I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
pu
t
er
E
ng
ineering
(
I
J
E
CE
)
Vo
l.
1
4
,
No
.
5
,
Octo
b
er
20
2
4
,
p
p
.
5
9
2
6
~
5
9
4
0
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
1
4
i
5
.
pp
5
9
2
6
-
5
9
4
0
5926
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
Photo
v
o
ltaic
po
w
er pre
di
ction usi
n
g
deep
lea
rning
m
o
dels:
recent
adv
a
nces a
nd new
insig
hts
B
a
s
m
a
Sa
a
d,
Asm
a
a
E
l H
a
n
na
ni,
Abdel
ha
k
Aqqa
l,
Ra
hh
a
l Erra
t
t
a
hi
La
b
o
r
a
t
o
r
y
o
f
I
n
f
o
r
ma
t
i
o
n
T
e
c
h
n
o
l
o
g
i
e
s,
N
a
t
i
o
n
a
l
S
c
h
o
o
l
o
f
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s
o
f
E
l
J
a
d
i
d
a
,
C
h
o
u
a
i
b
D
o
u
k
k
a
l
i
U
n
i
v
e
r
s
i
t
y
,
El
J
a
d
i
d
a
,
M
o
r
o
c
c
o
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ma
y
1
3
,
2
0
2
4
R
ev
is
ed
J
u
l 4
,
2
0
2
4
Acc
ep
ted
J
u
l 9
,
2
0
2
4
Artifi
c
ial
in
telli
g
e
n
c
e
(AI)
a
n
d
it
s
a
p
p
li
c
a
ti
o
n
a
c
ro
ss
v
a
rio
u
s
d
o
m
a
in
s
h
a
v
e
sp
a
rk
e
d
sig
n
ifi
c
a
n
t
in
tere
st,
with
e
a
c
h
d
o
m
a
in
p
re
se
n
ti
n
g
d
isti
n
c
t
c
h
a
ra
c
teristics
a
n
d
c
h
a
ll
e
n
g
e
s.
I
n
t
h
e
re
n
e
wa
b
le
e
n
e
r
g
ies
se
c
to
r
,
a
c
c
u
ra
te
p
re
d
ictio
n
o
f
p
o
we
r
o
u
t
p
u
t
f
ro
m
p
h
o
to
v
o
lt
a
ic
(
P
V)
p
a
n
e
ls u
sin
g
A
I
is
c
ru
c
ial
fo
r
m
e
e
ti
n
g
e
n
e
rg
y
d
e
m
a
n
d
a
n
d
f
a
c
il
it
a
ti
n
g
e
n
e
rg
y
m
a
n
a
g
e
m
e
n
t
a
n
d
sto
ra
g
e
.
Th
e
field
o
f
d
a
ta
a
n
a
ly
sis
h
a
s
g
ro
wn
ra
p
id
l
y
in
re
c
e
n
t
y
e
a
rs,
with
p
re
d
ictiv
e
m
o
d
e
ls
b
e
c
o
m
in
g
i
n
c
re
a
sin
g
ly
p
o
p
u
lar
fo
r
f
o
re
c
a
stin
g
a
n
d
p
re
d
ic
ti
o
n
tas
k
s.
Ho
we
v
e
r,
th
e
a
c
c
u
ra
c
y
a
n
d
re
li
a
b
il
it
y
o
f
t
h
e
se
m
o
d
e
ls
d
e
p
e
n
d
h
e
a
v
il
y
o
n
th
e
q
u
a
li
t
y
o
f
d
a
ta,
d
a
ta
p
re
p
r
o
c
e
ss
in
g
,
m
o
d
e
l
lea
rn
i
n
g
a
n
d
e
v
a
lu
a
ti
o
n
.
I
n
t
h
is
c
o
n
tex
t,
t
h
is
p
a
p
e
r
a
ims
to
p
ro
v
i
d
e
a
n
in
-
d
e
p
th
re
v
iew
o
f
p
re
v
io
u
s
re
se
a
r
c
h
a
n
d
re
c
e
n
t
p
r
o
g
re
ss
in
P
V
so
l
a
r
p
o
we
r
fo
re
c
a
stin
g
a
n
d
p
re
d
ictio
n
b
y
id
e
n
ti
f
y
in
g
a
n
d
a
n
a
ly
z
in
g
th
e
m
o
st
imp
a
c
ti
n
g
fa
c
t
o
rs.
Th
e
fi
n
d
i
n
g
s
o
f
t
h
e
li
tera
tu
re
re
v
iew
a
re
th
e
n
u
se
d
t
o
imp
lem
e
n
t
a
b
e
n
c
h
m
a
rk
fo
r
P
V
p
o
we
r
p
re
d
ictio
n
u
sin
g
d
e
e
p
lea
rn
in
g
m
o
d
e
ls
i
n
d
iffere
n
t
c
li
m
a
tes
a
n
d
P
V
p
a
n
e
l
s.
Th
e
a
im
o
f
imp
lem
e
n
ti
n
g
th
is
b
e
n
c
h
m
a
rk
is
to
g
a
in
in
si
g
h
t
s
in
t
o
th
e
c
h
a
ll
e
n
g
e
s
a
n
d
o
p
p
o
rtu
n
it
ies
o
f
P
V
p
o
we
r
p
re
d
icti
o
n
a
n
d
t
o
im
p
ro
v
e
th
e
a
c
c
u
ra
c
y
,
re
li
a
b
il
it
y
a
n
d
e
x
p
lain
a
b
il
it
y
o
f
p
re
d
ictiv
e
m
o
d
e
ls i
n
th
e
fu
tu
re
.
K
ey
w
o
r
d
s
:
Ar
tific
ial
in
tellig
en
ce
Data
p
r
ep
r
o
ce
s
s
in
g
Dee
p
lear
n
in
g
Fo
r
ec
asti
n
g
h
o
r
izo
n
L
iter
atu
r
e
r
ev
iew
Ph
o
to
v
o
ltaic
p
o
wer
f
o
r
ec
asti
n
g
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
B
asm
a
Saad
L
ab
o
r
ato
r
y
o
f
I
n
f
o
r
m
atio
n
T
e
ch
n
o
lo
g
ies,
Natio
n
al
Sch
o
o
l o
f
Ap
p
lied
Scien
ce
s
o
f
E
l Jad
id
a,
C
h
o
u
aib
Do
u
k
k
ali
Un
iv
e
r
s
ity
E
l Jad
id
a,
Mo
r
o
cc
o
E
m
ail: saad
.
b
@
u
cd
.
ac
.
m
a
1.
I
NT
RO
D
UCT
I
O
N
R
en
ewa
b
le
en
er
g
y
is
ex
p
er
ie
n
cin
g
i
n
cr
ea
s
in
g
d
em
an
d
g
l
o
b
ally
,
d
r
iv
en
b
y
en
v
i
r
o
n
m
e
n
t
al
co
n
ce
r
n
s
an
d
th
e
d
esire
f
o
r
en
er
g
y
s
ec
u
r
ity
an
d
ec
o
n
o
m
ic
d
e
v
elo
p
m
e
n
t.
So
lar
en
er
g
y
d
er
iv
e
d
f
r
o
m
p
h
o
to
v
o
ltaic
(
PV)
p
an
els
is
a
p
r
o
m
in
en
t
g
r
ee
n
an
d
s
u
s
tain
ab
le
en
e
r
g
y
s
o
u
r
c
e,
r
eq
u
i
r
in
g
m
in
im
al
m
ain
te
n
an
ce
an
d
o
f
f
er
in
g
a
lo
n
g
er
life
s
p
an
co
m
p
a
r
ed
to
th
er
m
al
en
e
r
g
y
.
T
h
er
e
ar
e
tw
o
ty
p
es
o
f
PV
s
o
lar
p
o
wer
:
I
n
p
u
t
p
o
wer
(
s
o
lar
ir
r
ad
ian
ce
)
an
d
o
u
t
p
u
t
p
o
we
r
(
s
o
lar
p
o
wer
)
.
So
lar
i
r
r
ad
ian
c
e
co
r
r
esp
o
n
d
s
to
th
e
lig
h
t
r
ec
eiv
ed
f
r
o
m
th
e
s
u
n
an
d
is
m
ea
s
u
r
ed
in
s
p
ac
e
o
r
o
n
th
e
PV
p
a
n
el
u
s
in
g
watts
o
r
jo
u
les
p
er
s
q
u
ar
e
m
eter
.
T
h
r
ee
v
ar
ian
ts
o
f
th
is
en
er
g
y
e
x
is
t.
E
x
tr
ater
r
estrial
ir
r
ad
ian
ce
(
E
I
)
,
g
lo
b
al
h
o
r
izo
n
tal
ir
r
ad
ian
ce
(
GHI
)
,
d
ir
ec
t
n
o
r
m
al
ir
r
a
d
ian
c
e
(
DNI
)
an
d
d
if
f
u
s
e
h
o
r
iz
o
n
tal
ir
r
ad
ia
n
ce
(
DHI
)
[
1
]
.
So
lar
PV
p
o
wer
r
ef
er
s
t
o
th
e
p
o
we
r
g
e
n
er
ated
b
y
th
e
p
an
el.
I
t
is
m
ea
s
u
r
ed
u
s
in
g
wa
tts
,
jo
u
les,
watts
p
er
s
q
u
ar
e
m
eter
o
r
jo
u
les
p
e
r
s
q
u
ar
e
m
eter
.
Pre
cise
p
r
ed
ictio
n
o
f
p
o
we
r
g
en
e
r
ated
b
y
PV
p
a
n
els
is
p
iv
o
tal
f
o
r
ef
f
icien
t
P
V
p
o
wer
m
a
n
ag
em
en
t,
tr
an
s
m
is
s
io
n
an
d
s
to
r
ag
e.
Acc
u
r
ate
p
r
e
d
ictio
n
s
en
ab
le
g
r
id
o
p
e
r
ato
r
s
an
d
e
n
er
g
y
p
r
o
v
id
er
s
to
p
r
o
ac
tiv
ely
b
alan
ce
s
u
p
p
ly
an
d
d
em
an
d
,
o
p
tim
ize
en
er
g
y
s
to
r
ag
e
s
y
s
tem
s
an
d
en
h
an
ce
o
v
er
all
g
r
id
s
tab
ilit
y
.
Ho
wev
er
,
tr
ad
itio
n
al
p
r
ed
ictio
n
m
eth
o
d
s
o
f
ten
s
tr
u
g
g
le
to
ca
p
tu
r
e
th
e
n
u
an
ce
s
o
f
s
o
lar
en
er
g
y
d
y
n
a
m
ics.
Her
ein
lies
th
e
m
o
tiv
atio
n
f
o
r
h
ar
n
ess
in
g
ar
tific
ial
in
tellig
en
ce
(
AI
)
in
s
o
lar
p
o
wer
p
r
e
d
ictio
n
.
AI
’
s
ca
p
ac
ity
to
an
aly
ze
v
ast
an
d
co
m
p
le
x
d
atasets
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
P
h
o
to
vo
lta
ic
p
o
w
er
p
r
ed
ictio
n
u
s
in
g
d
ee
p
lea
r
n
in
g
mo
d
els:
r
ec
en
t
a
d
va
n
ce
s
a
n
d
…
(
B
a
s
ma
S
a
a
d
)
5927
ad
ap
t
to
ch
an
g
in
g
co
n
d
itio
n
s
an
d
u
n
c
o
v
er
in
tr
icate
p
atter
n
s
is
th
e
d
r
iv
in
g
f
o
r
ce
b
eh
in
d
t
h
is
m
o
tiv
atio
n
.
I
t
is
wo
r
th
n
o
tin
g
th
at
s
o
lar
en
e
r
g
y
f
o
r
e
ca
s
tin
g
an
d
p
r
e
d
ictio
n
a
r
e
two
d
if
f
er
en
t
ter
m
s
.
Fo
r
ec
a
s
tin
g
is
u
s
ed
wh
en
p
r
ed
ictin
g
th
e
o
u
tp
u
t
at
tim
e
+
b
ased
o
n
th
e
in
p
u
t
at
tim
e
.
I
n
p
r
ed
ictio
n
,
in
p
u
t
an
d
o
u
tp
u
t
d
ata
ar
e
b
o
th
at
tim
e
.
T
h
ese
ter
m
s
ar
e
u
s
e
d
in
ter
ch
an
g
ea
b
ly
in
th
is
s
tu
d
y
an
d
in
th
e
liter
atu
r
e.
I
n
th
is
r
esear
ch
,
we
will
f
o
cu
s
o
n
s
o
lar
PV
p
o
wer
g
en
er
atio
n
(
o
u
tp
u
t
p
o
wer
)
p
r
e
d
ictio
n
an
d
co
n
d
u
ct
a
c
o
m
p
r
e
h
e
n
s
iv
e
r
ev
iew
o
f
th
e
liter
atu
r
e.
W
e
will
d
elv
e
in
t
o
v
ar
io
u
s
asp
ec
ts
,
in
clu
d
in
g
d
ata
s
o
u
r
ce
s
,
in
p
u
t
p
atter
n
s
,
d
ata
p
r
ep
r
o
ce
s
s
in
g
tech
n
iq
u
es
an
d
m
o
d
elin
g
m
et
h
o
d
s
em
p
lo
y
ed
b
y
r
esear
ch
er
s
to
ac
h
iev
e
ef
f
ec
tiv
e
a
n
d
ac
cu
r
ate
p
r
ed
ictio
n
s
.
W
e
will a
ls
o
p
r
o
v
id
e
a
cr
itical
an
a
ly
s
is
o
f
p
r
ev
io
u
s
s
tu
d
ies,
h
i
g
h
l
ig
h
tin
g
th
eir
s
tr
en
g
th
s
an
d
lim
itatio
n
s
.
I
n
th
e
ex
ten
s
iv
e
b
o
d
y
o
f
liter
a
tu
r
e
o
n
PV
p
o
wer
p
r
ed
ictio
n
,
d
iv
er
s
e
d
atasets
h
av
e
b
ee
n
u
s
e
d
,
lead
in
g
to
a
m
u
ltit
u
d
e
o
f
co
n
cl
u
s
io
n
s
.
E
ac
h
au
th
o
r
d
r
aws
th
eir
o
wn
in
s
ig
h
ts
,
o
f
ten
in
f
lu
en
ce
d
b
y
s
p
ec
if
ic
d
atasets
,
m
o
d
els,
d
ata
p
r
e
p
r
o
ce
s
s
in
g
m
eth
o
d
s
an
d
in
p
u
t
p
atter
n
s
.
T
h
i
s
d
iv
er
s
ity
o
f
f
i
n
d
in
g
s
u
n
d
er
s
co
r
es
th
e
n
ee
d
f
o
r
a
u
n
if
ied
b
e
n
ch
m
ar
k
th
at
co
m
p
r
eh
en
s
iv
ely
ac
co
u
n
ts
f
o
r
all
f
ac
to
r
s
af
f
ec
tin
g
PV p
o
wer
p
r
ed
i
ctio
n
.
Ad
d
itio
n
ally
,
p
r
o
p
o
s
ed
s
y
s
tem
s
f
o
r
p
r
e
d
ictin
g
PV
p
o
wer
ar
e
o
f
ten
n
o
t
th
o
r
o
u
g
h
ly
in
ter
p
r
eted
o
r
an
aly
ze
d
to
id
en
tif
y
wh
er
e
th
e
m
o
d
els
p
er
f
o
r
m
well
an
d
wh
er
e
th
ey
d
o
n
o
t.
T
h
e
g
en
er
a
lizab
ilit
y
o
f
t
h
ese
m
o
d
els
ac
r
o
s
s
d
if
f
er
en
t
clim
ate
co
n
d
itio
n
s
h
as a
ls
o
n
o
t b
ee
n
a
d
eq
u
ately
e
x
p
lo
r
e
d
.
T
h
u
s
,
th
e
p
r
im
a
r
y
o
b
jectiv
e
o
f
th
is
p
a
p
er
is
to
estab
lis
h
a
n
ex
p
er
im
en
tal
f
r
am
ewo
r
k
th
at
r
ig
o
r
o
u
s
ly
co
n
s
id
er
s
th
ese
in
f
lu
en
tial
f
ac
to
r
s
in
PV
p
o
wer
f
o
r
ec
asti
n
g
.
B
y
d
o
in
g
s
o
,
we
aim
to
v
alid
a
te
an
d
s
u
b
s
tan
tiate
th
e
h
y
p
o
th
eses
an
d
co
n
cl
u
s
io
n
s
p
u
t
f
o
r
th
in
th
e
ex
is
tin
g
l
iter
atu
r
e,
p
r
o
v
id
i
n
g
a
m
o
r
e
c
o
m
p
r
eh
e
n
s
iv
e
an
d
u
n
if
ied
p
er
s
p
ec
tiv
e
o
n
th
e
f
iel
d
.
