fati̇h onur hocaoğlu res_2015
TRANSCRIPT
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 1/80
MODELLING WIND AND PV DATA ANDSIZE OPTIMIZATION OF HYBRID
SYSTEMS USING ADVANCED
TECHNIQUES
Fatih Onur Hocaoglu
RES Winter School 2015
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 2/80
Outline
• Introduction to Wind/PV systems
• Wind Speed Modeling
• Solar Radiation Modeling• Hybrid System Sizing
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 3/80
Introduction to Wind/PV systems
WL ≤ WPV + WG + WB
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 4/80
Why Wind Speed Modeling
• Design of a proper wind energy system
(requires the prediction of wind speed
statistical parameters)
• Design of wind sensitive structures
• Air pollution studies
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 5/80
Model: Markov Processes
0 1ij
p
1
1ij
j
p
11 12 13 1
21 22 23 2
1 2 3
. . .
. . .
. . . . . . .
. . . . . . .
. . .
n
n
n n n nn
p p p p
p p p p
A
p p p p
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 6/80
Wind Speed Modeling using
Markov Approach
, 1, 2,...,ij
ij
ij
j
m p i j n
m
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 7/80
Wind Speed Modeling using
Markov Approach0.61 0.29 0.08 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.18 0.47 0.28 0.06 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.04 0.20 0.46 0.23 0.06 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00
0.01 0.05 0.20 0.45 0.22 0.06 0.01 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.06 0.26 0.42 0.20 0.04 0.01 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.01 0.07 0.28 0.41 0.18 0.04 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.01 0.01 0.07 0.29 0.39 0.21 0.02 0.01 0.00 0.00 0.00
0.00 0.00 0.00 0.01 0.02 0.09 0.30 0.38 0.17 0.02 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.01 0.04 0.12 0.33 0.39 0.08 0.02 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.01 0.03 0.18 0.27 0.37 0.08 0.05 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.18 0.09 0.23 0.41 0.09 0.00
0.00 0.00 0.00 0.00 0.08 0.08 0.00 0.00 0.15 0.31 0.08 0.31 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.00 0.00 0.00 0.00
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 8/80
Syntetic generation of wind speed
data
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 9/80
Wind Speed Modeling using
Markov Approach0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 ......
0.00 0.61 0.18 0.10 0.05 0.03 0.02 0.00 0.00 0.00 0.00 0.00 ......
0.00 0.25 0.24 0.24 0.14 0.06 0.04 0.01 0.00 0.00 0.00 0.00 ......
0.00 0.12 0.18 0.28 0.23 0.11 0.05 0.03 0.01 0.01 0.00 0.00 ......
0.00 0.05 0.11 0.15 0.23 0.19 0.12 0.07 0.02 0.03 0.00 0.00 ......
0.00 0.02 0.04 0.10 0.18 0.30 0.16 0.10 0.04 0.02 0.01 0.00 ......
0.00 0.01 0.03 0.05 0.10 0.17 0.28 0.17 0.10 0.04 0.02 0.01 ......
0.00 0.00 0.00 0.02 0.04 0.10 0.20 0.25 0.18 0.11 0.05 0.02 ......
0.00 0.00 0.00 0.00 0.02 0.06 0.12 0.23 0.25 0.13 0.10 0.05 ......
0.00 0.00 0.00 0.00 0.01 0.02 0.05 0.12 0.22 0.25 0.18 0.07 ......
0.00 0.00 0.00 0.00 0.00 0.01 0.03 0.05 0.13 0.22 0.23 0.16 ......
0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.04 0.05 0.15 0.22 0.21 ......
...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ......
