PERAMALAN RADIASI SINAR MATAHARI JANGKA SANGAT PENDEK PADA PORTABLE SOLAR WATER PUMP MENGGUNAKAN METODE DECOMPOSITION-FEED FORWARD NEURAL NETWORK

VERY SHORT-TERM SOLAR RADIATION FORECASTING ON PORTABLE SOLAR WATER PUMP USING DECOMPOSTION-FEED FORWARD NEURAL NETWORK METHOD

ABSTRAK

Portable Solar Water Pump merupakan energi terbarukan yang dapat mengkonversi radiasi matahari ke dalam bentuk  energi listrik. Selanjutnya Portable Solar Water Pump digunakan untuk irigasi pertanian tanaman tembakau  Radiasi matahari mendorong terjadinya peningkatan penyerapan photovoltaic, yang menyebabkan peningkatan kinerja life time baterai. Sel surya mempunyai nilai efisiensi tinggi apabila foton yang berasal dari sinar matahari bisa diserap sebanyak-banyaknya.Penelitian ini bertujuan mendesain Portable Solar Water Pump  Lalu meramalkan radiasi sinar matahari menggunakan metode Hybrid Decomposition – Feed Forward Neural Network. Dan mendapatkan nilai eror menggunakan MSE sebesar 1,16, RMS 1,07 , MAD 8,6 dan Mape sebesar 0,9 %. Dengan nilai radiasi tertinggi sebesar 1338.4 W/.

Kata kunci : Portable Solar Water Pump, Decomposition,Feed Forward Neural Network, Decomposition-Feed Forward Neural Network

ABSTRACT

Portable Solar Water Pump is a renewable energy that can convert solar radiation into electrical energy. Furthermore, the Portable Solar Water Pump is used for agricultural irrigation of tobacco plants. Solar radiation encourages an increase in photovoltaic absorption, which causes an increase in battery life time performance. Solar cells have a high efficiency value if as many photons from sunlight can be absorbed as possible. This research aims to design a Portable Solar Water Pump and then predict solar radiation using the Hybrid Decomposition - Feed Forward Neural Network method. And get an error value using MSE of 1.16, RMS 1.07, MAD 8.6 and Mape of 0.9%. With the highest radiation value of 1338.4 W/m^2.

Keywords: Portable Solar Water Pump, Decomposition, Feed Forward Neural Network, Decomposition-Feed Forward Neural Network