W
e
also
co
n
d
u
cted
a
p
o
s
t
-
p
r
ed
ictio
n
an
aly
s
is
to
id
en
tify
P
V
p
an
els wh
er
e
th
e
m
o
d
el
p
e
r
f
o
r
m
s
well
an
d
w
h
e
r
e
it
d
o
es
n
o
t.
Ad
d
itio
n
ally
,
we
ex
am
in
ed
th
e
m
o
d
el’
s
ab
il
ity
to
g
e
n
er
alize
its
r
esu
lts
ac
r
o
s
s
d
if
f
er
en
t c
lim
ate
co
n
d
itio
n
s
.
T
h
is
s
tu
d
y
o
f
f
er
s
a
n
im
p
o
r
ta
n
t
o
p
p
o
r
tu
n
ity
to
ad
v
an
ce
t
h
e
u
n
d
e
r
s
tan
d
in
g
an
d
id
en
tif
y
k
n
o
wled
g
e
g
ap
s
in
th
e
d
o
m
ain
o
f
s
o
lar
PV
p
o
wer
p
r
ed
ictio
n
.
Sev
er
al
k
ey
p
o
in
ts
d
is
tin
g
u
is
h
o
u
r
wo
r
k
f
r
o
m
p
r
ev
io
u
s
ly
p
u
b
lis
h
ed
r
ev
iew
s
tu
d
ies
in
th
e
f
ield
:
i
)
T
h
is
wo
r
k
co
m
p
r
eh
e
n
s
iv
ely
co
n
s
id
er
s
all
f
ac
to
r
s
th
at
ca
n
im
p
ac
t
s
o
lar
PV
en
er
g
y
p
r
ed
ictio
n
,
in
cl
u
d
in
g
d
ata
s
o
u
r
ce
s
,
in
p
u
t
v
a
r
iab
les,
f
o
r
ec
asti
n
g
h
o
r
izo
n
s
,
d
a
ta
r
eso
lu
tio
n
,
d
at
a
p
r
ep
r
o
ce
s
s
in
g
tech
n
iq
u
es
an
d
p
r
ed
ictiv
e
m
o
d
els
;
ii
)
W
e
ex
clu
s
iv
ely
r
ev
iew
ar
ticles
p
u
b
lis
h
ed
in
th
e
last
d
ec
ad
e
an
d
cite
o
n
ly
th
o
s
e
th
a
t
ar
e
d
ir
ec
tly
r
elev
an
t
t
o
o
u
r
s
t
u
d
y
;
an
d
iii
)
An
ex
p
e
r
im
en
tal
s
etu
p
is
d
ev
elo
p
e
d
to
co
m
p
a
r
e
th
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
m
o
s
t
ac
cu
r
ate
lear
n
in
g
m
o
d
els
in
th
e
liter
atu
r
e
with
in
th
e
m
o
s
t
co
m
m
o
n
l
y
u
s
ed
f
o
r
ec
asti
n
g
h
o
r
iz
o
n
s
in
t
wo
d
if
f
er
e
n
t
clim
ates.
T
h
e
im
p
ac
t
o
f
d
ata
r
eso
lu
tio
n
an
d
i
n
p
u
t
v
ar
iab
les
is
also
in
v
esti
g
ated
.
T
h
e
ex
p
er
im
en
ta
l
s
etu
p
also
in
clu
d
es
a
p
o
s
t
-
p
r
ed
ictio
n
an
aly
s
is
o
f
m
o
d
el
p
er
f
o
r
m
an
ce
ac
r
o
s
s
1
1
d
if
f
er
en
t
ty
p
es
o
f
PV
p
an
els
an
d
a
cr
o
s
s
-
clim
ate
co
m
p
ar
is
o
n
.
T
o
th
e
b
est
o
f
o
u
r
k
n
o
wle
d
g
e,
th
is
is
th
e
f
ir
s
t
s
tu
d
y
th
at
co
m
p
r
eh
en
s
iv
ely
e
x
am
in
es
all
im
p
ac
tin
g
f
ac
t
o
r
s
f
r
o
m
d
ata
s
o
u
r
ce
s
to
m
o
d
el
le
ar
n
in
g
a
n
d
in
clu
d
es
a
d
etailed
ex
p
er
i
m
en
tal
an
aly
s
is
.
T
h
is
p
ap
er
is
o
r
g
an
ized
as
f
o
llo
ws:
s
ec
tio
n
2
s
h
o
ws
a
d
etailed
liter
atu
r
e
r
ev
iew
o
f
PV
p
o
wer
p
r
ed
ictio
n
/f
o
r
ec
asti
n
g
,
co
v
e
r
i
n
g
asp
ec
ts
s
u
ch
as
d
ata
s
o
u
r
ce
s
,
in
p
u
t
p
atter
n
s
,
d
ata
p
r
ep
r
o
ce
s
s
in
g
tech
n
iq
u
es
an
d
p
r
ed
ictio
n
m
o
d
els.
I
n
s
ec
t
io
n
3
,
we
p
r
esen
t
th
e
e
x
p
er
im
en
tal
s
etu
p
th
at
d
elv
es
in
to
t
h
e
asp
ec
ts
d
is
cu
s
s
ed
in
th
e
p
r
ec
ed
in
g
s
ec
tio
n
.
Su
b
s
eq
u
en
tly
,
s
ec
tio
n
4
d
elv
es
in
to
th
e
d
is
cu
s
s
io
n
o
f
th
e
r
esu
lts
.
Ultim
ately
,
th
e
co
n
clu
d
in
g
s
ec
tio
n
en
ca
p
s
u
lates th
e
f
in
d
in
g
s
an
d
p
r
o
v
id
es in
s
ig
h
ts
in
to
th
e
f
u
tu
r
e
d
ir
ec
tio
n
s
o
f
th
is
s
tu
d
y
.
2.
L
I
T
E
R
AT
U
RE
R
E
VI
E
W
T
h
e
liter
atu
r
e
r
ev
iew
will
co
n
ce
n
tr
ate
o
n
th
e
p
r
ed
ictio
n
o
f
s
o
lar
PV
p
o
wer
g
en
er
ati
o
n
(
o
u
t
p
u
t
p
o
wer
)
,
with
a
s
p
ec
if
ic
em
p
h
asis
o
n
ef
f
ec
tiv
e
a
n
d
ac
c
u
r
ate
p
r
ed
ictio
n
s
.
T
h
is
co
m
p
r
e
h
en
s
iv
e
r
ev
iew
will
ex
am
in
e
v
ar
i
o
u
s
asp
ec
ts
,
in
clu
d
in
g
d
ata
s
o
u
r
ce
s
,
in
p
u
t
p
att
er
n
s
,
d
ata
p
r
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
es
an
d
m
o
d
elin
g
m
eth
o
d
s
em
p
lo
y
e
d
b
y
r
esear
c
h
er
s
to
ac
h
iev
e
th
ese
p
r
ed
icti
o
n
s
.
T
h
r
o
u
g
h
th
is
d
etailed
liter
atu
r
e
r
ev
iew,
we
will
p
r
o
v
id
e
a
c
o
m
p
r
e
h
en
s
iv
e
an
d
u
p
-
to
-
d
ate
o
v
e
r
v
iew
o
f
th
e
s
tate
-
of
-
th
e
-
a
r
t
in
s
o
lar
PV
p
o
wer
g
e
n
er
atio
n
p
r
ed
ictio
n
.
T
h
is
e
n
d
ea
v
o
r
aim
s
to
o
f
f
er
v
alu
a
b
le
in
s
ig
h
ts
f
o
r
f
u
t
u
r
e
r
esear
ch
an
d
th
e
ad
v
a
n
ce
m
en
t
o
f
r
o
b
u
s
t
an
d
ac
cu
r
ate
p
r
ed
ictiv
e
m
o
d
el
s
with
in
th
e
d
o
m
ain
o
f
s
o
lar
e
n
er
g
y
a
p
p
licatio
n
s
.
2
.
1
.
Da
t
a
s
o
urce
s
PV
p
o
wer
p
r
ed
ictio
n
is
v
er
y
d
ep
en
d
e
n
t
o
n
tim
e
an
d
wea
th
er
co
n
d
itio
n
s
.
C
o
n
s
eq
u
e
n
tly
,
r
esear
ch
er
s
tr
y
co
llect
d
ata
f
r
o
m
d
iv
e
r
s
e
s
o
u
r
ce
s
,
s
u
ch
as
m
eteo
r
o
l
o
g
ical
s
tatio
n
s
[
2
]
,
n
u
m
er
ical
wea
th
er
p
r
ed
ictio
n
(
NW
P)
m
o
d
els
[
3
]
,
web
s
ites
a
n
d
s
er
v
ices w
ith
p
u
b
lic,
p
r
iv
at
e
o
r
lim
ited
ac
ce
s
s
[
4
]
o
r
th
r
o
u
g
h
th
e
d
e
p
lo
y
m
e
n
t
o
f
s
en
s
o
r
s
an
d
p
h
y
s
ical
PV
p
lan
ts
[
2
]
.
T
h
is
d
ata
co
llectio
n
r
esu
lts
in
a
s
er
ies
o
f
u
n
iq
u
e
d
ata
r
ec
o
r
d
s
,
d
is
tin
g
u
is
h
ed
b
y
th
eir
r
esp
ec
ti
v
e
m
ea
s
u
r
em
en
t
tim
estam
p
s
.
T
h
er
e
ar
e
two
ty
p
es
o
f
d
atab
a
s
es:
u
n
i
-
v
ar
iate
an
d
m
u
ltiv
ar
iate.
Un
i
-
v
ar
iate
d
ata
b
ases
co
n
s
id
er
tim
e
as
th
e
o
n
ly
f
ac
to
r
th
at
ca
n
a
f
f
ec
t
th
e
P
V
p
o
wer
p
r
ed
ictio
n
,
s
o
o
n
ly
tim
e
an
d
PV
p
o
wer
a
r
e
th
e
in
p
u
t
v
a
r
iab
les
in
th
is
ca
s
e.
I
n
co
n
tr
ast,
m
u
ltiv
ar
iate
d
atab
ases
in
clu
d
e
ad
d
itio
n
al
f
ac
to
r
s
s
u
ch
as
wea
th
er
d
ata
(
e.
g
.
,
h
u
m
id
ity
,
air
te
m
p
er
atu
r
e,
p
r
ess
u
r
e,
p
r
ec
ip
itat
io
n
)
an
d
PV
p
an
el
ch
ar
ac
ter
is
tics
.
No
tab
ly
,
8
6
%
o
f
th
e
d
atab
ases
u
s
ed
in
t
h
e
r
e
v
iewe
d
ar
ticles f
all
in
to
th
e
m
u
ltiv
ar
iate
ca
teg
o
r
y
.
T
h
ese
m
u
ltiv
ar
iate
d
ata
b
ases
ca
n
b
e
f
u
r
th
e
r
ca
teg
o
r
ized
in
t
o
in
teg
r
ated
an
d
s
ep
ar
ated
s
u
b
s
ets:
i
)
I
n
teg
r
ate
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
4
,
No
.
5
,
Octo
b
e
r
2
0
2
4
:
5
9
2
6
-
5
9
4
0
5928
d
atab
ases
co
m
b
in
e
wea
th
er
d
ata
with
h
is
to
r
ical
PV
p
o
wer
o
u
tp
u
t,
wh
ile
ii
)
Sep
ar
ated
d
atab
ases
ar
e
ex
clu
s
iv
ely
d
ed
icate
d
t
o
wea
t
h
er
d
ata
o
r
PV
p
o
wer
d
ata.
G
en
er
ally
,
r
esear
ch
e
r
s
u
n
if
y
th
e
tim
e
s
tep
s
b
etwe
en
r
ec
o
r
d
s
to
m
er
g
e
b
o
th
s
ep
ar
ated
d
atab
ases
in
to
an
in
teg
r
ated
o
n
e.
T
ab
le
1
r
ep
r
esen
ts
p
u
b
licly
av
ailab
le
in
teg
r
ated
d
atasets
u
s
ed
b
y
au
th
o
r
s
f
o
r
p
r
ed
ictin
g
th
e
s
o
lar
p
o
wer
g
e
n
er
ated
b
y
p
h
o
to
v
o
ltaic
p
an
els.
Of
f
ice
o
f
s
cien
tific
an
d
tech
n
ical
in
f
o
r
m
atio
n
(
OSTI
)
d
ata
[
5
]
is
o
n
e
o
f
th
e
s
m
allest
d
atase
ts
in
ter
m
s
o
f
th
e
m
ea
s
u
r
em
en
t
p
er
io
d
,
co
v
er
i
n
g
ap
p
r
o
x
im
ately
o
n
e
y
ea
r
f
o
r
ea
ch
lo
ca
tio
n
.
Ho
wev
er
,
it
ex
ce
ls
in
ter
m
s
o
f
lo
ca
tio
n
s
an
d
in
p
u
t
v
ar
iab
les
with
3
lo
ca
tio
n
s
an
d
m
o
r
e
th
a
n
4
3
in
p
u
t
v
ar
iab
les.
I
n
th
e
ex
is
tin
g
liter
atu
r
e,
th
e
p
u
b
licly
av
ailab
le
d
atasets
ar
e
ty
p
ically
lim
ited
in
s
ize,
r
ar
ely
ex
ten
d
in
g
b
ey
o
n
d
o
n
e
y
ea
r
o
f
m
ea
s
u
r
em
en
ts
.
Attem
p
tin
g
to
m
e
r
g
e
lar
g
e
-
s
ca
le
s
ep
ar
ated
d
atasets
,
s
u
ch
as
wea
th
er
an
d
PV
p
o
wer
d
atasets
,
d
o
es
n
o
t
n
ec
ess
ar
ily
lead
to
a
s
u
b
s
tan
tial
d
atab
ase
d
u
e
to
v
ar
iatio
n
s
in
d
ata
r
eso
lu
tio
n
b
etwe
e
n
th
e
two
s
o
u
r
ce
s
.
An
o
th
er
s
ig
n
if
ican
t
co
n
s
tr
ain
t
ar
is
es
f
r
o
m
th
e
h
ig
h
p
r
ed
ic
tio
n
er
r
o
r
s
ass
o
ciate
d
with
N
W
P
m
o
d
els.
A
s
a
r
esu
lt,
m
ea
s
u
r
ed
d
ata
h
as
p
r
o
v
en
to
b
e
h
ig
h
ly
r
eliab
le
f
o
r
g
en
er
atin
g
ac
c
u
r
ate
r
esu
lts
[
6
]
.
Giv
en
th
e
cr
itical
r
o
le
o
f
wea
th
er
f
o
r
ec
ast
d
ata
f
o
r
PV
p
o
wer
p
r
ed
ictio
n
s
in
r
ea
l
-
wo
r
l
d
ap
p
licatio
n
s
,
f
u
r
th
er
in
v
esti
g
atio
n
s
s
h
o
u
ld
b
e
co
n
d
u
cted
to
en
h
a
n
ce
th
e
p
r
ec
is
io
n
o
f
f
o
r
ec
ast
ed
wea
th
er
d
ata,
th
er
eb
y
im
p
r
o
v
in
g
th
e
o
v
er
all
ef
f
icien
cy
o
f
th
e
s
y
s
tem
.
T
ab
le
1
.
T
h
e
u
s
ed
p
u
b
lic
d
atas
ets in
th
e
liter
atu
r
e
D
a
t
a
s
e
t
D
a
t
a
r
e
so
l
u
t
i
o
n
P
e
r
i
o
d
Lo
c
a
t
i
o
n
N
u
mb
e
r
o
f
i
n
p
u
t
v
a
r
i
a
b
l
e
s
D
e
sert
k
n
o
w
l
e
d
g
e
A
u
s
t
r
a
l
i
a
(
D
K
A
)
d
a
t
a
[
7
]
5
m
i
n
2
0
1
6
-
N
o
w
A
l
i
c
e
(
2
0
1
6
-
N
o
w
)
Y
u
l
a
r
a
(
2
0
1
9
-
N
o
w
)
W
e
a
t
h
e
r
d
a
t
a
:
6
P
a
n
e
l
c
h
a
r
a
c
t
e
r
i
st
i
c
s:
1
1
O
S
TI
d
a
t
a
[
5
]
5
m
i
n
2
0
1
1
-
2
0
1
4
C
o
c
o
a
(
2
0
1
1
-
2
0
1
2
)
G
o
l
d
e
n
(
2
0
1
2
-
2
0
1
3
)
Eu
g
e
n
e
(
2
0
1
2
-
2
0
1
4
)
W
e
a
t
h
e
r
d
a
t
a
:
8
P
a
n
e
l
c
h
a
r
a
c
t
e
r
i
st
i
c
s:
1
7
S
u
n
La
b
d
a
t
a
[
8
]
1
m
i
n
2
0
1
4
-
2
0
1
7
P
o
r
t
u
g
a
l
W
e
a
t
h
e
r
d
a
t
a
:
8
P
a
n
e
l
c
h
a
r
a
c
t
e
r
i
st
i
c
s:
4
P
h
o
t
o
v
o
l
t
a
i
c
p
o
w
e
r
o
u
t
p
u
t
(PVO)
d
a
t
a
[
9
]
15
mi
n
2
0
1
8
-
2
0
1
9
C
h
i
n
a
W
e
a
t
h
e
r
d
a
t
a
:
6
P
a
n
e
l
c
h
a
r
a
c
t
e
r
i
st
i
c
s:
1
0
S
a
f
i
d
a
t
a
[
1
0
]
5
m
i
n
2
0
1
6
-
2
0
1
8
M
o
r
o
c
c
o
W
e
a
t
h
e
r
d
a
t
a
:
3
P
a
n
e
l
c
h
a
r
a
c
t
e
r
i
st
i
c
s:
2
Le
c
c
e
d
a
t
a
[
1
1
]
5
m
i
n
2
0
1
2
-
2
0
1
3
I
t
a
l
y
W
e
a
t
h
e
r
d
a
t
a
:
3
P
a
n
e
l
c
h
a
r
a
c
t
e
r
i
st
i
c
s:
7
S
O
LETE
d
a
t
a
[
1
2
]
5
m
i
n
2
0
1
8
-
2
0
1
9
D
e
n
mar
k
W
e
a
t
h
e
r
d
a
t
a
:
7
P
a
n
e
l
c
h
a
r
a
c
t
e
r
i
st
i
c
s:
5
2.
