Estimation of potential solar energy in MTI region (Malaysia, Thailand and Indonesia) based on linear, nonlinear and artificial neural network models
Solar radiation (SR) data offer information on the amount of the sun potential at a location on the earth during a specific time. These data are very important for designing sizing solar photovoltaic (PV) systems. Due to the high cost of installation and fitting troubles, these barriers cause lack...
Saved in:
| Format: | Thesis |
|---|---|
| Language: | English |
| Subjects: | |
| Online Access: | http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/77971/1/Page%201-24.pdf http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/77971/2/Full%20text.pdf http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/77971/4/Hudhaifa.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| id |
my-unimap-77971 |
|---|---|
| record_format |
uketd_dc |
| spelling |
my-unimap-779712023-03-06T00:30:58Z Estimation of potential solar energy in MTI region (Malaysia, Thailand and Indonesia) based on linear, nonlinear and artificial neural network models Syed Alwee Aljunid, Syed Junid, Prof. Dr. Solar radiation (SR) data offer information on the amount of the sun potential at a location on the earth during a specific time. These data are very important for designing sizing solar photovoltaic (PV) systems. Due to the high cost of installation and fitting troubles, these barriers cause lack of data and make data availability difficult. Prediction models for solar radiation are the key solution to substitute these important data and cover the missing from it. Therefore, there is a demand to develop alternative ways of predicting these data. The zone of Malaysia, Thailand and Indonesia (MTI), which are part of southeast Asia (SEA), is a huge area Had no model can cover all regions but only individual models assigned to particular countries. On the other hand, the zone (MTI) had practiced many types of modeling techniques for solar radiation prediction, with variation in its prediction attitude and results accuracy; hence, it is very important to implement a comparison between models in order to find the most accurate one. Best prediction model according to accuracy, need to be compared with other similar neighbor models within the same zone. This study presents linear, non-linear models as MTI linear and MTI nonlinear models in order to develop a standardization modeling technique in this zone and Artificial neural network (ANN) models has been implemented also in the same area to predict its global and diffuse solar radiation. The different models have been tested in different areas. These areas a r e classified as zone, region and globally. It is found that the zone and region models are accurate and could be used to predict solar radiation, which is an interested achievement. Nevertheless, global models have a high error percentage. The results showed that the ANN models are accurate in comparison with the nonlinear and linear models in which the mean absolute percentage error (MAPE) in calculating the solar energy in Malaysia by the ANN model is 5.3%, while the MAPE for the MTI nonlinear and linear models is 6.4%, 7.3% respectively. In addition, the root mean square error (RMSE) shows the following promising results, 7.2% for ANN model and 8.1%, 8.5% for the MTI nonlinear and linear models respectively. Finally, the mean bias error (MBE) comes up with these next results ANN model is -1.3%, the MTI nonlinear model is -1.1% and MTI linear model is -1.1%. Universiti Malaysia Perlis (UniMAP) Thesis en http://dspace.unimap.edu.my:80/xmlui/handle/123456789/77971 http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/77971/3/license.txt 8a4605be74aa9ea9d79846c1fba20a33 http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/77971/1/Page%201-24.pdf 946cc542424c80e75caa4c5838da2fee http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/77971/2/Full%20text.pdf e5ba46e19d7b890917c3ce2a4b347aa7 http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/77971/4/Hudhaifa.pdf a15a7fa8c894f2a15019d622c8429eac Universiti Malaysia Perlis (UniMAP) Solar energy Solar radiation Photovoltaic power systems Neural networks (Computer science) School of Computer and Communication Engineering |
| institution |
Universiti Malaysia Perlis |
| collection |
UniMAP Institutional Repository |
| language |
English |
| advisor |
Syed Alwee Aljunid, Syed Junid, Prof. Dr. |
| topic |
Solar energy Solar radiation Photovoltaic power systems Neural networks (Computer science) |
| spellingShingle |
Solar energy Solar radiation Photovoltaic power systems Neural networks (Computer science) Estimation of potential solar energy in MTI region (Malaysia, Thailand and Indonesia) based on linear, nonlinear and artificial neural network models |
| description |
Solar radiation (SR) data offer information on the amount of the sun potential at a location on the earth during a specific time. These data are very important for designing sizing solar photovoltaic (PV) systems. Due to the high cost of installation
and fitting troubles, these barriers cause lack of data and make data availability difficult. Prediction models for solar radiation are the key solution to substitute these important data and cover the missing from it. Therefore, there is a demand to develop alternative ways of predicting these data. The zone of Malaysia, Thailand and Indonesia (MTI), which are part of southeast Asia (SEA), is a huge area Had no model can cover all regions but only individual models assigned to particular countries. On
the other hand, the zone (MTI) had practiced many types of modeling techniques for solar radiation prediction, with variation in its prediction attitude and results accuracy; hence, it is very important to implement a comparison between models in
order to find the most accurate one. Best prediction model according to accuracy,
need to be compared with other similar neighbor models within the same zone. This
study presents linear, non-linear models as MTI linear and MTI nonlinear
models in order to develop a standardization modeling technique in this zone and
Artificial neural network (ANN) models has been implemented also in the same
area to predict its global and diffuse solar radiation. The different models have
been tested in different areas. These areas a r e classified as zone, region and
globally. It is found that the zone and region models are accurate and could be
used to predict solar radiation, which is an interested achievement. Nevertheless,
global models have a high error percentage. The results showed that the ANN
models are accurate in comparison with the nonlinear and linear models in which
the mean absolute percentage error (MAPE) in calculating the solar energy in
Malaysia by the ANN model is 5.3%, while the MAPE for the MTI nonlinear and
linear models is 6.4%, 7.3% respectively. In addition, the root mean square error
(RMSE) shows the following promising results, 7.2% for ANN model and 8.1%,
8.5% for the MTI nonlinear and linear models respectively. Finally, the mean bias
error (MBE) comes up with these next results ANN model is -1.3%, the MTI
nonlinear model is -1.1% and MTI linear model is -1.1%. |
| format |
Thesis |
| title |
Estimation of potential solar energy in MTI region (Malaysia, Thailand and Indonesia) based on linear, nonlinear and artificial neural network models |
| title_short |
Estimation of potential solar energy in MTI region (Malaysia, Thailand and Indonesia) based on linear, nonlinear and artificial neural network models |
| title_full |
Estimation of potential solar energy in MTI region (Malaysia, Thailand and Indonesia) based on linear, nonlinear and artificial neural network models |
| title_fullStr |
Estimation of potential solar energy in MTI region (Malaysia, Thailand and Indonesia) based on linear, nonlinear and artificial neural network models |
| title_full_unstemmed |
Estimation of potential solar energy in MTI region (Malaysia, Thailand and Indonesia) based on linear, nonlinear and artificial neural network models |
| title_sort |
estimation of potential solar energy in mti region (malaysia, thailand and indonesia) based on linear, nonlinear and artificial neural network models |
| granting_institution |
Universiti Malaysia Perlis (UniMAP) |
| granting_department |
School of Computer and Communication Engineering |
| url |
http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/77971/1/Page%201-24.pdf http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/77971/2/Full%20text.pdf http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/77971/4/Hudhaifa.pdf |
| _version_ |
1776104256447709184 |