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 10/80
Wind Speed Modeling using
Markov Approach
1
k
ik ij
j
P p
ik P is transition probability in the i th row at the k th state
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 11/80
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 12/80
Wind Speed Modeling using
Markov Approach0.61 0.90 0.98 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
0.18 0.64 0.92 0.98 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
0.04 0.23 0.69 0.92 0.98 0.99 1.00 1.00 1.00 1.00 1.00 1.00 1.00
0.01 0.06 0.26 0.72 0.93 0.99 1.00 1.00 1.00 1.00 1.00 1.00 1.00
0.00 0.00 0.06 0.32 0.75 0.95 0.99 1.00 1.00 1.00 1.00 1.00 1.00
0.00 0.00 0.01 0.08 0.36 0.77 0.95 0.99 1.00 1.00 1.00 1.00 1.00
0.00 0.00 0.01 0.02 0.09 0.38 0.76 0.97 0.99 1.00 1.00 1.00 1.00
0.00 0.00 0.00 0.01 0.03 0.12 0.42 0.80 0.98 1.00 1.00 1.00 1.00
0.00 0.00 0.00 0.00 0.02 0.05 0.17 0.50 0.89 0.97 1.00 1.00 1.00
0.00 0.00 0.00 0.00 0.00 0.01 0.04 0.22 0.49 0.86 0.95 1.00 1.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.18 0.27 0.50 0.91 1.00 1.00
0.00 0.00 0.00 0.00 0.08 0.15 0.15 0.15 0.31 0.62 0.69 1.00 1.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.00 1.00 1.00 1.00
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 13/80
Wind Speed Modeling using
Markov Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 14/80
Wind Speed Modeling using
Markov Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 15/80
Wind Speed Modeling using
Markov Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 16/80
Wind Speed Modeling using
Markov Approach
Observed DataGenerated Data from
Model-1
Generated Data from
Model-2
Min 0.00 0.02 0.00
Max 12.47 11.44 12.77
Mean 3.52 3.16 3.61
Median 3.35 2.95 3.47
Std 2.11 1.91 2.02
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 17/80
Wind Speed Modeling using
Markov Approach
0
5
10
15
20
1 3 5 7 9 11 13 15 17
Wind Speed (m/s)
Pro
bability(%)
Observed data Generated data from Model-1
Generated data from Model-2
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 18/80
Variation: Hidden Markov Models
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 19/80
Hidden Markov Models
'fatih'
< , fatih, , 0.3> == P( )=0.32 3 2 3
s s s s
E ={<x, ‘1’, x, 0.32> < x, ‘0’, x, 0.52> < x, ‘0’, y, 0.16> < y, ’1’, x, 1>}
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 20/80
Hidden Markov Models
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 21/80
Hidden Markov Models
The Problems that should be solved are,
1-) What is the probability of observed data O1,O2 ,…,OT ?Calculate P(O1,O2 ,…,OT |model )
2-) What state is most likely if we know the observation sequence? Findthe state sequence
q1,q2,…qT S.T P(q1,q2,…qT | O1,O2 ,…,OT , model )
3-) Given some data, how do we learn a good HMM to describe the
data
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 22/80
Viterbi Algorithm
1,
1
1, 1, 1
(1)
( ) ( ) , arg max ( ( )) ( )
( ) arg max ( )
t
t
i
c wi j k
i k ik
t t
s
s
t t s j P t P s s
t P s w
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 23/80
Wind Speed Modeling using
Hidden Markov Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 24/80
Wind Speed Modeling using
Hidden Markov Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 25/80
Wind Speed Modeling using
Hidden Markov Approach
( ).
