THREE LAYERED FEED-FORWARD NEURAL NETWORK BASED ESTIMATION OF OUTPUT POWER AND ENERGY ON PHOTOVOLTAIC (PV) MODULES
| Main Authors: | Syafaruddin, Wardi, Gassing, Zaenab Muslimin, Zulkifli Tahir, Engin Karatepe, Takashi Hiyama |
|---|---|
| Format: | Article |
| Bahasa: | eng |
| Terbitan: |
, 2013
|
| Subjects: | |
| Online Access: |
http://repository.unhas.ac.id/handle/123456789/6444 |
Daftar Isi:
- The paper presents the application of three layered feed-forward neural network for predicting the output power and energy generation of photovoltaic (PV) modules. The neural network structure is designed to determine the estimated output power of PV modules based on the input of open circuit voltage and cell temperature. The open circuit voltage is one of the electrical parameters output from PV modules based I-V characteristic models from Sandia National Laboratory. The estimated energy of PV generation is obtained by connecting the estimated output power with simple mathematical integration process. The proposed method is tested to PV modules from different manufacturers and technologies, such as monocrystalline Silicon, Cadmium Telluride (CdTe) and triple junction amorphous Silicon. The simulation results show the high accuracy prediction compared to the conventional data measurements.