Artificial intelligent platform to increase the ...

Artificial intelligent platform to increase the performance of energy

production of solar panels

By

Mohammed Ali Alghareeb

Supervised:

Dr.Walid Fahs & Dr.Jamal Haydar

1

Outline

Introduction

Problem statement

The factors effect on Solar cell

Weather forecasting system

Artificial neural network

Methodology

Result and testing

System advantages

Conclusion

2

Introduction

We use artificial intelligence

neural network (ANN) to

solve the dust problem on the

solar cell surface by

depending on the factors that

reduce the performance of

the solar cell .

3

Problem statement

4

The main problem to reduce the performance of the solar cell is dust

18,98 15,2 15,1

18,05 18 17,3 19,1 16,24 16

18,29 18,3 17,9

0

5

10

15

20

25

25/2/2018 23/3/2018 7/4/2018 5/5/2018 28/5/2018 24/6/2018

Open Circuit voltage Curve of clean and unclean Solar Panel

Unclean Clean

This chart shows the effect of the dust on Voltage

5

Dust Test

The factors effect on Solar cell

Temperature

Humidity

Weather

6

Temperature

19,9 19,74 18,16 17,48 16,35

0

5

10

15

20

25

15 C 25 C 30 C 40 C 50C

Temoerature

Voltage VS Temperature

Voltage [V]

7

This chart shows the effect of the Temperature on Voltage

17,1

16,72

16,53 16,45 16,41

16,33 16,32

15,8

16

16,2

16,4

16,6

16,8

17

17,2

25 30 35 40 45 50 55

Humidity

Voltage VS Humidity

This chart shows the effect of the Humidity on Voltage

Humidity

8

Day Monocrystalline

Multi crystalline

Sunny Day Higher power Higher power

Cloudy Day Low power Low power

9

Weather

This table shows the effect of the weather on solar cell

Weather

Weather forecasting system

Sunny

Good

Partially Cloudy

Bad

Rain

Very bad

10

Weather forecasting system

Five years data to predict the weather

11

Weather forecasting

Artificial neural network

12

Biological neuron model

A computer system similar to the human neural system

Methodology

13

Artificial Intelligence System to Predict Dust on Solar Cell surface

Result and testing

14

We apply on (Arduino, raspberry pi, Matlab) the results of the test lead to

the error rate (0.25) the below chart shows that

MATLAB simulation

System advantage

Water: The system uses 75% less water than manual-cleaning methods.

Automatically management: This system can be used in any place (forestry,

desert, mountains, hills snow area)

Less energy: Due to less times of cleaning.

smart and autonomous: Sensors help in receiving the data that are related to

weather, then start the cleaning process keeping in mind the weather

condition

Remote management: the cleaner’s system help monitoring and manage the

cleaning of solar panels from any place.

15

Conclusion

we find If solar cells

are not cleaned for a

month, solar cells

performance will be

reduced to 50%

which means we

must clean the

surface of the solar

cell always.

16

17