2
.
I
np
ut
p
a
t
t
er
ns
I
n
p
u
t
p
atter
n
s
r
ef
er
to
t
h
e
in
p
u
t
v
ar
iab
les,
f
o
r
ec
asti
n
g
h
o
r
izo
n
s
an
d
d
ata
r
eso
lu
tio
n
.
W
ith
in
th
e
liter
atu
r
e,
a
p
leth
o
r
a
o
f
v
ar
iab
les
ar
e
em
p
lo
y
ed
,
f
allin
g
in
to
two
p
r
in
cip
al
ca
teg
o
r
i
es:
m
eteo
r
o
lo
g
ical
(
wea
th
er
)
an
d
p
a
n
el
ch
ar
ac
te
r
is
tics
v
ar
iab
les.
T
h
e
m
o
s
t
u
s
ed
wea
th
er
v
ar
iab
les
in
th
e
liter
atu
r
e
in
clu
d
e
am
b
ien
t
tem
p
er
atu
r
e,
s
o
lar
ir
r
ad
iatio
n
,
win
d
s
p
ee
d
a
n
d
h
u
m
id
ity
.
Pan
el
ch
ar
ac
ter
is
tics
v
ar
iab
les
h
av
e
g
ain
ed
r
ec
en
t
p
r
o
m
in
en
ce
b
ec
au
s
e
th
ey
h
av
e
s
h
o
wn
an
i
n
ter
esti
n
g
im
p
ac
t
o
n
PV
p
o
wer
p
r
e
d
ictio
n
b
y
im
p
r
o
v
in
g
th
e
ac
cu
r
ac
y
o
f
th
e
m
o
d
els
[
1
]
,
[
1
3
]
.
Am
o
n
g
th
ese,
p
a
n
el
tem
p
er
atu
r
e
s
tan
d
s
o
u
t
as
th
e
m
o
s
t
f
r
eq
u
en
tly
u
s
ed
v
ar
iab
le,
s
h
o
wca
s
in
g
a
c
o
m
p
ar
ab
le
in
f
lu
e
n
ce
o
n
s
o
lar
ce
ll
p
er
f
o
r
m
an
ce
as
am
b
ie
n
t
tem
p
er
atu
r
e
[
1
4
]
.
T
h
e
af
o
r
em
en
tio
n
ed
in
p
u
t
v
ar
iab
le
s
ar
e
h
ig
h
ly
im
p
ac
tin
g
m
o
d
el
p
er
f
o
r
m
an
ce
.
Ho
wev
er
,
s
o
m
e
v
ar
iab
les
m
ay
n
o
t
b
e
av
ailab
le
d
u
e
to
th
e
n
o
n
-
m
ee
t
o
f
d
ata
q
u
ality
th
r
esh
o
ld
s
o
r
th
e
s
en
s
o
r
’
s
h
ig
h
co
s
ts
.
E
s
tab
lis
h
in
g
an
ef
f
icien
t
f
ea
tu
r
e
en
g
i
n
ee
r
in
g
s
tr
ateg
y
b
ec
o
m
es
im
p
er
ativ
e
to
ex
tr
ac
t
th
e
ca
u
s
al
r
elatio
n
s
h
ip
b
etwe
en
in
p
u
t
v
ar
iab
les
an
d
th
e
tar
g
et
v
ar
i
ab
le
(
i.e
.
,
th
e
im
p
ac
t
o
f
ea
ch
v
ar
iab
le
s
ep
ar
ately
)
.
C
o
n
s
eq
u
en
tly
,
m
is
s
in
g
v
ar
iab
les
ca
n
b
e
e
f
f
ec
tiv
ely
s
u
b
s
titu
ted
with
s
u
itab
le
alter
n
ativ
es,
wh
ile
ess
en
tial
v
ar
iab
les
ca
n
b
e
em
p
h
asized
.
T
h
e
f
o
r
ec
ast h
o
r
i
zo
n
r
ef
er
s
to
th
e
f
u
t
u
r
e
p
er
io
d
f
o
r
w
h
ich
t
h
e
m
o
d
el
g
en
er
ates
p
r
e
d
ictio
n
s
.
T
h
er
e
ar
e
f
o
u
r
ty
p
es
o
f
f
o
r
ec
ast
h
o
r
izo
n
:
i
)
n
o
wca
s
tin
g
o
r
r
ea
l
-
tim
e
f
o
r
ec
asti
n
g
,
wh
ich
aim
s
to
p
r
o
v
id
e
ac
cu
r
ate
f
o
r
ec
asts
f
o
r
a
v
er
y
s
h
o
r
t
p
er
io
d
o
f
tim
e
s
p
an
n
i
n
g
m
in
u
tes,
ii
)
s
h
o
r
t
-
ter
m
f
o
r
ec
asti
n
g
,
wh
ich
is
lim
ited
to
s
ev
er
al
h
o
u
r
s
,
a
d
ay
o
r
at
m
o
s
t,
a
wee
k
,
iii
)
m
ed
iu
m
-
ter
m
f
o
r
ec
asti
n
g
,
ap
p
licab
le
f
o
r
f
o
r
ec
a
s
ts
ex
ten
d
in
g
u
p
to
s
ev
er
al
wee
k
s
an
d
iv
)
lo
n
g
-
t
er
m
f
o
r
ec
asti
n
g
,
a
d
d
r
ess
in
g
f
o
r
ec
asts
s
p
an
n
in
g
m
o
n
th
s
o
r
ev
en
y
ea
r
s
.
T
h
e
f
o
r
ec
ast
h
o
r
iz
o
n
p
atter
n
d
ep
e
n
d
s
o
n
th
e
p
r
o
d
u
ce
r
an
d
c
o
n
s
u
m
er
n
ee
d
s
.
Fo
r
ex
am
p
le,
n
o
wca
s
tin
g
an
d
s
h
o
r
t
-
ter
m
f
o
r
ec
asti
n
g
ar
e
wid
ely
u
s
ed
f
o
r
elec
tr
icity
m
ar
k
etin
g
,
p
r
icin
g
,
r
ea
l
-
tim
e
m
o
n
ito
r
in
g
an
d
ec
o
n
o
m
ic
lo
ad
d
is
p
atch
.
I
n
c
o
n
tr
ast,
lo
n
g
-
t
er
m
f
o
r
ec
asti
n
g
is
u
s
ed
f
o
r
m
an
ag
i
n
g
p
o
wer
tr
an
s
m
is
s
io
n
,
r
atio
n
in
g
an
d
d
is
tr
ib
u
tio
n
.
W
ith
in
th
e
liter
atu
r
e,
th
e
m
o
s
t
co
m
m
o
n
ly
u
s
ed
ter
m
s
ar
e
“
s
h
o
r
t
”
a
n
d
“
n
o
wca
s
tin
g
”
.
Ap
p
r
o
x
im
ately
2
2
%
o
f
th
e
r
e
v
iewe
d
ar
ticles
o
p
t
f
o
r
1
-
d
a
y
ah
ea
d
f
o
r
ec
asti
n
g
(
1
D)
,
wh
ile
1
3
%
f
av
o
r
1
-
h
o
u
r
ah
ea
d
f
o
r
ec
asti
n
g
(
1
H)
.
T
h
e
f
o
r
ec
asti
n
g
h
o
r
izo
n
is
r
elate
d
to
d
ata
r
eso
lu
tio
n
,
also
k
n
o
wn
as
tim
estep
,
d
ata
s
am
p
lin
g
,
in
ter
v
al
o
r
tim
e
h
o
r
izo
n
.
I
t
r
e
f
er
s
to
th
e
tim
e
in
ter
v
al
b
etwe
en
two
co
n
s
ec
u
tiv
e
d
ata
s
am
p
les.
I
t
ca
n
b
e
eq
u
al
to
o
r
less
th
an
th
e
f
o
r
ec
asti
n
g
h
o
r
izo
n
.
Fo
r
ex
am
p
le,
in
th
e
co
n
tex
t
o
f
a
1
-
d
ay
ah
ea
d
f
o
r
ec
asti
n
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
P
h
o
to
vo
lta
ic
p
o
w
er
p
r
ed
ictio
n
u
s
in
g
d
ee
p
lea
r
n
in
g
mo
d
els:
r
ec
en
t
a
d
va
n
ce
s
a
n
d
…
(
B
a
s
ma
S
a
a
d
)
5929
h
o
r
izo
n
,
3
2
%
o
f
r
ev
iewe
d
ar
ti
cles
s
elec
t
h
o
u
r
ly
r
eso
lu
tio
n
,
2
5
%
o
p
t
f
o
r
1
5
m
in
u
tes
r
eso
l
u
tio
n
(
1
5
m
in
)
a
n
d
1
7
% a
d
h
er
e
to
d
aily
r
eso
lu
tio
n
.
Mo
d
el
ac
cu
r
ac
y
is
also
r
elate
d
to
f
o
r
ec
asti
n
g
h
o
r
izo
n
an
d
d
ata
r
eso
lu
tio
n
.
I
n
a
p
r
ev
io
u
s
liter
atu
r
e
r
ev
iew
[
1
5
]
,
r
esear
ch
er
s
co
n
cl
u
d
ed
t
h
a
t
an
in
cr
ea
s
e
in
th
e
f
o
r
ec
asti
n
g
h
o
r
izo
n
d
u
r
atio
n
lead
s
to
an
in
cr
ea
s
e
in
th
e
er
r
o
r
m
etr
ic
.
Ho
wev
er
,
it
is
n
o
tewo
r
th
y
th
at
s
u
ch
a
c
o
n
c
lu
s
io
n
m
ay
n
o
t
b
e
u
n
iv
er
s
ally
ap
p
licab
le,
as
d
e
m
o
n
s
tr
ated
b
y
J
eb
li
et
a
l.
[
1
6
]
.
T
h
eir
s
tu
d
y
in
d
icate
d
th
a
t,
wh
ile
u
s
in
g
th
e
r
an
d
o
m
f
o
r
est
m
o
d
el,
as
th
e
f
o
r
ec
asti
n
g
h
o
r
izo
n
in
cr
ea
s
e
s
,
th
e
er
r
o
r
m
etr
ic
ten
d
s
to
d
ec
r
ea
s
e.
T
h
u
s
,
th
e
r
elatio
n
s
h
ip
b
etwe
en
f
o
r
ec
as
tin
g
h
o
r
iz
o
n
an
d
m
o
d
el
ac
cu
r
ac
y
m
a
y
v
ar
y
d
ep
e
n
d
in
g
o
n
th
e
s
p
ec
if
ic
ch
ar
ac
ter
is
tics
o
f
th
e
d
ataset
a
n
d
th
e
p
r
ed
ictiv
e
m
o
d
el
b
ein
g
em
p
lo
y
e
d
.
Fu
r
th
er
in
v
esti
g
atio
n
an
d
an
aly
s
is
ar
e
r
eq
u
ir
ed
to
f
u
ll
y
u
n
d
er
s
tan
d
t
h
e
im
p
licatio
n
s
o
f
t
h
e
f
o
r
ec
a
s
tin
g
h
o
r
izo
n
o
n
m
o
d
el
a
cc
u
r
ac
y
ac
r
o
s
s
v
ar
io
u
s
s
ce
n
ar
io
s
.
Data
r
eso
lu
tio
n
is
a
cr
u
cial
f
ac
to
r
in
f
lu
en
ci
n
g
th
e
ac
cu
r
ac
y
o
f
PV
p
o
wer
p
r
ed
ict
io
n
,
as
h
ig
h
lig
h
ted
in
th
e
s
tu
d
y
b
y
Saad
et
a
l.
[
1
7
]
.
T
o
p
r
o
v
id
e
t
h
e
m
o
d
el
with
m
o
r
e
d
etailed
in
f
o
r
m
atio
n
,
it is
ad
v
is
ab
le
to
u
tili
ze
s
m
all
d
ata
r
eso
lu
tio
n
.
B
y
em
p
lo
y
in
g
f
in
er
-
g
r
ai
n
ed
d
ata,
th
e
m
o
d
el
ca
n
ca
p
tu
r
e
a
n
d
in
ter
p
r
et
s
u
b
tle
v
ar
iatio
n
s
in
th
e
in
p
u
t v
ar
iab
les,
u
ltima
te
ly
lead
in
g
to
m
o
r
e
p
r
ec
is
e
an
d
r
eliab
le
p
r
ed
ictio
n
s
.
2
.
3
.
Da
t
a
prepro
ce
s
s
ing
Data
p
r
ep
r
o
ce
s
s
in
g
is
a
f
u
n
d
am
en
tal
s
tep
in
d
ata
s
cien
ce
p
r
o
jects,
wh
er
e
th
e
u
ltima
te
g
o
al
is
to
p
r
ep
ar
e
t
h
e
d
ata
f
o
r
r
ec
o
g
n
itio
n
b
y
p
r
ed
ictiv
e
m
o
d
els
an
d
en
h
an
ce
d
ata
q
u
ality
.
T
h
is
p
r
o
ce
s
s
in
v
o
lv
es
v
ar
io
u
s
tech
n
iq
u
es
to
clea
n
,
tr
an
s
f
o
r
m
,
an
d
o
r
g
a
n
ize
th
e
d
ata.
T
h
e
m
o
s
t
u
s
ed
tech
n
iq
u
es
f
o
r
PV
p
o
wer
d
ata
p
r
ep
r
o
ce
s
s
in
g
in
r
ev
iewe
d
a
r
ticles
ar
e
o
r
g
an
ized
with
in
f
iv
e
s
u
b
-
s
tep
s
:
d
ata
clea
n
in
g
,
d
ata
en
r
ich
m
en
t
,
f
ea
tu
r
es im
p
o
r
ta
n
ce
,
d
ata
f
ilter
in
g
an
d
d
ata
n
o
r
m
aliza
tio
n
.
2
.
3
.
1
.
Da
t
a
c
lea
nin
g
Du
e
to
th
e
in
co
r
r
ec
t
f
o
r
m
ats,
er
r
o
n
eo
u
s
v
alu
es,
o
r
tr
an
s
m
is
s
io
n
er
r
o
r
s
,
o
u
tlier
s
(
ab
n
o
r
m
al
d
ata)
o
r
Nan
v
alu
es
(
m
is
s
in
g
d
ata)
c
o
u
ld
b
e
f
o
u
n
d
an
d
alter
in
a
cc
u
r
ately
th
e
m
o
d
el’
s
lear
n
in
g
.
T
h
er
ef
o
r
e,
d
ata
im
p
u
tatio
n
s
tr
ateg
ies
b
ec
o
m
e
ess
en
tial.
Fo
r
ex
am
p
le,
Z
h
u
et
a
l.
[
1
8
]
r
ep
lace
d
o
u
tlier
s
wit
h
th
e
p
r
ev
io
u
s
tim
e
p
o
in
t
v
alu
e,
d
em
o
n
s
tr
atin
g
t
h
e
im
p
ac
t
o
f
d
ata
clea
n
in
g
o
n
m
o
d
el
tr
ain
in
g
a
n
d
th
e
s
u
b
s
eq
u
e
n
t
en
h
an
ce
m
e
n
t
in
p
r
ed
ictio
n
ac
c
u
r
ac
y
.