in
n
oo
i j
o ij
n p s s
n p
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 26/80
Wind Speed Modeling using
Hidden Markov Approach
Min Max Mean Median Std
Measured 0 11.4000 3.5277 3.3500 2.1279
Model Generated 0 11.9523 3.3191 1.8253 3.1638
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 27/80
Wind Speed Modeling using
Hidden Markov Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 28/80
Wind Speed Modeling using
Hidden Markov Approach
Min Max Mean Median Std
Measured 0.11 8.2954 3.5277 3.4525 1.3917
Model Generated 0.2224 8.7660 3.5964 3.6058 1.5205
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 29/80
Solar Radiation Data Analysis
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 30/80
Solar Radiation Data Analysis
R R 2
x44
– x34
0.911 0.830
x44 –
x24 0.894 0.799
x44
– x14
0.898 0.806
x44
– x43
0.938 0.879
x44
– x42
0.807 0.651
x44
– x14
0.630 0.397
x44
– x33
0.870 0.760
x44
– x22
0.740 0.548
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 31/80
Solar Radiation Data Analysis
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 32/80
Solar Radiation Data Analysis
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 33/80
Solar Radiation Data Analysis
0 5 10 15 20 250
50
100
150
200
250
300
350
400
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 34/80
Optimal Coefficient Linear
Prediction Filters
,i j x , 1i j
x
1,i j x 1, 1 ?i j x
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 35/80
Optimal Coefficient Linear
Prediction Filters
1 111 12 13
21 22 23 2 2
31 32 33 3 3
a r R R R
R R R a r
R R R a r
1 2 3
0a a a
1a R r
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 36/80
Optimal Coefficient Linear
Prediction Filters
x11
x22 .
.
. .
.
xm1 . . . xmn
x12. . . . .x1n
.
.
1D-Filter 1
x11 x12
x22 .
.
. .
.
xm1 . . . xmn
. . . . .x1n
.
.
x11 x12
x22 .
.
. .
.
xm1 . . . xmn
.. .x1n
.
.
x11
x21 x22 .
.
. .
.
xm1 . . . xmn
. . . . .x1n x11
x21 x22 .
. . . . .x1n x11
x21 x22 .
.
. .
x41
xm1 . . . xmn
. . . . .x1n
x13
x31 x32 ... .
. . .
xm1 . . . xmn
x31
1D-Filter 4
1D-Filter 2
1D-Filter 5
1D-Filter 3
1D-Filter 6
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 37/80
Optimal Coefficient Linear
Prediction Filters
x11
x21
x12
x22 .
.
. .
.
xm1 . . . xmn
. . . . .x1n
x22 x21
x13
.
. .
.
xm1 . . . xmn
x11 x12 . x1n
2D-Filter 2 2D-Filter 3
x21
x12
.
. .
.
xm1 . . . xmn
x11 . x1n
2D-Filter 1
x22
2
1 1
1( ( , ) ( , ))
N M
i j
RMSE Rad i j Rad i j NM
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 38/80
Testing the Performance of 2D
Approach using OCLPF
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 39/80
Testing the Performance of 2D
Approach using NNs
Input Layer Output Layer Hidden Layer
i1 o
Input Layer Output Layer Hidden Layer
i1
i2
o
Input Layer Output Layer Hidden Layer
i2
i3
i1
o
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 40/80
Testing the Performance of 2D
Approach using NNs
0 10 20 30 40 50 60 70 80 90 10010
-2
10-1
100
Asama
P e r f o r m a n s
Performans: 0.0153904
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 41/80
Results
OCLPF
NNs
Train Data Test Data
RMSE RMSE RMSE
1-D Filter1/NN1 58.53 0.94 58.80 0.93 62.64 0.94
1-D Filter2/NN2 45.02 0.96 41.86 0.97 39.79 0.98
1-D Filter3/NN3 44.36 0.96 44.70 0.96 43.73 0.97
1-D Filter4/NN4 67.05 0.91 69.30 0.90 74.32 0.91
1-D Filter5/NN5 61.94 0.92 63.95 0.92 67.63 0.92
1-D Filter6/NN6 76.49 0.88 60.83 0.93 63.64 0.93
2-D Filter1/NN7 46.56 0.96 45.58 0.96 47.98 0.96
2-D Filter2/NN8 40.23 0.97 35.42 0.98 34.57 0.98
2-D Filter3/NN9 39.37 0.97 42.15 0.97 43.15 0.97
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 42/80
It is obvious from the literature that prediction performance increases in case of
different meteorological parameters are employed as inputs of models.
(Zarzalejo et al., 2005; Yorukoglu and Celik, 2006; Psiloglou and Kambezidis, 2007; Li et al.,
2010; Kaplanis and Kaplani, 2010; Pandey and Soupir, 2012; Inman et al., 2013;......)
