Predicting PV cell performance by artificial neural network

Other Title(s)

التنبؤ بكفاءة الخلايا الشمسية باستخدام نظام الشبكة العصبية

Dissertant

al-Masfih, Suhaib Ibrahim

Thesis advisor

Abu Zayd, Mahmud
al-Dhiyabat, Salah

University

Mutah University

Faculty

Faculty of Engineering

Department

Mechanical Engineering Department

University Country

Jordan

Degree

Master

Degree Date

2017

English Abstract

The input layer of ANN used in this work comprises the operating parameters that affect PV cell performance including solar irradiation, temperature, hourly clearness index ,wind speed, relative humidity, hour angle, and number of day, the hourly performance data of a PV panels collected in the period from 13:00 PM April 14/ 2017 to 19:00 PM May /19 2017 were randomly separated into two parts ;two thirds of the data was used for ANN training while the remaining third was used for ANN validation.

The best results were obtained when ANN comprise with two layers with 8 and 5 nodes, respectively; therefore 7-8-5-1 ANN structure has been adopted in this work and has been used to predict the PV efficiency at different outdoor conditions The results showed that ANN with two hidden layers has given the best modeling results with R2=0.9646 and the mean absolute percentage error (MAPE) equals 7.36% and 15.77% for training and validation, respectively.

The model showed that the general trend was that both polar N-S and full tracking systems have given the highest output power and performance ratio.

The improvement in cell performance when using tracking systems was in the range from less than 1% to 24 % while both full tracking and polar N-S tracking gives the highest improvement

Main Subjects

Civil Engineering

No. of Pages

54

Table of Contents

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-780443