Sev
er
al
s
tu
d
ies
in
[
1
9
]
–
[
2
1
]
r
esp
ec
tiv
ely
ex
clu
d
e
d
Nan
v
al
u
es
o
r
r
ep
lace
d
th
em
u
s
in
g
lin
ea
r
in
ter
p
o
latio
n
o
n
d
ata
f
r
o
m
th
e
co
r
r
esp
o
n
d
in
g
tim
e
o
f
s
im
ilar
d
ay
s
.
Ho
wev
er
,
W
an
g
et
a
l
.
[
2
2
]
a
n
d
Ng
u
y
en
et
a
l.
[
6
]
r
em
o
v
ed
w
h
o
lly
o
u
tlier
s
.
2
.
3
.
2
.
Da
t
a
enrichm
ent
D
a
t
a
e
n
r
ic
h
m
e
n
t
e
n
h
a
n
c
es
d
a
ta
q
u
a
l
i
t
y
b
y
a
d
d
i
n
g
e
x
t
r
a
v
a
r
ia
b
l
e
s
,
i
n
c
l
u
d
i
n
g
w
e
at
h
e
r
t
y
p
e
a
n
d
s
e
as
o
n
c
l
a
s
s
i
f
ic
a
t
i
o
n
[
2
3
]
,
[
2
4
]
.
W
e
a
t
h
e
r
t
y
p
e
c
l
a
s
s
i
f
i
c
at
i
o
n
r
e
l
i
es
o
n
v
a
r
i
a
b
l
e
s
l
i
k
e
g
r
o
u
n
d
r
a
d
i
a
t
i
o
n
i
n
t
e
n
s
it
y
a
n
d
h
i
s
t
o
r
i
c
al
r
a
d
i
at
i
o
n
r
a
ti
o
s
.
S
e
as
o
n
c
l
as
s
i
f
i
c
a
ti
o
n
d
i
v
i
d
es
t
h
e
y
e
a
r
i
n
t
o
f
o
u
r
p
e
r
i
o
d
s
t
o
i
m
p
r
o
v
e
d
a
t
a
q
u
a
l
i
t
y
a
n
d
a
s
s
e
s
s
m
o
d
e
l
p
e
r
f
o
r
m
a
n
c
e
.
H
o
w
e
v
e
r
,
a
d
d
i
n
g
w
e
a
t
h
e
r
t
y
p
e
b
as
e
d
s
o
l
el
y
o
n
s
o
l
a
r
i
r
r
a
d
i
a
n
c
e
m
a
y
b
e
u
n
r
e
l
i
a
b
l
e
,
a
s
o
t
h
e
r
f
a
c
t
o
r
s
l
i
k
e
c
l
o
u
d
a
m
o
u
n
t
a
n
d
h
u
m
i
d
i
t
y
a
r
e
c
r
u
c
i
a
l
.
M
o
r
e
o
v
e
r
,
c
l
i
m
a
t
e
c
h
a
n
g
e
c
a
n
c
a
u
s
e
r
a
p
i
d
w
e
a
t
h
e
r
f
l
u
c
t
u
a
t
i
o
n
s
w
i
t
h
i
n
a
d
a
y
.
C
a
r
ef
u
l
c
o
n
s
i
d
e
r
a
ti
o
n
o
f
t
h
e
s
e
f
a
ct
o
r
s
i
s
e
s
s
e
n
t
ia
l
d
u
r
i
n
g
d
a
t
a
p
r
e
p
r
o
c
e
s
s
i
n
g
.
2
.
3
.
3
.
F
ea
t
ures Im
po
rt
a
nce
Featu
r
es
im
p
o
r
tan
ce
s
er
v
es tw
o
p
r
im
ar
y
p
u
r
p
o
s
es: d
ata
r
e
d
u
ctio
n
an
d
m
o
d
el
o
u
tp
u
t e
x
p
lai
n
ab
ilit
y
.
−
Data
r
ed
u
ctio
n
: it
is
a
d
ata
p
r
ep
r
o
ce
s
s
in
g
ap
p
r
o
ac
h
aim
s
to
r
ed
u
ce
d
ata
d
im
en
s
io
n
ality
an
d
im
p
r
o
v
e
m
o
d
el
ef
f
icien
cy
b
y
lim
itin
g
th
e
n
u
m
b
er
o
f
u
s
ed
v
ar
ia
b
les.
T
h
e
tech
n
iq
u
es
u
s
ed
in
th
e
liter
atu
r
e
i
n
clu
d
e
Pear
s
o
n
co
r
r
elatio
n
[
2
]
,
[
2
5
]
wh
e
r
e
au
th
o
r
s
r
em
o
v
e
th
e
v
ar
iab
les
w
ith
lo
w
v
alu
es
o
f
th
e
co
r
r
elatio
n
co
ef
f
icien
t.
Prin
cip
al
co
m
p
o
n
e
n
t
an
aly
s
is
(
PC
A)
wh
ich
m
ap
s
th
e
d
ata
i
n
to
a
lo
wer
-
d
im
en
s
io
n
al
s
p
ac
e
[
2
6
]
.
Step
wis
e
an
d
s
tan
d
ar
d
r
eg
r
ess
io
n
[
1
9
]
,
[
2
7
]
a
r
e
u
s
ed
to
s
elec
t
th
e
in
p
u
t
v
ar
iab
les
th
at
h
ig
h
ly
af
f
ec
t
th
e
PV
p
o
wer
o
u
tp
u
t.
Alm
eid
a
et
a
l.
[
2
8
]
an
d
Saad
et
a
l.
[
2
9
]
f
ig
u
r
ed
o
u
t
th
at
all
p
r
ed
icto
r
s
ar
e
im
p
o
r
tan
t
to
g
et
m
o
r
e
ac
cu
r
ate
p
r
e
d
ictio
n
s
.
−
E
x
p
lain
ab
le
ar
tific
ial
in
tellig
e
n
ce
(
XAI
)
:
it
is
a
p
o
s
t
-
h
o
c
ex
p
lan
atio
n
tech
n
iq
u
e,
em
p
l
o
y
ed
to
g
ain
in
s
ig
h
ts
in
to
th
e
co
n
tr
ib
u
tio
n
s
o
f
in
p
u
t v
ar
iab
les
an
d
p
r
o
v
i
d
e
ex
p
la
n
a
tio
n
s
f
o
r
b
lac
k
-
b
o
x
p
r
ed
ictiv
e
m
o
d
els.
XAI
is
p
ar
ticu
lar
ly
p
r
ev
alen
t
in
en
er
g
y
an
d
p
o
wer
s
y
s
tem
s
ap
p
licati
o
n
s
,
o
f
ten
u
s
ed
with
in
tr
ee
-
b
a
s
ed
m
o
d
els
lik
e
ex
tr
em
e
g
r
ad
ien
t
b
o
o
s
tin
g
(
XGBo
o
s
t)
.
Sh
ap
ely
ad
d
itiv
e
ex
p
lan
atio
n
(
SHAP)
an
d
lo
ca
l
in
ter
p
r
etab
le
m
o
d
el
a
g
n
o
s
tic
e
x
p
lan
atio
n
s
(
L
I
ME
)
a
r
e
two
wid
ely
u
s
ed
XAI
tech
n
iq
u
es
i
n
th
is
d
o
m
ain
[
3
0
]
.
I
n
th
e
co
n
tex
t
o
f
PV
p
o
we
r
p
r
ed
ic
tio
n
,
Ku
zlu
et
a
l.
[
3
1
]
c
o
m
p
ar
ed
SHAP,
L
I
ME
an
d
E
L
I
5
m
eth
o
d
s
f
o
r
ex
p
lain
in
g
a
r
an
d
o
m
f
o
r
est
(
R
F)
r
eg
r
ess
o
r
.
T
h
ey
f
o
u
n
d
th
at
am
o
n
g
v
a
r
io
u
s
m
eth
o
d
s
,
SHAP
s
tan
d
s
o
u
t
a
s
th
e
o
n
ly
ap
p
r
o
ac
h
ca
p
a
b
le
o
f
p
r
o
v
id
i
n
g
c
o
m
p
r
e
h
en
s
iv
e
ex
p
lan
atio
n
s
b
y
co
n
s
id
er
i
n
g
all
p
o
ten
tial
p
r
ed
ictio
n
s
th
r
o
u
g
h
e
x
h
au
s
ti
v
e
co
m
b
in
atio
n
s
o
f
in
p
u
t
v
a
r
iab
les.
I
n
ter
esti
n
g
ly
,
th
e
in
t
eg
r
atio
n
o
f
XAI
tech
n
iq
u
es
with
d
ee
p
lear
n
in
g
(
DL
)
m
et
h
o
d
s
in
th
e
PV
p
o
wer
p
r
ed
ictio
n
d
o
m
ain
r
em
ai
n
s
u
n
ex
p
lo
r
ed
in
PV p
o
wer
p
r
ed
ictio
n
co
n
te
x
t.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
4
,
No
.
5
,
Octo
b
e
r
2
0
2
4
:
5
9
2
6
-
5
9
4
0
5930
2
.
3
.
4
.
Da
t
a
f
ilte
ring
Data
f
ilter
in
g
is
u
s
ed
f
o
r
s
ig
n
al
p
r
o
ce
s
s
in
g
.
T
h
e
wa
v
elet
tr
an
s
f
o
r
m
er
(
d
ec
o
m
p
o
s
itio
n
)
i
s
th
e
m
o
s
t
co
m
m
o
n
l
y
u
s
ed
tech
n
iq
u
e.
W
ith
o
v
er
1
4
wav
elet
tr
an
s
f
o
r
m
er
f
am
ilies
av
ailab
le,
s
elec
tin
g
th
e
m
o
s
t
s
u
itab
le
o
n
e
d
e
p
en
d
s
o
n
t
h
e
d
esire
d
v
ar
iab
les
to
b
e
ex
tr
ac
te
d
.
T
h
e
Sy
m
m
let
-
5
f
am
ily
is
co
n
s
id
er
ed
c
o
n
v
e
n
ien
t
b
y
C
h
ian
g
et
a
l
.
[
3
2
]
,
wh
ile
Z
h
u
et
a
l.
[
3
3
]
u
s
e
th
e
D
M
ey
er
wav
elet
f
am
ily
.
Ad
d
itio
n
ally
,
s
o
m
e
au
t
h
o
r
s
o
p
t
f
o
r
th
e
m
ax
im
u
m
o
v
er
lap
d
is
cr
ete
wav
elet
tr
an
s
f
o
r
m
(
MO
DW
T
)
an
d
Ma
llat
ap
p
r
o
ac
h
es
[
3
4
]
,
[
3
5
]
.
Alth
o
u
g
h
wav
elet
d
ec
o
m
p
o
s
itio
n
is
a
p
o
ten
t
m
eth
o
d
f
o
r
an
aly
zin
g
tim
e
s
er
ies
d
ata
an
d
ca
p
tu
r
i
n
g
f
r
eq
u
e
n
cy
-
r
elate
d
p
atter
n
s
,
its
in
teg
r
atio
n
with
lear
n
in
g
m
o
d
els
in
tr
o
d
u
ce
s
s
p
ec
if
ic
ch
allen
g
es.
No
tab
ly
,
it
ca
n
lead
to
an
escalatio
n
in
d
ata
d
im
en
s
io
n
al
ity
an
d
r
esu
lt in
i
n
f
o
r
m
atio
n
lo
s
s
d
u
r
in
g
th
e
tr
a
n
s
f
o
r
m
atio
n
p
r
o
ce
s
s
.
2
.
3
.
5
.
Da
t
a
no
rm
a
liza
t
io
n
No
r
m
aliza
tio
n
is
cr
u
cial
f
o
r
r
ea
l
-
wo
r
ld
d
ata
with
v
a
r
y
in
g
r
an
g
es
to
p
r
e
v
en
t
b
ias
in
lin
ea
r
p
r
ed
ictiv
e
m
o
d
els.
R
escalin
g
d
ata
b
etwe
en
0
an
d
1
o
r
b
etwe
en
-
1
an
d
1
en
s
u
r
es th
at
all
v
ar
iab
les h
av
e
eq
u
al
im
p
o
r
tan
ce
.
T
h
e
m
in
-
m
a
x
s
ca
lar
m
eth
o
d
s
u
b
tr
ac
ts
th
e
m
in
im
u
m
v
alu
e
an
d
d
iv
id
es
it
b
y
th
e
r
a
n
g
e
[
3
6
]
,
[
3
7
]
.
E
lam
i
m
et
a
l.
[
3
8
]
u
s
ed
an
alter
n
ate
f
o
r
m
u
la
b
ased
o
n
ac
tu
al,
m
in
im
u
m
an
d
m
a
x
im
u
m
v
alu
es
.
B
ess
a
et
a
l.
[
3
9
]
em
p
lo
y
ed
th
e
clea
r
s
k
y
m
o
d
el
f
o
r
PV
p
o
wer
n
o
r
m
aliza
tio
n
.
Nev
er
th
eless
,
n
o
p
r
e
v
io
u
s
s
tu
d
y
h
as
in
v
esti
g
ated
th
e
im
p
o
r
tan
ce
an
d
i
n
f
lu
en
ce
o
f
d
ata
n
o
r
m
aliza
tio
n
i
n
m
o
d
e
l le
ar
n
in
g
.
I
t
is
im
p
o
r
tan
t
to
n
o
te
t
h
at
th
e
ap
p
licatio
n
o
f
ce
r
tain
d
ata
p
r
ep
r
o
ce
s
s
in
g
s
u
b
-
s
tep
s
m
en
tio
n
ed
ab
o
v
e
m
ay
n
o
t
alwa
y
s
b
e
n
ec
ess
ar
y
t
o
ac
h
iev
e
ac
cu
r
ate
p
r
ed
ictio
n
s
.
Fo
r
ex
a
m
p
le,
d
ata
n
o
r
m
aliza
tio
n
m
ay
h
o
l
d
litt
le
s
ig
n
if
ican
ce
f
o
r
s
o
m
e
lear
n
in
g
m
o
d
els
(
e.
g
.
,
tr
ee
-
b
ased
m
o
d
els
wh
er
e
th
e
s
p
litt
in
g
d
ec
is
io
n
s
ar
e
b
ased
s
o
lel
y
o
n
th
e
o
r
d
er
in
g
o
f
v
alu
es
wit
h
in
ea
ch
f
ea
tu
r
e
a
n
d
n
o
t
o
n
t
h
e
ab
s
o
lu
te
s
ca
le
o
f
th
e
f
ea
tu
r
e)
.
Data
e
n
r
ich
m
e
n
t
m
ay
ca
u
s
e
d
ata
r
ed
u
n
d
a
n
cy
,
wh
er
e
th
e
s
am
e
p
atter
n
m
ay
e
x
is
t
in
m
u
ltip
le
f
ea
t
u
r
es.
On
th
e
o
th
er
h
a
n
d
,
d
ata
clea
n
in
g
is
a
cr
itical
s
tep
,
co
n
tr
ib
u
tin
g
to
th
e
cr
ea
tio
n
o
f
a
r
eliab
le
d
ataset
an
d
en
h
an
cin
g
its
q
u
ality
.
Hen
ce
,
clea
n
s
in
g
tech
n
iq
u
es
an
d
s
tr
ateg
ies
s
h
o
u
ld
b
e
ca
r
ef
u
lly
s
elec
ted
esp
ec
ially
d
ea
lin
g
with
a
s
u
b
s
tan
tial
n
u
m
b
er
o
f
m
is
s
in
g
s
am
p
les
o
r
o
u
tlier
s
.
Fu
r
th
er
r
esear
ch
is
n
ee
d
ed
to
p
r
o
v
id
e
in
s
ig
h
ts
in
to
d
ata
e
n
r
ich
m
en
t
a
n
d
d
ata
f
ilter
in
g
s
tr
ateg
ies an
d
h
o
w
to
av
o
id
d
ata
r
e
d
u
n
d
an
cy
as we
ll a
s
in
f
o
r
m
atio
n
lo
s
s
.
Ad
d
itio
n
ally
,
d
ata
r
ed
u
ctio
n
m
eth
o
d
s
u
s
ed
in
t
h
e
s
tu
d
ied
ar
ticles,
s
u
ch
as
PC
A
an
d
c
o
r
r
elatio
n
an
aly
s
is
,
ar
e
s
till
l
im
ited
in
ter
m
s
o
f
d
etec
tin
g
ca
u
s
ality
r
elatio
n
s
h
i
p
s
b
etwe
en
in
p
u
t
a
n
d
o
u
tp
u
t
v
ar
iab
les
an
d
th
eir
ap
p
licab
ili
ty
m
ay
v
a
r
y
ac
r
o
s
s
d
if
f
er
en
t sit
u
atio
n
s
,
esp
ec
ially
in
f
ea
tu
r
e
e
n
g
in
ee
r
in
g
s
ce
n
ar
i
o
s
.
2
.
4
.