Multi Dimensional Filter Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 43/80
It is obvious from the literature that prediction performance increases in case of
different meteorological parameters are employed as inputs of models.
(Zarzalejo et al., 2005; Yorukoglu and Celik, 2006; Psiloglou and Kambezidis, 2007; Li et al.,
2010; Kaplanis and Kaplani, 2010; Pandey and Soupir, 2012; Inman et al., 2013;......)
Multi Dimensional Filter Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 44/80
It is obvious from the literature that prediction performance increases in case of
different meteorological parameters are employed as inputs of models.
(Zarzalejo et al., 2005; Yorukoglu and Celik, 2006; Psiloglou and Kambezidis, 2007; Li et al.,
2010; Kaplanis and Kaplani, 2010; Pandey and Soupir, 2012; Inman et al., 2013;......)
Zi,j Zi,j+1
Zi+1,j
Ži+1,j+1=
?
How can we use this template?
Multi Dimensional Filter Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 45/80
Multi Dimensional Filter Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 46/80
1, 1,1 , ,1 1 , 1,1 2 1, ,1 3 , ,2 1 , 1,2 2 1, ,2 3 , ,3 1
, 1,3 2 1, ,3 3
i j i j i j i j i j i j i j i j
i j i j
Z Z a Z a Z a Z b Z b Z b Z c
Z c Z c
1, 1,1
1, 1,1 1, 1,1
i ji j i j
Z Z
1 2 3 1 2 3 1 2 3
0
a a a b b b c c c
11 12 13 14 15 16 17 18 19
21 22 23 24 25 26 27 28 29
31 32 33 34 35 36 37 38 39
41 42 43 44 45 46 47 48 49
51 52 53 54 55 56 57 58 59
61 62 63 64 65 66 67 68 69
71 72 73 74 75 76
81 82 83 8
91 92 93
r r r r r r r r r
r r r r r r r r r
r r r r r r r r r
r r r r r r r r r
r r r r r r r r r
r r r r r r r r r
r r r r r r
r r r r
r r r
1 1
2 2
3 3
1 4
2 5
3 6
77 78 79 1 7
4 85 86 87 88 89 2 8
94 95 96 97 98 99 3 9
a r
a r
a r
b r
b r
b r
r r r c r
r r r r r c r
r r r r r r c r
1.a R r
+1,+1,12
=1
=1
Multi Dimensional Filter Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 47/80
Performance Evaluat ion :
March 1, 2012 – Febuary
1, 2013
Multi Dimensional Filter Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 48/80
Deney Sonuçları:
1 Mart 2012 – 28 Şubat
0 2000 4000 6000 80000
200
400
600
800
1000
1200
1400
Zaman (Saat)
Ex
traterrestrialIşınım(W/m
)
0 2000 4000 6000 80000
200
400
600
800
1000
1200
Zaman (Saat)
G
ü n e ş I ş ı n ı m ı ( W / m 2 )
0 2000 4000 6000 8000 10000-20
-10
0
10
20
30
40
50
Zaman (Saat)
Sıcaklık
(0C
)
0 1000 2000 3000 4000 5000 6000 7000 8000 9000-300
-200
-100
0
100
200
300
Zaman (Saat)
Extraterrestrialışınımınbirincide
recedentürevi
0 2000 4000 6000 8000 10000-800
-600
-400
-200
0
200
400
600
800
Zaman (Saat)
G ü n e ş ı ş ı n ı m ı n ı n b i r i n c i d e r e
c e d e n t ü r e v i
0 1000 2000 3000 4000 5000 6000 7000 8000 9000-15
-10
-5
0
5
10
15
20
Zaman (Saat)
S ı c a k l ı ğ ı n
b i r i n c i d e r e c e d
e n
t ü r e v i
Multi Dimensional Filter Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 49/80
Performance Evaluat ion :
Multi Dimensional Filter Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 50/80
Performance Evaluat ion :
Data Employed
Mevcut very ile 1 saat sonrasının
tahmini (Filtre 1-4-7)
Model 1 Solar Radiation R(i+1)=a∗R(i)
Data Employed
Mevcut very ile 1 saat sonrasının
