Estimation of Hourly Solar Radiation on Horizontal Surface Using GAMF (Genetic Algorithm Modified Fuzzy) (Case Study in Surabaya)

Indonesia Malaysia Research Consortium Seminar 2018

Setyawan, Erma Hakim^1^2 ; Abadi, Imam^1 ; Kusumawarni, Sartika Arie^1

Engineering Physics Departement, Faculty of Industrial Technology, Sepuluh Nopember Institute of Technology, Surabaya

60111, Indonesia^1

SMAN 1 Porong, Sidoarjo, Indonesia^2

关键词: Adaptive neuro-fuzzy;    Extreme learning machine;    Horizontal surfaces;    Installed capacity;    Photovoltaic performance;    Photovoltaic systems;    Renewable energy source;    Solar radiation estimation;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/588/1/012024/pdf
DOI  :  10.1088/1757-899X/588/1/012024

来源: IOP

【 摘 要 】

The biggest renewable energy sources in Indonesia is solar energy. The installed capacity of Solar Power System in 2017 is still very far from the target. Solar radiation is very infulental on the photovoltaic performance in generating energy. The need for solar radiation estimation has become important in the design of photovoltaic system. Previous research has been done using Extreme Learning Machine (ELM) and Adaptive Neuro Fuzzy (ANFIS) methods. In this research, Genetic Algorithm Modified Fuzzy (GAMF) method used to estimate the solar radiation. Meteorogical dates used in this research are temperature, humidity, wind velocity and wind direction. There are 2 types of datas that are BMKG data and measurement data. Three experimental variations of input variation were performed for each data. For BMKG data the best estimation result is obtained when using humidity, temperature and wind velocity as variation inputs with RMSE and MAE of 145.19 and 72. While for the best result estimation result data obtained when applying humidity, temperature, wind speed and wind direction as variation inputs with values of RMSE and MAE were 1.44 and 0.65.