P
re
dict
iv
e
m
o
dels
Pre
d
ictiv
e
m
o
d
els
lear
n
in
g
is
an
ess
en
tial
s
tep
in
t
h
e
ar
tifi
cial
in
tellig
en
ce
d
o
m
ain
.
I
t
ai
m
s
to
f
in
d
h
id
d
en
d
ata
p
atter
n
s
in
d
ata
an
d
g
en
er
ate
p
r
e
d
ictio
n
s
b
y
ap
p
ly
in
g
co
m
p
lex
m
ath
em
ati
ca
l
f
u
n
ctio
n
s
(
i.e
.
,
m
o
d
els),
ea
ch
with
its
u
n
iq
u
e
d
ata
in
p
u
t,
en
co
d
in
g
a
n
d
n
o
r
m
aliza
tio
n
r
eq
u
ir
e
m
en
ts
.
I
n
th
e
s
tu
d
ied
ar
ticles,
f
o
u
r
ty
p
es o
f
lear
n
in
g
m
o
d
els
ar
e
u
s
ed
:
2
.
4
.
1
.
P
er
s
is
t
ence
m
o
dels
T
h
ese
s
er
v
e
as
a
b
aselin
e
o
r
r
ef
er
en
ce
m
o
d
el,
d
eliv
er
in
g
g
o
o
d
r
esu
lts
f
o
r
s
h
o
r
t
tim
e
h
o
r
izo
n
s
an
d
d
ata
with
lo
w
v
ar
ia
b
ilit
y
[
4
0
]
.
A
n
aiv
e
v
er
s
io
n
ass
u
m
es
co
n
s
tan
t
atm
o
s
p
h
er
ic
co
n
d
itio
n
s
b
etwe
en
tim
e
an
d
+
1
,
m
ak
in
g
p
o
wer
at
tim
e
eq
u
a
l
to
p
o
wer
at
+
1
an
d
r
ely
in
g
o
n
r
ea
l
o
b
s
er
v
atio
n
s
at
tim
e
as
in
p
u
t.
W
h
ile
an
im
p
r
o
v
e
d
v
er
s
io
n
co
n
s
id
er
s
clea
r
s
k
y
c
o
n
d
itio
n
s
,
e
s
p
ec
ially
in
n
o
n
-
s
tatio
n
ar
y
tim
e
s
er
ies
[
4
1
]
.
I
f
th
e
ex
p
ec
ted
o
u
t
p
u
t p
o
wer
at
th
e
c
lear
s
k
y
co
n
d
itio
n
eq
u
als ze
r
o
,
th
e
p
o
wer
at
tim
e
+
1
will m
atch
th
e
clea
r
s
k
y
co
n
d
itio
n
p
o
wer
at
tim
e
+
1
.
T
h
is
m
o
d
el
q
u
an
tifie
s
th
e
wo
r
s
t
ac
ce
p
tab
le
p
er
f
o
r
m
an
ce
o
f
d
ee
p
lear
n
in
g
,
m
ac
h
in
e
lear
n
in
g
o
r
s
tatis
tical
m
o
d
els.
2
.
4
.
2
.
Sta
t
is
t
ica
l
m
o
dels
W
ell
-
s
u
ited
f
o
r
s
m
all
d
atasets
an
d
lim
ited
in
p
u
t
p
ar
am
eter
s
an
d
h
a
v
e
b
ee
n
ex
ten
s
iv
ely
e
m
p
lo
y
ed
i
n
tim
e
s
er
ies
p
r
ed
ictio
n
,
p
ar
ticu
lar
ly
f
o
r
PV
p
o
wer
o
u
t
p
u
t
p
r
e
d
ictio
n
.
No
ta
b
le
ex
a
m
p
les
in
c
lu
d
e
au
t
o
r
eg
r
ess
iv
e
in
teg
r
ated
m
o
v
in
g
av
er
a
g
e
(
AR
I
MA
)
[
4
2
]
,
au
to
r
e
g
r
ess
iv
e
m
o
v
in
g
av
e
r
ag
e
(
AR
MA
)
[
4
3
]
an
d
s
ea
s
o
n
al
au
to
r
eg
r
ess
iv
e
in
teg
r
ated
m
o
v
in
g
av
er
a
g
e
(
SAR
I
MA
)
[
4
4
]
.
Ho
wev
er
,
th
ese
m
o
d
els
m
ay
f
ac
e
ch
allen
g
es
in
ad
eq
u
ately
ca
p
tu
r
in
g
co
m
p
lex
n
o
n
lin
ea
r
r
elatio
n
s
h
i
p
s
in
h
er
en
t
in
ce
r
tain
tim
e
s
er
ies
d
ata.
Ad
d
itio
n
ally
,
th
eir
ab
ilit
y
to
h
an
d
le
in
tr
icate
s
ea
s
o
n
al
p
atter
n
s
o
r
tr
en
d
s
m
ay
b
e
lim
ited
w
h
en
c
o
m
p
a
r
ed
t
o
th
e
ca
p
ab
ilit
ies
o
f
s
o
m
e
m
ac
h
in
e
lear
n
in
g
an
d
d
ee
p
lear
n
in
g
m
o
d
els.
As
s
u
ch
,
th
e
ad
o
p
tio
n
o
f
d
ee
p
lear
n
in
g
m
o
d
els
in
th
is
d
o
m
ain
h
as
g
a
r
n
er
e
d
atten
tio
n
d
u
e
to
t
h
eir
p
o
te
n
tial
to
ad
d
r
ess
th
ese
lim
itatio
n
s
an
d
o
f
f
er
m
o
r
e
r
o
b
u
s
t
p
r
ed
ictio
n
ca
p
ab
ilit
ies [
3
6
]
,
[
4
1
]
.
2
.
4
.
3
.
M
a
chine
lea
rning
m
o
dels
As
o
p
p
o
s
ed
to
s
tatis
tical
m
o
d
els,
th
ese
m
o
d
els
ca
n
h
an
d
le
v
a
s
t
d
ata
an
d
co
m
p
lex
p
atter
n
s
.
T
h
er
e
is
a
wid
e
v
ar
iety
o
f
m
ac
h
i
n
e
lear
n
in
g
m
o
d
els,
b
u
t
in
th
is
r
ev
iew
,
we
will
o
n
ly
d
is
cu
s
s
th
e
m
o
s
t
u
s
ed
an
d
r
eliab
l
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
P
h
o
to
vo
lta
ic
p
o
w
er
p
r
ed
ictio
n
u
s
in
g
d
ee
p
lea
r
n
in
g
mo
d
els:
r
ec
en
t
a
d
va
n
ce
s
a
n
d
…
(
B
a
s
ma
S
a
a
d
)
5931
o
n
es,
n
o
ta
b
ly
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SVM)
,
lin
ea
r
r
eg
r
ess
io
n
(
L
R
)
an
d
r
an
d
o
m
f
o
r
est
(
R
F)
.
SVM
h
as
g
ain
ed
p
r
o
m
i
n
en
ce
d
u
e
to
it
s
im
p
r
ess
iv
e
p
er
f
o
r
m
an
ce
a
n
d
a
b
ilit
y
to
u
n
co
v
er
n
o
n
-
li
n
ea
r
d
e
p
en
d
e
n
cies.
R
ec
en
tly
,
s
ev
er
al
s
tu
d
ies h
av
e
f
o
u
n
d
th
at
d
ee
p
lear
n
i
n
g
m
o
d
els o
f
ten
o
u
tp
e
r
f
o
r
m
SVM
[
1
6
]
,
[
2
4
]
,
[
4
5
]
.
2
.
4
.
4
.
Dee
p
lea
rning
m
o
dels
Sit
u
ated
with
in
th
e
m
ac
h
in
e
l
ea
r
n
in
g
f
ield
,
d
ee
p
lea
r
n
in
g
u
tili
ze
s
ar
tific
ial
n
eu
r
o
n
s
to
ex
tr
ac
t
d
ata
p
atter
n
s
,
m
ir
r
o
r
i
n
g
th
e
h
u
m
an
b
r
ain
’
s
f
u
n
ctio
n
in
g
.
Pr
o
m
in
en
t
d
ee
p
lear
n
in
g
m
o
d
els
in
clu
d
e
class
ical
ar
tific
ia
l
n
eu
r
al
n
etwo
r
k
s
(
ANN)
o
r
m
u
ltil
ay
er
p
er
ce
p
tr
o
n
(
ML
P)
a
n
d
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
(
L
STM
)
.
4
1
%
o
f
th
e
r
ev
iewe
d
ar
ticles
r
eg
ar
d
ML
P
as
th
e
m
o
s
t
ac
cu
r
ate
alg
o
r
ith
m
b
y
co
m
p
ar
in
g
it
with
m
ac
h
i
n
e
lear
n
in
g
m
o
d
els,
with
L
STM
also
d
ee
m
e
d
ef
f
ic
ien
t
[
1
7
]
,
[
4
6
]
.
C
o
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
(
C
NN)
,
wh
ile
less
co
m
m
o
n
,
h
as
d
em
o
n
s
tr
ated
lo
wer
p
r
ed
ictio
n
er
r
o
r
s
in
ce
r
tain
ca
s
es
[
3
4
]
,
[
4
7
]
.
E
n
s
em
b
le
m
o
d
els,
o
r
h
y
b
r
id
m
o
d
els,
co
m
b
in
e
m
u
ltip
le
lear
n
i
n
g
m
o
d
els,
eith
er
o
f
th
e
s
am
e
o
r
d
if
f
er
en
t
ty
p
es,
o
f
f
er
in
g
en
h
an
ce
d
ac
cu
r
ac
y
in
p
r
ed
ictin
g
s
o
lar
e
n
er
g
y
p
r
o
d
u
c
tio
n
[
4
8
]
,
[
4
9
]
.
2
.
5
.
E
v
a
lua
t
io
n
t
ec
hn
iqu
e
s
E
v
alu
atio
n
tech
n
iq
u
es
p
lay
a
cr
u
cial
r
o
le
in
v
alid
atin
g
f
o
r
e
ca
s
tin
g
m
o
d
els,
esp
ec
ially
in
th
e
ca
s
e
o
f
PV
p
o
wer
f
o
r
ec
asti
n
g
,
wh
ich
f
alls
u
n
d
er
s
u
p
e
r
v
is
ed
lear
n
in
g
,
a
f
o
r
m
o
f
r
eg
r
ess
io
n
.
T
o
as
s
ess
th
e
ac
cu
r
ac
y
o
f
r
eg
r
ess
io
n
m
o
d
els,
we
s
h
o
u
ld
q
u
an
tify
an
d
o
b
s
er
v
e
h
o
w
clo
s
e
th
e
esti
m
ated
(
p
r
ed
icted
)
v
alu
es
ar
e
to
th
e
o
b
s
er
v
ed
(
r
ea
l)
v
al
u
es
b
y
c
alcu
latin
g
r
esid
u
als
u
s
in
g
d
i
f
f
er
en
t
m
ath
em
atica
l
tec
h
n
iq
u
es.
T
h
e
liter
atu
r
e
p
r
ed
o
m
i
n
an
tly
r
elies
o
n
r
o
o
t
m
ea
n
s
q
u
ar
e
d
er
r
o
r
(
R
MSE
)
an
d
m
ea
n
a
b
s
o
lu
te
er
r
o
r
(
MA
E
)
m
etr
ics
(1
)
a
n
d
(
2
)
,
o
f
ten
co
m
p
lem
e
n
ted
b
y
g
r
a
p
h
ical
r
e
p
r
esen
tati
o
n
s
o
f
p
r
ed
icted
v
er
s
u
s
ac
tu
a
l
v
alu
es.
R
MSE
,
a
co
m
m
o
n
l
y
u
s
ed
m
etr
ic,
ten
d
s
to
b
e
s
en
s
itiv
e
to
o
u
tlier
s
as
it
p
en
alize
s
lar
g
e
r
esid
u
als.
On
th
e
o
th
er
h
an
d
,
MA
E
o
f
f
er
s
a
m
o
r
e
b
alan
ce
d
v
iew
b
y
c
o
n
s
id
er
in
g
th
e
av
e
r
ag
e
ab
s
o
lu
te
d
if
f
er
en
c
e
b
etw
ee
n
p
r
ed
icted
an
d
ac
tu
al
v
alu
es,
g
iv
in
g
less
i
m
p
o
r
tan
ce
to
ex
tr
e
m
e
er
r
o
r
s
.
T
h
e
g
r
ap
h
ical
r
ep
r
esen
tatio
n
o
f
ac
tu
al
v
er
s
u
s
p
r
ed
icted
v
alu
es
r
em
ain
s
a
v
a
lu
ab
le
m
eth
o
d
f
o
r
ass
ess
in
g
th
e
clo
s
en
ess
b
etwe
en
p
r
e
d
ictio
n
s
an
d
ac
tu
al
d
ata
,
b
u
t it
ca
n
b
ec
o
m
e
ch
allen
g
in
g
to
in
ter
p
r
et
with
e
x
ten
s
iv
e
test
d
ata.
=
√
1
∗
∑
(
−
)
²
1
(
1
)
=
1
∗
∑
|
(
−
)
|
1
(
2
)
Ho
wev
er
,
wh
ile
th
ese
m
etr
ics
ef
f
ec
tiv
ely
e
v
alu
ate
m
o
d
el
p
e
r
f
o
r
m
a
n
ce
,
th
e
y
m
ay
n
o
t
s
er
v
e
as
a
f
air
b
asis
f
o
r
co
m
p
ar
in
g
s
y
s
tem
s
ac
r
o
s
s
d
if
f
er
en
t
p
r
ed
ictio
n
h
o
r
izo
n
s
a
n
d
d
atasets
.
T
o
e
n
ab
le
eq
u
itab
le
co
m
p
a
r
is
o
n
s
,
n
o
r
m
alize
d
m
etr
ics
s
u
ch
as
n
o
r
m
alize
d
m
ea
n
ab
s
o
lu
te
er
r
o
r
(
NM
AE
)
,
n
o
r
m
alize
d
r
o
o
t
m
ea
n
s
q
u
ar
ed
er
r
o
r
(
NR
MSE
)
an
d
m
ea
n
a
b
s
o
lu
t
e
p
er
ce
n
tag
e
er
r
o
r
(
MA
PE)
ar
e
in
tr
o
d
u
ce
d
(
3
)
,
(
4
),
an
d
(
5
)
.
Am
o
n
g
t
h
ese,
NM
AE
s
tan
d
s
o
u
t a
s
th
e
m
o
s
t f
r
eq
u
e
n
tly
ap
p
lied
m
etr
ic
in
th
e
liter
atu
r
e.
=
(
3
)
=
(
4
)
=
1
∗
∑
|
(
−
)
|
1
(
5
)
W
h
er
e
is
th
e
av
er
ag
e,
t
h
e
m
e
d
ian
o
r
t
h
e
s
tan
d
ar
d
d
ev
iatio
n
o
f
th
e
p
o
wer
g
e
n
er
ated
in
th
e
t
est s
et.
2
.
6
.
G
a
p
identif
ica
t
io
n
T
h
r
o
u
g
h
th
e
co
m
p
r
eh
en
s
iv
e
an
aly
s
is
o
f
th
e
cu
r
r
en
t
s
tate
-
of
-
th
e
-
a
r
t
in
PV
s
o
lar
en
e
r
g
y
p
r
o
d
u
ctio
n
p
r
ed
ictio
n
,
s
ev
er
al
c
r
itical
g
ap
s
h
av
e
em
er
g
ed
,
h
i
g
h
lig
h
tin
g
ar
ea
s
wh
er
e
f
u
r
th
e
r
r
esear
c
h
a
n
d
d
ev
elo
p
m
en
t
a
r
e
ess
en
tial:
−
T
em
p
o
r
al
d
ata
s
p
an
:
a
s
ig
n
if
ic
an
t
p
o
r
tio
n
o
f
p
u
b
licly
a
v
ailab
le
d
atasets
,
ar
o
u
n
d
7
3
%,
h
as
li
m
ited
tem
p
o
r
al
co
v
er
ag
e,
o
f
ten
n
o
t
ex
ten
d
i
n
g
b
ey
o
n
d
o
n
e
y
ea
r
.
T
h
is
d
ef
icie
n
cy
u
n
d
er
s
co
r
es
th
e
u
r
g
en
t
n
e
ed
f
o
r
ex
te
n
s
iv
e
d
atasets
s
p
an
n
in
g
m
u
ltip
le
y
e
ar
s
,
id
ea
lly
in
c
o
r
p
o
r
atin
g
co
m
p
r
eh
en
s
iv
e
wea
th
er
a
n
d
p
a
n
el
ch
ar
ac
ter
is
tic
d
ata.
−
NW
P
m
o
d
el
r
eliab
ilit
y
:
th
e
h
i
g
h
p
r
ed
ictio
n
er
r
o
r
s
ass
o
ciate
d
with
NW
P
m
o
d
els
r
aise
c
o
n
ce
r
n
s
ab
o
u
t
t
h
eir
s
u
itab
ilit
y
as
p
r
im
ar
y
d
ata
s
o
u
r
ce
s
.
