South
Africa,
Saudi
Arabia,
and
India.
To
achieve
this,
utilizes
two
satellite
data
sources—Himawari-8
Japan
Metosat
Second
Generation-MSG
rest
countries—and
spanning
time
between
January
2022
April
2024.
A
robust
methodology
determining
albedo
parameters
specific
to
was
developed.
During
cloudy
days,
estimates
provided
by
generally
exceeded
in
all
countries.
Conversely,
on
clear
there
a
tendency
underestimation,
as
indicated
median
values
mean
bias
(MB)
most
The
model
slightly
underestimates
daily
radiation
values,
with
MB
ranging
from
−27.5
+10.2
W·m−2.
Notably,
root
square
error
(RMSE)
days
is
significantly
lower,
24.8
108.7
W·m−2,
compared
which
RMSE
lie
75.3
180.2
In
terms
R2
both
satellites
show
strong
correlations
estimated
actual
value
consistently
above
0.86
monthly
scale
over
92%
points
falling
within
±2
standard
deviations.