tahmini (Filtre 1-4-7)
Model 2 Temperature- Solar Radiations R(i+1)=a∗R(i)+b∗S(i)
Model 3
Extraterrestrial Radiations- Solar
Radiations
R(i+1)=a∗R(i)+b∗E(i)
Model 4
Future Extraterrestrial
Radiations- Solar Radiations
R(i+1)=a∗R(i)+b∗E(i+1)
Model 5
First order derivative oftemperature- Solar Radiations
R(i+1)=a∗R(i)+b∗( ∂S(i) )/∂t
Model 6
First order derivative of
temperature- Solar Radiation
R(i+1)=a∗R(i)+b∗( ∂E(i) )/∂t
Model 7
First order derivative of
radiation- Solar radiation
R(i+1)=a∗R(i)+b∗( ∂R(i) )/∂t
Model 8
First order derivative of future
extraterrestrials- Solar radiation
R(i+1)=a∗R(i)+b∗( ∂E(i+1) )/∂t
Multi Dimensional Filter Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 51/80
Performance Evaluat ion :
Kullanılan veriler Mevcut very ile 1 saat sonrasının
tahmini (Filtre 1-4-7)
Model 9
Sıcaklık-Extraterrestrial Işınım-
Güneş Işınımı R(i+1)=a∗R(i)+b∗E(i)+c∗S(i)
Model 10
Sıcaklık-Gelecek ExtraterrestrialIşınım- Güneş Işınımı
R(i+1)=a∗R(i)+b∗E(i+1)+c∗S(i)
Model 11
Geçmiş-Şimdiki ExtraterrestrialIşınım- Güneş Işınımı
R(i+1)=a∗R(i)+b∗E(i-1)+c∗E(i)
Model 12
Şimdiki-Gelecek Extraterrestrial
Işınım- Güneş Işınımı R(i+1)=a∗R(i)+b∗E(i)+c∗E(i+1)
Model 13
Extraterrestrial Işınım veSıcaklığın 1. mertebeden türevi –
Güneş Işınımı R(i+1)=a∗R(i)+b∗∂E(i)/∂t+c∗∂S(i)/∂t
Model 14
Gelecek Extraterrestrial Işınım veSıcaklığın 1.mertebeden türevi-Güneş Işınımı
R(i+1)=a∗R(i)+b∗∂E(i+1)/∂t+c∗∂S(i)/∂t
Multi Dimensional Filter Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 52/80
Resul ts:
RMSE(W/m2) MBE(W/m2)
Filter-1 Filter-2 Filter-3 Filter-1 Filter-2 Filter-3
Model-1 103,056 106,13 71,617 -8,7688 -9,4984 -15,4211
RMSE(W/m2) MBE(W/m2)
Filter-7 Filter-8 Filter-9 Filter-7 Filter-8 Filter-9
Model-2
101,896
104,941
69,974
5,9120
0,7507
-1,0896
Model-3
98,102
92,387
61,064
0,8737
5,1521
4,1404 Model-4 77,766 98,786 63,871 -8,9170 0,6794 7,7531
Model-5 90,512 105,799 68,820 7,1283 -8,9790 -1,6505
Model-6
65,858
105,838
60,990
4,2371
-9,5205
0,4748
Model-7 85,346 104,909 67,558 13,5672 -8,3142 -5,5875
Model-8 65,556 105,815 61,780 -0,7808 -9,5960 1,2491
RMSE(W/m2) MBE(W/m2)
Filter-4 Filter-5 Filter-6 Filter-4 Filter-5 Filter-6
Model-9 97,922 92,387 60,842 1,3655 4,9215 4,7141
Model-10 76,603 98,607 62,504 -7,1460 4,8505 1,3850
Model-11 65,187 92,107 X 0,5263 5,7969 X
Model-12 63,390 93,261 X -5,5989 1,3302 X
Model-13 65,667 105,247 60,739 4,1965 -0,7403 3,9799
Model-14
64,767
105,685
61,118
-1,0290
-8,8602
6,2795
Multi Dimensional Filter Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 53/80
Model-3 98,102
Model-4 77,766
Resul ts:
Model-2 101,896
Model-5
90,512
Model-3 98,102
Model-6 65,858
Extraterrestrial ışınımın mevcut değeri yerinegelecek değerinin kullanıldığı durum
Sıcaklık yerine türevinin kullanıldığı durum
Extraterrestrial ışınım yerine türevinin kullanıldığıdurum
Model-9
97,922
Model-10 76,603
Model-11 65,187
Model-12 63,390
Model-13 65,667
Model-14
64,767
R(i+1)=a R(i)+b E(i)+c S(i)
R(i+1)=a
R(i)+b
E(i+1)+c
S(i)
R(i+1)=a
R(i)+b
E(i-1)+c
E(i)
R(i+1)=a
R(i)+b
E(i)+c
E(i+1)
R(i+1)=a
R(i)+b
∂E(i)/∂t+c
∂S(i)/∂t
R(i+1)=a R(i)+b ∂E(i+1)/∂t+c ∂S(i)/∂t
Multi Dimensional Filter Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 54/80