T
h
e
r
eliab
ilit
y
o
f
m
ea
s
u
r
ed
d
ata
in
ac
h
iev
in
g
ac
cu
r
ate
r
esu
lts
h
as
b
ee
n
co
n
s
is
ten
tly
d
em
o
n
s
tr
ated
in
th
e
liter
atu
r
e.
T
o
p
r
e
d
ict
PV
p
o
wer
g
e
n
er
atio
n
ac
c
u
r
ate
ly
in
r
ea
l
wo
r
l
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
4
,
No
.
5
,
Octo
b
e
r
2
0
2
4
:
5
9
2
6
-
5
9
4
0
5932
co
n
d
itio
n
s
,
th
er
e’
s
a
n
ee
d
t
o
en
h
an
ce
th
e
r
eliab
ilit
y
o
f
f
o
r
ec
asted
wea
th
er
d
ata,
g
i
v
en
its
in
h
er
en
t
in
ac
cu
r
ac
ies co
m
p
ar
ed
to
m
ea
s
u
r
ed
wea
th
er
d
ata.
−
I
n
p
u
t
v
ar
iab
les
an
d
f
ea
tu
r
e
e
n
g
in
ee
r
in
g
:
wh
ile
th
e
liter
atu
r
e
p
r
o
v
id
es
a
m
u
ltit
u
d
e
o
f
i
n
p
u
t
v
a
r
iab
les,
u
n
d
er
s
tan
d
i
n
g
th
e
ca
u
s
al
r
elatio
n
s
h
ip
s
b
etwe
en
th
ese
v
ar
iab
les
is
n
o
t
alwa
y
s
clea
r
.
A
ls
o
,
s
o
m
e
v
ar
iab
les
ar
e
o
f
ten
o
m
itted
d
u
e
to
d
ata
q
u
ality
lim
itatio
n
s
o
r
h
ig
h
s
en
s
o
r
co
s
ts
.
A
m
o
r
e
c
o
m
p
r
e
h
en
s
iv
e
ex
p
lo
r
atio
n
o
f
ca
u
s
ality
r
elatio
n
s
h
ip
s
ca
n
h
elp
id
en
tify
s
u
b
s
titu
te
v
ar
ia
b
les
f
o
r
m
is
s
in
g
d
ata
an
d
p
r
i
o
r
itize
ess
en
tia
l
v
ar
iab
les f
o
r
ac
c
u
r
ate
p
r
ed
ictio
n
s
.
−
Data
p
r
ep
r
o
ce
s
s
in
g
im
p
ac
t
a
s
s
es
s
m
en
t:
d
ata
p
r
ep
r
o
ce
s
s
in
g
f
o
r
PV
p
o
wer
p
r
ed
ictio
n
m
o
d
els
p
r
esen
ts
an
o
th
er
ar
ea
o
f
co
n
ce
r
n
.
T
h
e
in
f
lu
en
ce
o
f
v
ar
io
u
s
d
ata
p
r
ep
r
o
ce
s
s
in
g
tech
n
iq
u
es
with
in
d
if
f
er
en
t
p
r
ed
ictiv
e
m
o
d
els r
eq
u
ir
es th
o
r
o
u
g
h
ev
alu
atio
n
.
−
Pre
d
ictiv
e
m
o
d
el
p
e
r
f
o
r
m
an
ce
:
wh
ile
d
ee
p
lear
n
in
g
m
o
d
els
h
av
e
g
ain
e
d
p
r
o
m
in
en
ce
,
th
er
e
is
a
n
ee
d
f
o
r
well
-
d
ef
in
ed
c
r
iter
ia
to
g
u
id
e
m
o
d
el
s
elec
tio
n
b
ased
o
n
s
p
ec
if
ic
d
atasets
an
d
ap
p
licatio
n
s
.
Fu
r
th
er
m
o
r
e,
th
er
e
is
a
clea
r
d
em
an
d
f
o
r
g
en
er
alize
d
p
r
ed
ictiv
e
m
o
d
els
th
at
ca
n
ac
cu
r
ately
f
o
r
ec
ast
PV
p
o
wer
u
n
d
er
d
iv
er
s
e
co
n
d
itio
n
s
,
in
clu
d
in
g
v
ar
io
u
s
wea
th
er
s
ce
n
ar
i
o
s
,
clim
ates
an
d
p
an
el
ty
p
es.
A
d
d
itio
n
ally
,
th
e
im
p
lem
en
tatio
n
o
f
p
o
s
t
-
p
r
ed
ic
tio
n
an
aly
s
is
is
ess
en
tial f
o
r
ac
h
iev
in
g
e
x
p
lain
ab
le
AI
s
o
l
u
tio
n
s
.
−
C
o
m
p
r
eh
en
s
iv
e
b
en
ch
m
a
r
k
:
g
iv
en
t
h
e
d
iv
e
r
s
e
d
atasets
u
s
ed
in
p
r
i
o
r
r
esear
c
h
with
in
d
if
f
er
e
n
t
ex
p
er
im
en
tal
s
etu
p
s
as m
en
tio
n
ed
in
T
a
b
le
2
,
wh
ic
h
h
av
e
y
i
eld
ed
a
m
u
ltit
u
d
e
o
f
d
is
p
ar
ate
co
n
clu
s
io
n
s
,
th
e
d
ev
elo
p
m
e
n
t
o
f
a
co
m
p
r
eh
e
n
s
iv
e
b
en
ch
m
ar
k
is
im
p
er
ativ
e.
Su
ch
a
b
en
ch
m
a
r
k
s
h
o
u
ld
s
y
s
tem
atica
lly
co
n
s
id
er
all
f
ac
to
r
s
in
f
lu
en
cin
g
PV
p
o
wer
p
r
ed
ictio
n
,
p
r
o
v
i
d
in
g
a
u
n
if
ied
a
n
d
h
o
lis
tic
p
er
s
p
ec
tiv
e
o
n
th
is
cr
itical
f
ield
.
I
n
co
n
cl
u
s
io
n
,
id
en
tif
y
in
g
t
h
e
s
e
g
ap
s
in
th
e
cu
r
r
en
t
s
tate
o
f
th
e
ar
t
u
n
d
er
s
co
r
es
th
e
n
ee
d
f
o
r
f
u
r
th
er
r
esear
ch
an
d
d
e
v
elo
p
m
e
n
t
in
th
e
f
ield
o
f
s
o
lar
PV
p
o
wer
p
r
ed
ictio
n
.
Ad
d
r
ess
in
g
th
ese
g
ap
s
will
n
o
t
o
n
l
y
co
n
tr
ib
u
te
t
o
m
o
r
e
ac
cu
r
ate
an
d
r
eliab
le
p
r
ed
ictio
n
m
o
d
els
b
u
t
also
en
h
a
n
ce
th
e
p
r
ac
tical
a
p
p
licatio
n
s
o
f
s
o
lar
en
er
g
y
tec
h
n
o
lo
g
y
.
T
h
e
s
u
b
s
eq
u
en
t
s
ec
tio
n
s
p
r
o
v
id
e
a
n
e
x
p
er
im
en
tal
an
aly
s
is
ad
d
r
ess
in
g
s
o
m
e
o
f
th
ese
ch
allen
g
es
an
d
g
a
p
s
.
E
s
p
ec
ially
th
e
n
ee
d
f
o
r
a
c
o
m
p
r
e
h
en
s
iv
e
b
en
c
h
m
ar
k
an
d
m
o
d
el
o
u
tco
m
e
ex
p
lan
atio
n
(
p
o
s
t
-
p
r
ed
ictio
n
an
aly
s
is
)
.
T
ab
le
2
.
T
h
e
ac
cu
r
ate
s
y
s
tem
s
in
th
e
liter
atu
r
e
A
r
t
i
c
l
e
P
e
r
i
o
d
F
o
r
e
c
a
st
i
n
g
h
o
r
i
z
o
n
I
n
p
u
t
v
a
r
i
a
b
l
e
s
D
a
t
a
p
r
e
p
r
o
c
e
ss
i
n
g
P
r
e
d
i
c
t
i
v
e
mo
d
e
l
Ev
a
l
u
a
t
i
o
n
(
N
M
A
E)
W
a
n
g
e
t
a
l
.
2
0
1
7
[
3
4
]
1Y
7
5
mi
n
P
V
p
o
w
e
r
D
a
t
a
n
o
r
ma
l
i
z
a
t
i
o
n
D
a
t
a
f
i
l
t
e
r
i
n
g
C
N
N
3
.
8
2
%
Ese
y
e
e
t
a
l
.
2
0
1
8
[
3
5
]
1Y
1D
S
o
l
a
r
r
a
d
i
a
t
i
o
n
/
h
u
mi
d
i
t
y
t
e
m
p
e
r
a
t
u
r
e
/
c
l
o
u
d
c
o
v
e
r
W
i
n
d
s
p
e
e
d
/
P
r
e
ssu
r
e
D
a
t
a
c
l
e
a
n
i
n
g
D
a
t
a
f
i
l
t
e
r
i
n
g
S
V
M
4
.
2
2
%
Li
et
a
l
.
2
0
1
9
[
3
7
]
1M
15
mi
n
P
V
p
o
w
e
r
D
a
t
a
n
o
r
ma
l
i
z
a
t
i
o
n
R
N
N
1
.
5
4
%
W
a
n
g
e
t
a
l
.
2
0
2
0
[
4
6
]
5Y
1D
P
V
p
o
w
e
r
D
a
t
a
n
o
r
ma
l
i
z
a
t
i
o
n
D
a
t
a
e
n
r
i
c
h
me
n
t
LSTM
2
.
7
8
%
N
g
u
y
e
n
e
t
a
l
.
2
0
2
1
[
6
]
1Y
5
mi
n
S
o
l
a
r
r
a
d
i
a
t
i
o
n
A
mb
i
e
n
t
t
e
m
p
e
r
a
t
u
r
e
W
i
n
d
s
p
e
e
d
/
h
u
mi
d
i
t
y
D
a
t
a
c
l
e
a
n
i
n
g
LSTM
3
.
4
9
%
3.
E
XP
E
R
I
M
E
N
T
A
L
SE
T
UP
T
h
is
s
ec
tio
n
m
eticu
lo
u
s
ly
c
o
n
s
id
er
s
all
f
ac
to
r
s
in
f
lu
e
n
cin
g
ac
cu
r
ate
s
o
lar
p
o
we
r
p
r
e
d
i
ctio
n
.
W
e
em
p
lo
y
th
e
in
teg
r
ate
d
m
u
ltiv
ar
iate
d
ataset,
OSTI
d
ata
[
5
]
,
as
o
u
r
d
ata
s
o
u
r
ce
.
F
o
llo
win
g
liter
atu
r
e
r
ec
o
m
m
en
d
atio
n
s
,
we
co
n
s
id
er
all
av
ailab
le
p
r
ed
icto
r
s
.
W
e
ch
o
o
s
e
f
o
r
ec
asti
n
g
h
o
r
izo
n
s
o
f
5
m
i
n
u
tes,
1
5
m
in
u
tes,
1
h
o
u
r
an
d
1
d
a
y
.
Data
p
r
ep
r
o
ce
s
s
in
g
p
r
io
r
itiz
es
d
ata
clea
n
in
g
as
th
e
m
o
s
t
cr
itical
s
tep
.
Fo
r
m
o
d
elin
g
,
we
u
tili
ze
th
e
th
r
e
e
h
ig
h
ly
ac
cu
r
ate
lear
n
in
g
m
o
d
els:
ML
P,
C
NN
,
an
d
L
STM
.
T
o
ev
al
u
ate
th
ese
m
o
d
els,
we
u
s
e
th
e
NM
AE
as
o
u
r
ad
o
p
ted
m
etr
ic.
T
h
e
ex
p
e
r
im
en
ts
wer
e
c
o
n
d
u
cted
u
s
in
g
Py
th
o
n
p
r
o
g
r
a
m
m
in
g
lan
g
u
ag
e.
Fo
r
d
ata
p
r
ep
r
o
ce
s
s
in
g
,
we
u
ti
lized
th
e
lib
r
ar
ies
Nu
m
Py
[
5
0
]
,
Pan
d
as
[
5
1
]
,
Ma
tp
lo
tlib
[
5
2
]
an
d
Seab
o
r
n
[
5
3
]
.
T
e
n
s
o
r
Flo
w
[
5
4
]
a
n
d
Sci
k
it
-
lear
n
[
5
5
]
wer
e
em
p
l
o
y
ed
f
o
r
m
o
d
el
lear
n
in
g
an
d
ev
alu
atio
n
.
T
h
e
wo
r
k
s
tatio
n
s
p
ec
if
icatio
n
s
ar
e
as f
o
llo
ws:
˗
Pro
ce
s
s
o
r
I
n
tel(
R
)
Xeo
n
(
R
)
Si
lv
er
4
2
1
4
R
C
PU @
2
.
4
0
GHz
2
.
3
9
GHz
;
˗
R
AM
in
s
talled
6
4
.
0
GB
;
˗
T
y
p
e
o
f
s
y
s
tem
: 6
4
-
b
it
o
p
er
ati
n
g
s
y
s
tem
,
x
6
4
-
b
ased
p
r
o
ce
s
s
o
r
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
P
h
o
to
vo
lta
ic
p
o
w
er
p
r
ed
ictio
n
u
s
in
g
d
ee
p
lea
r
n
in
g
mo
d
els:
r
ec
en
t
a
d
va
n
ce
s
a
n
d
…
(
B
a
s
ma
S
a
a
d
)
5933
3
.
1
.
Da
t
a
d
escript
io
n
T
h
e
d
ataset
u
s
ed
in
th
is
s
tu
d
y
,
OSTI
d
ata
[
5
]
,
en
co
m
p
ass
es
t
h
r
ee
lo
ca
tio
n
s
with
5
-
m
in
u
te
r
eso
lu
tio
n
.
I
n
o
u
r
an
al
y
s
is
,
we
s
elec
t
o
n
ly
E
u
g
e
n
e
(
Or
e
g
o
n
)
an
d
C
o
c
o
a
(
Flo
r
id
a)
lo
ca
tio
n
s
d
u
e
to
m
o
r
e
t
h
an
3
0
%
o
f
m
is
s
in
g
d
ata
in
th
e
Go
ld
en
(
Fl
o
r
id
a)
lo
ca
tio
n
.
E
u
g
en
e
r
ep
r
esen
ts
a
Ma
r
in
e
W
e
s
t c
lim
ate
an
d
C
o
co
a
r
ep
r
esen
ts
a
Su
b
tr
o
p
ical
clim
ate.
As
r
ep
o
r
ted
in
[
5
6
]
,
[
5
7
]
,
b
o
th
E
u
g
en
e
an
d
C
o
co
a
f
o
llo
w
a
ca
te
g
o
r
izatio
n
a
p
p
r
o
ac
h
b
ased
o
n
th
e
m
ea
n
h
o
u
r
ly
tem
p
er
atu
r
e
o
f
th
e
d
ay
o
v
e
r
th
e
y
ea
r
s
f
r
o
m
1
9
8
6
to
2
0
1
6
.
Day
s
ar
e
ass
ig
n
e
d
s
p
ec
if
ic
lab
els
ac
co
r
d
i
n
g
to
th
e
r
a
n
g
e
o
f
tem
p
er
atu
r
e
v
alu
es.
I
n
E
u
g
en
e,
t
h
is
ap
p
r
o
ac
h
r
esu
lts
in
th
e
class
if
icatio
n
o
f
d
ay
s
in
to
f
iv
e
ty
p
es:
c
o
ld
,
co
o
l,
co
m
f
o
r
tab
le,
war
m
an
d
h
o
t
.
I
n
C
o
co
a
d
ay
s
ar
e
ca
te
g
o
r
ize
d
in
to
th
r
ee
ty
p
es:
co
m
f
o
r
tab
le
,
war
m
an
d
h
o
t
.
T
h
is
d
iv
er
g
e
n
ce
in
class
if
icatio
n
r
e
f
lects
th
e
d
is
tin
ct
clim
ate
ch
ar
ac
ter
is
tics
o
f
ea
ch
lo
ca
ti
o
n
.
T
h
e
d
ataset
co
v
e
r
s
1
1
P
V
p
an
els
f
r
o
m
6
d
if
f
er
e
n
t
tech
n
o
lo
g
ies
in
clu
d
in
g
m
u
lti
-
cr
y
s
tallin
e
s
ilico
n
,
s
in
g
le
cr
y
s
tallin
e
s
ilico
n
,
am
o
r
p
h
o
u
s
s
ilico
n
tan
d
em
an
d
tr
i
p
le
-
j
u
n
ctio
n
,
ca
d
m
iu
m
tellu
r
id
e,
co
p
p
er
in
d
iu
m
g
alli
u
m
s
elen
id
e
an
d
Am
o
r
p
h
o
u
s
s
ilico
n
/cr
y
s
tallin
e
s
ilico
n
o
r
h
eter
o
ju
n
ctio
n
with
in
tr
in
s
ic
th
in
-
lay
er
PV
m
o
d
u
l
es.
