Скачать или смотреть 78 - PV Power Prediction based on Artificial Neural Network Optimized by Genetic Algorithm

Khadija Lmesri, Samira Chabaa, Mohammed Ali Jallal, Nasima El Assri, Sihame Nachat and Abdelouhab Zeroual
Code: (S3104_ID078)
Paper Title : PV Power Prediction based on Artificial Neural Network Optimized by Genetic Algorithm

Presented in:
9th International Renewable and Sustainable Energy Conference #IRSEC21
IEEE Conference
Online, November 23-27, 2021
Workshop 3 - The Advances in Renewable Resources Forecasting and their Economic
https://www.med-space.org/irsec21

Presented by :
Khadija Lmesri
Physics department Faculty of Sciences Semlalia, Cadi Ayyad University
Marrakesh, Morocco

Abstract - Hourly solar photovoltaic (PV) power measurements in a specific period and site are very important for designing and supervising PV systems. However, in most cases these measures are absent. The focus of this research study is to offer for PV systems engineers an efficient model based on Artificial Neural Network (ANN). To predict the output power of a PV system installed in Agadir city (320°25’39”N 9°35’53”W), Morocco. The architecture has a great impact on the convergence of the Neural Network. This architecture is optimized using Genetic Algorithm (GA). The aim of this approach is to find the optimal number of hidden layers, the activation function and the initial vector of weights. The attained results confirm the accuracy of the proposed model to predict the hourly PV power output in case of missing real values. Consequently, electrical power engineers can adopt this prediction model to design and supervise the PV systems and integrated it into the power grid."