Resul ts:
0 200 400 600 800 1000 12000
100
200
300
400
500
600
700
800
900
1000
Ölçülen Güneş Işınım Değeri (W/m2)
TahminEdilenGüneşIşınımDeğeri(W/m
2 )
Multi Dimensional Filter Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 55/80
Performance Evaluat ion :
05
1015
2025
0
100
200
300
400-1000
-500
0
500
1000
SaatGün
TahminHatası(W/m
2 )
Multi Dimensional Filter Approach
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 56/80
Results:
It is possible to use different meteorological parameters ralated with solar
radiaion data in case proposed multi dimensional approach is used.
14 different models with 9 different templates are proposed.
It is obtained that future and past values of extraterrestrial radiation carry
important information for the prediction model.
It is observed that the performance of the prediction is increased in case the
first and second order derivatives of solar radiation values are employed.
In solar radiation prediction one day before values also carry information about
prediction. On the other hand one hour before values carry much moreinformation.
Multi Dimensional Filter Approach
Analytical Modeling Approach for
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 57/80
Analytical Modeling Approach for
Solar Radiation Data
Novel Analytical Modeling
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 58/80
Novel Analytical Modeling
Approach for Solar Radiation Data
Novel Analytical Modeling
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 59/80
Novel Analytical Modeling
Approach for Solar Radiation Data
2 2( )( ) x b c g x ae
2 2 2 21 1 2 2( ) ( )
1 2( ) x b c x b c
g x a e a e
Novel Analytical Modeling
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 60/80
Novel Analytical Modeling
Approach for Solar Radiation Data
a b c
D1 285.6000 12.9500 2.7280
D2 226.7000 13.4000 2.8410
D3 301.8000 12.9500 2.7710 D4 235.1000 12.1700 2.8930
D5 201.1000 14.5600 1.8690
D6 229.7000 12.7800 3.0010
D7 134.3000 13.2700 3.2470
D8 122.9000 12.9200 3.4460
D9 298.9000 12.1100 2.8850
D10 103.1000 12.3000 3.3970
Novel Analytical Modeling
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 61/80
Novel Analytical Modeling
Approach for Solar Radiation Data
1 Source Gauss 2 Source Gauss
D1 0.9907 0.9981
D2 0.982 0.9971
D3 0.9881 0.9986
D4 0.9596 0.9949
D5 0.9242 0.9974
D6 0.9136 0.9889
D7 0.9293 0.9907
D8 0.9561 0.9964
D9 0.9882 0.9899
D10 0.9155 0.9978
Novel Analytical Modeling
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 62/80
Novel Analytical Modeling
Approach for Solar Radiation Data
Novel Analytical Modeling
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 63/80
Novel Analytical Modeling
Approach for Solar Radiation Data
Novel Analytical Modeling
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 64/80
Novel Analytical Modeling
Approach for Solar Radiation Data
1-Source
Gauss
2-source
Gauss
RMSE 57.20 105.25
A novel modeling approach using
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 65/80
A novel modeling approach using
extraterrestrial solar radiations
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 66/80