T
h
is
d
ataset
co
m
p
r
is
es
o
v
er
4
3
v
ar
iab
les,
en
co
m
p
ass
in
g
wea
th
er
d
ata,
p
an
el
ch
ar
ac
ter
is
tics
,
s
en
s
o
r
s
u
n
ce
r
tain
ties
an
d
I
V
cu
r
v
es
d
ata.
R
i
g
o
r
o
u
s
q
u
ality
ass
u
r
an
ce
tech
n
iq
u
es
ar
e
ap
p
lied
to
r
em
o
v
e
d
ata
n
o
t
m
ee
tin
g
esta
b
lis
h
ed
q
u
ality
s
tan
d
ar
d
s
.
T
h
e
d
ata
is
p
ar
titi
o
n
ed
in
to
th
r
ee
s
ets:
a
tr
ain
in
g
s
et
co
m
p
r
is
in
g
7
0
%,
a
v
alid
atio
n
s
et
co
m
p
r
is
in
g
1
0
%,
an
d
a
t
est
s
et
co
m
p
r
is
in
g
2
0
%
o
f
th
e
d
ata.
T
h
e
s
p
ec
if
ic
d
ay
s
ch
o
s
en
f
o
r
ea
c
h
s
u
b
s
et
a
r
e
d
etailed
i
n
o
u
r
p
r
ev
i
o
u
s
wo
r
k
[
1
]
.
T
h
is
m
eth
o
d
allo
ws
t
h
e
m
o
d
el
to
b
e
tr
ain
e
d
an
d
test
ed
u
s
in
g
d
ata
s
p
an
n
in
g
th
e
en
tire
y
ea
r
,
f
r
o
m
J
an
u
a
r
y
to
Dec
em
b
er
.
3
.
2
.
Da
t
a
p
re
pro
ce
s
s
ing
I
n
th
is
s
tu
d
y
,
we
r
em
o
v
e
v
ar
iab
les
th
at
m
ay
ca
u
s
e
d
ata
leak
ag
e
f
r
o
m
th
e
co
llecte
d
d
ataset.
Par
ticu
lar
ly
th
o
s
e
ass
o
ciate
d
with
PV
p
o
wer
g
e
n
er
atio
n
,
s
u
ch
as
v
o
ltag
e
g
en
er
ated
b
y
th
e
PV
s
y
s
tem
o
r
I
V
cu
r
v
e
d
ata.
R
esu
ltin
g
in
th
e
s
elec
tio
n
o
f
s
p
ec
if
ic
v
ar
iab
les,
in
clu
d
in
g
i)
wea
th
er
v
a
r
iab
les
(
h
u
m
id
ity
,
tem
p
er
atu
r
e,
p
r
ec
ip
itatio
n
,
p
r
ec
ip
itatio
n
p
r
io
r
m
ain
ten
a
n
ce
an
d
p
r
ess
u
r
e)
,
ii)
p
a
n
el
ch
ar
ac
ter
is
tics
(
ar
ea
,
v
o
ltag
e,
c
u
r
r
en
t,
m
ain
ten
a
n
ce
,
tem
p
er
atu
r
e,
n
u
m
b
e
r
o
f
ce
l
ls
s
er
ies
an
d
p
ar
allel
ce
ll
s
tr
in
g
s
)
an
d
iii)
tim
e
v
ar
iab
les.
T
h
e
tar
g
et
is
n
o
r
m
al
ized
b
y
d
iv
id
i
n
g
it
b
y
t
h
e
ar
ea
o
f
t
h
e
p
a
n
el,
r
esu
ltin
g
in
t
h
e
u
n
it
W
/m
²,
th
e
a
r
ea
o
f
th
e
p
an
el
is
r
em
o
v
ed
t
o
av
o
id
u
n
n
ec
ess
ar
y
d
ata
c
o
m
p
lex
ity
.
W
e
h
an
d
le
m
is
s
in
g
v
al
u
es
b
y
r
e
m
o
v
in
g
d
a
y
s
with
m
o
r
e
th
an
th
r
ee
m
is
s
in
g
v
ar
iab
les
(
r
ep
r
esen
tin
g
o
n
ly
th
r
ee
d
ay
s
)
.
Fo
r
t
h
e
p
r
ec
ip
itatio
n
v
ar
ia
b
le
co
n
tain
in
g
ap
p
r
o
x
i
m
ately
3
%
m
is
s
in
g
v
alu
es,
we
f
ill
in
m
is
s
in
g
v
alu
es
o
f
en
tire
d
ay
s
u
s
in
g
th
e
n
ea
r
est
d
ay
’
s
p
r
ec
ip
itatio
n
(
s
et
to
0
)
.
I
n
c
ases
wh
er
e
o
n
ly
h
alf
-
d
ay
p
r
ec
ip
itatio
n
v
alu
es
ar
e
m
is
s
in
g
,
we
u
s
e
th
e
last
m
ea
s
u
r
ed
v
al
u
e
to
f
ill
th
e
g
a
p
s
.
T
im
e
an
d
m
ain
ten
an
ce
v
ar
i
ab
les
ar
e
en
co
d
ed
.
T
im
e
is
d
i
v
id
ed
in
t
o
f
o
u
r
s
u
b
-
v
ar
iab
les,
r
ep
r
esen
tin
g
th
e
m
o
n
th
,
d
a
y
,
h
o
u
r
an
d
m
in
u
te,
wh
ile
th
e
m
ain
ten
an
ce
f
ea
tu
r
e
is
co
n
v
er
ted
in
to
f
o
u
r
co
lu
m
n
s
,
r
ep
r
esen
tin
g
th
e
h
o
u
r
o
f
m
ain
ten
a
n
ce
s
tar
t,
m
in
u
te
o
f
m
ain
ten
an
ce
s
tar
t,
h
o
u
r
o
f
m
ain
ten
an
ce
e
n
d
an
d
m
in
u
te
o
f
m
ai
n
ten
an
ce
e
n
d
.
3
.
3
.
P
re
dict
iv
e
m
o
dels
s
elec
t
io
n
T
h
e
lear
n
in
g
m
o
d
els
u
s
ed
in
th
is
s
tu
d
y
in
clu
d
e
ML
P,
L
ST
M
,
an
d
C
NN.
Fo
r
m
o
d
el
o
p
ti
m
izatio
n
,
th
e
Gr
id
s
ea
r
ch
ap
p
r
o
ac
h
[
2
9
]
is
a
d
o
p
ted
,
wh
ic
h
ex
h
a
u
s
tiv
ely
ev
alu
ates
v
ar
io
u
s
co
m
b
in
atio
n
s
o
f
h
y
p
er
p
ar
a
m
eter
s
to
s
elec
t
th
e
b
est
-
p
er
f
o
r
m
in
g
r
eg
r
ess
o
r
.
W
h
ile
Gr
id
s
ea
r
c
h
in
cr
ea
s
es
th
e
co
m
p
u
tatio
n
a
l
tim
e
an
d
r
eso
u
r
ce
r
eq
u
ir
em
e
n
ts
d
u
e
to
its
ex
h
au
s
tiv
e
n
atu
r
e,
it
r
em
ain
s
th
e
o
p
tim
al
s
tr
ateg
y
f
o
r
t
h
o
r
o
u
g
h
ly
e
x
p
lo
r
in
g
all
p
o
ten
tial
p
a
r
am
eter
c
o
n
f
i
g
u
r
at
io
n
s
.
As
m
en
tio
n
ed
in
t
h
e
Fig
u
r
e
1
,
th
e
in
p
u
t
v
ec
to
r
(
)
is
n
o
t
t
h
e
s
am
e
f
o
r
all
p
r
ed
ictiv
e
m
o
d
els.
Fo
r
ML
P
m
o
d
el
th
at
is
n
o
t
a
b
le
to
h
an
d
le
m
u
ltid
im
en
s
io
n
al
d
ata,
th
e
Ag
g
r
eg
atio
n
p
r
o
ce
s
s
is
ad
o
p
ted
an
d
s
u
m
m
a
r
y
s
tatis
tics
ar
e
u
tili
ze
d
.
Fo
r
h
u
m
id
ity
an
d
tem
p
er
at
u
r
e
v
ar
ia
b
les,
in
clu
d
in
g
th
e
tem
p
er
atu
r
e
o
f
th
e
p
an
el,
n
ew
v
ar
iab
les
ar
e
g
e
n
er
ated
b
y
ex
tr
ac
tin
g
m
in
im
u
m
,
m
ax
im
u
m
an
d
av
er
a
g
e
v
alu
es
,
as
th
ese
v
ar
iab
les
ex
h
ib
it
s
ig
n
if
ican
t
f
lu
ctu
atio
n
s
th
r
o
u
g
h
o
u
t
th
e
d
ay
.
Fo
r
p
r
ess
u
r
e,
o
n
ly
th
e
av
er
ag
e
v
al
u
e
is
co
n
s
id
er
ed
.
Pre
ci
p
itatio
n
d
ata
is
ag
g
r
eg
ated
b
y
co
n
s
id
er
in
g
th
e
ac
cu
m
u
lated
v
alu
e
o
v
e
r
th
e
co
u
r
s
e
o
f
th
e
f
o
r
ec
asti
n
g
h
o
r
izo
n
.
T
h
e
tr
ai
n
in
g
in
p
u
t
d
ata
is
r
ep
r
esen
ted
in
th
e
s
h
ap
e
o
f
(
×
)
,
wh
er
e
d
en
o
tes
th
e
n
u
m
b
er
o
f
tr
ain
i
n
g
s
am
p
les
a
n
d
d
en
o
tes
th
e
n
u
m
b
er
o
f
v
a
r
iab
les.
I
n
th
e
c
ase
o
f
L
STM
an
d
C
NN
m
o
d
els,
th
e
d
ata
is
s
tr
u
ctu
r
ed
as
(
×
)
,
wh
er
e
r
ep
r
esen
ts
th
e
n
u
m
b
e
r
o
f
s
am
p
les,
r
ep
r
esen
ts
th
e
s
eq
u
en
ce
len
g
th
an
d
r
ep
r
esen
ts
th
e
n
u
m
b
er
o
f
v
ar
iab
les.
Data
ag
g
r
e
g
atio
n
s
tr
ateg
ies
ar
e
n
o
t
r
eq
u
i
r
ed
with
in
L
STM
an
d
C
NN
m
o
d
els,
as th
ey
in
h
e
r
en
tly
h
an
d
le
m
u
ltid
im
en
s
io
n
a
l d
ata.
Fo
r
C
NN,
all
d
is
tin
ct
5
-
m
in
u
te
s
am
p
les o
f
ea
ch
h
o
r
izo
n
a
r
e
co
n
s
id
er
ed
(
(
)
)
,
wh
ile
L
STM
in
co
r
p
o
r
at
es
th
e
p
r
ev
io
u
s
h
o
r
izo
n
d
ata
(
(
−
1
)
)
in
ad
d
itio
n
to
th
e
cu
r
r
e
n
t
s
am
p
les.
T
h
e
o
p
tim
al
s
eq
u
en
ce
len
g
th
f
o
r
L
STM
is
d
eter
m
in
ed
to
b
e
2
,
with
th
e
n
u
m
b
er
o
f
v
ar
iab
les
s
et
to
(
∗
)
,
wh
er
e
r
ep
r
esen
ts
th
e
in
p
u
t
v
a
r
iab
les’
len
g
th
a
n
d
M
d
en
o
tes
t
h
e
n
u
m
b
e
r
o
f
5
-
m
i
n
u
te
in
ter
v
als in
ea
ch
h
o
r
izo
n
.
3
.
4
.
E
v
a
lua
t
i
o
n m
et
rics
I
n
o
u
r
ca
s
e,
we
h
a
v
e
ch
o
s
en
t
o
u
s
e
th
e
MA
E
(
2
)
,
R
MSE
(
1
)
an
d
NM
AE
(
3
)
as
ev
alu
atio
n
m
etr
ics
.
Ou
r
ev
alu
atio
n
c
r
iter
ia
wer
e
d
esig
n
ed
with
two
p
r
im
ar
y
o
b
jectiv
es:
i)
to
ev
alu
ate
p
r
ed
ictio
n
ac
cu
r
ac
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
4
,
No
.
5
,
Octo
b
e
r
2
0
2
4
:
5
9
2
6
-
5
9
4
0
5934
co
m
p
r
eh
e
n
s
iv
ely
an
d
ii)
to
e
n
s
u
r
e
th
at
h
ig
h
er
PV
p
o
wer
v
a
lu
es,
r
ep
r
esen
tativ
e
o
f
p
ea
k
c
o
n
d
itio
n
s
,
wer
e
n
o
t
u
n
d
u
ly
p
en
alize
d
in
o
u
r
ass
es
s
m
en
t.
MA
E
an
d
NM
AE
p
la
y
p
iv
o
tal
r
o
les
as
d
ec
is
iv
e
m
e
tr
ics
f
o
r
ev
alu
ati
n
g
th
e
m
o
d
els
ac
r
o
s
s
d
if
f
er
en
t
clim
ates
an
d
h
o
r
izo
n
s
.
MA
E
p
r
o
v
id
es
in
s
ig
h
ts
in
to
th
e
m
ag
n
itu
d
e
o
f
er
r
o
r
s
p
r
o
d
u
ce
d
b
y
th
e
m
o
d
el,
o
f
f
er
in
g
a
clea
r
v
iew
o
f
its
p
er
f
o
r
m
a
n
ce
.
On
th
e
o
th
er
h
a
n
d
,
NM
AE
en
ab
les
co
m
p
r
eh
e
n
s
iv
e
ass
ess
m
en
t
s
o
f
v
ar
io
u
s
lear
n
in
g
m
o
d
els,
f
ac
ilit
atin
g
co
m
p
ar
is
o
n
s
o
f
th
eir
p
er
f
o
r
m
an
ce
ac
r
o
s
s
d
iv
er
s
e
f
o
r
ec
asti
n
g
h
o
r
iz
o
n
s
,
clim
ates
an
d
PV
p
an
el
tech
n
o
lo
g
ies.
T
h
is
m
etr
ic
is
p
ar
ticu
lar
ly
s
u
itab
le
f
o
r
o
u
r
s
p
ec
if
ic
s
ce
n
ar
io
as
it
allo
ws
u
s
to
ef
f
ec
tiv
ely
e
v
alu
ate
d
ata
f
r
o
m
d
i
f
f
er
en
t
p
r
e
d
ictio
n
h
o
r
i
zo
n
s
an
d
co
m
p
ar
e
it
with
o
th
er
m
o
d
els d
o
cu
m
en
te
d
in
th
e
liter
atu
r
e
th
at
o
f
te
n
e
m
p
lo
y
d
is
tin
ct
d
atasets
.
Fig
u
r
e
1
.
T
h
e
wo
r
k
f
lo
w
o
f
p
r
e
d
ictio
n
p
r
o
ce
s
s
f
o
r
ML
P,
L
ST
M
,
an
d
C
NN
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
4
.
1
.
I
m
pa
ct
o
f
f
o
re
ca
s
t
ing
ho
rizo
ns
a
nd
pre
dict
iv
e
m
o
dels
I
n
th
is
s
ec
tio
n
,
we
d
elv
e
in
to
th
e
im
p
ac
t
o
f
f
o
r
ec
asti
n
g
h
o
r
i
zo
n
s
o
n
PV
p
o
wer
p
r
ed
ictio
n
,
u
tili
zin
g
h
ig
h
ly
ac
cu
r
ate
lear
n
in
g
m
o
d
els
an
d
co
m
m
o
n
f
o
r
ec
asti
n
g
h
o
r
izo
n
s
f
r
o
m
ex
is
tin
g
liter
at
u
r
e.
Ou
r
o
b
jectiv
e
is
to
p
r
e
d
ict
PV
p
o
wer
g
en
er
atio
n
ac
r
o
s
s
two
g
e
o
g
r
a
p
h
ically
d
is
tin
ct
lo
ca
tio
n
s
with
v
a
r
y
in
g
clim
ates,
in
v
o
lv
in
g
1
1
p
an
els.
W
e
in
co
r
p
o
r
ate
d
a
ta
f
r
o
m
all
p
a
n
els
d
u
r
in
g
th
e
tr
ain
in
g
,
v
alid
atio
n
an
d
test
in
g
p
h
ases
o
f
m
o
d
el
d
ev
elo
p
m
e
n
t.
