Artificial neural network technique to predict the power output of photovoltaic for monocrystalline and polycrystalline
The demand for energy is predicted to rise rapidly in the near future as a result of population development and industrialization around the world. However, increased use of fossil fuels is responsible for the majority of environmental pollution and greenhouse gas emissions, which are widely b...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English English |
| Published: |
2022
|
| Subjects: | |
| Online Access: | http://eprints.uthm.edu.my/6966/1/24p%20ABDOU%20MANI%20MOHAMED.pdf http://eprints.uthm.edu.my/6966/2/ABDOU%20MANI%20MOHAMED%20COPYRIGHT%20DECLARATION.pdf http://eprints.uthm.edu.my/6966/3/ABDOU%20MANI%20MOHAMED%20WATERMARK.pdf http://eprints.uthm.edu.my/6966/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The demand for energy is predicted to rise rapidly in the near future as a result of
population development and industrialization around the world. However, increased
use of fossil fuels is responsible for the majority of environmental pollution and
greenhouse gas emissions, which are widely believed to be the primary drivers of
global warming and are contributing to it. This project represent the design of artificial
neural network model (ANN) that predict the power output of the photovoltaic (PV)
for monocrystalline and polycrystalline. The objectives of this project is to develop
ANN model, to evaluate power and efficiency of two different photovoltaic panel. The
data was collectedfrom 5 May 2018 to 6 May 2020. However, the input parameters are
metreological data is used as input for ANN model. The voltage produced by
polycrystalline is much more lager than monocrystalline voltage. In contrast,
monocrystalline PV panel tend to have a higher current values compared to
polycrystalline PV pnael. Mean square error (MSE) training of this model was equal
to MSE testing and MSE validation. It means the data of model have been learning
very well during training and zero means that it has an overestamte the prediction of
the network. It is clear that the two models have a very good fit curve of the data as
the correaltion coefficient, R value is equal to 1. However, the atual and predeictd
values show a similarity in trends for both PV modules. The estimated voltage, current
and power when compared to the actual value has no significant differences. Overall,
polycrystalline panel has a better performance and the efficiency was 0.999% and
0.997% for moncrystalline. |
|---|