Extraterrestrial Formulation
2sin( ) / 24 I C R
sin( ) sin( )sin( ) cos( ) cos( )cos( ) L D L D h
15( 12)ch h
2D Representation of
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 67/80
2D Representation of
Extraterrestrial Data
Modeling Solar Radiations from
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 68/80
Modeling Solar Radiations from
Extraterrestrial Radiations
Modeling Solar Radiations from
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 69/80
Modeling Solar Radiations from
Extraterrestrial Radiations
Novel Hybrid (Wind/PV) Sizing
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 70/80
Novel Hybrid (Wind/PV) Sizing
Algorithm
WL ≤ WPV + WG + WB
Novel Hybrid (Wind/PV) Sizing
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 71/80
Novel Hybrid (Wind/PV) Sizing
Algorithm
PV m m f pc P A P I
0
ci
r ci r
r ci
w r ci
ci co
V V P V V V
V V
P P V V V
V V ve V V
Novel Hybrid (Wind/PV) Sizing
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 72/80
Novel Hybrid (Wind/PV) Sizing
Algorithm
( ) ( 1) ( ( ) ( ) )Wg Gw Pv GP LDSOC t SOC t N E t N E t E
( ) Max
SOC t SOC
( ) Min
SOC t SOC
/ f LLP T T
Novel Hybrid (Wind/PV) Sizing
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 73/80
Novel Hybrid (Wind/PV) Sizing
Algorithm
( , , ) ( , ) Pv Wg bat Pv Wg Q N N N F N N
Novel Hybrid (Wind/PV) Sizing
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 74/80
Novel Hybrid (Wind/PV) Sizing
Algorithm
• For any number of Pv ( ) and any number ofwind generator ( ) are
calculated and versus t are
plotted This curve is named as characteristic
curve of remaining-deficient energies.
• For all time t , the following sum is calculated for
accumulation of the characteristic curve
• The time periods in which the sum in Eq. Above
is negative are counted.
Pv N
Wg N ( ) ( ) ( )
Gw GP LDt
E t E t E t
( ) ( ) ( )Gw GP LD
E t E t E t
1
( ( ) ( ) ( ))Gw GP LD
t
E t E t E t
Novel Hybrid (Wind/PV) Sizing
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 75/80
Novel Hybrid (Wind/PV) Sizing
Algorithm
• The collected count from previous step is divided to
the total considered period of time to find LLP
• Here constitutes a subset of the longest
time period satisfying the inequality in the numerator.
1
#( ( ( ) ( ) ( ))) 0Gw GP LD
t
E t E t E t
LLP T
1, 2,...,T
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 76/80
Capacity Calculation Illustration
-30
-20
-10
0
10
20
30
40
50
60
Remaining/D
eficit
Energies(W)
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31
t (Day)
Novel Hybrid (Wind/PV) Sizing
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 77/80
Novel Hybrid (Wind/PV) Sizing
Algorithm
• Where OC is the optimum capacity of batteries, N is
the length of x[t] sequence for which is greaterthan zero, and NUM is the length of N for the
maximum accumulation
0
arg max [ ] 0 N
N t
NUM x t
0
[ ]
N
t
OC x t
0 [ ]
N
t
x t
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 78/80
E i l S d
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 79/80
Experimental Study
8/18/2019 Fatıh Onur Hocaoğlu Res_2015
http://slidepdf.com/reader/full/fatih-onur-hocaoglu-res2015 80/80
Thank You
Fatih Onur Hocaoglu