T
h
e
m
etr
ics
s
u
m
m
ar
ized
in
T
a
b
les
3
,
4
,
an
d
5
o
f
f
e
r
d
is
tin
ct
p
er
s
p
ec
tiv
es
o
n
m
o
d
el
p
er
f
o
r
m
an
ce
,
allo
win
g
u
s
to
ass
es
s
th
eir
ef
f
ec
tiv
en
ess
f
r
o
m
v
ar
io
u
s
an
g
les.
W
h
en
ev
al
u
atin
g
th
e
im
p
ac
t
o
f
th
e
f
o
r
ec
asti
n
g
h
o
r
izo
n
,
NM
AE
em
er
g
es
as
th
e
m
o
s
t
s
u
it
ab
le
m
etr
ic.
A
s
ig
n
if
ican
t
an
d
p
o
s
itiv
e
co
r
r
elatio
n
ex
is
ts
b
etwe
en
th
e
f
o
r
ec
asti
n
g
h
o
r
izo
n
a
n
d
th
e
NM
AE
m
etr
i
c
ac
r
o
s
s
C
NN,
L
STM
,
an
d
ML
P
m
o
d
els
f
o
r
b
o
th
clim
ates.
As
th
e
f
o
r
ec
asti
n
g
h
o
r
izo
n
ex
ten
d
s
,
NM
AE
in
cr
ea
s
es,
p
r
im
ar
ily
d
u
e
to
th
e
g
r
o
w
in
g
v
o
lu
m
e
o
f
d
ata
av
ailab
le
f
o
r
m
o
d
el
tr
ai
n
in
g
.
Fo
r
a
c
o
m
p
r
e
h
en
s
iv
e
ass
ess
m
en
t
o
f
th
e
b
est
m
o
d
el
p
e
r
f
o
r
m
an
ce
,
it
b
ec
o
m
es
ev
id
en
t
th
at
L
STM
o
u
tp
er
f
o
r
m
s
o
th
er
m
o
d
els
in
ter
m
s
o
f
MA
E
an
d
NM
AE
,
n
o
tab
ly
in
th
e
1
5
-
m
in
u
te
an
d
5
-
m
in
u
te
f
o
r
ec
asti
n
g
h
o
r
izo
n
s
a
cr
o
s
s
b
o
th
E
u
g
e
n
e
a
n
d
C
o
c
o
a
clim
ates
(
co
n
s
id
er
in
g
t
h
e
co
n
f
id
en
ce
in
te
r
v
al)
.
I
t
ac
h
iev
es
im
p
r
ess
iv
e
NM
AE
v
alu
es
o
f
1
.
6
9
%
an
d
1
.
3
9
%
f
o
r
5
-
m
i
n
u
tes
p
r
e
d
ictio
n
s
an
d
1
.
9
4
%
an
d
1
.
5
6
%
f
o
r
15
-
m
in
u
te
p
r
e
d
ictio
n
s
in
th
e
s
e
r
esp
ec
tiv
e
clim
ates.
Ho
wev
er
,
th
e
R
MSE
v
alu
es
d
o
n
o
t
s
h
o
w
s
ig
n
if
ican
t
d
is
p
ar
ities
am
o
n
g
th
e
m
o
d
els
.
T
h
e
s
ce
n
ar
io
ch
an
g
es
wh
e
n
d
ea
lin
g
with
h
o
u
r
ly
a
n
d
d
aily
p
r
ed
ictio
n
s
,
as
L
STM
’
s
p
er
f
o
r
m
an
ce
is
clo
s
er
to
o
r
wo
r
s
e
t
h
an
C
NN
b
y
tak
in
g
in
t
o
co
n
s
id
er
atio
n
all
m
etr
ics.
T
h
is
d
is
cr
ep
an
cy
ca
n
b
e
attr
ib
u
te
d
to
t
h
e
am
o
u
n
t
a
n
d
c
o
m
p
le
x
ity
o
f
tr
ain
in
g
d
ata.
L
STM
lev
er
ag
es
p
r
ev
io
u
s
o
b
s
er
v
atio
n
s
,
p
r
o
v
i
n
g
ch
allen
g
in
g
d
u
e
to
lim
ited
d
ata.
I
n
c
o
n
tr
ast,
C
NN
r
elies
s
o
lely
o
n
wea
th
er
d
ata
f
o
r
th
e
s
p
ec
if
ic
d
ay
,
s
im
p
lify
in
g
th
e
d
ata’
s
co
m
p
lex
ity
an
d
y
i
eld
in
g
b
etter
r
esu
lts
in
th
ese
s
ce
n
ar
io
s
.
Ou
r
ex
p
er
im
en
tal
f
in
d
in
g
s
p
r
o
v
i
d
e
a
v
alu
ab
le
co
m
p
ar
is
o
n
t
o
ex
is
tin
g
liter
atu
r
e,
as
d
etailed
in
T
ab
le
2
.
Sp
ec
if
ically
,
o
u
r
L
STM
m
o
d
e
l
ac
h
iev
ed
NM
AE
v
alu
es
o
f
1
.
3
9
%
an
d
1
.
6
9
%
f
o
r
th
e
5
-
m
i
n
u
te
h
o
r
izo
n
in
th
e
C
o
co
a
an
d
E
u
g
en
e
lo
ca
tio
n
s
,
r
esp
ec
tiv
ely
,
s
h
o
wca
s
in
g
im
p
r
o
v
em
en
ts
o
v
er
th
e
b
est NM
AE
o
f
3
.
4
9
% r
ep
o
r
ted
in
p
r
ev
io
u
s
s
tu
d
ies.
Ad
d
itio
n
a
lly
,
o
u
r
C
NN
m
o
d
els
f
o
r
th
e
d
aily
h
o
r
iz
o
n
in
b
o
th
C
o
co
a
a
n
d
E
u
g
en
e
(
2
.
6
1
%
an
d
2
.
6
7
%)
o
u
tp
e
r
f
o
r
m
ed
th
e
ex
is
tin
g
d
aily
s
y
s
tem
s
d
o
cu
m
en
ted
in
th
e
liter
atu
r
e
(
4
.
2
2
%)
.
R
eg
ar
d
in
g
th
e
1
5
-
m
in
u
te
h
o
r
iz
o
n
,
o
u
r
r
esu
lts
u
s
in
g
L
STM
in
C
o
co
a
an
d
E
u
g
en
e
(
1
.
6
3
%
an
d
1
.
8
1
%)
a
r
e
co
m
p
ar
ab
le
to
t
h
e
ac
cu
r
ate
1
5
-
m
in
u
te
h
o
r
izo
n
s
y
s
tem
(
1
.
5
4
%)
r
e
p
o
r
te
d
in
th
e
li
ter
atu
r
e
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
P
h
o
to
vo
lta
ic
p
o
w
er
p
r
ed
ictio
n
u
s
in
g
d
ee
p
lea
r
n
in
g
mo
d
els:
r
ec
en
t
a
d
va
n
ce
s
a
n
d
…
(
B
a
s
ma
S
a
a
d
)
5935
T
ab
le
3
.
T
h
e
NM
AE
(
%)
o
f
le
ar
n
in
g
m
o
d
els in
d
if
f
er
en
t
h
o
r
i
zo
n
s
an
d
clim
ates
C
l
i
ma
t
e
R
e
g
r
e
ss
o
r
Eu
g
e
n
e
C
o
c
o
a
1D
1H
1
5
mi
n
5
m
i
n
1D
1H
1
5
mi
n
5
m
i
n
M
LP
1
3
.
4
8
±
1
.
6
9
8
.
0
3
±
0
.
3
1
7
.
8
2
±
0
.
1
2
7
.
1
5
±
0
.
0
3
1
6
.
1
6
±
1
.
5
1
7
.
9
8
±
0
.
2
7
.
1
5
±
0
.
0
9
6
.
5
8
±
0
.
0
5
LSTM
3
.
8
6
±
0
.
3
1
2
.
1
0
±
0
.
0
7
1
.
9
4
±
0
.
0
3
1
.
6
9
±
0
.
0
1
4
.
2
7
±
0
.
5
4
2
.
0
9
±
0
.
0
6
1
.
5
6
±
0
.
0
2
1
.
3
9
±
0
.
0
1
C
N
N
2
.
6
1
±
0
.
1
6
2
.
1
2
±
0
.
0
5
2
.
0
0
±
0
.
0
2
1
.
8
2
±
0
.
0
1
2
.
6
7
±
0
.
1
7
2
.
1
2
±
0
.
0
6
1
.
8
1
±
0
.
0
2
1
.
5
2
±
0
.
0
1
T
ab
le
4
.
T
h
e
MA
E
(
W
/m
²)
o
f
lear
n
in
g
m
o
d
els in
d
if
f
er
en
t h
o
r
izo
n
s
an
d
clim
ates
C
l
i
m
a
t
e
R
e
g
r
e
ss
o
r
Eu
g
e
n
e
C
o
c
o
a
1D
1H
1
5
mi
n
5
m
i
n
1D
1H
1
5
mi
n
5
m
i
n
M
LP
5
4
8
.
9
9
±
6
8
.
7
2
7
.
2
2
±
1
.
0
0
7
.
2
9
±
0
.
1
2
2
.
4
5
±
0
.
0
2
7
3
1
.
2
6
±
6
8
.
5
1
3
0
.
2
4
±
0
.
7
5
7
.
4
4
±
0
.
0
9
2
.
6
0
±
0
.
0
2
LSTM
1
5
7
.
4
3
±
1
2
.
7
5
7
.
1
0
±
0
.
2
4
1
.
8
1
±
0
.
0
2
0
.
5
8
±
0
.
0
0
5
1
9
3
.
1
4
±
2
4
.
5
2
7
.
9
3
±
0
.
2
4
1
.
6
3
±
0
.
0
2
0
.
5
5
±
0
.
0
0
5
C
N
N
1
0
6
.
2
9
±
6
.
7
6
7
.
2
0
±
0
.
1
6
1
.
8
6
±
0
.
0
2
0
.
6
2
±
0
.
0
0
5
1
2
1
.
0
1
±
7
.
7
3
8
.
0
5
±
0
.
2
2
1
.
8
8
±
0
.
0
3
0
.
6
0
±
0
.
0
0
5
T
ab
le
5
.
T
h
e
R
MSE
(
W
/m
²)
o
f
lear
n
in
g
m
o
d
els in
d
if
f
er
e
n
t
h
o
r
izo
n
s
an
d
clim
ates
C
l
i
m
a
t
e
R
e
g
r
e
ss
o
r
Eu
g
e
n
e
C
o
c
o
a
1D
1H
1
5
mi
n
5
m
i
n
1D
1H
1
5
mi
n
5
m
i
n
M
LP
1
0
4
9
.
7
6
±
5
3
5
.
5
3
5
3
.
0
4
±
1
7
.
5
5
1
2
.
5
8
±
2
.
6
9
3
.
6
1
±
0
.
5
1
0
9
9
.
4
1
±
5
1
1
.
7
3
4
3
.
3
5
±
1
0
.
7
1
0
.
2
9
±
1
.
8
1
3
.
6
9
±
0
.
5
LSTM
2
2
8
.
8
0
±
1
2
2
.
1
1
1
3
.
1
5
±
7
.
1
8
2
.
7
8
±
0
.
5
4
0
.
8
7
±
0
.
1
5
3
5
1
.
6
2
±
3
0
2
.
2
4
1
2
.
8
2
±
5
.
1
7
2
.
4
3
±
0
.
5
7
0
.
9
1
±
0
.
2
1
C
N
N
1
3
8
.
0
0
±
5
0
.
1
9
1
0
.
4
1
±
2
.
3
2
.
7
3
±
0
.
5
0
0
.
9
2
±
0
.
1
5
1
5
2
.
4
3
±
5
5
.
8
3
1
2
.
4
4
±
4
.
1
4
2
.
7
9
±
0
.
6
3
0
.
9
5
±
0
.
2
1
4
.
2
.
I
m
pa
ct
o
f
da
t
a
re
s
o
lutio
n
T
o
in
v
esti
g
ate
th
e
im
p
ac
t
o
f
d
ata
r
eso
lu
tio
n
o
n
PV
p
o
wer
f
o
r
ec
asti
n
g
,
L
STM
m
o
d
el
wa
s
em
p
lo
y
ed
an
d
ev
alu
ate
d
at
d
if
f
e
r
en
t
tem
p
o
r
al
r
eso
lu
tio
n
s
to
p
r
e
d
ict
h
o
u
r
ly
g
e
n
er
ated
PV
p
o
we
r
.
5
m
in
u
tes,
1
5
m
i
n
u
tes
an
d
1
h
o
u
r
wer
e
th
e
s
elec
ted
r
eso
lu
tio
n
s
.
T
h
e
r
esu
lts
,
d
ep
ict
ed
in
th
e
Fig
u
r
e
2
,
r
e
v
ea
l
in
tr
i
g
u
in
g
in
s
ig
h
ts
in
t
o
th
e
r
elatio
n
s
h
ip
b
etwe
en
d
ata
r
eso
lu
tio
n
a
n
d
f
o
r
ec
asti
n
g
a
cc
u
r
ac
y
.
As
th
e
g
r
a
p
h
d
em
o
n
s
tr
ates,
th
e
NM
AE
v
alu
es
ex
h
ib
it
d
is
tin
ct
tr
en
d
s
ac
r
o
s
s
v
ar
y
in
g
r
eso
lu
tio
n
s
.
A
5
-
m
in
u
te
r
eso
lu
tio
n
r
esu
lts
in
h
ig
h
ly
ac
cu
r
ate
p
r
ed
ictio
n
s
with
a
lo
w
NM
A
E
o
f
a
p
p
r
o
x
im
ately
2
.
1
0
%
with
a
tig
h
t
u
n
ce
r
tain
t
y
m
ar
g
in
o
f
0
.
0
7
.
T
h
is
f
in
e
r
-
g
r
ain
ed
d
ata
r
eso
lu
tio
n
y
ield
e
d
th
e
m
o
s
t
ac
c
u
r
ate
p
r
ed
ictio
n
s
.
T
r
an
s
itio
n
in
g
to
a
1
5
-
m
in
u
t
e
r
eso
lu
tio
n
s
lig
h
tly
in
cr
ea
s
es
th
e
NM
AE
to
ab
o
u
t
2
.
2
3
%,
ac
co
m
p
an
ied
b
y
a
s
lig
h
t
r
is
e
in
u
n
ce
r
tain
ty
.
Fin
ally
,
at
a
1
-
h
o
u
r
r
eso
lu
tio
n
,
t
h
e
NM
AE
f
u
r
th
e
r
ex
p
a
n
d
s
to
ap
p
r
o
x
im
ately
2
.
2
6
%,
with
a
g
r
ea
ter
le
v
el
o
f
u
n
ce
r
tain
ty
o
f
0
.
1
3
.
T
h
ese
f
in
d
in
g
s
h
ig
h
lig
h
t
th
e
tr
ad
e
-
o
f
f
b
etwe
en
r
eso
lu
tio
n
a
n
d
f
o
r
ec
asti
n
g
ac
c
u
r
ac
y
.
Hig
h
er
-
r
eso
lu
tio
n
d
ata
,
s
u
ch
as
5
-
m
i
n
u
te
in
te
r
v
als,
lead
s
to
m
o
r
e
p
r
ec
is
e
p
r
ed
ic
tio
n
s
,
wh
ile
co
ar
s
er
r
eso
lu
ti
o
n
s
s
ac
r
if
ice
s
o
m
e
ac
cu
r
ac
y
f
o
r
r
e
d
u
ce
d
co
m
p
u
ta
tio
n
al
d
em
an
d
s
.
Fig
u
r
e
2
.
T
h
e
NM
AE
o
f
d
if
f
er
en
t r
eso
lu
tio
n
s
f
o
r
1
-
h
o
u
r
f
o
r
e
ca
s
tin
g
h
o
r
izo
n
u
s
in
g
th
e
L
STM
4
.
3
.
I
m
pa
ct
o
f
inp
ut
v
a
ria
bles
T
o
d
elv
e
d
ee
p
e
r
in
to
th
e
i
n
f
lu
en
ce
o
f
in
p
u
t
v
ar
ia
b
les
o
n
t
h
e
p
r
ed
ictio
n
o
f
PV
p
o
wer
u
s
in
g
an
L
STM
m
o
d
el,
we
h
ar
n
ess
ed
th
e
SH
AP
to
o
l
to
d
is
s
ec
t
ea
ch
f
ea
tu
r
e’
s
co
n
tr
ib
u
tio
n
to
th
e
o
v
er
a
r
ch
in
g
p
r
ed
ictio
n
.
As
m
en
tio
n
ed
in
Fig
u
r
e
3
,
th
e
r
e
s
u
lts
f
o
r
b
o
t
h
E
u
g
en
e
a
n
d
C
o
co
a
d
em
o
n
s
tr
ate
th
e
s
ig
n
i
f
ican
ce
o
f
all
p
r
e
d
icto
r
s
Evaluation Warning : The document was created with Spire.PDF for Python.